Advances in Rheology

PHYSICS RESEARCH AND TECHNOLOGY
ADVANCES IN RHEOLOGY RESEARCH
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
Additional books in this series can be found on Nova’s website

under the Series tab.
Additional e-books in this series can be found on Nova’s website

under the eBooks tab.
ADVANCES IN RHEOLOGY RESEARCH
MARÍA DOLORES TORRES PÉREZ

EDITOR
Copyright © 2018 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or
transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical
photocopying, recording or otherwise without the written permission of the Publisher.
We have partnered with Copyright Clearance Center to make it easy for you to obtain permissions
to reuse content from this publication. Simply navigate to this publication’s page on Nova’s
website and locate the “Get Permission” button below the title description. This button is linked
directly to the title’s permission page on copyright.com. Alternatively, you can visit
copyright.com and search by title, ISBN, or ISSN.
For further questions about using the service on copyright.com, please contact:
Copyright Clearance Center
Phone: +1-(978) 750-8400 Fax: +1-(978) 750-4470 E-mail: info@copyright.com.
NOTICE TO THE READER

The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or
implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of
information contained in this book. The Publisher shall not be liable for any special,
consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or
reliance upon, this material. Any parts of this book based on government reports are so indicated
and copyright is claimed for those parts to the extent applicable to compilations of such works.
Independent verification should be sought for any data, advice or recommendations contained in
this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage
to persons or property arising from any methods, products, instructions, ideas or otherwise
contained in this publication.
This publication is designed to provide accurate and authoritative information with regard to the
subject matter covered herein. It is sold with the clear understanding that the Publisher is not
engaged in rendering legal or any other professional services. If legal or any other expert
assistance is required, the services of a competent person should be sought. FROM A
DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE
AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS.
Additional color graphics may be available in the e-book version of this book.
Library of Congress Cataloging-in-Publication Data
ISBN: H%RRN
Published by Nova Science Publishers, Inc. † New York

CONTENTS
Preface vii
Chapter 1 Rheology and 3D Printing 1
María Dolores Torres Pérez
and Javier Seijo
Chapter 2 Rheological Characterisation of Injection Grouts
using Rotational Rheometry 13
Luis G. Baltazar, Fernando M. A. Henriques
and Maria T. Cidade
Chapter 3 Liquid CO2-Based Gel Fluid 43
Weixiang Cui, Mingyue Cui
and Chunpeng Wang
Chapter 4 Rheology Applied to Food Product Design 63
Anabela Raymundo, Ana Paula Batista
and Isabel Sousa
Chapter 5 Mistibility and Viscoelastic Properties of
Poly(Styrene-Co-Acrylonitrile) Blends 119
Yuji Aoki
Chapter 6 Rheology of Honey 175
Clara Tovar, María Shantel Rodríguez-Flores,
Olga Escuredo and María del Carmen Seijo
vi Contents
Chapter 7 Unusual Nonlinear Rheological Behavior

of Branched Polymers: Multiple Stress Overshoots
in Shear Startup - Experiments and Theory 193
Federico Sebastiano Grasso, Salvatore Coppola,
Fabio Bacchelli, Giovanni Ianniruberto
and Giuseppe Marrucci
Chapter 8 Viscosity and Viscoelasticity of Baby Foods 219
María Dolores Alvarez, Wenceslao Canet
and Beatriz Herranz
Chapter 9 Rheology as a Tool for Food Development 289
Suzana Caetano da Silva Lannes
Chapter 10 Using the Laplace Transform in Rheology Kinetic Modelling 309
I. T. Smykov
Chapter 11 The Design of Vibration Absorbers using the Rheological
Properties of Viscoelastic Materials 347
Juan Carlos Jauregui-Correa
Chapter 12 A Viscoelastic Fluid Due to a Nonlinear Accelerating
Elastic Sheet 381
U. S. Mahabaleshwar
Chapter 13 An MHD Boundary Layer Viscoelastic Fluid Flow
Past a Stretching/Shrinking Sheet in a Porous Medium 395
P. N. Vinay Kumar
Chapter 14 Thermal Radiation Effect on Fully Developed Laminar
Mixed Convection Flow in a Vertical Porous Stratum
by using Differential Transform Method 409
Patil Mallikarjun and U. S. Mahabaleshwar
About the Editor 433
Index 435
PREFACE
The authors of this volume discuss current advances in rheology research for
different application fields. The chapters include discussions on rheology and 3D
printing; rheological characterization of injection grouts using rotational rheometry;
analysis of rheological properties of 100% liquid CO2 based gel fracturing fluid; rheology
applied to food product design; miscibility and viscoelastic properties of poly(styrene-co-
acrylonitrile) blends; rheology of honey; unusual nonlinear rheological behavior of
branched polymers: multiple overshoots in stress growth, experiments and theory;
viscosity and viscoelasticity of baby foods; rheology as an instrument for food
development; using Laplace Transform in rheology kinetic modeling; design of vibration
absorbers using the rheological properties of viscoelastic materials; a viscoelastic fluid
due to a non-linear accelerating elastic sheet; an MHD boundary layer viscoelastic fluid
flow over a stretching sheet in a porous medium; and thermal radiation effect on fully
developed laminar mixed convection flow in a vertical porous stratum by using
differential transform method.
Chapter 1 - This mini review, covering scientific papers published from 2014 until
now, deals with the critically relevant role of the thermo-rheological properties of the
materials in 3D printability. Monitoring of the thermo-rheological behaviour of the
materials during processing is a key factor in their manufacturing. For this purpose,
relations between materials mechanical characteristics and 3D printability are studied.
The current trends in thermo-mechanical characterizations of printable materials are
presented. This review covers a broad range of printable materials and final applications
from food to non-food industries. All 3D printing applications are a huge challenge, with
potential applications in industrial design, aerospace, architecture, medical, tissue
engineering, and even food. Three types of printing materials (i.e., natively printable
material, non-printable traditionally materials, and alternative materials) are described.
Special emphasis was put on describing and discussing the quality of 3D printing. The
majority of works on 3D printing for non-food applications has been performed with oil-
viii María Dolores Torres Pérez
based printing materials, which are still used because they have a higher softening point
and make more flexible models that will bend before they break. Possible future
tendencies in new biopolymer materials and the improvement of their themo-rheological
features for 3D printing were also discussed in this review.
Chapter 2 - The overall goal of this chapter is to contribute to the understanding and
improvement of the rheological performance of injection grouts for consolidation of
stone masonry buildings. It is a well known fact that old masonries such as heritage or
ordinary buildings represent a large portion of the construction area in many European
cities. They frequently present a bad state of conservation and have particular technical
needs and specific weaknesses that require special approaches. Grout injection is an
interesting technique to repair and consolidate such stone masonries. It consists of
introducing a suspension (grout) into the masonry inner core in order to upgrading the
cohesion of the masonry and increasing its monolithic behaviour. Therefore, it is essential
to evaluate the effectiveness of the grouts in terms of keeping their adequate flow
capacity. For this purpose, rheology appears as a useful tool in the controlling and
optimisation of grouts design. Notwithstanding, during rheological measurements many
disturbing effects may arise, such as the wall slip, which can play a significant role
leading to an inaccurate rheological parameters determination. Despite the large number
of research related to wall slip on several fluids with yield stress, the wall slip is still a
concern for rheological measurements of grouts. Thus, this study aims to determine the
degree of wall slip during rheological measurements of natural hydraulic lime-based
grouts for stone masonry consolidation. For the determination of wall slip stress
controlled measurements at various gaps and different shear stresses were performed.
Overall, the results revealed that wall slip is function of the shear stress and the grout
composition. Results from this research can help to define an adequate measurement
protocol to correctly infer the rheological properties of injection grouts based on
rotational rheometry.
Chapter 3 - Liquid carbon dioxide (CO2) fracturing technology is one of the waterless
fracturing technologies. It has the advantages of avoiding the use of water, reducing
damage, and increasing production by improving the reservoir pressure. However, the
low viscosity of liquid CO2 and the difficulty to pump proppant are the main technical
challenges. No conventional thickening agent can provide the required thickening used in
100% liquid CO2. In this paper, the problem of thickening gelled liquid CO2 has been
solved by developing a new surfactant and thickening agent. Surface active agent can
make the liquid CO2 gel by forming wormlike micelle or cylinder micelles structure. The
cylinder micelles enhance the viscosity through entanglement inter-actions above a
critical concentration. Rheology and friction properties of liquid CO2 gel are evaluated
through high pressure pipeline rheological experimental device. The viscosity of liquid
CO2 gel fluid changes under different pressures (10 to 20 MPa), temperatures (0 to
100℃) and different shear rates. The experiment result shows that the effective viscosity
Preface ix
of liquid CO2 gel fluid is between 7.654 ~ 20.012mPa.s. The liquid CO2 was thickening
with viscosity increase at 86 to 498 times under the same conditions. Compared with the
pure CO2, the viscosity enhancement ratio of liquid CO2 based gel was at 86 to 498 times,
and the mixed system of liquid CO2 based gel showed the characteristic of shear thinning.
Pressure of CO2 thickening of cross-linked liquid mixture viscosity has little effect, the
effective viscosity only has a small amplitude increase; the effective viscosity of liquid
mixtures was decreased with the increase of temperature, which has an exponentially
decreasing trend. Considering all the factors, the effects of shear rate and temperature
have the greatest impact on rheological properties of fracturing fluid. Temperature on the
rheological parameters influence is bigger, whether under liquid or supercritical
condition, the liquid CO2 fracturing fluid flow index n increases with the increase of
temperature, and consistency coefficient K decreases with increase in temperature
decreases, the effect of pressure and thickening agent ratio on the rheological parameters
are smaller. Liquid CO2 based gel fracturing fluid provides a new fracturing system with
less formation damage to fracturing operation. This chapter will be beneficial to all
engineers and technologists who are currently working at waterless fracturing
applications.
Chapter 4 - Innovation is recognised as the key driver of economic growth. Trends in
food industry deal with a permanent need to develop new food products, adjusted to the
consumer demands and, in the near future, to scarcity of food resources. Concepts of food
as health and wellness promoters, “free-from” products, the use of alternative ingredients
such as new protein sources, or the use of by-products in food formulations are current
topics that act as driving forces for innovation. Currently, food product development
methodologies are generally based on the chemical and nutritional properties,
complemented with a sensory validation, carried out in the final stages of the
development process. However, the structure of food proved to be determinant for the
food appeal and strongly impacts consumers’ acceptance. It is well known that products
with the same chemical composition can present very different structures, resulting
differently perceived texture and sensory properties. Therefore, the use of rheological
tools to design new formulations and to optimise new industrial processes, from the
earliest stages of the development phases, proves to be a promising methodology. In the
creative process, the food macromolecules (proteins and polysaccharides) are the major
players for the creation of relevant food structures such as foams, emulsions and gels. For
the purpose, successful case studies will be presented and described with a focus on the
use of rheological techniques as the tools for the design of new food products based on
alternative macromolecules. The development of gluten-free or vegetarian products by
using alternative proteins and polysaccharides, as well as the use of food industry by-
products as source of these structuring biopolymers, along with the structural
implications of adding protein and/or fibre-rich healthy ingredients (e.g. microalgae,
psyllium) are some of the examples where rheology tools can give a powerful insight and
x María Dolores Torres Pérez
contribute to decision-making on product development stages. It will be emphasized how

the incorporation of new ingredients induces changes in the viscosity profile and other
rheological parameters, related to the consistency of the fluid food products or to the
texture of the solid or soft-solid foods. The use of small amplitude oscillatory
measurements will also be discussed in order to access the changes of the internal food
structure promoted by changes on processing procedures or by the inclusion of new food
ingredients on the food formulation. The impact of these variables on the development of
the food structure will be discussed based on the linear viscoelastic properties. The use of
sensory and instrumental texture measurements, in different experimental modes and
applied to different foodstuffs, will be discussed. The possibility to correlate instrumental
texture parameters with sensory evaluation will be considered as the strategy to bring the
products characteristics of the developed products to the consumer’s requirements.
Chapter 5 - This article gives a review of the miscibility and viscoelastic properties of
poly(styrene-co-acrylonitrile) (SAN) blends. SAN is miscible with poly(methyl
methacrylate) (PMMA), poly(-caprolactone) (PCL), poly(-methyl styrene)-co-
acrylonitrile (MSAN), poly[styrene-co-(maleic anhydride)] (SMA), poly[styrene-co-(N-
phenylmaleimide)] (SMI), and so on. The miscibility of SAN is mainly due to the
presence of unfavorable interaction between styrene (S) and acrylonitrile (AN)
comprising the copolymer. First, the authors explain the miscibility of SAN with PMMA,
SMA, and SMI. The phase behavior of MSAN/SAN blends is also explained in relation
to slow phase separation due to low mobility. The  parameters for monomer (i)-
monomer (j) interaction are evaluated from the miscibility map. Second, the authors
report the viscoelastic properties in the glass transition region. Time-temperature
superposition principle can be applied over the temperature range measured for miscible
SMA/SAN blends, but cannot be applied for immiscible blends although the principle
can be applied below or above the Tg of high Tg component. Temperature dependence of
the shift factor aT can be expressed in WLF equation denoted by aT = 8.86(T 
Ts)/(101.6  T  Ts), when the Ts is chosen to be fitted in the data. Third, the authors
report the viscoelastic properties of SAN blends in the terminal flow region. Miscible
blends exhibit the terminal tails (G2 and G and behaves as a viscous fluid
having G G at low frequencies. On the other hand, immiscible blends do not show
the terminal tails and have the long time relaxation due to a inhomogeneous structure.
Lastly, the authors discuss compositional dependences of the zero-shear viscosity 0 and
plateau modulus GN0 of miscible blends are discussed in relation to entanglement
couplings. It is emphasized that the viscoelastic properties of miscible blends should be
compared at an iso-free volume condition. Then, the entanglement molecular weights in
the blends are identical with those in the component (co)polymers. The plateau modulus
GN0 of miscible blends is identical with those in the component (co)polymers and varies
linearly with the blend composition.
Preface xi
Chapter 6 - Honey is a complex fluid with viscoelastic properties predominantly

similar to those of liquids but with certain elastic component. Properties of honey are
influenced by composition, temperature, sugar crystals and principally moisture content.
In most research works, honey is reported to be a Newtonian liquid, but the presence of
natural colloids contributed to their time dependant behaviour (thixotropy). For the
present study two kinds of measurements have been made: first, from small amplitude
oscillatory shear (SAOS) tests, to determine the viscoelastic moduli in the linear
viscoelastic (LVE) range, when storage (G) and loss (G) moduli are independent on the
shear strain/stress applied. SAOS tests allow evaluate the time-dependence of G and G
providing the mechanical properties of honey at rest. Second, it has been obtained the
flow curves which give the relationship between the shear viscosity and the shear rate to
obtain the flow curves, showing honey changes in response to continuous deformation. In
the samples analysed, the power law model indicated a weak shear thinning behaviour. In
addition, it has been assayed the time-dependence of flow behaviour using a step test
with three intervals: reference, high-shear interval and regeneration interval to test
thixotropy. These experimental proofs have been completed applying upward ramp, hold
time, and downward ramp to determine the hysteresis area and verify the possible
thixotropy. Moreover, the viscoelastic behaviour has been analysed by transient tests
which provide creep and creep-recovery curves, using stress values (0) (during creep
stage), in the LVE range. These tests serve to compare the viscosity values obtained from
flow curves, and at the same time to obtain a percentage of elasticity after removing load
at the rest phase. Temperature (T) has a most significant role on the viscosity of honey its
temperature dependence was examined applying a constant shear rate (50 s-1) and an
upward T-ramp, at constant heating rate (1°C/min) from 20 to 50 °C. In all samples
studied the viscosity values followed an Arrhenius relationship and the activation energy
(Ea) was obtained.
Chapter 7 - Rheological properties of polymers for elastomer applications are
strongly dependent on their molecular structure. Most commercially available elastomers
are entangled polydisperse, branched polymers with an unknown molecular structure.
From the experimental point of view, rheological measurements can be carried out in the
linear and nonlinear viscoelastic regimes. Both type of measurements can be very fruitful
in characterizing polymeric materials and, if combined, they can disclose peculiar
physical properties providing a deep insight into molecular structure. That is because the
two regimes deal with very different aspects of molecular dynamics. As in the literature
the linear viscoelastic regime is widely explored, in this chapter the authors choose to
focus on nonlinear rheological measurements for a number of elastomers, i.e.,
Polybutadiene (BR), styrene-butadiene copolymers (SBR) and nitrile butadiene (NBR)
rubber, which differ significantly in branching content. For such materials, single point
data (e.g., Melt Flow Index or Mooney Torque) are unable to describe the rheological
features of polymers with complex branching pattern. A significant example is the case of
xii María Dolores Torres Pérez
emulsion-polymerized elastomers, characterized by broad Molar Mass Distribution and

branched architecture, both depending on the conversion degree and temperature attained
in the reactors. Furthermore, polymers with a highly branched molecular structure can
display multiple overshoots during shear startup experiments at constant shear rate. Such
overshoots can be related to complex molecular dynamics, such as chain orientation,
chain stretching, and branchpoint withdrawal, the extent of the latter being strongly
dependent on the previous deformation history. It is well known that shear startup data
collected by means of a Mooney-type viscometer can display stress overshoots,
particularly prominent for randomly branched materials. Noticeably, similar results were
recently obtained also for model branched polymers tested with a strain controlled
rheometer equipped with a cone-partitioned plate geometry. The extremely slow
attainment of a steady state, and the high degree of nonlinearity are related to a
significant branching content of the high molecular weight fraction of the material.
Although scientifically intriguing, as well as relevant for polymer processing, such
phenomena have been seldom investigated in the rheological literature. Only recently,
this phenomenology has been the subject of careful experimental works, as well as of
attempts of theoretical understanding. The aim of this work is to present some original
results together with a short review of some of the landmark experimental results and
theoretical understanding gained in the last years on this topic.
Chapter 8 - Traditionally, infants have been supplemented by ready-to-eat semisolid
formulated foods Composition of these baby foods is being continually modified and
adapted with the aim of producing healthier foods. However, there are no recent
rheological studies of baby foods. A comprehensive study on flow and viscoelastic
behavior of four current commercial types of Spanish vegetable-based infant purees was
carried out. The effect of a wide range of temperatures, which are encountered during
processing and storage of baby foods, on rheological properties is documented. Flow
behavior of vegetable-based infant purees was analyzed at different temperatures (5-65
°C) giving particular attention to their time-dependent properties in a shear rate range (5-
200 s–1). Power law model parameters describing flow behavior of samples depended on
kind of infant puree, its water content and measurement temperature. Arrhenius model
was used to explain temperature effect on apparent viscosity at 50 s–1. Infant purees
exhibited thixotropic behavior for all temperatures tested. For the same temperature,
differences in hysteresis loop magnitudes were observed among purees, being more
noticeable at lower temperatures. Two models were used to describe the time-dependent
behavior, namely Weltman model, and second-order structural kinetic model. For all
infant purees, the initial shear stress and the extent of thixotropy increased and decreased
significantly with increases in shear rate and temperature. The breakdown rate of puree
associations also accelerated at higher shear rates, but no trend was observed with
temperature. In turn, the viscoelastic properties of the four vegetable-based infant purees
were evaluated in temperature range of 5 to 80 °C. Samples behaved like weak gels, with
Preface xiii
the exception of rice and chicken puree at 35, 50 and 65 °C, which behaved like a
macromolecular solution. At 5-65 °C, storage modulus (G') and storage loss (G") could
be well described by a power function of the frequency (R2 > 0.92), and the dependency
of their parameters with temperature was modeled by Arrhenius’s model and quadratic
functions. Approach named ‘weak gel model’ was also applied to the baby foods, and
both G' and G" values decreased with an increase in temperature between 5 and 50 °C.
Master curves were obtained applying time-temperature superposition (TTS) principle,
and horizontal shift factor was sensitive to choice of viscoelastic property being selected
for shifting procedure. At 5-50 °C, infant purees behaved like thermoreologically
complex materials. Rheological measurements were well supported by particle size
distributions. Controlling temperature, moisture content and particle size distribution may
be employed to optimize applicability, stability, and sensory of baby foods. Results
provide a useful fingerprint of rheological behavior of these products. Generated data and
findings can be extrapolated to other infant systems for developing new products for
convenience to newborn consumers and manufacturers.
Chapter 9 - Food designers always thing in a way to find the best setting of
ingredients to manage the formulations. Process and formulations input conditions to a
product with good properties. The interactions between ingredients can lead to a several
physical situations. When small stresses are applied to the product it should feel an
impression that is related to a high viscosity. At a certain stress, the yield stress, the
viscosity starts to decrease rapidly. The lower the yield stress the easier the product can
be distributed on the surface, and it can determines the thickness of the layer of the
product on the surface. Rheology, as a tool for evaluating liquid, semi solid and solid
products, can contribute to an optimal result of structural arrangement of the formulated
product. Therefore, this chapter aims to discuss some rheological concepts to help
formulators find the more accurate way to get the best profile of obtained products in
selected examples.
Chapter 10 - The Chapter aims to provide a modern state of art in rheology,
balancing theoretical principles and their practical application. In the Chapter the utility
of direct and inverse Laplace transform for known rheological model descriptions is
demonstrated. Laplace transform considered to be very powerful math tool, useful in
different fields of science and technology. With increasing complexity of technical
problems Laplace transform gives real support in solution of complex problems with very
simple approaches. During modeling with the use of Laplace transform the studying
object is defined as the object under technological impact: input and interfering signals
evoke output reaction. It is demonstrated that kinetic changes in rheological parameters
during time can be considered as transfer function (transient response) for the
investigated object changing from one state to another under influence of input signals.
As the examples of Laplace transform practical use, some our results are provided
concerning the kinetic changes in rheological parameters during fermentative phase sol-
xiv María Dolores Torres Pérez
gel transition in the milk during cheese making. The methodological significance of the
Chapter is provided by examples of deep interconnections between pure theoretical
models and practically important systems.
Chapter 11 - The design of vibration absorbers, in general, combine materials with
high rigidities and very elastic materials. The main purpose of these designs is to reduce
the transmission of vibration waves along structures and diminish the damage in
machinery. There are many methods for reducing the transmission of vibration waves;
but the most common are passive vibration isolators. These isolators are constructed with
a metal base and an elastomer that acts as a damper. Vibration isolation is the most
widely used technique for vibration and noise protection, because it breaks the
transmitted wave by modifying the structural stiffness of the system, and by adding a
higher damper. The design of these devices represents a big challenge, even though these
devices have been used for decades, the theoretical models and numerical solutions are
valid only within limited geometries and specific designs. There are two major
limitations: one is the development of an appropriate elastomer model that relates the
rheological properties with the actual dynamic behavior, and the numerical formulation
of the dynamic model that includes the elastomeric properties and the rigid behavior of
the metal components.
There have been different approaches for solving these limitations, but the analytical
results are far from the experimental result, except in well-known cases. In this chapter, a
detail analysis of the different formulations for design vibration absorbers is presented. In
this analysis, the definition of the dynamic behavior of viscoelastic materials based on
their rheological properties is included and three numerical formulations are presented,
namely: Finite Differences and Finite Elements. The results are illustrated with practical
examples and the basis of more sophisticated tools that allows the design of complex
vibration isolators are included.
Chapter 12 – The present chapter deals with the steady, laminar boundary layer flow
of non-Newtonian fluid flow due to a nonlinear accelerating sheet. The accelerating of
the sheet is presumed to be atleast a quadratic function in the axial distance. The study
encompasses two classes of viscoelastic fluids namely, Walters’ liquid B and second
order liquid. Similarity solution is used to obtain the velocity profiles which are governed
by a highly nonlinear differential equation. Analytical expressions are obtained for stream
function and velocity profiles as functions of the accelerating and viscoelastic parameters.
The flow filed is influenced by the physical parameters like Chandrasekhar number,
nonlinear accelerating parameter and mass transpiration parameter. The results so
obtained are in line with the classical works with nonlinear accelerating to enhance the
exciting of the liquid as it moves away from the slit. When the liquid is at rest, the effect
of nonlinear stretching parameter is null, the expression for the stream function coincides
with that of Rajagopal (1984) and Siddheshwar and Mahabaleshwar (2005). The presence
of viscoelastic and nonlinear accelerating sheet parameter has significant impact on
Preface xv
physical properties of the laminar boundary layer flows due to a nonlinear accelerating
sheet. This physical phenomenon has got wide applications in the industrial processes
such as cooling of the liquid in the extrusion processes and fabrication of glass,
manufacturing of polymer fibers by means of the metal spinning, coiling of wires and
other allied industrial processes.
Chapter 13 - The present chapter deals with the steady, laminar boundary layer flow
of an incompressible viscoelastic fluid past a stretching/shrinking sheet in the porous
medium and magnetohydrodynamics (MHD). The flow of viscoelastic fluid past a
stretching /shrinking sheet is widely considered for mathematically modeling the problem
due to its significant applications in the industrial processes such as wire drawing,
manufacture of polymer fibres and plastic fibres, in the glass blowing and to increase the
performance in the lubricants production. This flow problem is mathematically modeled
into a nonlinear system of partial differential equations and which are transformed via
similarity transformation reducing it to Navier-stokes equation of dimension two into
system of ordinary differential equations. These equations are solved exactly for
analytical solution. The presence of MHD greatly influences the flow geometry due to the
Lorentz force acting against the flow field. In addition to this, the presence porosity in the
medium, results in the Darcy drag force which further supports the opposition to the flow
field. This combined effect is much advantageous as these two forces acts as one. Also,
the outcome depends on this physical parameter which fits into practical problems of
chemical and metallurgical industries. The effects of various physical parameters like
Chandrasekhar number (Q), inverse Darcy number (Da-1) and suction/injection (Vc)
parameters on the flow geometry are discussed graphically. The results so obtained are in
line with that of the classical works and the limiting cases are provided to give good
comparision.
Chapter 14 - An analysis is carried out to study the flow and heat transfer to analyze
the effect of thermal radiation on combined free and forced convection flow in a parallel-
plate vertical channel taking into account the effect of porous medium with symmetric
and asymmetric wall heating conditions. Three types of thermal boundary conditions are
described. These thermal boundary conditions are isothermal- isothermal, isoflux-
isothermal and isothermal-isoflux for the left-right walls of the channel. The results are
represented by graphically for various governing parameters such as porous parameter,
the ratio of Grashof number to Reynolds number, radiation parameter, Brinkman number
and perturbation parameter for equal and different wall temperatures. The coupled non
linear governing equations are also solved analytically using regular perturbation method.
It is also found that both the analytical and differential transform method solutions agree
very well for small values of the perturbation parameter. In the present study, the DTM
and perturbation method were used to find analytical solutions of mixed convection
problem in the presence of porous medium, radiation effect and viscous dissipation. It
was found that DTM is powerful method for solving problems consisting of systems of
xvi María Dolores Torres Pérez
non linear differential equations. A comparison was made between perturbation method
solutions, available published results and the present approximate solutions. The
numerical results indicate that the DTM gives more accurate in comparison to shown
methods. The method has been applied directly without requiring linearization,
discretization or perturbation. The obtained results certify the reliability of the algorithm
and give it a wider applicability to non linear differential equations. Also some major
observations are: the velocity and temperature increases for mixed convection parameter
and perturbation parameter where as it decreases for increasing the porous medium and
radiation parameter.
In: Advances in Rheology Research ISBN: 978-1-53612-875-8
Editor: María Dolores Torres Pérez © 2018 Nova Science Publishers, Inc.
Chapter 1
RHEOLOGY AND 3D PRINTING
María Dolores Torres Pérez* and Javier Seijo

Department of Chemical Engineering and Biotechnology,
New Museums Site, University of Cambridge, Cambridge, UK
ABSTRACT
This mini review, covering scientific papers published from 2014 until now, deals
with the critically relevant role of the thermo-rheological properties of the materials in
3D printability. Monitoring of the thermo-rheological behaviour of the materials during
processing is a key factor in their manufacturing. For this purpose, relations between
materials mechanical characteristics and 3D printability are studied. The current trends in
thermo-mechanical characterizations of printable materials are presented. This review
covers a broad range of printable materials and final applications from food to non-food
industries. All 3D printing applications are a huge challenge, with potential applications
in industrial design, aerospace, architecture, medical, tissue engineering, and even food.
Three types of printing materials (i.e., natively printable material, non-printable
traditionally materials, and alternative materials) are described. Special emphasis was put
on describing and discussing the quality of 3D printing. The majority of works on 3D
printing for non-food applications has been performed with oil-based printing materials,
which are still used because they have a higher softening point and make more flexible
models that will bend before they break. Possible future tendencies in new biopolymer
materials and the improvement of their themo-rheological features for 3D printing were
also discussed in this review.
Keywords: 3D printer, mechanical features, outstanding technologies, viscoelasticity
*
Corresponding Author Email: mariadolores.torres.perez@usc.es.
2 María Dolores Torres Pérez and Javier Seijo
INTRODUCTION
3D printing is more than just a novelty – it is the future of manufacturing industries

(Jia et al., 2016). The majority of work on 3D printing for non-food applications has been
made with oil-based printing materials, which are still used because they have a higher
softening point and make more flexible models that will bend before they break (Godoi et
al., 2016). All 3D printing applications are a huge challenge, with potential applications
in a broad range of fields such as industrial design, aerospace, architecture, medical,
pharmaceutical, tissue engineering, and even food (Sun et al., 2015; Jonathan and Karim,
2016). This emerging technology has the potential to revolutionize not only the non-food
industries but also the food production by boosting culinary creativity, food
sustainability, and nutritional customizability (Tian et al., 2016; Torres et al., 2014,
2016), but still many research is necessary to extend their versatility and overcome the
scientific and technical barriers, transforming the challenge in reality.
Monitoring of the thermorheological behaviour of the printable materials during
processing is a key factor in their manufacturing (Abbadessa et al., 2016). An optimum
thermorheological characterization of different mixtures commonly used in 3D printing is
crucial to select the best processing conditions and extend the understanding framework
of these materials applications (Jonathan and Karim, 2016). Note here that the
preparation of inks with proper rheology performance is critically relevant for extrusion
based 3D printing (Zhong et al., 2017). Latter authors explained that in order to make the
ink mixture printable, proper rheology performance is required. Namely, the ink should
possess shear-thinning properties, allowing smooth extrusion from the nozzle under
shear, and the extruded ink must be strong enough to keep its shape and withstand the
weight of postextruded inks at rest.
In a comprehensible review (Aho et al., 2015), it was indicated that the melt
processability of polymers is highly dependent on their rheological characteristics.
Additionally, latter authors stated that rheological measurements should be considered as
a more central part of the material characterization tool box when selecting suitable
candidates for 3D printing. Recently, Chen et al., (2017) stated that the ability to control
the materials rheology yields a flexible manufacturing route to fabricating 3D metal parts
with full density and complex geometries. Note here that specialy important is the
comprehensive study of the rheology of bubbly liquids in order to obtain light printable
structures or aerated foodstuff with optimal mechanical properties (Torres et al., 2015).
Nevertheless, the relations between the thermorheological properties of hydrogels and 3D
materials printability have not been extensively studied (Huijun et al., 2016). It should be
highlighted that some recent works have suggested, as a critically relevant tool, the use of
Computational Fluid Dynamics (CFD) models to establish a connection between the
paste thermorheology, printing parameters and the line profile (Lee and Lee, 2016;
Leppiniemi et al., 2017).
Rheology and 3D Printing 3
This rapidly developing technology has huge potential in order to improve the
population wellbeing, being critically relevant the selection of suitable printable materials
and the corresponding thermomechanical characterization. In this context, the main
objectives of this review were to summary the different types of printable materials, the
current trends in the corresponding thermomechanical characterizations and the
thermorheological features of 3D printable materials for different applications, covering
scientific papers published from 2014 until now. Specific emphasis will be given to the
relations between thermorheological behaviour and 3D printable materials quality. Future
trends focused on alternative materils will be also proposed.
TYPES OF PRINTING MATERIALS: NATIVELY PRINTABLE,

AND NON-PRINTABLE TRADITIONALLY MATERIALS
Oil-based printring materials have been extensively used in 3D-printing for non-food
applications because their adequate mechanical features (Godoi et al., 2016). For
applications as those from food industry, where oil-based materials are not suitable,
notable efforts have been made to obtain/pre-process materials adequate for 3D printing,
and raise their thermal stability features during post-processing. In this sense, in a recent
comprehensive review (Sun et al., 2015), the available printing materials were classified
into two categories based on their printability characteristics: natively printable and non-
printably traditionally materials.
In above review, it was described that natively printable materials such as hydrogel,
cake frosting, cheese, hummus or chocolate could be extruded smoothly from 3D printing
syringe. Sugars, starch, or mashed potatos blends were also tested as powder materials for
3D printer. In this line, it was indicated that a number of sugar teeth were fabricated for
demonstration. Other authors (Gong et al., 2014), indicated that some traditional foods
were tested for printability study, being the most successful material pasta dough,
analysed by viscosity, consistency and solidifying features. Above authors stated that
foodstuff made by natively printable materials can be fully controlled on taste, nutritional
properties, and texture. Other composite formulations like batters and protein pastes
usually need a post-deposition cooking process. This will make food product structures
more difficult to keep their shapes. Those natively printable materials, which are stable
enough to hold the shape after deposition, do not require further post processing and,
consequently can be reserved for medical and space applications.
Another foodstuff such as rice, meat, fruits or vegetables, commonly consumed, are
not printable by nature. In above review, it was indicated that to improve their extrusion
capability, the addition of hydrocolloids to these solid materials has been extensively
studied for many applications. It was reported elsewhere (Lille et al., 2017) the use of
simple additives and ingredients to modify traditional food recipes and created complex
geometries and novel formulations. Even though semi-solid liquids and solid foods have
already been manipulated to become printable by gastronomic tricks, it is difficult to test
and modify the whole list. One possible solution was to employ a small ingredients group
in order to create a platform with several freedom degrees on texture and flavor (Tian et
al., 2016). By fine tuning hydrocolloids’ content, a very broad textures range can be
reached. Several authors (see as e.g., Lille et al., 2017) stated that experimented food
texture employing two polymeric systems, and explored structural requirements for post-
processing materials such as protein pastes and cake mixtures.
CURRENT TRENDS IN THERMO-MECHANICAL CHARACTERIZATIONS

OF PRINTABLE MATERIALS
Rheological measurements, in terms of steady-shear flow curves (apparent viscosity),

oscillatory data (viscous and elastic moduli), creep and recovery curves or temperature
sweeps, have been extensively used to characterize materials and select suitable
candidates for 3D-printing (Tubío et al., 2016). In this work, aqueous colloidal ZnO inks
with tailored rheological properties based on above rheological measurements have been
developed for 3D printing of periodic structures. In general, shear-thinning behavior
allows flow through fine nozzles without clogging. Viscoelastic behavior enables printing
of self-supporting features. In order to determine the longest possible print time of the
uncured ink material after preparation, it should be examined the time-dependent
rheology of the ink at room temperature (Doss et al., 2014). Recently, in a comprehensive
review (Aho et al., 2015) it has been indicated that the melt processability of polymers
and their mixtures is greatly dependent on the rheological characteristics of these
systems. Moreeover, latter authors stated that rheological measurements should be
considered as a more central part of the material characterization tool box when selecting
suitable candidates for melt processing by three dimensional 3D printing or other new
emerging technologies. Above authors explained that the polymer processing industry
offers established methods, platforms, and models for rheological characterization, and
they can often be readily applied in different fields from food to non-food industries as in
the pharmaceutical manufacturing. They highlighted that measured and calculated
rheological parameters together with thermal and mechanical material data are essential
for the process simulations, which are also becoming increasingly important. In
aforementioned review, the authors gave an overview to the basics of rheology and
collected many studies examples where rheological measurements have been utilized in
setting up or evaluating extrusion processes. In this line, authors provided numerous
examples of different experimental set-ups available for rheological measurements,

discussing in detail each of their typical application area, advantages and restrictions.
In a completely different field, Rodríguez et al., (2017) used several rheological
measurements to determine suitability of silk-based biopolymer blends for 3D printing in
order to obtain soft tissue implants. Bio ink formulations were determined to be optimal
based on: i) flow behavior that would allow for stable extrusion over the shear rate ranges
expected during printing, ii) physiologically relevant melting (30-40 ºC) and deposition
temperatures (20 - 25 ºC), and iii) formation of insoluble material when incubated under
physiological conditions. Unconfined compression testing has been also conducted to
estimate the mechanical properties of the material. Above rheological measurements
allowed obtaining a broad range of physical and mechanical properties suitable for tissue
reconstruction.
Other authors, Chen et al., (2017) introduce an alternative approach to additive
manufacturing of metals for which new rheological approaches are necessary. Their
approach has been realized through controlling the microstructure of semi-solid alloy
slurries to access an appropriate rheological behavior, namely a well-suited viscosity
range and a shear-thinning characteristic, which allow steady state printing of 3D metallic
structures. A facile control of the microstructural features during solidification can be
mediated via manipulating the rheological protocols.
Overall, it seems to be clear that rheological measurements should be considered as a
standard part of the material characterization when selecting adequate candidates for
processes like 3D printing. Adequately designed processing solutions will provide the
basis for safe and robust manufacturing of innovative final products.
THERMORHEOLOGICAL FEATURES OF 3D PRINTABLE MATERIALS:

QUALITY PARAMETERS
Thermorheological features are critically revelant for all processing technologies, but
especially in emerging technologies as 3D printing, as indicated in several works. Some
of the most representatives for different fields are summarized below.
Huijun et al., (2016) investigated comprehensively the rheological characteristics and
the 3D printability of alginate-based hydrogels. These authors studied the influence of
CaCl2 amount and alginate content on the gelation features of alginate in aqueous
solution. Later authors reported that the critical CaCl2 content at the gel point rose
linearly with increasing alginate content, showing that much more CaCl 2 are required to
cross-link alginate chains into gel networks at a larger alginate content. In this work, it
was highlighted that the alginate/CaCl2 hydrogels exhibited shear-thinning behaviour,
however the shear-thinning or thixotropic features of alginate/CaCl2 hydrogels were not
significant enough for 3D printing. In contrast, the alginate hydrogels thixotropic feature
and recovery time were pointed out as critically relevant to control the printability. The
authors stated that the thixotropic characteristics are essential since indicate how rapidily
and how much hydrogel viscosity can recover after printing. In the same line, the
recovery time is also important since it is the time given to the hydrogel for recovering its
viscosity during a 3D printing process. In this work, the authors found that the alginate
hydrogel is not an ideal biomaterial for 3D printing with an extrusion-based printer owing
to its low ability of viscosity recovery, however the viscosity recovery of the alginate
hydrogel could be improved by the presence of a small amount of graphene oxide. The
authors defined some measureable parameters to describe and discuss the 3D printing
quality. To sum up, the works provides a new approach to the analysis of 3D printability
of a hydrogel.
In a contemporary work (Zhu et al., 2016), an extrusion-based 3D printing platform
was nicely developed to fabricate ultratough polyion complex hydrogels into various 3D
structures. The viscous complex solution was straight extruded out of nozzle into water,
and tough physical hydrogels with fixed 3D structures were reached after sol−gel
transition caused by dialyzing out of salt and counterions of ultratough polyion complex
solution. The printability of the complex solution was studied by adjusting the material
formulation and printing parameters. The printed PIC structures exhibited excellent
mechanical properties in terms of extensibility, strength, and toughness, evaluated
rheologically. The authors indicated that the 3D printing technology based on sol−gel
transition provides an easy approach to fabricate tough ultratough polyion hydrogels into
complex structures, which may be used in other types of tough physical hydrogels as
well, and where the thermorheological characterization is essential. The authors pointed
out that above study is of immediate use in promoting the practical applications of tough
hydrogels in structural biomaterials, soft actuators, and so forth.
In another contemporary work (Tang et al., 2016), nanocellulose/CdS quantum dot
composites were fabricated by controlling the carboxylate content of the nanocellulose
and the molar ratio of Cd2+/–COOH. New photoelectric inks were prepared based on the
composites, in which the CdS quantum dots acted as the pigment and the nanocellulose
as the binder. The rheological characterization was critically relevant in this work, as the
photoelectric ink was stable with excellent fluidity (viscosity 30.3 mPa s) and rheology.
Consequently, the authors remarked that above ink could be applied to screen-printing
and three-dimensional 3D printing.
Another authors (Feildin et al., 2016) were able to reach the much-coveted goal of
reliable, complex ceramic parts with low porosity and high strength using robocasting as
a 3D printing technique. In this study a robust hydrogel formulation was optimised for
use as the extrusion paste for robocasting. The paste’s rheological features were
determined and the printing process was optimised with the objective of attaining dense
monolithic ceramic parts. The authors indicated that the pastes showed shear thinning
behaviour with yield stresses that can reach values above 1 kPa and depend mostly on
their solid content and the particle size distribution. In the paper, it was indicated that it is
possible to formulate printable Al2O3and SiC inks with solid contents as high as 40 vol%
with suitable flexural strengths. It should be remarked that the sources of strength
limiting defects were identified by rheology and related to the printing process.
Other relevant study about rheology and 3D printing has been published by (Boetker
et al., 2016). These authors prepared custom-made feedstock material for 3D printing that
could be used for varying the release of API from 3D printed objects. In this work,
rheological measurements indicated that the flow features were sensitive to the content of
undissolved particles. This work showed the importance of rheological measurements as
well as the potential of custom-made, drug loaded feedstock materials for 3D printing of
precision pharmaceutical drug products for controlled release.
Later on, Zhong et al., (2017) reported the extrusion based 3D printing of
geopolymer/GO nanocomposites for the first time. These authors reported that the
presence of GO can notably modify the rheology of geopolymer precursors, and enable
the 3D printing of geopolymer, which suggests a strong GO-HGPP interaction. The
authors explained this fact by the assumption that a thick layer of water film is formed in
the confined space between GO and ASOP, which contribute greatly the complex
modulus of the blend. In the work it was indicated that the obtained geopolymer/GO
nanocomposites exhibited adequate mechanical performance. They remarked that their
strategy maked it possible to explore the employment of GO as rheology property
modifier by encapsulation, and enable 3D printing materials that was un-printable
previously, and thus offer opportunity to create novel and innovative 3D printing
materials.
Other relevant work has been also recently on this topic (Wang et al., 2017a). These
authors presented a study on the importance of functional characteristics of surimi gel for
an additive layer manufacturing process. Their outcomes indicated that the rheological
characteristics of surimi gel with 1.5 g/100 g NaCl is suitable for printing which is related
to the possible protein structural modifications by the presence of salt. These authors
highlighted improvements in gel strength and network structure with increasing sodium
chloride addition. Besides material properties, the outcomes of process optimisation for
3D printing indicated that the nozzle height, nozzle diameter, the nozzle movement rate
and the extrusion speed were also essential parameters affecting the geometry accuracy of
the 3D surimi printing. The authors indicated thanks to rheological characterizations that
2.0 mm nozzle diameter and 5.0 mm nozzle height were the optimal features to provide
adequate bonding between layers to maintain proper geometry.
In a contemporary work, Wang et al., (2017b) studied the influence of cellulose
nanofibrils (CNF) amount and a coupling agent on the rheological characterisitics and
flexural features of CNF-PP polypropylene composites for 3D printer filaments.
Rheological measurements indicated that elastic modulus, complex viscosity, viscosity,
and transient flow shear stress of polypropylene were rose by around 33%, 25%, 15%,
and 27% at the chosen frequency and shear rates after 10 wt% cellulose nanofibrils was
added into the propylene. In this work, it was also observed that creep strain was reduced
with the presence of 10 wt% CNF because of increased rigidity. The damping factor and
stress relaxation time remained the same even at 10 wt% CNF addition as the CNF–PP
interaction is weak. The presence of coupling agent rose the complex viscosity, elastic
modulus, viscosity, transient flow shear stress, and creep strain of CNF-filled PP and
dropped the stress relaxation of these composites. They also indicated that the flexural
strength and modulus of PP were increased by 5.9% and 26% after 10 wt% CNF was
added into PP. To sum up, the above authors indicated that presence of CNF into PP,
through a “fast masterbatch production process” method, marginally changed the
rheological characteristics from a practical consideration, which is a critically relevant
achievement. The authors indicated that small variation in rheological characteristics at
lower shear rate brought by CNF make the resulting PP composites filament process
friendly to material extrusion devices.
In another work (Hamilton et al., 2017), it was indicated that the compatibility of
commercially available foodstuff, marmite and vegemite, can be used to fabricate 3D
structures, like eye-catching food designs and edible circuitry onto bread substrates. The
authors indicated the relevance of characterising the rheological features of the food
products in retaining the structural integrity of the printed materials. They also explained
that the electrical conductivity of both products can be used to print edible circuits on
“breadboards.” The authors pointed out that the importance of rheological
characterization in the development of food processing techniques that have the potential
to be used in a broad of applications from space missions, aged-care facilities to
healthcare.
Following with the food field, Lille et al., (2017) evaluated the applicability of
extrusion-based 3D printing technology for food pastes formulated with starch, protein
and fiber-rich materials, as a starting point in the development of healthy, customized
snack products. The authors evaluated the starch-, cellulose nanofiber-, milk powder-,
oat- and faba bean protein-based materials printability and their mixtures using
rheological experiments. In this work, it was indicated that rheological measurements
revealed that the shape stability after printing was linked with the paste yield stress.
Rheology was also critically relevant during postprocessing to select the optimum drying
conditions. Extrusion-based 3D printing, jointly with rheological characterization, are
two promising tools for producing healthy, structured foods, although further research is
needed for optimising the thermomechanical features of the printed materials.
FUTURE TRENDS
Further research into new materials, as biopolymers, suitable for 3D printing and the
corresponding thermorheological characterization is essential in order to improve the
current mechanical properties of available products and extend the possibilities towards
innovative applications. The future of biopolymer materials lies in demonstrating that
plant-based materials from natural, underused and renewable sources can outperform
their traditional, oil-based counterparts. Biopolymers have recently gained attention for
new manufacturing technologies as 3D printing, because they are much easier to work
with during processing, and are food safe and odour free (Lille et al., 2017). Namely, 3D
biopolymer-based food printing offers a great potential for preparing customizable foods
with improved nutritional features through enriched foods, tailored to the needs of
specific target groups like people with food intolerances, elder people, athletes or
astronauts (Godoi et al., 2016). The social experience in the design freedom a new food is
also critically relevant, since social media will support recipes and cooking pre- and post-
fun communities.
REFERENCES
Aho, J., Boetker, J. P., Baldursdottir, S., Rantanen, J. 2015. Rheology as a tool for
evaluation of melt processability of innovative dosage forms. International Journal
of Pharmaceutics, 494, 623–642.
Abbadessa, A., Blokzijl, M. M., Mouser, V. H. M., Marica, P., Malda, J., Hennink, W.
E., Vermonden, T. 2016. A thermo-responsive and photo-polymerizable
chondroitinsulfate-based hydrogel for 3D printing applications. Carbohydrate
Polymers, 149, 163-174.
Boetker, J., Water, J. J., Aho, J., Arnfast, L., Bohr, A., Rantanen, J. 2016. Modifying
release characteristics from 3D printed drug-eluting products. European Journal of
Pharmaceutical Sciences, 90, 47–52.
Chen, W., Thornley, L., Coe, H. G., Tonneslan, S. J. Vericella, J. J., Zhu, C., Duoss, E.
B., Hunt, R. M., Wight, M. J., Apelian, D., Pascall, A. J., Kuntz, J. D., Spadaccini, C.
M. 2017. Direct metal writing: Controlling the rheology through microstructure.
Applied Physical Letters, 110, 094104; doi: 10.1063/1.4977555.
Duoss, E. B., Weisgraber, T. H., Hearon, K., Zhu, C., Small, W., Metz, T. R., Vericella,
J. J., Barth, H. D., Kuntz, J. D., Maxwell, R. S., Spadaccini, C. M., Wilson, T. S.
2014. Three-Dimensional Printing of Elastomeric, Cellular Architectures with
Negative Stiffness. Advanced Functional Materials, 21, 1-9.
Feilden, E., García-Tuñón, B., Giuliani, F., Saiz, E., Vandeperre, L. 2016. Robocasting of
structural ceramic parts with hydrogel inks. Journal of the European Ceramic
Society, 36, 2525-2533.
Genina, N., Holländer, J., Jukarainenc, H., Mäkilä, E., Salonen, J., Sandler, N. 2016.
Ethylene vinyl acetate (EVA) as a new drug carrier for 3D printed medical drug
delivery devices. European Journal of Pharmaceutical Sciences, 90, 53-63.
Godoi, F. C., Prakash, S., Bhandari, R. 2016. 3D printing technologies applied for food
design: Status and prospects. Journal of Food Engineering, 179, 44-54.
Gong, J., Shitara, M., Serizawa, R., Makino, M., Kabir, M. H., Furukawa, H. (2014). 3D
printing of meso-decorated gels and foods. Materials Science Forum, 783, 1250-
1254.
Hamilton, C. A., Alici, G., Panhuis, M. 2017. 3D printing Vegemite and Marmite:
Redefining “breadboards.” Journal of Food Engineering, http://dx.doi.org/10.1016/
j.jfoodeng.2017.01.008.
Huijun, L., Sijun, L., Lin, L. 2016. Rheological study on 3D printability of alginate
hydrogel and effect of graphene oxide. Journal of Bioprinting, 2, 54-66.
Jia, F., Wang, X., Mustafee, N., Hao, L. 2016. Investigating the feasibility of supply
chain-centric business models in 3D chocolate printing: A simulation study.
Technological Forecasting and Social changes, 102, 202-213.
Jonathan, G., Karim, A. 2016. 3D printing in pharmaceutics: A new tool for designing
customized drug delivery systems. International Journal of Pharmaceutics, 499, 376-
394.
Leppiniemi, J., Lahtinen, P., Paajanen, A., Mahlberg, R., Metsä-Kortelainen, S.,
Pinomaa, T., Pajari, H., Vikholm-Lundin, I., Pursula, P., Hytönen, V. P. (2017). 3D-
printable bioactivated nanocellulose-alginate hydrogels. ACS Applied Mater
Interfaces, 9, 21959-21970.
Lee, J., Lee, K. 2016. Block-based inner support structure generation algorithm for 3D
printing using fused deposition modeling; The International Journal of Advanced
Manufacturing Technology, doi:10.1007/s00170-016-9239-3.
Lille, M., Nurmela, A., Nordlund, E., Metsa-Kortelainen, S., Sozer, N. 2017.
Applicability of protein and fiber-rich food materials in extrusion-based 3D printing.
Journal of Food Engineering, http://dx.doi.org/10.1016/j.jfoodeng. 2017.04.034.
Rodriguez, M. J., Brown, J., Giordano, J., Lin, S. J., Omenetto, F. G., Kaplan, D. L. 2017.
Silk based bioinks for soft tissue reconstruction using 3-Dimensional (3D) printing
with in vitro and in vivo assessments. Biomaterials, 117, 105-115.
Sun, J., Peng, Z., Zhou, W., Fuh, J. Y. H., Hong, G. S., Chiu, A. 2015. A review on 3D
printing for customized food fabrication. Procedia Manufacturing, 1, 308-319.
Tang, A., Liua, Y., Wang, Q., Chen, Q., Liu, W., Fang, Z., Wang, L. 2016. A new
photoelectric ink based on nanocellulose/CdS quantum dots forscreen-printing.
Carbohydrate Polymers, 148, 29–35.
Tian, J., Bryksa, B. C., Yada, R. Y. 2016. Feeding the world into the future – food and
nutrition security: the role of food science and technology. Frontiers in Life Science,
doi: 0.1080/21553769.2016.1174958.
Torres, M. D, Azevedo, G., Hilliou, L. 2016. Phase diagrams of hybrid carrageenans
extracted from Ahnfelptiosis devoniensis and Chondrus crispus. Carbohydrate
Polymers, 136, 449-458.
Torres, M. D, Hallmark, B., Wilson, D. I. 2015. Effect of bubble volume fraction on the
shear and extensional rheology of bubbly liquids based on guar gum (a Giesekus
fluid) as continuous phase. Journal of Food Engineering, 146, 129-142.
Torres, M. D., Raymundo, A., Sousa, I. 2014. Effect of sucrose, stevia and xylitol on
rheological properties of gels from blends of chestnut and rice flour. Carbohydrate
Polymers, 98, 249-256.
Tubío, C. R., Guitián, F., Gil, A. 2016. Fabrication of ZnO periodic structures by 3D
printing. Journal of the European Ceramic Society, 36, 3409–3415.
Wang, L., Zhang, M., Bhandari, B., Yang, Ch. 2017a. Investigation on fish surimi gel as
promising food material for 3D printing. Journal of Food Engineering,
http://dx.doi.org/10.1016/j.jfoodeng.2017.02.029.
Wang, L., Gardner, D. J., Bousfield, D.W. 2017b. Cellulose nanofibril-reinforced
polypropylene composites for material extrusion: rheological properties. Polymer
Engineering and Science, DOI 10.1002/pen.24615.
Zhong, J., Zhou, G. X., He, P. G., Yang, Z. H., Jia, D. C. 2017. 3D printing strong and
conductive geo-polymer nanocomposite structures modified by graphene oxide.
Carbon, 117, 421-26.
Zhu, F., Cheng, L., Yin, J., Wu, Z. L., Qian, J., Fu, J., Zheng, Q. 2016. 3D Printing of
Ultratough Polyion Complex Hydrogels. ACS Applied Materials Interfaces, 8,
31304−31310.
Chapter 2
RHEOLOGICAL CHARACTERISATION OF INJECTION

GROUTS USING ROTATIONAL RHEOMETRY
Luis G. Baltazar1,*, Fernando M. A. Henriques1

and Maria T. Cidade2
1
Departamento de Engenharia Civil, Faculdade de Ciências e Tecnologia,
Universidade NOVA de Lisboa, Portugal
2
Departamento de Ciência dos Materiais e Cenimat/I3N,
Faculdade de Ciências e Tecnologia, Universidade NOVA de Lisboa, Portugal
ABSTRACT
The overall goal of this chapter is to contribute to the understanding and

improvement of the rheological performance of injection grouts for consolidation of
stone masonry buildings. It is a well known fact that old masonries such as heritage or
ordinary buildings represent a large portion of the construction area in many European
cities. They frequently present a bad state of conservation and have particular technical
needs and specific weaknesses that require special approaches. Grout injection is an
interesting technique to repair and consolidate such stone masonries. It consists of
introducing a suspension (grout) into the masonry inner core in order to upgrade the
cohesion of the masonry and increasing its monolithic behaviour. Therefore, it is essential
to evaluate the effectiveness of the grouts in terms of keeping their adequate flow
capacity. For this purpose, rheology appears as a useful tool in the controlling and
optimisation of grouts design. Notwithstanding, during rheological measurements many
disturbing effects may arise, such as the wall slip, which can play a significant role
leading to an inaccurate rheological parameters determination. Despite the large number
of research related to wall slip on several fluids with yield stress, the wall slip is still a
concern for rheological measurements of grouts. Thus, this study aims to determine the
degree of wall slip during rheological measurements of natural hydraulic lime-based
*
Corresponding Author Email: luis.baltazar@fct.unl.pt.
14 Luis G. Baltazar, Fernando M. A. Henriques and Maria T. Cidade
grouts for stone masonry consolidation. For the determination of wall slip stress
controlled measurements at various gaps and different shear stresses were performed.
Overall, the results revealed that wall slip is function of the shear stress and the grout
composition. Results from this research can help to define an adequate measurement
protocol to correctly infer the rheological properties of injection grouts based on
rotational rheometry.
Keywords: rotational rheometry, injection grouts, masonry, natural hydraulic lime
INTRODUCTION
Stone Masonry
During many centuries stone masonry walls were used as a constructive technique
and they represent the vast majority of old buildings and dwellings in many urban centres
in Portugal as well as across Europe (Figure 1). Old stone masonry walls present peculiar
mechanical behaviour which is different from that of brick masonry walls or regular
stone masonry in new buildings (Binda et al. 2006). Masonry is a general term that
includes a large variety of materials and construction techniques. Old stone masonry can
be seen as a heterogeneous element, made of stones, mortar and other kinds of
unbounded material. Nevertheless, it is not only the nature of materials that affect their
characteristic and eventual state of damage; the constructive solution adopted plays also
an important role (Van Rickstal 2000).
Over the years a classification of stone masonry together with a description of its
constructive evolution has been carried out. An extensive research started in the early
nineties by Binda and her collaborators at the Polytechnic of Milan with the purpose of
developing a masonry data-base for mathematical models to detect the load-carrying
capacity of the masonry and, therefore, the study of vulnerability of historic centres in
Italy; it allowed the identification of a different morphology of stone masonry (Binda et
al. 1997; Penazzi et al. 2001; Binda et al. 2003). Considering the types of masonry
geometries found and their subsequent classification based on the number of different
layers and on the type of constraints between them (Binda 2000), it was possible to define
four main classes of sections as represented in Figure 2.
Concerning the multi-leaf section, and in particular the three-leaf type, whose internal
leaf (inner core) has a generally high thickness (higher than 50cm), such a wall is
traditionally made with a completely loose internal core - all kinds of unbounded material
can be found (such as sand and small stones) filling the inner core - held by two separate
external leaves. These external leaves work as framework for the internal core, which in
turns plays the important role of load-bearing capacity. The description of the
particularities of the other stone masonry cross sections is beyond the scope of this
Rheological Characterisation of Injection Grouts using Rotational Rheometry 15
chapter, but is available in Binda et al. (2000) or Syngellakis (2014) for the interested
reader.
(a) (b)
Figure 1. Examples of stone masonry buildings in Portugal: (a) Santa Clara monastery and (b)
Jeronimos monastery.
(a) (b) (c) (d)
Figure 2. Classification of the cross sections of stone masonry walls: (a) a single leaf, (b) two leaves
without connection, (c) two leaves with connection, (d) three leaf.
Figure 3. Deformation and failure of a multi-leaf masonry wall.
A multi-leaf masonry wall is usually characterized by its constructive weaknesses,

such as scarce or no connection through the thickness and presence of voids and cracks
which lead to instability of the masonry under both horizontal and vertical loads. These
constructive peculiarities facilitate the collapse of the external leaves that may lead to the
global overturning of the wall (Penazzi et al. 2001). This mechanism of collapse has
frequently occurred during seismic events. In fact, masonry walls are prone to detach
from the rest of the building and to fail out-of-the plane by disaggregating into different
parts (see Figure 3).
In general, the worst weakness of multi-leaf masonry walls is the lack of monolithic
behaviour, which is frequent when the wall is made of small pebbles or by two external
layers even well ordered but not mutually connected. Due to these constructive
particularities the wall can become more brittle, particularly when external loads act in a
horizontal direction; however, the same problem can happen under vertical loads if they
act eccentrically. Based on the above mentioned, it can be said that the performance of
old stone masonry relies on their monolithic behaviour rather than on the mechanical
properties of individual elements. It means that the good condition of the connections
between elements is essential towards masonry integrity. In addition, the masonry
conservation frequently requires the need for consolidation in order to increase or
maintain the masonry shear and flexural strength. Nevertheless, the choice and design of
the consolidation technique must start from the study and a thorough understanding of the
real nature and behaviour of masonry.
Grout Injection as Consolidation Technique
Grout injection or grouting is a process of introducing a fluid mixture into the

masonry for the consolidation and improving of mechanical properties of the whole
masonry system. Grout is a suspension of binder particles in water which has a
considerable fluidity in order to be poured or pumped into pores, voids and cracks within
masonry. According to several authors (Van Rickstal 2000; Toumbakari 2002; Baronio et
al. 2003), the grout injection technique aims to fill large and small voids and cracks,
therefore upgrading the cohesion of masonry, increasing also its monolithic behaviour
(by establishing adhesion between layers badly connected), and the load capacity. To
these ends, the grout should be well designed in order to achieve maximum performance.
It means that the flow properties of fresh grout are as important as its properties in the
hardened state, since they govern the ability of the grout to fill the voids. The
optimization of grout injectability involves the control of its fresh properties, such as the
rheological ones so that the grout may be pumped and flow inside the masonry.
Furthermore, it is a technique which can improve the stability without changing the
appearance of the building or monument, since the grout is introduced/injected into the
internal core of masonry and it does not damage the aesthetic outlook of the building.
Concerning the main principles that explain the grout injection efficacy, it can be said
that the filling of voids eliminates the concentration of vertical compressive stresses
around the lateral edges of the voids and reestablishes the adhesion forces between the
elements of the masonry. Moreover, the mechanical properties of the grout itself improve
the compressive strength of the mortar (which is either non-existing or very deteriorated)
and consequently the compressive strength of the whole masonry.
Grout injection can be performed with perforated pipes drilled in the masonry
(Figure 4). There are four different methods to carry out grouting in masonry walls:
manually, by pump, gravity or vacuum. The selection of the best method must be done
according with the context of the specific case and condition of masonry wall. For
instance the pouring of grout by means of a vacuum system may be useful where fine
fractures and small voids are present. This is so because the creation of a complete
vacuum system in masonry with huge voids or even with a long and complex fracture
pattern is often difficult to carry out on site. On the other hand, the gravity grouting
technique is advisable for much damaged masonry elements. It consists of letting the
grout enter the masonry, with the help of the force of gravity. Therefore, it is only
effective if the voids inside the masonry are large enough, once the pressure associated is
very low. The hand grouting technique can be seen as a small scale procedure that can be
adapted to local consolidation and it can be an alternative when no other facilities are
available. Normally, it is carried out in association with repointing solution. Nevertheless,
according to Van Rickstal (2000), the pumped grouting is actually the most common
technique due to its superior efficiency, especially when large volumes of grout are
needed.
(a) (b)
Figure 4. Grout injection: (a) detail of the injection pipe and (b) grout filling the masonry.
Grout injection can be pointed out as the most appropriate strengthening technique
for old masonry walls, especially when there are requirements regarding the preservation
of the original appearance. It should be noted, however, that grouting is an irreversible
technique; meaning that special attention to the selection of injection grouts, which must
be compatible with the in situ materials in order to ensure a successful consolidation

intervention, must be carried out. The selection of materials for grouts generally depends
on a range of factors, such as the desired rheological properties, performance
characteristics and compatibility with the original materials (Zajadacz and Simon 2006;
Valluzzi 2009). Compatibility between the new repair materials and the original materials
of the masonry is actually a huge concern because the grout must be compatible with the
original materials and at same time reinstating the integrity of the whole masonry system
without unintended consequences. Some authors suggest that to design compatible grouts
the analysis of the original material should be undertaken, following the example of
repair mortars (Ferragni et al. 1985). However, trying to match the composition of
original mortars presents a problem in the case of injection grouts because it is very
probable that the same composition will not produce working properties desirable for an
injection grout, such as rheological ones. This means that an adequate choice of grout
components (specially the binder) is critical to the success of a consolidation process. A
consolidation intervention should not cause any damage to the structural and architectural
authenticity of the building. The properties of hydraulic lime binder tend to give an
acceptable early strength besides the fact that the lime provides the wanted physical and
chemical compatibility to the historical materials. Moreover, taking into account the lack
of information about the rheological behaviour of hydraulic lime-based grouts this
chapter is limited to natural hydraulic lime binder, and other binders are not further
discussed.
RHEOLOGY OF CEMENTITIOUS SUSPENSIONS
Grouts can be seen as a colloidal suspension consisting of suspended binder particles

in water. The efficacy of the grouts depends mainly on the injectability properties.
Therefore, it is essential to evaluate the effectiveness of the grouts in terms of keeping
their adequate fluidity. The control of fluidity can be done through rheological
measurements. Currently, rheology is often used as a tool in the control and design of
several cementitious materials, such as cement-based pastes, mortars, concretes, and
grouts (Roussel 2007; Roussel et al. 2010; Baltazar et al. 2013a; Cardoso et al. 2014;
Baltazar et al. 2017a; 2017b).
The rheology of cementitious-based grouts is dominated by the combination of van
der Waals forces, electrostatic forces, hydrodynamics and chemical reactions producing
calcium silicate hydrates (starting as soon as the water is added to the binder powder).
The result is a complex microstructure that will be readily affected by flow. In the
following topics, several factors that influence the rheology of cementitious grouts are
presented.
Interactions Influencing the Rheology of Grouts
In cementitious grouts, colloidal interparticle forces and coagulation processes

dominate its rheology since the major parts of the particles are of colloidal size. A colloid
is a suspension where the disperse phase particles have at least one dimension in the
range 1nm - 1µm (Russel et al. 1989). There are several interparticle forces acting on
particles in a suspension; these forces can have a significant effect on the microstructure
and on various rheological properties of the suspension. According to Barnes (2000) four
kinds of forces coexist in a colloidal suspension, namely:
 Colloidal forces
 Brownian randomising forces
 Viscous forces
 Gravitational forces
Colloidal forces arise from interactions between the particles and can result in
repulsion or attraction between particles, due to different mechanisms as Van der Waals
forces (attraction) and electrostatic charges (repulsion). So, if the net result of all the
forces is attraction, the particles tend to flocculate, whilst overall repulsion means that
they remain separate (i.e., deflocculated). Van der Waals forces are always attractive and
it dominates at long intermolecular distances, while at short distances it is the steep Born
repulsion that originates from the overlapping of the electron clouds of the molecules. In
general, the forces of attraction and repulsion between atoms and/or molecules have
inverse power dependence to the distance among them.
The balance between the Van der Waals and the repulsive forces resulted in the
development of a theory, denoted the DLVO theory, which was proposed by Derjaguin
and Landau, and, independently, Verwey and Overbeek (1947). This theory combines the
effect of the Van der Waals attraction and the electrostatic repulsion due to the so called
double layer of counter-ions (Yang et al. 1997). When ions are present in a system that
contains an interface (solid/liquid), there will be a variation in the ion density near the
interface. When a mineral particle is immersed in a liquid, a surface charge appears
which is due to one or more of the following reasons:
 Ionisation: which means a rupture of ionic links at the surface of the mineral and
dispersion of the ions in the suspension liquid
 Adsorption: physical or chemical fixation of ions at the particle surface
 Hydration: initiation of processes (after reaction such as dissolution, hydrolysis)
Binder particles (like cement or hydraulic lime) already contain surface charges at the
anhydrous state. When these particles are dispersed in water, which is a much ionised
medium, the new surface charge that will be develop on the binder particles will depend
on the preservation of the electroneutrality of the suspension. This means that a
compensation of the particles surface charges will occur by ions with the opposite sign,
which necessarily comes from the water. The result is the concentration of ions around
the particles and the formation of a layer. This layer is composed of two parts (thus the
name double layer), one very well fixed on the surface of the particle called Stern layer
and a second, more loosely fixed, called diffuse layer. There is however a small
difference between the structure of the double layer of binder particle and of other
minerals. This is due to the fact that hydraulic binder surface, when contacted with water,
begins to react chemically and thereby generates both surface charge and ions in
considerable quantities, which are released into the water (Toumbakari 2002). Detailed
information on this subject can be found in (Cheng 1987; Billberg 2006; Mewis and
Wagner 2012).
In addition, the concept of zeta-potential appears in the literature often associated
with colloidal systems and can be also applied to cementitious suspensions (Ferrari et al.
2010; Ran et al. 2010; Vikan 2005). Zeta-potential is the electric potential at the interface
between stern and diffuse layer (Yang et al. 1997) and some studies have shown that
attraction is associated to positive zeta-potential and repulsion to negative one. Very
often, admixtures like high range water reducers are used to improve the dispersion of
cementitious suspension. The mechanism of action of those admixtures is by means (not
only) of electrostatic charges. So, one way to study the effect of high range water
reducers upon cementitious systems is through zeta-potential measurements. The
presence of high range water reducer causes an increase of the electrostatic repulsive
forces resulting in the mutual repel and dispersion of the binder particles and therefore an
increase of the absolute magnitude of zeta-potential can be expected (Ferrari et al. 2010).
The action of the high range water reducer will be addressed again in the materials
section.
Brownian forces are characterized by thermal movements in random directions of all
atoms and molecules, for example in the continuum phase, which will make them
constantly collide with each other. This results in that the particles themselves increasing
their kinetic energy and starting to move randomly. The direction of each individual
particle changes constantly and the path will be zigzag-shaped, which has an important
influence on the suspension properties. For instance, the Brownian motion favours the
microstructure build-up when suspension is at rest. Viscous forces action over particles is
proportional to the difference of local velocity between the particle and the surrounding
liquid phase although, due to their inertia, it will be the large size particles present in
suspension that, are more susceptible to viscous forces. It should be noted, however, that
the influence of particle size on type of force acting on the particles is of the utmost
importance. This means that surface forces such as Van der Waal forces, electrostatic
forces and Brownian motion can only dominate over gravitational and inertial forces if
the particle size is small enough (i.e., colloidal size). These mechanisms are much less
important for bigger particles in suspension (such as sand or flocs of particles) in which
gravitational force, buoyancy and stirring dominate (Shaw 1992).
In suspensions (such as injection grouts), when at rest (or under very low shear rates),
Brownian and/or colloidal forces dominate. It means that the flocculated state of grout
particles (namely colloidal ones), when at rest, is dominated by electrostatic forces and
Brownian motion and its yield stress reflects the forces that hold the particles together.
However, when flocculated particles are sheared (shear rate increases) the flocs are
progressively broken down to the primary particles. This increase of flow velocity will
also lead to an increase of overall repulsion, due to liquid pressure, meaning that particles
remain deflocculated. So, at higher shear rates, the viscous and stirring forces start to
control the rheological behaviour of the suspension. When shear rate is reduced, or even
stopped, a suspension microstructure will start to rebuild and, consequently, the particles
will coagulate and flocculate into agglomerates. According to Barnes (1997) the driving
force to rebuild the suspension microstructure is Brownian motion and since these forces
are high for small size particles, the rate of microstructure change is a function of particle
size. Hence, suspensions with large particles recover their viscosity (after cessation of
shear) slower than suspensions of small particles. In the same way, large size particle
suspensions will break down faster under shearing. As can be understood by the above
mentioned, many mechanisms behind the thixotropy of particle suspensions are the same
found in the colloidal domain. Thus, all these concepts and considerations are very
important in the design of thixotropic suspensions/materials, such as injection grouts for
masonry consolidation. More detailed information on colloidal suspension rheology can
be found in (Barnes 1997; Yang et al. 1997; Wallevik 2009; Billberg 2006; Mewis and
Wagner 2012).
MATERIALS
Binder
Most researchers in the field of conservation of old buildings indicate that they try to
select the binder to be used in the grout based on their similarities to the original
materials (Baglioni et al. 1997; Bras 2011; Valluzzi 2009). Notwithstanding, there are
however different opinions about the type of binder to use in a grouting intervention in an
old building. Many authors defend the idea of using hydraulic lime as binder (Ferragni et
al. 1982, 1985), while others support the use of traditional materials and argue that simple
air lime is adequate in most circumstances (Ballantyne 1996). However, both agree that
lime based grouts are without any doubt the most compatible with the original materials
for the consolidation of old masonry. Regarding air lime, there are several downsides to
using this binder. The most highlighted disadvantage of using air lime as a grout binder is
that it requires exposure to carbon dioxide (CO2) in the air to set and, with minimal
exposure to air inside a wall, carbonation can only proceed very slowly, so the
development of strength and durability will also be slow.
On the other hand, hydraulic lime is commonly used for grouts because, like air lime,
it is likely to be compatible with original lime-based materials (Ferragni et al. 1982,
1985; Giorgio 2001). The main advantage over air lime is that they set in the absence of
air, and hence is particularly suitable for grouting internal voids. However, there can also
be some disadvantages to using hydraulic lime as binder, for instance, it can be
excessively strong (lower elasticity than the original materials) which may end up leading
to some mechanical incompatibility (Toumbakari and Van Gemert 1998). This
disadvantage can be avoided, however, if a proper grout design is made, since the
hardened performance depends on the composition used (Baltazar et al. 2015a). In
literature several research works (Ballantyne 1996; Kenyon 1996; Oldenbourg 1996) on
cement-based grouts can be found; however these grouts are mainly used for
consolidation of modern cement-based systems or concrete structural elements, since it is
known that cement does not correspond to the ideal binding agent to be used in most old
masonry buildings. According to Collepardi (1990), the use of cement based grouts to
restore old masonries is an example of incorrect restoration intervention, because the
interaction of cement with the pre-existing materials may end by causing more harm than
good. Reasons for such degradation are beyond the scope of this chapter, but can be
found in corresponding literature (Collepardi 1990).
A consolidation intervention should not cause any damage to the structural and
architectural authenticity of the building. The properties of hydraulic lime binder tend to
give an acceptable early strength besides the fact that the lime provides the required
physical as and chemical compatibility to the historical materials. Moreover, taking into
account the lack of information about the rheological behaviour of hydraulic lime grouts,
this chapter is limited to natural hydraulic lime (NHL) binder. The natural hydraulic lime
adopted is labelled as NHL5 and produced according to the European standard EN459
(EN459-1:2010). The chemical and physical properties of NHL are listed in Table 1 and
2, respectively. The grain size distribution of NHL5 is presented in Figure 5.
Table 1. Chemical composition of natural hydraulic lime
Al2O3 CaO Fe2O3 MgO MnO SiO2 SiC SO3 SrO K2O
2.00 85.00 2.00 1.00 0.03 8.00 0.01 1.00 0.05 0.70
Table 2. Physical properties of natural hydraulic lime
Label Compression resistance Setting time Specific gravity Fineness (Blaine)

at 28 days (MPa) (g/cm3) (cm2)/g
NHL5 5.5 Start 2h End 6h 2.7 9400
Figure 5. Grain size distribution of natural hydraulic lime.
Admixture
The voids and cracks through which the grout has to penetrate can be relatively
small. Such grouts need a good fluidity in order to achieve low viscosity fluidizers (also
known as high range water reducers or super plasticizers) which can be incorporated into
grout mixtures in limited amounts to improve or modify the flow properties. However,
most of high range water reducers (HRWR) are produced to control the fluidity of cement
grouts, mortars and concrete rather than hydraulic lime-based grouts. Very few studies on
the influence of HRWR on the behaviour of hydraulic lime grouts, particularly
concerning rheological behaviour and stability, can be found. It means that there is a need
to study the HRWR effectiveness over the hydraulic lime based grouts in order to
contribute to the improvement of knowledge affecting both the scientific community and
industry. HRWR can improve the injectability of grouts, reduce segregation and the water
content required to achieve the desired working properties. This means that the water
content is reduced to an acceptable level without damaging the required fluidity. In
addition to the decrease in viscosity HRWR, may improve fresh stability (Jolicoeur and
Simard 1998) and durability (Baglioni et al. 1997; Toumbakari and Van Gemert 1998).
As consequence of reduced water content a higher mechanical strength and lower
shrinkage can be obtained (Van Rickstal et al. 2003).
The latest generation of HRWR is based on carboxylic polymers (PC) or carboxylic

ethers (PE) backbone and is the most commonly used nowadays. They may vary widely
in composition depending on the choice of monomers. HRWR are composed by pendent
groups, also known as side chains that are grafted onto the backbone at frequent intervals.
The backbone provides anchoring to the lime particles while the long side chain provides
the steric hindrance effect resulting from the extension of this side chain away from the
binder particles surface. In other words, the dispersion mechanism happens when the
HRWR is adsorbed on the surface of the binder particles and the steric repulsion is
exerted when the binder particles come closer but cannot touch each other due to the long
side chains in the graft copolymer (see Figure 6).
It is commonly accepted that variations in the nature and relative proportions of the
different monomers in the copolymer yield a group of products having different ranges of
functional properties. For instance, Yamada et al. (2000) studied the dispersing properties
of polycarboxylate type HRWR with polyether side chains and with different degrees of
polymerization of the side chain. Polymers with longer side chains and lower degrees of
backbone polymerization were found to have the highest dispersing power. More
recently, Winnefeld et al. (2007) studied the relation between the structural parameters of
polycarboxylate-ether type with their performance in cementitious systems. They found
that the charge density of the polymer increases with decreasing side chain density and
with decreasing side chain length, leading to an increasing amount of free carboxylic
groups. This means that it is expected that the absorption of the polymers is enhanced
with increasing charge density.
Figure 6. Representation of a suspension dispersed through coupled mechanisms exercised by

polycarboxylate-based HRWR.
Another important effect associated to the use of HRWR is called depletion

phenomenon. Depletion is caused by surplus polymer molecules which stay in the water
phase between the particles and prevent them from getting close enough to form
agglomerates. These surplus polymer molecules in the liquid phase can result from: (i) an
excessive initial dosage of HRWR and/or (ii) as the hydration reaction proceeds, the
amount of free water decreases, and so does the distance between the hydration surfaces
of the neighbouring cement (hydrate) particles. As the interparticle space becomes
smaller, the polymer concentration becomes higher. The concentration of polymers
confined in this volume may create a substantial osmotic pressure effect that tends to
expel the polymers from the confined interparticle space. This effect would induce a
particle-particle attraction (Chandra and Bjornstrom 2002) meaning that a higher HRWR
concentration may lead to reverse effect (Banfill 2011). Thus, considering the mentioned
above and previous researches (Baltazar et al. 2013b; Jorne et al. 2015) a dosage of
HRWR of 1.0 wt% was adopted and maintained in all grout mixtures. A polycarboxylate
ether based HRWR, conforming to ASTM C494 (ASTM C494-2005) Type F was used. It
had a specific gravity, pH, chloride content, charge and solid content of 1.05, 8, <0.10%,
anionic and 28-32%, respectively.
Water
Typically, the liquid present in injection grouts is water. The main role of water in
grouts is to act as a dispersive medium for the particles. In hydraulic binder, such as
hydraulic lime, it also contributes to the chemical reactions. The proportion present in the
initial grout mix is important; high water content gives improved flow and injectability
(Rosquoët et al. 2003; Toumbakari et al. 1999; Eriksson et al. 2004;), since with a larger
amount of water and the binder particles are less likely to clog but this also leads to less
stable suspensions with segregation and bleeding as well as with more potential for
cracking (Flatt and Girardet 2000). High water content increases the total porosity and
causes reduced mechanical strength once the grout has hardened. The initial water
content is strongly influenced by the type of binder and pozzolans used, since an increase
of specific surface area leads to a higher water demand. It means that the initial water
content must allow plenty of binder hydration and a proper injection but without
compromising all the other properties of the grout. This means that a compromise
between injectability, segregation and bleeding must be found.
Different water/binder (w/b) ratio used to produce grouts can be found in literature
since the type of grouting intervention also affects the range of w/b ratio to adopt in the
grout composition. According to Miltiadou (1990) the hydraulic binder-based grouts for
the consolidation of masonry structures have w/b ratios in the rage 0.5-1.5. In the same
field, Bras (2011) studied the optimization of hydraulic lime grout with a w/b ratio in the
range 0.6-0.8. In the case of grouts for pos-tensioning tendons, the w/b ratios found range
of 0.35 to 0.42, while grouts for soil and rock injection can have a w/b ratio between 1
and 2. In this study, different w/b ratios (0.45, 0.50 and 0.55) were used in view of the
results reported in our previous studies (Jorne et al. 2014; Baltazar et al. 2015a).
EXPERIMENTAL PROCEDURES
Grout Preparation
The hydraulic lime grouts were prepared at room temperature of 20 ± 5ºC and a
relative humidity of of 50 ± 5%. The grout mixtures were prepared in the laboratory in
batches of 3 litres and the components mixed using a high shear mixer equipped with
helicoidal blade. The mixing procedure adopted was obtained in previous research using
the design of experiments method (Baltazar et al. 2012): the whole lime is added to 70%
of total constitutional water and mixed for 10 minutes. Then, the remaining water (with
diluted HRWR) was added within 30 s. Each grout was passed through a sieve with 1.18
mm (no. 16 ASTM) prior to the extraction of three samples, from each mix, to be used on
the rheological measurements.
Rheological Measurements
The rheological characterization of cementitious suspensions can be challenging

because of the need for suitable devices as well as appropriate measurement procedures.
Parallel-plates geometry is often used in rotational rheometers to quantify the properties
of suspensions such as cementitious grouts (Figure 7). This geometry is very convenient
since the gap height can be easily changed, even without reloading the sample.
Additionally, this geometry is also suitable for suspensions containing large particles.
Nevertheless, during measurements with this geometry, a top water-rich layer on the
sample can be developed due to settlement of the particles. The formation of this water-
rich layer causes a lubricating action, making the rotation of the upper plate easier and,
therefore, not representative of the bulk material resistance, which lead to an inaccurate
rheological parameters determination.
(a) (b)
Figure 7. Parallel plate geometry (a) loaded with grout sample (b) schematic representation of
geometry.
In suspensions with yield stress and at low shear rates the flow measured may be
totally affected by wall slip (Struble and Sum, 1995). This slippage phenomenon was
initially shown in rheological measurements of cement paste by Wesche et al. (1973).
Meanwhile other studies have focused this phenomenon in cementitious pastes as well as
for other types of suspensions (Lapasin et al. 1983; Buscall et al. 1993; Barnes, 1995) but
using geometries different than the one used in this study. In this context, this chapter
aims to determine the degree of wall slip during rheological measurements of NHL-based
grouts using a rotational rheometer and, therefore, contribute to define an adequate
measurement protocol and data analysis approach.
Firstly, we will focus on the measurement of rheological properties of NHL-based
grouts using a rotational rheometer equipped with parallel smooth plates and the accuracy
of two rheological models will be analysed. Then, the presence of wall slip will be
checked based on the Mooney analysis (Mooney, 1931), in order to quantify the wall slip
phenomenon and to show the real influence of wall slip on the rheological properties.
Moreover, easy solutions to minimize the influence of wall slip on rheological
measurements will be proposed.
The rheological measurements were performed with a Bohlin Gemini HR nano
rotational rheometer (Malvern, UK). The tests were carried out with parallel-plate
geometry. The diameter of the geometry was 40 mm and the gap was 2 mm. To avoid the
undesirable influence from mechanical histories, fresh samples were subjected to a pre-
shearing at an identical shear rate of 1 s-1 for 1 min and left standing for an additional 1
min before measurements took place. Then, the sample was subjected to a stepped ramp
with different shear rates starting from 1 to100 s-1. Each shear rate was applied long
enough in order to ensure the attendance of the steady state.
Yield Stress and Plastic Viscosity
In the frame of rheology of cementitious suspension knowledge, there are different

analytical models to calculate the rheological properties of grouts. Nevertheless,
according to previous researches (Jorne et al. 2015; Baltazar et al. 2015b) the Bingham
model (Eq. 1) and the modified Bingham model (Eq 2) are the most suitable models to
describe the rheological behaviour of NHL-based grouts.
τ = τ0 + ηP × γ̇ (1)
where: 𝜏 is the shear stress (Pa), 𝜏0 is the yield stress (Pa), ηP is the plastic viscosity
(Pa.s) and γ̇ is the shear rate (s-1). τ0 indicates the minimum force to overcome the
internal friction so that the grout starts flowing. The ηP indicates the strength of the
internal friction between the particles once flow is initiated. The smaller is the τ0 and ηP
better is the injectability of the grout.
τ = τ0 + ηP × γ + c ̇ × γ̇ 2 (2)
where c is a constant with no physical meaning.
Wall Slip
The best way to confirm the presence of wall slip is taking measurements with
smooth and serrated parallel-plates. Notwithstanding, procedures exist to determine the
amount of wall slip without the need of both geometries. According to Mooney analysis
(Mooney, 1931; Yoshimura and Prudhomme 1988), the wall slip can be characterized by
the slip velocity (Vs). It is possible to determine Vs from the bulk shear rate or the true
shear rate ( 𝛾̇ ) and the measured shear rate ( 𝛾̇𝑚 ) by performing stress controlled
measurements at different gap distances (h). If slip occurs then the Vs will depend only
on the applied shear stress (σ) but not on the gap. Thus by varying the gap and keeping
the stress constant it is possible to determine the Vs and the true shear rate using Eq 3.
2.𝑉𝑠
𝛾̇𝑚 = ℎ
+ 𝛾̇ (3)
Stress controlled measurements at various gaps of 1 mm, 1.5 mm, 2 mm and 3 mm

using constant applied stresses of 10 Pa, 15 Pa and 20 Pa were conducted. The measured
shear rate was then plotted against the inverse gap and a linear regression model fitted to
the data (Eq. 3) giving the Vs and 𝛾̇ as the gradient of the linear regression and the y-axis
intersect, respectively.
Figure 8. Flow curve for the NHL grout with (a) w/b ratio = 0.45; (b) w/b ratio = 0.50; (c) w/b ratio =
0.55 (symbols = experimental data; lines = models fitted to the data).
RESULTS AND DISCUSSION
Yield Stress and Plastic Viscosity
The experimentally observed shear stress versus shear rate profile determined for
grouts tested is shown in Figure 8. It shows that the flow curves are not totally linear;
instead they apparently present a Newtonian initial behaviour that is followed by a non-
linear flow i.e., shear-thinning behaviour. In Figure 8(a) the sharp change in trend at low
shear rates is highlighted, which is believed not to represent the flow of grout sample, but
rather a slip phenomenon at plate-grout interface. This flow curve shape has already been
reported, for other fluids with yield stress, by other researchers as being a consequence of
wall slip.
The rheological models (Eq. 1 and Eq. 2) and the correlation coefficients obtained are
presented in Figure 8. From the results presented, it can be noted that the highest
correlation coefficients were obtained with the modified Bingham model, which is a
consequence of the better fit of the second order polynomial equation to the shear-
thinning behaviour of the grout mixtures. Nevertheless, it is believed that yield stress
calculated by using the modified Bingham model is lower than the true yield stress.
According to some researchers (Nehdi and Rahman, 2004; Banfill, 2006) the slippage
can contribute to underestimation of yield stress values. Based on these statements it is
mandatory to quantify the slip.
Wall Slip
The wall slip is caused by several factors acting simultaneously, such as gravitational
forces and both electrostatic and steric effects between the geometry walls and the
suspension particles. All these phenomena, together with the smooth shearing surface of
the upper plate, make it really difficult to determine an absolute value of the rheological
parameters, especially of the yield stress. In order to determine the presence of wall slip,
stress controlled measurements at different gaps were performed. Examples of measured
shear rate versus inverse gap is presented in Figure 9 for grout with w/b ratio = 0.50 at
different shear stresses. As previously mentioned, to estimate the slip velocity a linear
model fit (Eq. 3) was applied to the data with gradient of the line equal to 2.Vs and the
intercept equal to the true shear rate. From Figure 9, it can be seen that for the shear stress
of 10 Pa a constant shear rate at each gap was obtained, which means no wall slip
occurred, while the higher shear stresses show a slight gradient (slip velocity) with higher
shear rates reported at bigger gaps which is attributed to wall slip.
Figure 9. Measured shear rate vs 1/gap at different shear stress for the grout with w/b of 0.50.
Figure 10. Measured shear rate vs 1/gap at different shear stress for the grout with w/b of 0.55.
The profiles of measured shear rate versus gap obtained for the w/b ratio of 0.55 is
presented in Figure 10. For this water/binder ratio the slip velocity was estimated to be
12.8 mm/s, 16.1 mm/s and 17.1 mm/s and the true shear rate 43.4 s-1, 92.7 s-1 and 110.9 s-
1
for the shear stress of 10 Pa, 15 Pa and 20 Pa, respectively. This is quite higher than the
measured shear rate range of 39.8 to 19.9 s-1, 79.6 to 54.8 s-1 and 99.6 to 74.7 s-1 for each
shear stress, which suggest a significant degree of wall slip.
From the results presented, it is clear that the degree of slip velocity (i.e., wall slip)
increases with increasing shear stress. Moreover, the slip velocity also increases with
increasing w/b ratio, which means that grouts with higher wall slip are associated with
greater tendency to segregation (higher amount of free water). This can be explained by
the uniform settlement of binder particles, leaving a thin layer of water over the surface,
which compromises the rheological measurements.
Ways to Minimize the Wall Slip Effects
The authors decided to eliminate the slippage effects by performing rheological

measurements with an upper plate with emery paper glued (grit 120) (as shown in
Figure 11). This solution changes the surface roughness of the upper plate which is
expected to improve the adhesion between the plate and the grout sample.
Figure 12 shows the flow curves obtained with the smooth and rough plate for the
w/b ratio tested. There is a clear difference between the results obtained with rough and
smooth plates. The measurements performed with rough plate led to greater shear
stresses, consequence of the roughness of the plate surface that contact with the sample.
Additionally, the flow curve for the measurement with rough plate shows a behaviour
close to a Bingham fluid (without the sharp change in trend observed in the other one)
and quite marks the presence of a yield point, from which the fluid starts flowing.
To better understand the influence of the plate’s roughness in the rheological
parameters the Bingham model and the modified Bingham model were fitted to the flow
curves. The rheological properties including the yield stress and the plastic viscosity of
the grout samples are listed in Table 3. The correlation coefficients (R2) obtained from
both models are also given.
(a) (b)
Figure 11. Geometry detail: (a) smooth plate and (b) rough plate with emery paper glued.
Figure 12. Flow curve with smooth and rough plate for the grout: (a) w/b ratio = 0.45; (b) w/b ratio =
0.50 and (c) w/b ratio = 0.55.
Table 3. Rheological properties measured with rough and smooth plate
Measurements with rough plate Measurements with smooth plate

Modified Bingham model
w/b τ0 (Pa) ηP (Pa.s) R2 τ0 (Pa) ηP (Pa.s) R2
0.45 47.96 1.30 0.987 13.64 1.35 0.947
0.50 11.69 0.29 0.999 2.67 0.33 0.957
0.55 4.93 0.20 0.996 0.77 0.21 0.981
Bingham model
w/b τ0 (Pa) ηP (Pa.s) R2 τ0 (Pa) ηP (Pa.s) R2
0.45 42.73 0.78 0.956 20.07 0.67 0.872
0.50 8.51 0.19 0.995 4.07 0.19 0.909
0.55 5.42 0.14 0.986 1.58 0.13 0.948
Figure 13. Example of fitting the Bingham model without the initial trend of the flow curve.
From the results presented in Table 3 it is clear that the modified Bingham model is
the one that best expresses the behaviour of the grout samples regardless of the geometry
roughness. Moreover, it can be noted that the roughness of the plate has a higher
influence on the yield stress than on the plastic viscosity. The lowest yield stress values
were obtained in measurements made with smooth plates. Taking as an example the grout
made of w/b = 0.55, a yield stress was obtained, 84.3% lower than the one obtained with
the rough plate (for the modified Bingham model). There is no doubt that the slippage
leads to an underestimation of the yield stress.
Therefore, in order to mitigate the effect of slippage when smooth plates are used, the
initial trend of the flow curve is not considered in model fitting (i.e., shear rates below 40
s-1) as shown in Figure 13. So that the yield stress value is determined as the intercept of
Eq. 1 at zero-shear rate.
Figure 14. Influence of the measurement approach in the rheological properties: (a) yield stress values;
(b) plastic viscosity.
The grout rheological parameter as function of plate roughness is shown in Figure 14.
As can be seen, the approach of smooth plate without the initial data allows obtaining
yield stress values mostly similar to the ones obtained with rough plate. In contrast, for
the smooth plate considering the entire flow curve the yield stress values are too low as a
result of the slippage as previously demonstrated. Regarding the plastic viscosity values,
it can be observed that the differences between approaches are reduced; therefore, it can
be concluded that the wall slip phenomenon has little influence over the plastic viscosity.
CONCLUSION
In this chapter, the influence of wall slip on rheological measurements of natural

hydraulic lime grouts using a rotational rheometer equipped with parallel plates was
demonstrated. In the light of the achieved results, the following conclusions can be
pointed out:
1) Based on the Mooney analysis it was found that the slip velocity increases with
increasing water/binder ratio, which means that grouts with higher wall slip are
associated with greater tendency to segregation. Thus, a proper formulation of
the grout is recommended in order to avoid grouts with a higher tendency to
segregate which promote slippage.
2) For rheological measurements of injection grouts some aspects have to be taken
into account in order to avoid causing significant errors in the rheological
parameters; in particular on the yield stress values, since the wall slip has a
higher influence on this parameter than on the plastic viscosity.
3) When serrated plate geometry is not available the authors propose to perform
rheological measurements with a smooth plate with emery paper glued in order to
minimize the wall slip phenomenon. This way the adhesion between the plate
and the grout sample is improved. The flow curve for the measurement with
rough plate shows a behaviour closer to the one of a Bingham fluid.
4) Another approach to eliminate the wall slip is to ignore the initial section of the
flow curve in the model fitting. The yield stress values obtained with this
approach showed good agreement with the ones obtained with rough plate,
indicating the validity of the measurements carry out with smooth plates despite
the wall slip at low shear rates.
ACKNOWLEDGMENTS
This work is funded by National Funds through FCT - Portuguese Foundation for
Science and Technology, Reference UID/CTM/50025/2013 and FEDER funds through
the COMPETE 2020 Programme under the project number POCI-01-0145-FEDER-
007688. The first author would like to thank the financial support from FCT/MCTES
under the scholarship SFRH/BPD/108427/2015.
REFERENCES
ASTM C494. (2005). Standard specification for chemical admixtures for concrete. ASTM
International, West Conshohocken, PA, US.
Baglioni, P. et al. (1997). New Autogenous Lime-Based Grouts Used in the Conservation
of Lime-Based Wall Paintings. Studies in Conservation 42(1):43–54.
Ballantyne, A. (1996). Stabilising the Substrate: Grouting – the Options. Anne
Ballantyne’s Contribution on 28th October 1995. Conservation News 60 60:43–45.
Baltazar, L. G., Henriques, F. M. A. and Jorne, F. (2012). Optimisation of Flow
Behaviour and Stability of Superplasticized Fresh Hydraulic Lime Grouts through
Design of Experiments. Construction and Building Materials 35:838–45.
Baltazar L. G., Henriques F. M. A. Jorne F., Cidade M. T. (2013a). Performance
improvement of hydraulic lime based grouts for masonry consolidation – An
experimental study. WIT Transactions on the Build Environment 131: 417-430.
Baltazar, L. G., Henriques, F. M. A., Jorne, F., Cidade, M. T. (2013b). The use of
rheology in the study of the composition effects on the fresh behaviour of hydraulic
lime grouts for injection of masonry walls. Rheologica Acta 52(2): 127–138.
Baltazar, L. G., Henriques, F. M. A., Cidade, M. T. (2015a). Contribution to the design of
hydraulic lime-based grouts for masonry consolidation. Journal of Civil Engineering
and Management 21(6): 698-709.
Baltazar, L. G., Henriques, F. M. A., Cidade, M. T. (2015b). Experimental study and
modeling of rheological and mechanical properties of NHL grouts. Journal of
Materials in Civil Engineering 27(12): 04015055.
Baltazar, L. G., Henriques, F. M. A., Rocha, D., Cidade, M. T. (2017a). Experimental
characterization of injection grouts incorporating hydrophobic silica fume. Journal of
Materials in Civil Engineering 29(10): 04017167.
Baltazar, L. G., Henriques, F. M. A., Cidade, M. T. (2017b). Grouts with improved
durability for masonry consolidation: An experimental study with non-standard
specimens. Key Engineering Materials 747: 480–487.
Banfill, P. F. G. (2006). Rheology of Fresh Cement and Concrete. Rheology Reviews. 61–
130.
Barnes, H. A. (1995). A review of the slip (Wall depletion) of polymer solutions,
emulsions and particle suspensions in viscometers; its cause, character, and cure.
Journal of Non-Newtonian Fluid Mechanics 56: 221-251.
Barnes, H. A. (1997). Thixotropy - A Review. Journal of Non-Newtonian Fluid
Mechanics (70):1–33.
Barnes, H. A. (2000). A Handbook of Elementary Rheology. Institute of Non-Newtonian
Fluid Mechanics, University of Wales.
Baronio, G. et al. (2003). Characterisation of the Materials Used in the Construction of
the Noto Cathedral. Construction and Building Materials 17:557–71.
Billberg, P. (2006). Form Pressure Generated by Self-Compacting Concrete — Influence

of Thixotropy and Structural Behaviour at Rest Peter Billberg School of Architecture
and the Built Environment Division of Concrete Structures Royal Institute of
Technology. PhD Thesis. School of Architecture and the Built Environment,
Stockholm, Sweden.
Binda, L. and Saisi, A. (2001). State of the Art of Research on Historic Structures in
Italy. Department of Structural Engineering, Politecnico of Milão, Italy.
Binda, L. et al. (1997). Repair and Investigation Techniques for Stone Masonry Walls.
Construction and Building Materials 11(3):133–42.
Binda, L., Penazzi, D. and Saisi, A. (2003). Historic Masonry Buildings: Necessity of a
Classification of Structures and Masonries for the Adequate Choice of Analytical
Models. Proc. of VI International Symposium Computer Methods in Structural
Masonry-STRUMAS, Rome, Italy 168–73.
Binda, L., Saisi, A. and Tedeschi, C. (2006). Compatibility of Materials Used for Repair
of Masonry Buildings: Research and Applications. Fracture and Failure of Natural
Building Stone 167–82.
Binda, L., Saisi, A. and Tiraboschi, C. (2000). Investigation Procedures for the Diagnosis
of Historic Masonries. Construction and Building Materials 14(4):199–233.
Bras, A. (2011). Grout Optimization for Masonry Consolidation. PhD Thesis.
Universidade Nova de Lisboa, Lisbon.
Buscall, R., McGowan, J. I., Morton-Jones, V. (1993). The rheology of concentrated
dispersion of weakly attracting colloidal particles with and without wall slip. Journal
of Rheology 37 (4): 621-641.
Cardoso, F. A., John, V. M., Pileggi, R. G., Banfill, P. F. G. (2014). Characterisation of
rendering mortars by squeeze-flow and rotational rheometry. Cement Concrete
Research 57: 79–87.
Chandra, S. and Bjornstrom, J. (2002). Influence of Superplasticizer Type and Dosage on
the Slump Loss of Portland Cement Mortars — Part II. Cement and Concrete
Composites 32:1613–19.
Cheng, D. (1987). Thixotropy. International Journal of Cosmetic Science 9:151–91.
Collepardi, M. (1990). Degradation and Restoration of Masonry Walls of Historical
Buildings. Materials and Structures 23(2):81–102.
EN 459-1. (2010). Building lime - Part 1: Definitions, specifications and conformity
criteria. European Committee for Standardization.
Eriksson, M., Friedrich, M. and Vorschulze, C. (2004). Variations in the Rheology and
Penetrability of Cement-Based Grouts—an Experimental Study. Cement and
Concrete Research 34(7):1111–19.
Ferragni, D. et al. (1982). Essais de Laboratoire Sur Des Coulis a Base de Ciment.
Mortars, Cements and Grouts used in the Conservation of Historic Buildings:
Symposium, 3–6 November 1981, Rome, ed. ICCROM, Rome 185–206.
Ferragni, D. et al. (1985). In Situ Consolidation of Wall and Floor Mosaics by Means of
Injection Grouting Techniques. Mosaics: Conservation in situ, Aquileia 1983,
ICCROM (3):83–101.
Ferrari, L. et al. (2010). Interaction of Cement Model Systems with Superplasticizers
Investigated by Atomic Force Microscopy, Zeta Potential, and Adsorption
Measurements. Journal of colloid and interface science 347(1):15–24.
Flatt, R. J. and Girardet, F. (2000). Injectable Slurries for the in Situ Conservation of
Pavement Mosaics. Proceedings of the 9th International Congress on Deterioration
and Conservation of Stone, Venice, June 19–24, ed. Vasco Fassina, Elsevier,
Amsterdam 2:297–305.
Giorgio, C. (2001). Restoring the Basilica of St. Francis of Assisi. CRM - Cultural
Resources Management 24(8):26–29.
Jolicoeur, C. and Simard, M. A. (1998). Chemical Admixture-Cement Interactions:
Phenomenology and Physico-Chemical Concepts. Cement and Concrete Composites
20(2-3):87–101.
Jorne, F., Henriques, F. M. A, Baltazar, L. G. (2014). Evaluation of consolidation of
different porous media with hydraulic lime grout injection. Journal of Cultural
Heritage 16: 438-451.
Jorne, F., Henriques, F. M. A, Baltazar, L. G. (2015). Influence of superplasticizer,
temperature, resting time and injection pressure on hydraulic lime grout injectability.
Correlation analysis between fresh grout parameters and grout injectability. Journal
of Building Engineering 4: 140-151.
Kenyon, C. (1996). Stabilising the Substrate: Grouting – the Options: Review of Stone
Section Meeting, British Museum, 28th October 1995. Conservation News 59:71–73.
Lapasin, R., Papo, A., Rajgelj, S. (1983). Flow behavior of fresh cement pastes; a
comparison of different rheological instruments and techniques. Cement and
Concrete Research 13: 346-349.
Mewis, J. and Wagner, N. J. (2012). Colloidal Suspension Rheology. Cambridge
University Press.
Mooney, M. (1931). Explicit formulas for slip and fluidity. Journal of Rheology. 2: 210–
222
Nehdi, M., Rahman, M. A. (2004). Estimating rheological properties of cement pastes
using various rheological models for different test geometry, gap and surface friction.
Cement and Concrete Research 34(11): 1993–2007.
Oldenbourg, C. (1996). Stabilising the Substrate: Grouting – the Options: Christoph
Oldenbourg’s Contribution on 28th October 1995. Conservation News 60:45.
Penazzi, D. et al. (2001). Repair and Strengthening of Historic Masonry. International
Congress More than Two Thousand Years in the History of Architecture
Safeguarding the Structure of our Architectural Heritage, Bethlehem (Palestine)
2:1-6.
Ran, Q. et al. (2010). Adsorption Mechanism of Comb Polymer Dispersants at the

Cement/water Interface. Journal of Dispersion Science and Technology 31(6):790–
98.
Rosquoët, F. et al. (2003). Experimental Study of Cement Grout. Cement and Concrete
Research 33(5):713–22.
Roussel, N. (2007). Rheology of fresh concrete: From measurements to predictions of
casting processes. Materials and Structures 40(10): 1001–1012.
Roussel, N., Lemaître, A., Flatt, J. R., Coussot, P. (2010). Steady state flow of cement
suspensions: A micromechanical state of the art. Cement and Concrete Research
40(1): 77–84.
Russel, W. B., Saville, D. A., Schowalter, W. R. (1989). Colloidal Dispersions.
Cambridge University Press.
Shaw, D. J. (1992). Introduction to Colloid & Surface Chemistry, 4th Edition. Reed
Educational and Professional Publishing Limited, Great Britain.
Struble, L. and Sun, G. K. (1995). Viscosity of Portland cement pastes as a function of
concentration. Advanced Cement Based Materials 2: 62–69.
Syngellakis, S. (2014). Heritage Masonry - Materials and Structures. Wessex Institute of
Technology, UK.
Toumbakari, E. E. and Van Gemert, D. (1998). Injection Grouts for Ancient Masonry:
Strength Properties and Microstructural Evidence. Compatible Materials for the
Protection of European Cultural Heritage, ed. G. Biscontin, A. I. Moropoulou, M.
Erdik and J. Delgado Rodrigues, Technical Chamber of Commerce, Athens 56:191–
200.
Toumbakari, E. E., et al. (1999). Effect of Mixing Procedure on Injectability of
Cementitious Grouts. Cement and Concrete Research 29(6):867–72.
Toumbakari, E. E. (2002). Lime-Pozzolan-Cement Grouts and Their Structural Effects on
Composite Masonry Walls. PhD Thesis, Katholieke Universiteit Leuven.
Valluzzi, M. R. (2009). Requirements for the Choice of Mortar and Grouts for
Consolidation of Three-Leaf Stone Masonry Walls. RILEM International Workshop
Repair Mortars for Historic Masonry, Delft University of Technology, The
Netherlands, 26–28 January 2005, ed C. Groot, RILEM Publications SARL 382–97.
Van Rickstal, F. (2000). Grout Injection of Masonry, Scientific Approach and Modeling.
PhD Thesis, Katholieke Universiteit Leuven.
Van Rickstal, F. et al. (2003). Development of Mineral Grouts for Consolidation
Injection. Consolidation of Masonry, Advances in Materials Science and restoration,
ed. D. Van Gemert, Verlag, Freiburg 1:61–75.
Verwey, E. J. W. (1947). Theory of the Stability of Lyophobic Colloids. The Journal of
Physical Chemistry 51:631–36.
Vikan, H. (2005). Rheology and Reactivity of Cementitious Binders with Plasticizers.
PhD Thesis Norwegian University of Science and Technology, Norway.
Wallevik, J. (2009). Introductions to Rheology of Fresh Concrete. Course compendium,

ICI Rheocenter course, 17th-18th August 2009, Reyjavik, Iceland.
Wesche, K., Flaatten, H., Vom Berg, W. (1973). On the influence of separations on the
rheological investigations of cement paste and mortar in the rotation rheometer, in:
A. M. Neville, D. Slater (Eds.), Fresh Concrete, Important Properties and their
Measurement, vol. 2, RILEM, Leeds, England 1 -11.
Winnefeld, F. et al. (2007). Effects of the Molecular Architecture of Comb-Shaped
Superplasticizers on Their Performance in Cementitious Systems. Cement and
Concrete Composites 29(4):251–62.
Yamada, K. et al. (2000). Effects of the Chemical Structure on the Properties of
Polycarboxylate-Type Superplasticizer. Cement and Concrete Research 30(2):197–
207.
Yang, M., Neubauer, C. M., Jennings, H. (1997). Interparticle Potential and
Sedimentation Behavior of Cement Suspensions. Advanced Cement Based Materials
5:1–7.
Yoshimura, A. and Prudhomme, R. K. (1988). Wall slip corrections for Couette and
parallel disk viscometers. Journal of Rheology 32(1): 53–67.
Zajadacz, K. and Simon, S. (2006). Grouting of Architectural Surfaces – the Challenge of
Testing. Theory and Practice in Conservation: Inter. Seminar, ed. J. Delgado
Rodrigues and J. Mimoso, National Laboratory of Civil Engineering, Lisbon 509–17.
Chapter 3
LIQUID CO2-BASED GEL FLUID
Weixiang Cui*, Mingyue Cui and Chunpeng Wang

Overseas Engineering and Technology Research Institute,
Research Institute of Petroleum Exploration and Development, Beijing, China
ABSTRACT
Liquid carbon dioxide (CO2) fracturing technology is one of the waterless fracturing
technologies. It has the advantages of avoiding the use of water, reducing damage, and
increasing production by improving the reservoir pressure. However, the low viscosity of
liquid CO2 and the difficulty to pump proppant are the main technical challenges. No
conventional thickening agent can provide the required thickening used in 100% liquid
CO2. In this paper, the problem of thickening gelled liquid CO 2 has been solved by
developing a new surfactant and thickening agent. Surface active agent can make the
liquid CO2 gel by forming wormlike micelle or cylinder micelles structure. The cylinder
micelles enhance the viscosity through entanglement inter-actions above a critical
concentration.
Rheology and friction properties of liquid CO2 gel are evaluated through high
pressure pipeline rheological experimental device. The viscosity of liquid CO 2 gel fluid
changes under different pressures (10 to 20MPa), temperatures (0 to 100℃ ) and different
shear rates. The experiment result shows that the effective viscosity of liquid CO 2 gel
fluid is between 7.654 ~ 20.012mPa.s. The liquid CO 2 was thickening with viscosity
increase at 86 to 498 times under the same conditions.
Compared with the pure CO2, the viscosity enhancement ratio of liquid CO2 based
gel was at 86 to 498 times, and the mixed system of liquid CO 2 based gel showed the
characteristic of shear thinning. Pressure of CO2 thickening of cross-linked liquid mixture
viscosity has little effect, the effective viscosity only has a small amplitude increase; the
effective viscosity of liquid mixtures was decreased with the increase of temperature,
which has an exponentially decreasing trend. Considering all the factors, the effects of
shear rate and temperature have the greatest impact on rheological properties of
*
Corresponding Author Email: cuiwx69@petrochina.com.cn.
44 Weixiang Cui, Mingyue Cui and Chunpeng Wang
fracturing fluid. Temperature on the rheological parameters influence is bigger, whether

under liquid or supercritical condition, the liquid CO 2 fracturing fluid flow index n
increases with the increase of temperature, and consistency coefficient K decreases with
increase in temperature decreases, the effect of pressure and thickening agent ratio on the
rheological parameters are smaller.
Liquid CO2 based gel fracturing fluid provides a new fracturing system with less
formation damage to fracturing operation. This chapter will be beneficial to all engineers
and technologists who are currently working at waterless fracturing applications.
Keywords: liquid CO2，waterless fracturing fluid, high pressure pipeline rheological test,
effective viscosity, friction
INTRODUCTION
Compared with traditional fracturing, pure liquid CO2 fracturing, a waterless

fracturing technology has the features of easy flowback, no solid residue and low damage
to reservoir, making it a good stimulation way for ultralow-permeability reservoirs and
water sensitive low permeability tight oil and gas formations [1], so it has attracted wide
attention. However, liquid CO2 fracturing technology still faces some issues, like low
viscosity of liquid CO2, poor sand carrying capacity, fluid loss and high injection
pressure [2-4]. If the viscosity of liquid CO2 fracturing fluid can be improved
significantly, CO2 fracturing will become an effective development technique for water
sensitive low permeability reservoirs and low pressure reservoirs, increasing oil and gas
production of these reservoir largely. Therefore, how to improve the viscosity of the CO2
fracturing fluid and keep the no damage feature of CO2 fracturing at the same time have
become key issues in the study of dry fracturing technology [5]. To increase the viscosity
of fracturing fluid in dry fracturing, the fracturing fluid for dry fracturing needs to be
modified. In this study, we have developed a new surfactant thickening agent to increase
the viscosity of CO2, and the rheological and friction properties of the thickened liquid
CO2 mixture system have been evaluated with high pressure pipeline rheological
experiment at high temperature and high pressure, to give us an insight into the viscosity
performance of the liquid CO2 fracturing fluid [6].
EXPERIMENT
Rheological properties of non-Newtonian fluid can be measured by a variety of

instruments, such as capillary rheometer, rotating cylinder viscometer and control stress
rheometer etc. They have their own characteristics, capillary rheometer is suitable for low
and high shear rate, particularly applicable to high shear rate range, and can test at rising
Liquid CO2-Based Gel Fluid 45
pressure and temperature [7], so it has become the first choice in studying rheological
properties of thickened cross-linking CO2 fracturing fluid under real fracturing conditions
(Note: during fracturing, the pumping rate was 2.5m3/min in 78mm tubing, and the shear
rate in the pipeline was about 894s-1). In order to characterize the rheological properties
of CO2 dry fracturing fluid, the power law model which would not produce large error
due to the uncertainties in the preparation, thermal stability and shear deformation of the
fracturing fluid was adopted [8].
The basic equation for laminar flow of a viscous fluid in a circular tube is:
D w
U  f   2 d
2 w3 0
(1)
Figure 1. Schematic diagram of the high-pressure fracturing fluid experiment system.
Where U is mean velocity of the fluid in the pipe, D is the diameter of the pipe,
 w is
wall shear stress. Equation (1) is transformed into:
1 8U 3 w
 w   f   2 d
4 D 0
(2)
Figure 2. Photo of the high-pressure fracturing fluid experiment system.
Equation (2) is the basic formula of tube type rheometer. The constitutive equation of
the power law fluid is:
1
  n
f     
k (3)
The equation (3) is substituted into the basic formula of tube rheometer (2), and by
taking integration and logarithm on both sides, we get:
 3n  1 
n
 8U 
lg  w  lg k    n lg  
 4n   D  (4)
Where: wall shear stress

 w is:
pD
w 
4L (5)
RHEOLOGICAL PROPERTIES
In actual fracturing with liquid CO2 fracturing fluid, supercritical high pressure liquid
CO2 is usually mixed with additives and then injected into the formation, when flowing
in pipeline, a number of factors can affect the rheological properties of CO2 fracturing
fluid, such as temperature, pressure, thickener dosage, shear rate macroeconomic factors
[9], and internal factors such as microstructure of thickening agent, type of surfactant, in
comparison, in engineering application, macroscopic factors have stronger effects on the
properties of fracturing fluid, therefore, in the experiment, the effects of temperature,
pressure, shear rate, thickener ratio on rheological properties of fracturing fluid were
investigated primarily.
Rheological Test of Liquid CO2 Fracturing Fluid
The experiment was conducted in tube of 8mm diameter. It can be seen from the
experiment results in Table 1 the liquid CO2 fracturing fluid has an effective viscosity
from 7.654 mPa.s to 20.012mPa.s, 95 to 498 times of the viscosity of pure liquid CO2
fracturing fluid, indicating the thickener we developed works well in increasing viscosity
of the fracturing fluid.
Table 1. Tested viscosity of liquid CO2 fracturing fluid under different conditions
Pressure Temperature Thickener Shear Fracturing Viscosity of Viscosity

MPa ℃ dosage rate fluid viscosity liquid CO2 times
% s-1 mPa·s mPa·s
10 15 1.0 220 15.986 0.0892 179.2
10 15 1.0 442 9.996 0.0892 112.1
20 15 1.25 219 16.669 0.1081 154.2
20 0 1.5 393 20.012 0.1313 152.4
20 30 1.5 393 8.461 0.0891 95.0
20 61 1.5 224 11.8 0.0236 498
20 70 1.5 393 8.567 0.0524 163.5
20 100 3.0 393 8.120 0.0372 218.3
Effect of Shear Rate on Effective Viscosity of Thickened CO2 Crosslinking

Fluid
The effective viscosity curves of thickened CO2 fracturing fluid with shear rate at 15℃
and 10 MPa and 20 MPa respectively are shown in Figure 4. (a) and (b), in which ГTH
indicates the dosage of thickener. It can be seen from the figures that the effective
viscosity of the thickened liquid CO2 fracturing fluid at different temperatures decreases
exponentially with the increase of shear rate [10], which fully proves the thickened liquid
CO2 fracturing fluid is a typical shear thinning non-Newton fluid, and also at the shear
rate of less than 500 s-1, the changes of its viscosity is much more drastic than that at the
of shear rate of more than 500 s-1. Under the experimental temperature and pressure
conditions, CO2 is in liquid form. Chen [11] and other researchers suggested that the CO2
properties were similar to organic solvents, as shown in Figure 3, in the thickened liquid
CO2 fracturing fluid system, hydrophilic groups of rod or wormlike micelles formed by
crosslinking are drawn to a small amount of water molecules, meanwhile, pro CO2
groups are attracted by the CO2 molecules, resulting in the increase of interaction force
between molecules, and increase of CO2 viscosity. The viscosity change of the whole
system mainly derives from changes in spatial structure of surfactant micelles and the
change in CO2 droplet shape. As the properties of CO2 liquid is more similar to those of
Newton fluid, namely CO2 as the external phase of the mixture system, the effect of
shearing on its viscosity is negligible, so the shear thinning of the mixture system is
mainly originated from the destruction of surfactant spatial structure caused by shearing
in which the space network structure is gradually broken into single micelles, in other
words, the volume density of micelle sieve increases, moreover, the flow direction of
micelles turns from isotropic before shearing to single gradually, resulting in reduction of
flow resistance and consequently decrease of fluid viscosity.
Figure 3. Structure of surfactant micelles in liquid CO2.
Effect of Temperature on Viscosity of CO2 Fracturing Fluid System
In the actual fracturing operation, often high pressure super cold liquid CO2 and
thickener are mixed and then injected into the formation through tubing. As the fracturing
fluid enters into the formation, the temperature gradually increases, the high pressure
super cold CO2 will turn to supercritical state, the CO2 plays completely different roles in
the two states, therefore, when examining the effect of temperature on the effective
viscosity of liquid CO2 fracturing fluid, we made experiments on two processes, and the
thickened CO2 cross-linking liquid system with CO2 in liquid and supercritical states
were defined as liquid system and supercritical mixture system respectively.
(a)
(b)
Figure 4. The variation of effective viscosity with shear rate.
Figure 5. Variation of effective viscosity with temperature.

First the effect of temperature on viscosity of liquid mixture is analyzed. Figure 5

shows the variation of effective thickness of the mixture system with 3% of thickening
agent with temperature (0 to 100℃) at 20 MPa, and shear rate of 393 s-1 to 446 s-1
respectively. It can be seen from the figure that the effective viscosity of the liquid
mixture system decreases exponentially with the increase of temperature. This is caused
by the influence of temperature on the rheological characteristics of surfactant, based on
Harris’s study, the influence of temperature on surfactant is in full agreement with the
Arrhenius relationship, that is, the relationship between temperature and viscosity meets
the relation   Av exp  E f / RgT  , the rise of temperature increases the motion activity
of the surfactant molecules, subsequently, the damage of surfactant micelle network
structure and breaking of micelles speed up, making the activation energy of surfactant
solution decrease, so the mixture system shows decrease in effective viscosity, that is to
say the viscosity variation of the liquid mixture with temperature is mainly caused by the
effect of temperature on the viscosity of surfactant.
It can be seen from figure 5 that when the temperature reaches a critical temperature
of 31℃, the mixture system in supercritical state increases in viscosity to certain degree,
then with the increase of temperature, the mixture system in supercritical state decreases
in viscosity slightly, and finally tends to be stable. The reason of this phenomenon is that
when the temperature is higher than the critical temperature, the CO2 changes from liquid
state to supercritical state, consequently, the mechanism generating viscosity of the
system mutates. It can be seen from the figure the mixture system changes phase state in
the temperature range of 30℃ to 40℃, slightly higher than the critical temperature, this is
because the CO2 phase change is a gradual process, when the temperature is slightly
higher than the critical temperature, on one hand because the phase change of CO2 will
cause change of micellar structure, making intermolecular interaction strengthen, and
thus viscosity increase, on the other hand, the rise of temperature will reduce activation
energy of the system, making the viscosity of the system decrease with the increase of
temperature, so combining the two aspects, before 40℃, the former takes the dominant
role, when temperature increases further, the latter begins to take the leading role.
Calculation and Influence Analysis of Rheological Parameters
From the above experimental results, it can be seen that pressure and temperature
both affect flow characteristics of the liquid CO2 fracturing fluid, but at different degrees.
Temperature has the strongest influence on rheological properties of the thickened CO2
crosslinking mixed system, when correlated with rheological parameters solely,
temperature and rheological parameters meets an exponential function, therefore, the
effect of temperature on rheological parameters has been fitted under given pressure and
thickening agent ratio.
The above parameter models were used to fit the experimental data, to get the
rheological parameter models of liquid CO2 fracturing fluid under different conditions:
P = 10MPa，ГTH = 3.0%:
 T T  
2
n  exp  1.557  1.7749  0.9907    (6)

 Tcr  Tcr  
 T 
k  exp  0.4489  0.933  (7)
 Tcr 
Where: the experimental temperature is 31.2℃. The average error of the above two
equations are 1.28/% and 0.3% respectively, and their scope of application is:
0℃  T  35℃, P = 10MPa, and ГTH = 3.0%.
P = 20MPa，ГTH = 3.0%:
 T T  
2
n  exp 1.6702  1.9053  1.0279   

 (8)
 Tcr  Tcr  
 T 
k  exp  0.3803  0.9289  (9)
 Tcr 
Figure 6. Curves of flow index of the liquid mixture system with temperature.
Where: the experimental temperature is 31.2℃. The average error of the two
equations are 1.3% and 2.97% respectively, and their scope of application is: 0℃  T 
35℃, P = 20MPa, ГTH = 3.0%.
The effect of temperature on rheological parameters of liquid CO2 fracturing fluid
were calculated and analyzed by the above formulas, Figure 7 and Figure 8 are the curves
of flow index and viscosity with temperature of the mixture system at a thickening agent
ratio of 3%, and pressure of 10MPa and 20MPa respectively. It can be seen from the
experimental results that the temperature has stronger influence on the rheological
parameters, with the increase of temperature, fluid flow index increases, and consistency
coefficient decreases, indicating the rise of temperature weakens the non-Newtonian
behavior of the mixture system. The working mechanism is the same with the effect of
temperature on viscosity of the liquid mixed system mentioned above.
Figure 8 and Figure 9 are the flow index and consistency coefficient curves with
temperature of the liquid CO2 fracturing fluid under supercritical state at the thickener
ratio of 3.0%, and pressure of 10MPa and 20MPa respectively. It can be seen from the
figures that under supercritical condition, with the increase of temperature, the flow index
ranges from 0.4 to 0.5, and the consistency coefficient changes in the range of 0.18 - 0.28,
clearly, the rheological parameters of the supercritical mixture system change with
temperature in the same pattern with those of the liquid mixture system, that is, with the
increase of temperature, flow index the supercritical mixture system increases, and
consistency coefficient of the system decreases, but the changes of rheological
parameters of this system are in smaller magnitude.
Figure 7. Curves of consistency coefficient of the liquid mixture system with temperature.
Figure 8. Flow index of the mixture system under supercritical condition with temperature.
FRICTION PROPERTIES
Fracturing fluid is a kind of high pressure fluid, its main role is to exceed the
formation stress and tensile strength of the surrounding rock into the formation, the
formation of an effective crack, and proppant transport in the cracks. So whether the
pressure of fracturing fluid in pipe flow in the process of drop accurate analysis of the
calculation, not only related to the reliability of the whole fracturing process, and can
provide accurate parameters of fracturing fluid in fracture fluid entrance, has a very large
impact on the success of fracturing and field construction of economy.
Figure 9. Curves of consistency coefficient of mixture system under supercritical condition with
temperature.
Friction Test Principle
The total pressure drop of the fluid passing through the straight pipe is composed of
three parts, namely, the frictional pressure drop, the pressure drop and the acceleration
pressure drop. The expression is as follows:
dP dPf dPg dPa

   
dz dz dz dz (10)
In the formula:
dP

dz —— Total pressure drop, Pa;
dPf
—— Frictional pressure drop, Pa;
dz
dP
—— Multiple potential drop, Pa;
dz
dP —— Acceleration pressure drop, Pa.
dz
Under the condition of low flow velocity, the pressure drop can be ignored. In the
horizontal pipeline, the pressure drop can be ignored. According to the Darcy formula,
the pressure drop can be expressed as:
L fu
2
p  p f  
D 2 (11)
In the formula:
p f
——Frictional pressure drop, Pa；
 ——Friction resistance coefficient ；
f
——Mixture density, kg·m-3。
Therefore, the coefficient of frictional resistance can be obtained：

 p f / L D
 f u2 / 2
(12)
Influence Law of Friction Coefficient
This section mainly introduces the friction resistance characteristics of the CO2 thick
liquid mixture at different temperature, pressure and flow rate. The experiment got
pressure drop under different conditions, by the formula (2-18) to calculate the
corresponding friction coefficient conditions, and then analyzes the friction resistance
coefficient of variation of velocity in different conditions are compared, and the fitted
formula between Blasius drag coefficient and Re number. Table 2 for the pressure drop
under the condition of CO2 crosslinked thickening liquid mixing systems at different
flow velocity value and Re number, by the table can be seen under the experimental
conditions, the pressure drop of CO2 thickening crosslinked fluid mixed system is
3660Pa, the minimum is 1760Pa, the friction coefficient is 0.6878, the minimum is
0.1255, under the influence of various factors make a specific analysis.
Figure 10 is a different temperature at 15℃ viscous friction coefficient curve, the
proportion of CO2 thickening agent crosslinked liquid mixtures can be seen, with the
increase of the proportion of the thickener, little change of friction coefficient, this is
because in the experimental conditions, CO2 is nearly similar to the Newton fluid in
liquid form in the mixed system, because of the large volume fraction of liquid CO 2
(97% ~ 98.5%), and the final mixture viscosity is mainly controlled by foreign minister
CO2 viscosity, thus changing thickener ratio caused by the viscosity increase is very
small, so the friction pressure gradient change of fracturing fluid is small, the friction
coefficient slightly increase.
Figure 11 shows that the frictional resistance coefficient of the mixed system of CO2
condensed liquid mixture decreases with the increase of temperature when the pressure is
20MPa and the ratio of the thickener is 3%. The effects of elevated temperature for the
fracturing fluid system is more performance for the influence on the nature of the
surfactant above, through the experimental study, the critical temperature is less than CO2,
CO2 thickening crosslinked fluid mixed system with the increase of temperature the
effective viscosity is reduced, the friction resistance will reduce the flow brought the tube,
thus the performance of the friction coefficient decreases with the increase of temperature.
Figure 12 is a thickening agent proportion is 1.5%, the temperature is 15℃, the
pressure is 10MPa, 15MPa, the friction coefficient curve, 20MPa can be seen, the friction
coefficient of CO2 thickening of cross-linked liquid mixed system with the increase in
pressure increased. As mentioned earlier, the pressure mainly through the solubility and
the internal spatial structure of surfactant system affect the viscosity of thickener, the
increase of pressure resulting in effective viscosity CO2 thickening liquid mixture
crosslinking increased, the frictional pressure drop of CO2 thickening liquid mixture
crosslinking increased slightly, showing friction the drag coefficient with the increase of
the pressure increased.
Table 2. Pressure drop and Re’ of liquid CO2 fracturing fluid system
at different flow rates
Conditions Thickener Flow rate Pressure Re’ λ

% m/s Pa
10MPa 1.5 0.220 1759 99 0.6459
15℃ 1.5 0.276 1911 144 0.4459
1.5 0.331 2052 192 0.3330
1.5 0.387 2141 252 0.2541
1.5 0.442 2209 318 0.2010
1.5 0.552 2668 411 0.1557
1.5 0.663 3103 510 0.1255
15MPa 1.5 0.220 1760 99 0.6465
15℃ 1.5 0.276 1918 143 0.4477
1.5 0.331 2135 185 0.3464
1.5 0.387 2295 235 0.2724
1.5 0.497 2683 331 0.1931
1.5 0.552 2915 376 0.1700
1.5 0.663 3398 466 0.1374
20MPa 2.0 0.220 1762 99 0.6474
15℃ 2.0 0.276 1924 143 0.4489
2.0 0.331 2169 182 0.3519
2.0 0.387 2393 225 0.2840
2.0 0.552 3172 346 0.1851
2.0 0.663 3660 432 0.1480
3.0 0.219 1856 93 0.6879
3.0 0.275 2054 133 0.4828
3.0 0.331 2280 173 0.3699
3.0 0.387 2440 221 0.2896
Figure 10. The variation curves of friction coefficient under different thickening agents,
Figure 11. Curves of friction coefficient at different temperatures.
Figure 12. Variation curves of friction coefficient under different pressures.
Mathematical Model of Friction Coefficient
Non Newtonian fluid tube frictional pressure drop calculation flow, there is no
mature theoretical analysis method, after 50s, although there are some research is also
limited to the smooth region. Because of the low viscosity of non-Newtonian fluid, the
Reynolds number of the flow in the pipeline is small, so the calculation of the smooth
zone generally adopts some empirical formulas which can meet the engineering
requirements.
It is difficult to describe quantitatively the relationship between the momentum
transfer of fluid phase, the friction force on the interface and the shear rate. Therefore, the
influence of the physical parameters and dynamic parameters on the friction coefficient
can be expressed by the following formula:
  f ( p, T ,  , u, D, L) (13)
The dimensionless effects of various factors can be obtained:
  f ( Re' ) (14)
That is to say, the calculation of the smooth region of non Newton fluid turbulence
can also be extended by the generalized Reynolds number, so that the formula and the
Newton fluid. In fact, the drag coefficient of the tube slip zone depends not only on the
generalized Reynolds number, but also on the two calculation methods.
A) Blasius type formula:
a
 (15)
Re' b
In the formula, a, B is a function of the flow index, corresponding to different values

of a and B can be obtained in the form of data fitting.
B) Semi empirical formula

Based on Carmen’s (Karman) formula and relevant experimental data, a semi
empirical formula for calculating the drag coefficient of the smooth region of non
Newtonian fluid turbulence is presented:
4.0   (1 2 )  0.4

n'
1
 lg  Re' ( )   1.2 (16)
 n' 0.75  4  n'
4
The theoretical calculation and experimental data of the formula (18) are basically in
agreement with the experimental data:
n '  0.36 ~ 1.0 ， Re  2900 ~ 36000 。

'
Figure 13. Fitting relation between friction coefficient and generalized Reynolds.
In this study selected the first empirical formula fitting form, friction coefficient of
CO2 thickening liquid mixtures were cross-linked building mathematical model, obtained
the relationship between friction coefficient CO2 thickening of cross-linked liquid mixed
system and the generalized Reynolds number:
  50.911Re'0.9538 (17)
Figure 13 shows the fitting relation between friction resistance coefficient and
generalized Reynolds number. The correlation coefficient is 0.99827 and the average
error is about 2.61%. The scope of the formula is:
93  Re′  554, 1.5%  ΓTH  3.0%, 0℃  t  30℃, 10MPa  P  20MPa.
APPLICATION
The new thickened liquid CO2 fracturing fluid has been used in 1 well of Jilin
Oilfield in China. The well is the first well treated by water free fracturing in Jinlin
Oilfield. The reservoirs are silt-fine sandstone, with a grain size of 0.1-0.5mm, porosity
of 3%-14.3%, on average 8.14%, and average permeability of 2.06×10-3μm2. The section
of 1935-1942.5m was treated with 19.8 m3 of proppant and 653.5m3 of liquid CO2. The
field test proves the thickened liquid CO2 fracturing fluid has good sand carrying
performance, providing a new means of volume fracturing for tight gas reservoirs.
Figure 14. Overview of the liquid CO2 fracturing in a Jilin Oilfield well site.
Figure. 15. Proppant concentration profile of the thickened liquid CO2 fracturing fluid.
By inputting net pressure, pumping rate and viscosity into Fracpro PT simulator, the
simulated fracture height was 11.7 meters and simulated fracture length was 180.9 meters.
The average concentration of proppant was 3.6 kg/m2.
Oil production rate after fracturing was 8.6 t/day, compared with the gas production
rate of 0.8 t/day on average of wells treated with crosslinking guar fracturing fluid in the
same area, it is nine times higher, proving the CO2 fracturing can increase production of
wells significantly.
(a) (b)
Figure 16. Stimulated reservoir volume of CO2 waterless fracturing fluid and water based fracturing
fluid: (a) CO2 waterless fracturing (b) water fracturing.
The results of microseismic monitoring downhole show the stimulated volume of

conventional hydraulic fracturing is 270 thousand m3, while the stimulated volume of
CO2 fracturing of the same layer in an adjacent well is 702 thousand m3 (see Figure.16),
2.6 times of the former, proving CO2 fracturing can increase controlled reserves of wells
significantly. To achieve the same fracturing effect of water base fracturing fluid, CO 2
fracturing can reduce liquid volume used to half of that used in water base fracturing
fluid and fracturing stages, in other words, 3000-8000m3 liquid CO2 can achieve the same
stimulated volume by 7600-20800m3 of water base fracturing fluid.
REFERENCES
[1] Harris P. C., Pippin P. M. High-rate foam fracturing: Fluid friction and perforation
erosion [J]. SPE Production & Facilities, 2000, 15(1): 27-32.
[2] Blauer R. E., Mitchell B. J. and Kohlhaas C. A. Determination of laminar,
turbulent, and transitional foam flow losses in pipes [C]. The annual SPE of AIME
California regional meeting, San Francisco, Apr 4, 1974.
[3] Harris P. C. and Heath S. J. Rheology of crosslinked foams [J], SPE Production &
Facilities, 1996, SPE 28512.
[4] Wang Kai. High pressure study of weak intermolecular interactions in micelles
[D]. Jilin: Jilin University, 2005.
[5] Glaser N., Adams D. J., Boeker A., Krausch G. Janus particles at liquid-liquid
interfaces. Langmuir 2006;22:5227–9. [2] Eastoe J., Hopkins-Hatzopoulos M.,
Dowding P. J. Action of hydrotropes and alkyl-hydrotropes. Soft Matter
2011;7:5917–25.
[6] Ahualli S., Iglesias G. R., Wachter W., Dulle M., Minami D., Glatter O.
Adsorption of anionic and cationic surfactants on anionic colloids: supercharging
and destabilization. Langmuir 2011; 27:9182–92.
[7] Esumi K., Ueno M., editors. Structure-performance relationships in surfactants.
2nd ed. New York: Marcel Dekker; 2003.
[8] Tabor R. F., Eastoe J., Dowding P. J. Surfactant adsorption and aggregation
kinetics: relevance to tribological phenomena. In: Biresaw G, Mittal KL, editors.
Surfactants in tribology. Boca Raton: CRC Press; 2008 [chapter 19].
[9] Shen Linhua. Study on the rheology and heat transfer mechanism of the foam
fracturing fluid in a vertical falling pipe [D]. Xi’an: Xi’an Jiao Tong University,
2006.
[10] Tabor R. F., Grieser F., Dagastine R. R., Chan D. Y. C. Measurement and analysis
of forces in bubble and droplet systems using AFM. J Colloid Interface Sci
2012;371:1–14.
[11] Chen Y., Pope T. L. Novel CO2-Emulsified Viscoelastic Surfactant Fracturing
Fluid System [C]. SPE paper 94603 presented at the SPE European Formation
Damage Conference, Sheveningen, The Netherlands, May 25-27 2005.
Chapter 4
RHEOLOGY APPLIED TO FOOD PRODUCT DESIGN
Anabela Raymundo*, Ana Paula Batista and Isabel Sousa

LEAF - Linking Landscape Environment Agriculture and Food Research Centre,
Instituto Superior de Agronomia, Universidade de Lisboa, Lisboa, Portugal
ABSTRACT
Innovation is recognised as the key driver of economic growth. Trends in food
industry deal with a permanent need to develop new food products, adjusted to the
consumer demands and, in the near future, to scarcity of food resources. Concepts of food
as health and wellness promoters, “free-from” products, the use of alternative ingredients
such as new protein sources, or the use of by-products in food formulations are current
topics that act as driving forces for innovation. Currently, food product development
methodologies are generally based on the chemical and nutritional properties,
complemented with a sensory validation, carried out in the final stages of the
development process. However, the structure of food proved to be determinant for the
food appeal and strongly impacts consumers’ acceptance. It is well known that products
with the same chemical composition can present very different structures, resulting
differently perceived texture and sensory properties. Therefore, the use of rheological
tools to design new formulations and to optimise new industrial processes, from the
earliest stages of the development phases, proves to be a promising methodology. In the
creative process, the food macromolecules (proteins and polysaccharides) are the major
players for the creation of relevant food structures such as foams, emulsions and gels. For
the purpose, successful case studies will be presented and described with a focus on the
use of rheological techniques as the tools for the design of new food products based on
alternative macromolecules. The development of gluten-free or vegetarian products by
using alternative proteins and polysaccharides, as well as the use of food industry by-
products as source of these structuring biopolymers, along with the structural
implications of adding protein and/or fibre-rich healthy ingredients (e.g., microalgae,
*
Corresponding Author: Instituto Superior de Agronomia, Tapada da Ajuda, 1349-017 Lisboa, Portugal.
Email: anabraymundo@isa.ulisboa.pt.
64 Anabela Raymundo, Ana Paula Batista and Isabel Sousa
psyllium) are some of the examples where rheology tools can give a powerful insight and
contribute to decision-making on product development stages. It will be emphasized how
the incorporation of new ingredients induces changes in the viscosity profile and other
rheological parameters, related to the consistency of the fluid food products or to the
texture of the solid or soft-solid foods. The use of small amplitude oscillatory
measurements will also be discussed in order to access the changes of the internal food
structure promoted by changes on processing procedures or by the inclusion of new food
ingredients on the food formulation. The impact of these variables on the development of
the food structure will be discussed based on the linear viscoelastic properties. The use of
sensory and instrumental texture measurements, in different experimental modes and
applied to different foodstuffs, will be discussed. The possibility to correlate instrumental
texture parameters with sensory evaluation will be considered as the strategy to bring
characteristics of the developed products to the consumer’s requirements.
Keywords: food-rheology, emulsions, gels, pasta, cookies, bread, cheese
1. INTRODUCTION
Food design is a complex activity and postulates a multidisciplinary approach.

Nutrition, chemistry, sensory perception, consumer’s acceptance, engineering, are some
of the uptakes to consider. However, the structure of food proved to be determinant of the
food appeal, mouthfeel, aroma, etc. and strongly impacts consumers’ acceptance. It is
well known that products with the same chemical composition can present very different
structures, resulting in differently perceived texture and sensory properties. Therefore, the
use of rheology tools to design new formulations and to optimise new industrial
processes, from the earliest stages of the development phases, proved to be an efficient
methodology.
In the creative process of food product development, the food macromolecules (Hill
et al. 1998) are the ingredients (mainly proteins and polysaccharides) fully responsible
for the relevant food structures (Blanshard & Mitchell 1988) such as foams, emulsions
and gels, soft solids or crisp bytes, although the presence of fat crystals and liquid fat are
very important in oral perception.
There are at least 2 cases as the main drivers for product development:
i. Responding to a consumer need
Developing a product highly convenient, with positive impact on health, fulfilling the
nutrition gaps or following recent trends of very active adults (Dias et al., 2015) or young
scholars or senior/retired active people (Rodriguez et al., 2003), or people with special
restraints like diabetics (Torres et al., 2013), or celiacs (Torres et al., 2014a),
corresponding to different characteristics and composition of the food products.
Rheology Applied to Food Product Design 65
ii. Responding to industrial efficiency
Industrial efficiency is most frequently related to lower costs or to produce more with
less (e.g., adding value to by-products) getting closer to concepts of sustainable
intensification or circular economy. In both cases the tactics are often centred on
formulation and processing:
a) Ingredient replacement.
The reasons for replacing an ingredient can vary from economic - lowering
formulation costs, to health promotion - using a functional ingredient, adding value and
differentiation to the new product.
b) Changing the processing line
For economic/environmental reasons, the use of low energy processing, often

combining multi-processes to be efficient and lower the energy consumption, or reasons
of other nature like increasing the lifetime of the product, can lead to changes in the
processing line.
When considering sustainable intensification and resource efficiency, the re-
introduction of by-products in the production stream (Tomé et al., 2014) is a challenge,
especially in the case of Food, where the safety and quality issues are highly demanding.
According to our practice, in the product development method, the first step is to
acquire all food products, available in the market, that are closely related to the food
product under development.
The next action is to create a data base with the relevant properties quantitatively
evaluated, the ones that are determinant in terms of consumer acceptance, like colour and
texture.
Since texture is a consequence of the internal structure of the food material, it is
mandatory to assess the structure of the food product. Microscopic techniques, although
being very powerful nowadays, give an image of the structure often very difficult to
interpret, most of the times not a highly representative image of the whole sample and
rarely quantifiable, therefore the quantification of the rheology behaviour in terms of one
indirect assessment of the food structure is fundamental. Using the rheology techniques
one can quantify:
I. the fracture profile of the crisp solids, or

II. the viscoelastic functions of the soft solids, and
III. the flow profile of the thick or thin liquids
After gathering all the data needed, testing all the related products commercially
available, one can do a basic statistical analysis to find the data distribution and focus on
the mid-range values.
These mid-range values are essential to define our targets. The aim is to optimise
food formulation or food processing with these values in mind.
Sometimes one needs to find out what are the variables that can modify the targeted
properties (Franco et al., 1998; Batista et al., 1999; Raymundo et al., 1998; 1999), and
Pareto-charts or similar statistical analysis can be of help, but usually one knows, from
the beginning, where to concentrate the efforts:
I. on formulation, or
II. on processing
Formulation is often consider to be the easier case in product development. When

modifying the formulation is related to the introduction of a functional ingredient with
positive impact on health, for example (Raymundo et al., 2002) or when the main
ingredient needs to be replaced for ethic believes like in the vegan diets (Arozarena et al.,
2001; Nunes et al., 2003) when an ingredient is replaced to reduce the costs, the
developed product has to fulfil all the specifications of the former product and that
facilitates the work. When the change in formulation relates to ingredients that impact the
structure, like the classic replacement of sugar; or the cross-off of gluten; these are the
difficult cases in product development, since their structure can be greatly affected. The
sugar replacement/cut out is a classic (Gaspar et al., 1998) namely in fruit juices is a case
study, where the need to introduce thickening agents, like polysaccharides, to compensate
for the sensory perception of the juice “body,” is well known.
In food processing, the simple introduction of low temperatures to increase the shelf
life of a product can be quite complex if the product is structure sensitive like soft paste
cheese (Alvarenga et al., 2011; 2013). Nowadays, with the recurrence of many processing
alternative techniques like extrusion-cooking, HHP (high hydrostatic pressure), OH
(ohmic heating), PEF (pulsed electric fields) and others, on one side and enzymatic
(Nascimento et al., 2006) and fermentation techniques on the other side, finding the right
processing variables that impact the relevant properties (Cabodevilla et al., 1994;
Raymundo et al., 2001; Alvarenga et al., 2008), when developing a food product, can be
quite complex.
In the classic processing layouts pH, Temperature, time, and moisture are commonly
the variables under consideration. However, when dealing with hydrocolloids, ions in
solution can make a difference (Gaspar et al., 1997; Torres et al., 2014b) as well as ion
strength. The responses, or measured properties, we need to target are the ones derived
from the data base and related to values of final aw (water activity) and pH for stability;
Lab* for colour; rupture stress and brittleness for hard solids; texture (Sousa et al., 1997)
storage and loss moduli vs frequency for soft solids and consistency and flow index or
viscosity for fluid foods.
With this matrix of variables and responses one can use the optimization tool of the
RSM (response surface methodology) to track how the variables determine the responses.
This tool is very useful since it includes the linear and non-linear relations as well as the
interactions between 2 or more variables, using a finite and manageable number of
experiments.
Following the same rationale, temperature can easily be replaced by pressure in the
case of HHP processing or by the intensity of the related effect in the other processing
methods. In the bio-processing areas, when enzymes and microbes are used, the
complexity increases and the know-how and experience of the team is crucial.
Using the response surfaces, derived from the RSM technique, or the respective
equations, one can go backwards and use the previously set targets to develop the product
with the right properties.
Some study cases, where this methodology is applied, are discussed in detail, further
on in this chapter. To be sure of the success of the food development it is recommended
to subject the developed food product to a sensory evaluation, which can be as simple as
a consumer panel with a hedonic response, before starting to launch for the market and
the marketing teams in the companies are well aware of this fact.
2. FOOD EMULSIONS
Emulsions are the basis of a huge range of food products and are generally stabilised
by either protein and/or surfactants. The market for food emulsions has grown in recent
years and it is expected to be around 5% between 2017 and 2020.
The increased demand for novel food emulsions, namely stable mayonnaises and
salad dressings, with a balanced nutritional composition, low-fat, egg-free, without
another allergenic proteins and sensory appealing, has been a trend. The appearance,
texture, flavour profile, and stability of emulsions is strongly influenced by their droplet
concentration (McClements, 2015), which is dependent on the interfacial properties of
the emulsifier systems - proteins and/or other surfactants.
The role of proteins as emulsifying agents have been extensively explored by the
food industry. They are natural, non-toxic, cheap and widely available, being interesting
ingredients to design novel foods. Proteins stabilise emulsions by forming viscoelastic
adsorbed layers, and their mechanical properties influence the emulsions stability
(McClements, 2015). The long term stability of an emulsion, desired as a commercial
requirement, can be achieved by selecting efficient emulsifier and stabilizing agents.
Emulsifiers are surface active molecules that facilitate the formation of emulsions by
lowering the oil-water interfacial tension and imparting stability by forming a protective
film around the dispersed droplets. Stabilizers are components that are able to improve
the stability of an emulsion, usually by increasing the viscosity of the continuous phase
(thickening agents) or by a colloidal protective action (Franco, 1995).
To design novel emulsion-based products by using mixtures of ingredients with
different interfacial activity and thickening capacity, the rheology techniques are relevant
tools, as it will be evidenced by the examples of applications discussed in this section.
2.1. Lupin Protein Emulsions
Recently, animal protein consumption has been substantially altered due to animal
diseases, cholesterol intake worries and strong demand for healthy food, along with ethic
orientations of denying animal intakes of any kind (Bollinger, 2001). The use of
vegetable proteins, especially from legume seeds, can be an interesting alternative to
animal protein based foods. Its high protein level and well-balanced amino-acid
composition makes them important sources of protein, with potential to replace meat and
dairy proteins, if necessary (van Vliet et al., 2002). Besides nutritional characteristics,
leguminous seeds such as lupin (Lupinus albus) or pea (Pisum sativum L.) protein
isolates present good functional properties, namely emulsion and gelling capacity, which
has led to a greater interest in this type of protein source as a promising food ingredient
(e.g., Guéguen, 2000; O´Kane et al., 2005).
The development of food emulsions, like mayonnaise or salad dressing, replacing the
traditionally used egg-yolk by lupin protein was previously studied by Raymundo et al.
(1998; 1999; 2001) and Franco et al. (1998). These studies were based on the texture and
rheology behaviour of the emulsions. The product development process started with the
characterisation of a wide range of commercial mayonnaises, in terms of texture
properties, linear viscoelastic behaviour and flow. These results generated a database that
was further used as a target for the development of lupin emulsions and the Response
Surface Methodology (RSM) was applied as a statistical tool for optimisation.
In terms of texture evaluation, Texture Profile Analysis (TPA) was performed for
different types of emulsions and it was stated that firmness and adhesiveness were the
most important variables to describe this type of products. On Figure 2.1, an example of a
texturogram, force versus time, obtained for lupin protein emulsions with two different
concentrations, can be observed. Firmness was considered as the maximum resistance to
the penetration of a cylindrical probe, recorded as the maximum force and adhesiveness,
characteristic of sticky materials, is defined as the resistance of the material when the
probe is recessing, corresponding to the first negative area of the graph.
Figure 2.1. Texturogram of lupin protein emulsions with 2 and 5% (w/w) of protein isolate.
The linear viscoelastic behaviour of the lupin emulsions was accessed using the small
amplitude oscillatory measurements (SAOS), considering a shear stress within the
viscoelastic linear region (Raymundo et al., 1998; 1999; 2001; Franco et al., 1998). An
example of the mechanical spectra obtained, can be observed on Figure 2.2, for
emulsions prepared with lupin protein subjected to different heat treatments, i.e., with
different levels of protein thermal denaturation – temperature ranged from 50 to 90ºC,
during 25ºC.
Figure 2.2. Evolution of the storage (G´), loss (G´´) moduli and Plateu modulus (G 0N) for emulsions
obtained with lupin protein submitted to different temperatures, during 25 min. (Raymundo et al., 1998
- adapted).
The interpretation of the mechanical spectra patterns gives information about the
degree of the emulsion structure and consequently on its stability. For denaturation at 70
and 90ºC, during 25 min, the occurrence of a well-developed plateau region, with a slight
frequency dependence for G’ can be observed and it is related with the formation of an
elastic structural network, made up from molecular entanglements with physical nature,
as a second level of structuring, acting as a reinforcement of the primary structure, which
confers high stability to the emulsion (Dickinson, 1998; Franco et al., 1995). For the
emulsions prepared with lupin protein treated at 50ºC, during 25min, a mechanical
spectrum with a markedly dependence on the oscillation frequency is observed. This
mechanical spectrum shape reflects a less structured internal organization and is
associated with a lower stability of the final product.
To apply the RSM tool, it is required to express each mechanical spectrum as a single
figure, to be used as dependent variable. The Plateau modulus (G0N) was considered to be
a measure of the density of the entanglements in the particulate network of oil droplets by
Barnes et al. (1989). This parameter can be estimated as the value of G’ obtained for the
minimum value of the loss tangent (tan = G’’/G’), according to Wu (1989). This figure
is commonly used to characterize different types of emulsions (Trujillo-Cayado et al.,
2016; Pires et al., 2012) and is an expedient way to compare different emulsion-systems,
obtaining indication about their stability.
The evaluation of droplet size distribution (DSD), accessed by laser diffraction, is
another approach to predict the emulsion stability. Average diameter can be calculated, as
Sauter Diameter (d32) that can also be used as a response for the RSM. On Figure 2.3 the
DSD curves and respective d32 are represented, for the same emulsions analysed in Figure
2.2. It has been found that emulsions with higher G0N values are associated with lower
droplet size values, which are related with a greater physical stability.
Figure 2.3. Droplet size distributions and respective Sauter diameter (d32) for emulsions obtained with
lupin protein submitted to different temperatures, during 25 min. (Raymundo et al., 1998 - adapted).
The design of emulsions based on alternative ingredients should ensure that its flow
behaviour is the expected for this type of products. Emulsions obtained from vegetable
proteins typically exhibit a shear-thinning behaviour, with a first Newtonian plateau
region for low shear rates and a second Newtonian plateau for higher shear rates
(Raymundo et al., 2001). Several rheological models can be adjusted to these curves, in
order to obtain their characteristic parameters. It has been verified that the Carreau
equation presents a good fit to the experimental data in most cases (Barnes, 1989;
Raymundo et al., 1999; Franco et al., 1998). Thus the zero shear rate limiting viscosity
(0), obtained from the first Newtonian plateau region can also be considered as a
response to be used in RSM. This parameter is directly related to the degree of the
emulsion structure and with the long term stability. Figure 2.4 presents flow curves of
emulsions prepared with lupin protein denatured at the previously referred conditions. It
is clear that the emulsion prepared with the protein isolate denatured under milder
conditions (50°C, 25 min) has lower values of 0, which is associated with a less
structured system, resulting from the low values of surface hydrophobicity of the
denatured lupin protein.
Figure 2.4. Evolution of the steady-state flow curves of emulsions obtained with lupin protein
submitted to different temperatures, during 25 min. (Raymundo et al., 1998 - adapted).
In the presented study, the denaturation conditions of the protein isolates that lead to
the formation of stable mayonnaises and salad dressings were optimized (Raymundo et
al., 1998). Raymundo et al. (1999; 2001) also evaluated the interaction of these isolates
with low molecular weight surfactants for the design of emulsions with different
applications, following the same approach. Franco et al. (1998) used the same approach
to optimize the processing conditions of these lupin protein emulsions. In conclusion,
“mayonnaises” and salad dressings egg-free, using lupin protein as emulsifier, where
developed including low fat versions with 30% reduction of oil (Raymundo et al. 2002).
2.2. Microalgae as an Ingredient for Food Emulsions
Microalgae are photosynthetic organisms that can be used as food ingredients due to
their well-balanced chemical composition, rich in proteins, essential amino acids,
polysaccharides, glucans, fibres, polyunsaturated fatty acids (PUFA), sterols among other
compounds (Gouveia et al., 2008a). Moreover, microalgae are extremely rich in natural
pigments, such as chlorophyll, phycobilliproteins (phycocyanin and phycoerythrin) and
carotenoids (e.g., -carotene, lutein astaxanthin, cantaxanthin, fucoxanthin) widely
recognized as having health-beneficent effects namely protection from oxidative stress
acting as antioxidants (Gouveia et al., 2008a). In fact, the microalgae growth conditions
can be manipulated (e.g., light, salinity and nutritional stresses) in order to induce
carotenoid accumulation in a process called carotenogenesis (Gouveia et al., 1996). The
use of microalgal biomass in food products such as oil-in-water (o/w) emulsions can
provide multiple benefits, from a nutritional point of view, as well as natural colouring
effect, especially desirable in egg-free vegetable “mayonnaises.”
However, the incorporation of food ingredients into an emulsion has implications on
its rheological behaviour and stability. Such influence is dependent upon the ability of
these ingredients to be disposed inside the droplets, at the interface and/or to increase the
viscosity of the continuous phase.
Taking as a starting point pea protein emulsions, similar to the ones developed with
lupin protein by Raymundo et al. (1998), microalgal biomass was incorporated into the
optimised system, in order to design food emulsions with alternative appealing colours
and added value in terms of health-beneficent effect (Gouveia et al., 2006).
The interactions of microalgae with emulsifier and stabilising agents at the surface of
the emulsion droplets as well as in the aqueous continuous phase (Dickinson, 2003;
Klinkerson et al., 2004) can contribute to a reinforcement of the emulsion structure
through the formation of physical entanglements (Clark et al., 1991; Riscardo et al.,
2003). However, if these constituents are incompatible, the emulsion may become less
stable, as in the case of surface viscoelasticity loss induced by the displacement of the
protein from the interface (Dickinson & Hong, 1995).
The ability of microalgal biomass to act as a pigment in food emulsions was
evaluated: Spirulina maxima, Chlorella vulgaris (green and carotenogenic) and
Haematococcus pluvialis (carotenogenic) biomass was added, at various concentrations,
to the aqueous continuous phase of the pea emulsions (Figure 2.5).
The addition of microalgal biomass originated emulsions with higher firmness than
the control emulsion (without biomass addition). For all microalgae, a linear positive
relation (R2>0.84) between biomass concentration and firmness was observed, although
this dependence assumes distinct features. It is also possible to observe the positive
impact of the carotenogenic microalgae (Chlorella orange and Haematococcus) on the
increase of the emulsion firmness and the reduced impact on firmness promoted by the
Spirulina addition. The increase in emulsion texture properties, through microalgal
biomass addition, can be due to an increase of the viscosity of the emulsion continuous
phase, by a stabilizing mechanism similar to the effect of a polysaccharide addition such
as xanthan gum or starch (Xie & Hettiarachchy, 1997; Raymundo et al., 2002). The
possibility of establishing interactions between pea protein and microalgal biomass,
should also be considered, since it can contribute to a structural reinforcement through
the formation of physical entanglements (Clark et al., 1992; Riscardo et al., 2003).
Figure 2.5. Firmness of pea o/w emulsions coloured with different concentrations of Spirulina maxima,
Chlorella vulgaris green, Chlorella vulgaris orange (carotenogenic) and Haematococcus pluvialis
biomass (Batista, 2012).
The microalgal biomass “fat mimetic” capacity was also investigated by Raymundo
et al. (2005). This effect deals with a cooperative action between the protein emulsifier
and the microalgal components, resulting in a reinforced emulsion structure. The total oil
content can thus be reduced, yielding emulsions with the same rheological and sensory
properties. The study was conducted by lowering the oil content from 65% to 50% at
constant pea protein content (3%) and microalgal biomass concentration (2%). The
emulsions were analysed in terms of texture, linear viscoelasticity, flow behaviour and
droplet size distribution (Raymundo et al., 2005). On Figure 2.6 the mechanical spectra
and the respective values of the plateau modulus for these emulsions are represented.
Figure 2.6. Mechanical spectra of o/w emulsions with 3% pea protein and 2% of Spirulina maxima
(Sp), Haematococcus pluvialis (Hp), Chlorella vulgaris green (Cv_gr) and orange (Cv_or) microalgal
biomass, for different oil content and respective values of G0N (Batista, 2012).
Through this approach it was possible to design low fat emulsions based on the
interaction between pea protein and microalgae biomass.
From the dynamic measurements, it was observed that, for a certain protein and
biomass concentration, higher oil content induced a reinforcement of the emulsion
structure, for all the microalgae studied. This was demonstrated by the increase of the
plateau modulus (G0N). For the emulsion enriched with Chlorella (green and orange), the
G0N value was significantly higher than for the standard. It was possible to produce
emulsions with 55% oil that were more structured than the emulsions with 65% oil and
no microalgal biomass addition. This is a relevant finding to obtain low-fat emulsions,
which represents a trend in terms of market demand in recent years.
2.3. Emulsions Stabilized by a Complex System of Animal and

Vegetable Proteins
The interaction between vegetable and animal proteins to stabilize food emulsions
(e.g., Tomé et al., 2014) is a topic with growing interest for the food industry. Graça et al.
(2016) used SAOS measurements to optimise protein based o/w emulsions stabilized by
different animal/vegetable blends ratios to develop a new meat product filling. Collagen
and pea proteins were selected, considering their recognised emulsifying activity. Pea
protein isolates have a similar behaviour to the previous referred lupin protein isolates
(section 2.1), in the stabilization of emulsions, forming structured and stable systems.
Collagen has been widely used in food industries as an ingredient to improve the
elasticity, consistency, and stability of foods, and can also be used as emulsifier (Surh et
al., 2006).
An interesting possibility is to convert oil-in-water emulsions into gels, to create
foods with improved sensory and texture profile and better preservation properties (Line
et al., 2005; Manoi & Rizvi, 2009). In many cases, the production of these food materials
involves a controlled heat treatment.
Based on the targeted characteristics of a filling product designed by the marketing
department of the company, oil-in-water emulsions were prepared, at room temperature,
with 9.0% of total protein, considering mixtures of collagen protein isolate (CP) and pea
protein isolate (PP) at different proportions.
The developing emulsion can be used as a filling in a product which goes through a
fast heating at high temperature, followed by controlled cooling to room temperature. In
this process it is desirable to have low viscosity emulsion in the initial phase, to facilitate
pumping (at 40°C), but which has the ability to form a gel upon cooling. The gel
formation based on the emulsion leads to an improvement on the product smoothness and
contributes to the increase of its conservation by lowering the water availability.
The development of the CP/PP emulsions was preformed based on their rheology
behaviour, simulating in the rheometer the main steps of the industrial processing. After
emulsification, at room temperature, the emulsions were subjected to a controlled heating
from 20 to 90ºC, at 0.5ºC/min, followed by a cooling step. From Figure 2.7 it is possible
to observe the impact of the heating and cooling process on the emulsions prepared with
9.0% (w/w) of total protein, with a ratio of 75% collagen and 25% pea protein. During
the heating cycle, a tendency to overlap and reverse the viscoelastic functions (G´ and
G”), from 20ºC to 90ºC, is observed. After the subsequent cooling cycle, the G´ and G”
values increase and remain practically constant, i.e., there is a strong build-up of system
structure with cooling. This might indicate the prevalence of a gel structure over the
emulsions structure on cooling. This type of thermal induced behaviour is well-adapted to
the industrial purpose, to obtain a final gel structure, specially designed to be used as a
filling.
Figure 2.7. Temperature sweeps from 20 to 90ºC (heating curves - round symbols) and back to 20ºC
(cooling curves – triangular symbols), at 0.5ºC/min, for samples with 9.0% (w/w) total protein, at a
ratio of 75% collagen and 25% pea protein. G’ – full symbols, G’’ – open symbols (Graça et al., 2016).
The rheological techniques used for the development of new emulsions with special
emphasis on the evaluation of the texture (firmness and adhesiveness), the oscillatory
measurements in the linear viscoelastic zone as well as the emulsions flow behavior
proved to be an useful tools in food product development.
3. FOOD GELS
Gels provide structure and stability to food products, entrapping water and food
ingredients such as flavours, sugar and oil in the gel network. Examples of food gels are
cheese, (hard) boiled eggs, jelly, custards, sausages, yogurt, tofu, etc. The development of
appropriate mechanical, structural and physicochemical properties in most foods is
mainly attained by using gelling agents. These agents are macromolecules, essentially
hydrocolloids that are usually biopolymers, mainly proteins and polysaccharides (e.g.,
gelatine, pectin, carrageenan, starch) which are easily dissolved or dispersed in water,
increase the viscosity of the solutions and when heated up and further cooled down
produce an intermolecular network, the gelling effect. Besides the macromolecules
crucial contribution to the structure development, water, lipids and flavour components
entrapped within the network all contribute to the overall sensory properties of foods with
a gel base.
Rheological studies can provide useful information on sol-gel and gel-sol transitions,
as well as on the kinetics of gel formation and matured gel properties. Generally, the
most used tests are stress, frequency, temperature and time sweeps under dynamic small-
amplitude oscillatory shear tests (SAOS), which should be conducted in the linear
viscoelastic range. A short description on these tests, and relevant derived parameters,
followed by case studies of novel gelled foods development will be presented in this
section.
3.1. Gel Point and Sol-Gel Transition
Gels can be formed through different mechanisms (e.g., aggregation, flocculation,

coagulation), involving several factors (e.g., temperature, pH, salinity, ions, pressure). In
all cases, there is a stage called sol-gel transition, which in physical terms corresponds to
the transition from the liquid to the solid state. This is not a thermodynamic transition but
a critical point where one characteristic length scale diverges (Lopes-da-Silva & Rao,
2007). There are many qualitative definitions of a gel point, which can be related to an
instant in time (gel time) or to a specific temperature (gel temperature). Several methods
have been used to estimate these parameters, as for example the time at which G' and G”
crossover (tan=1), at a given temperature. Some practical limitations may occur, when
for some biopolymer systems gelation begins from a sol-state characterized by a very low
viscosity, which is often below the resolution of the rheometer in SAOS experiments; and
the lowest detected value of G' may already be greater than G” (Gosal et al., 2004).
Extrapolation of the rapidly rising values of G' to the time or temperature axis, were
suggested as an accurate estimate of the gel point (Clark, 1991). The log discontinuity
method determines gel time as the time at which the modulus versus time log-log plot
shows a sudden slope increase (Gosal et al., 2004). Similarly, the gelation temperature is
considered as the temperature at which a frequency independent phase angle is obtained,
corresponding to a sharp rise in G’ and sharp fall of phase angle upon cooling (e.g.,
Verbeken et al., 2004).
3.2. Evaluation of Structure Development during Biopolymer Gelation
The gel state cannot be truly considered a state of equilibrium, because it evolves
with time. It represents a compromise between the polymer/polymer and polymer/solvent
interaction. Since most of the bonds responsible for the macromolecular gel structure are
weak interactions, such as hydrogen, electrostatic or hydrophobic bonds, they can suffer
variations within time (Nunes, 2006). The gradual development of the network structure
can be evaluated by dynamic rheological measurements, reflected by a progressive
increase in storage modulus (G'), with a tendency to reach a plateau value, the
equilibrium modulus. The beginning of the gelation process is dominated by the viscous
behaviour of the system (G">G') while the elastic behaviour dominates the final stages of
the experiment (G'>>G"). Both moduli increase as a result of the increasing density of
junction zones, but G' rises more sharply until it intercepts and then exceeds G". After an
initial rapid increase, resulting from the rapid formation of junction zones into the
polymer network, G' keeps increasing continuously as a result of the slower formation
and rearrangement of junction zones; values of G' eventually reach a pseudo plateau
region. In many biopolymer gelation processes, even after aging for long periods, G'
continues to increase steadily as a result of a continuous reorganization of the network in
a state close to equilibrium (Ross-Murphy, 1991; Rao & Cooley, 1993). In these cases,
exponential mathematical models can be applied to determine an equilibrium G’ value
(Nunes et al., 2003). Some examples of gel maturation kinetics, obtained by SAOS time
sweeps, at constant temperature and frequency, will be presented in sections 3.5.1 and
3.5.2.
3.3. Strong Gels and Weak Gels
Based on the macroscopic behaviour of gel systems, a practical and useful distinction
can be made between those systems which are free-standing as a consequence of the
development of the three-dimensional network, called “strong” or “true gels,” and those
characterized by a tenuous gel-like network which is easily broken when submitted to a
sufficiently high stress, called “weak gels” (Doublier et al., 1992). Clark & Ross-Murphy
(1987) differentiated gel type based on mechanical spectroscopy measurements. For
strong gels, the molecular rearrangements within the network over the time scales
analysed are very limited, G’ is higher than G” throughout the frequency range and
almost independent of frequency (Figure 3.1). In “weak gels” there is a stronger
dependence on frequency for the dynamic moduli, suggesting the existence of relaxation
processes occurring even at short time scales, and a lower difference between moduli
values, indicating that a lower percentage of the stored energy is recovered (Lopes-da-
Silva & Rao, 2007).
Figure 3.1. Typical dynamic viscoelastic behaviour (mechanical spectra) presented by a “strong” (a)
and a “weak” (b) gel (□ G’, ■ G”) (Batista, 2012).
3.4. Evaluation of Structure Loss during Melting/Softening of

Biopolymer Gels
The thermoreversible gelation of biopolymer solutions can be considered as an

equilibrium phenomenon, where a gel can fully recover its initial characteristics after
melting (heating) and re-cooling. However, important thermal hysteresis effects, related
to the non-equilibrium nature of the gel phase have been identified (e.g., Haque &
Morris, 1993). Gelatin and polysaccharides such as alginates, carrageenans and agarose
form thermoreversible gels, while gels formed by thermal denaturation of globular
proteins are irreversible. Application of thermal cycles is one of the methods that has
been used to study the thermal behaviour of biopolymer gels. For thermoreversible
biopolymer gels, important hysteresis effects, related to the non-equilibrium nature of the
gel phase can be identified by different methods, including dynamic rheology SAOS
(Haque & Morris, 1993), as represented in the example in Figure 3.2.
Figure 3.2. Evolution of viscoelastic moduli of -carrageenan gels, subjected to heating (from 5 to
80ºC) and immediate cooling (from 80 to 5ºC), at 0.5ºC/min. G’ (filled symbol), G” (open symbol)
(Nunes, 2006).
3.5. Case Studies for Mixed Polymer Gels
Natural biopolymers, such as globular proteins and polysaccharides, have been

widely used in the formulation of structured gelled food products (Doublier et al., 2000).
Most food products have a complex composition, comprising at least three biopolymer
types whose interaction determines the products’ structural and mechanical properties
(Tolstoguzov, 2003). Proteins and polysaccharides are the macromolecules responsible,

in great part, for the structure, mechanical and chemical properties of many food systems,
through their gelling, thickening and surface-stabilizing functional properties (Schmitt et
al., 1998). The behaviour in systems of two or more polymers is important because it is
different than that which would be exhibited in the absence of the other, and this
knowledge should help to formulate new foods as well as to develop innovative methods
to generate attractive food structures and textures (Lopes-da-Silva & Rao, 2007).
Therefore, considering that most food products are multicomponent systems, it is
important to consider not only the composition, structure and individual properties of
each component, but also the properties of the mixture (Tolstoguzov, 2003).
In the next subsections some case studies developed by our research group will be
presented to illustrate the development of novel mixed food gels, based upon rheology
tools. The research efforts and aims have been directed towards the development of
vegetarian and gluten-free products, as an alternative to dairy desserts, by exploring the
gelling behaviour of unconventional proteins and polysaccharides.
3.5.1. Pea Protein/-Carrageeenan/Starch Gels

In a case study develop vegan friendly desserts, pea protein isolate has been
successfully used in combination with -carrageenan and starch polysaccharides, to
develop strictly vegetable gelled desserts, as a vegan alternative to dairy-desserts (Nunes
et al., 2003; 2004; 2006a; 2006b).
The gelation process of these mixed systems was thermally induced, involving
extensive denaturation and/or conformation changes of the biopolymers and subsequently
the development of a gel network upon cooling. The thermal profile used on gel-setting
was determinant for the development of the gel structure which was reflected by its
rheological properties. Nunes et al. (2004) concluded that temperature favours protein
unfolding and biopolymers interaction, resulting in a significant increase on gel texture
and viscoelastic parameters (Figure 3.3). To produce well-structured gels with this
selection of polysaccharides, high temperatures, above 95ºC, were necessary. The
influence of cooling conditions on the gel-forming kinetics, final mechanical properties
and microstructure of this complex mixture was also evaluated (Nunes et al., 2006a). It
was shown that slower cooling rates (from 90ºC to 5ºC at -0.1ºC/min) allowed for more
complete phase separation with protein molecules structurally organised in larger
aggregates, resulting in a decrease of the storage modulus and texture parameters.
The texture and rheological parameters of these vegetable gels were compared to the
range of variation of each variable obtained for 12 commercial desserts, analysed as the
target standard (Figure 3.4) (Nunes et al., 2003; 2006b).
Figure 3.3. Maturation kinetic curves at 5ºC of pea/-carrageenan/starch gels subjected to different
prior heat treatments. A: 60ºC-15.5min (■); 66ºC-5.2min (▲). B: 80ºC-1min (▼); 80ºC-15.5min
(•);94ºC-5.2min (•); 100ºC-15.5min (■). G´ (filled symbol), G´´ (open symbol) (Nunes et al., 2004).
Figure 3.4. Firmness values for commercial desserts and pea protein/-carrageenan/starch gels
processes at different time and temperature of thermal treatment. The solid and dotted lines represent
the range of variation for commercial dairy and soya desserts, respectively (Nunes et al., 2006b).
Confocal laser scanning microscope (CLSM) images (Figures 3.5A,B) revealed that
interaction between pea protein and -carrageenan results in phase separation probably
due to a depletion-flocculation mechanism, which increases local polymer concentrations
(Nunes et al., 2006b). The two biopolymers formed two separate gelled networks, in
which -carrageenan acts as a continuous matrix dominating system’s rheological
behaviour. The water binding effect of the swollen starch granules, promoting the
concentration of pea protein and -carrageenan in their own phases, reinforced the gel
structure. The authors have proposed a model for pea protein-carrageenan/starch gel
systems, schematized in Figure 3.5C. SAOS rheological tests evidenced a synergetic
effect in the ternary system, considering that it was possible to obtain a well-structured
gel using much lower biopolymer concentrations than the ones required for the gelation
of each individual biopolymer.
a) b)
c)
Figure 3.5. CLSM images of pea protein (P)/-carrageenan (K)/starch (S) gels with different
composition: A) 2.0%P-0.15%K-2.5%S; B) 2.0%P-0.15%K-5%S (Nunes et al., 2006b). C) Schematic
representation of pea protein/-carrageenan/starch ternary gel system (Nunes, 2006).
3.5.2. Microalgae Gels

The above mentioned biopolymer gels, prepared with pea protein/-
carrageeenan/starch, served as model systems to study the addition of microalgae
biomass in gelled matrixes. As referred previously (section 2), microalgae can be used to
enhance the nutritional value of food due to their well-balanced chemical composition
and content in bioactive compounds (e.g., carotenoids, phycobilliproteins, omega-3 fatty
acids). Moreover, from a novel product development point of view, it is essential to
define the texture/sensory and rheology characteristics of these systems, since they reflect
the microstructural modifications that might arise as a consequence of microalgae
biomass addition to gel matrixes, which may distress the product stability.
Different microalgae species were added to the gel systems, which were
characterized in terms of maturation kinetic curves, monitored through SAOS time sweep
tests (Batista et al., 2008).
Figure 3.6. Maturation kinetic curves at 5ºC, of pea/-carrageenan/starch gels (■) with 0.75%
microalgal biomass addition: Chlorella vulgaris green ( ▲ ) and orange-carotenogenic ( ▼ ),
Haematococcus pluvialis (), Spirulina maxima (●), and Diacronema vlkianum (►). G’ (filled
symbol), G” (open symbol) (Batista et al., 2008).
The incorporation of microalgae on these mixed gel systems seemed to be beneficial,

especially for Haematococcus pluvialis (with an attractive “strawberry-pink” colour)
which promoted a structural reinforcement expressed by improved rheological properties
(Figure 3.6). This structural reinforcement action may be related to its high fat content
(41%), considering that fat droplets can act as active filler particles embedded in the gel
matrix as observed for milk gelled systems. However, the addition of Spirulina promoted
a drastic reduction on the gels rheology parameters. This should be related to a
thermodynamic incompatibility between the microalgae proteins and the other
components of the mixed gelled system, namely by hindering starch granules hydration
and gelatinization process (Batista et al., 2011).
Subsequently, the effect of different gel setting conditions (time/temperature
treatment, heating/cooling rates) on the linear viscoelastic properties of Spirulina and
Haematococcus gels was investigated (Batista et al., 2012). Increasing temperature (70-
90ºC, 5 min) resulted in more structured gels, while the effect of time (5-30 min, 90ºC)
was less pronounced. Higher values of the viscoelastic functions were achieved upon
heating and/or cooling the mixed biopolymer suspensions at lower rates (Figure 3.7).
However, the time length required for the process, and the subsequent costs involved,
should be considered.
The addition of microalgae biomass promoted some modifications on the gel
structure, although the response to gel setting conditions followed in general the
behaviour of the control gel. Haematococcus gels were less dependent on gel setting
conditions; while for Spirulina using lower heating/cooling rates and more extensive
thermal treatments enabled to overcome the structural drawback evidenced previously.
Figure 3.7. Evolution of G’ and G’’, of pea/-carrageenan/starch suspensions (■) with 0.75%
Haematococcus (●) and Spirulina (▲), during thermal treatment (90ºC/5min), performed at the same
heating and cooling rates (± 0.5, 1.0 and 5.0ºC/min). G’ (closed symbol), G” (open symbol), T (line)
(Batista et al., 2012).
The interaction of these microalgae with pea protein, -carrageenan and starch
biopolymers, in model binary (biopolymer/microalga) and ternary (pea protein/-
carrageenan/microalga; protein/starch microalga) gel systems was also studied in terms of
linear viscoelastic behaviour and fluorescence microscopy (Batista et al., 2011). The aim
of these studies was to clarify how these microalgae are arranged during the formation of
gel structure and how they interact with each biopolymer present in the complex mixed
gel system. For all the microalga/biopolymer systems studied, it was observed that
protein and polysaccharide biopolymers, alone or in binary combinations, are responsible
for the formation of the gel structure and resulting rheological behaviour. In most cases,
microalgae seem to be embedded in the gel’s network, causing denser microstructures
with improved rheological parameters. However, the fact that microalgal biomass
contains significant amounts of ions means that this should also have a significant impact
on the biopolymers’ interaction associated with the gelling mechanism, particularly in the
case of -carrageenan gels. Fluorescence optical microscopy proved to be a simple and
effective technique to observe these materials’ microstructures, with microalgae being
easily detected due to their natural pigments’ autofluorescence (Figure 3.8). The
microscopy results correlated well and supported the rheology findings.
a) b)
c)
Figure 3.8. Fluorescence microscopy images of 6% pea protein/3.75% starch gel systems (a) with
1.125% Spirulina (b) and Haematococcus (c) microalgal biomass addition (DAPI filter) (Batista et al.,
2011).
3.5.3. Chestnut and Rice Gels

The demand for high-quality gluten-free foodstuffs has considerably increased in
recent years due to an increase in diagnosed celiac patients. There is a large number of
alternative flour types to wheat flour such as corn, cassava, rice, soybean, chickpea or
amaranth flours as well as its blends (e.g., Marco & Rosell, 2008; Moreira et al., 2012).
Nevertheless, many gluten-free products available on the market are based on
formulations with low technological and nutritional quality, particularly compared with
their wheat counterparts (Mariotti et al., 2009).
Regarding this, our research group has devoted attention to the use of chestnut flour,
due to its organoleptic and nutritional characteristics (Moreira et al., 2012), namely, high
content of essential amino acids (4.0–7.0%), low fats (2.0–4.0%), high vitamin E, vitamin
B group, potassium, phosphorous and magnesium content, appreciable levels of dietary
fibre (5–15%), relatively high amount of sugar, mainly sucrose (20-32%) and starch (50-
60%) (Vasconcelos et al., 2010). However, due to limited technological functionality it
was decided to investigate their gelling behaviour in combination with rice flour, the
most commonly used gluten-free flour, in order to develop alternative desserts.
Texture studies (Figure 3.9) suggested that gels from mixtures of chestnut flour at
30% and rice flours (Rw- commercial rice meal, Rc - Carolino white rice meal and Ra-
Agulha white rice meal) at 20% showed an appropriate texture to develop gel-based new
desserts (Torres et al., 2014a).
These gels can be an interesting alternative to the gluten-free products market and
can provide additional economic value to wastes or byproducts coming from chestnut and
rice processing industry, adding sustainability to the rice milling industry and a
contribution to the circular economy trend. Different varieties of rice flours were tested,
with Carolino flour (Rc) providing much higher gel stability, as determined by
LUMISizer® (Torres et al., 2014a).
The effect of other ingredients relevant for consumer acceptance on the texture,
rheology, stability and sensory characteristics of the selected gels was also investigated in
detail (Torres et al., 2013; 2014b). The effect of different sweeteners was tested, being
observed that the presence of sucrose and xylitol strongly affected gluten-free gels
structure. Although all samples showed stable and strong gel properties, rheology
measurements indicated that the addition of sucrose involved a significant weakening of
the gels, whereas equivalent additions of stevia and xylitol resulted in a smaller softening
of these systems (Torres et al., 2013). Confocal microscopic images showed that the
sugar addition leads to a less aggregated structure with fracture lines well marked (Figure
3.10).
Figure 3.9. Firmness values of gels prepared using chestnut flour (Cf) and rice flours (●-Rw, ▲-Rc, ■-
Ra) at several ratios, for a total flour content of 30% (closed symbol) or 20% (open symbol) (Torres et
al., 2014a).
Figure 3.10. CLSM image of chestnut flour and whole rice flour gel in the presence of sucrose (Torres
et al., 2013).
The addition of salts, which can act as flavour enhancer and preserving agent, was
also studied in terms of the influence of cations concentration and ion valence, namely
Na+, K+ and Ca2+ on the gels rheological behaviour (Torres et al., 2014b). The presence
of chloride salts improved the mechanical properties of gluten-free gels and thereby the
samples structure was reinforced. The strength of the gels was enhanced in the presence
of CaCl2 followed by monovalent salts in the subsequent order, KCl and NaCl (Figure
3.11). The size of the ion impacted on the stabilization process, since ions with larger
radius as K+ have a more positive role in strengthening the gel structure.
Figure 3.11. Frequency sweeps of gels prepared with mixtures (50/50) of chestnut flour and whole rice
flour (●) in the presence of KCl () and CaCl2 (■). G’ (closed symbol), G” (open symbol) (Torres et
al., 2014b).
From the different case studies presented in this section it is clear that the food
product development of gelled-based foods, such as vegetarian and gluten-free desserts,
should be carefully designed. These systems are highly sensitive to processing conditions
(e.g., heating/cooling temperature profiles and rates) as well as to formulation issues,
from major structural biopolymers (and their interaction) to minor ingredients (e.g., salts
and sugars). The use of rheological tools, particularly SAOS, in combination with
microscope techniques, is fundamental to monitor and understand the structural changes
arising from the variation of processing and formulation parameters. A thorough rheology
study in the early stages of product development enables a deeper understanding of the
system under investigation, which can be later translated to adjust industrial processing
and formulation options, in order to attain final products with enhanced texture and
sensory properties, as well as physical stability along storage (e.g., resistance to syneresis
phenomena and thermoreversible behaviour).
4. CHEESE
Cheeses are milk-based food products that can be produced in a wide range of
flavours, textures and forms throughout the world, with more than 1000 varieties
identified (Fox & McSweeney, 2004). Cheese formation involves complex chemical,
enzymatic, biochemical and fermentative processes, from curd formation to ripening
aging. In all cases, cheese is produced by the casein protein coagulation due to
acidification (isoelectric point precipitation) or enzymatic rennet (chymosin) action. As a
result, a protein gel, in which fat is entrapped, is formed while the aqueous phase (whey)
is separated (Fox & McSweeney, 2004).
The structural properties of cheese can be described by its viscoelastic behaviour and
are mostly dictated by the properties of the continuous network formed by the protein
crosslinks (Zalazar et al., 2002). Rheology measurements using small amplitude
oscillatory stress (SAOS) allow the quantification of the elastic and viscous contributions
to the cheese properties (Alvarenga et al., 2011). Elastic response in cheese is primarily
due to the protein-protein bonds. In contrast, the viscous dissipation in cheese may be due
to the lubrication effect of fat on this protein matrix (Park, 2007). Cheese texture is
affected by factors that determine structure, such as milk composition, moisture content,
salt, pH and degree of proteolysis during ripening (Cunha et al., 2006). The firmness
(texture) and storage modulus (G′1Hz) have been identified by Alvarenga et al. (2008) as
the most adequate indicators of the cheese softening process during ripening.
Figure 4.1. Mechanical spectrum of non-frozen samples, NF (■) and samples frozen after 42 d of
ripening with fast freezing method and stored at -20°C, 20F42 (■) (Alvarenga et al., 2011).
The study of rheological and texture properties in different stages of cheese

formation and maturation can provide useful information on the impact of formulation
and/or processing changing variables.
Some studies from our group (Alvarenga et al., 2008; 2011; 2013) have focused on
evaluating the effect of freezing on the chemical and physical properties of raw ewe’s
milk semi-soft cheeses, namely Serpa cheese, at different stages of its maturation process
(e.g., ripening). It was observed that the effects on the cheeses chemical, physical and
rheological properties were less pronounced if the freezing was introduced at the end of
ripening (42 d), using a frozen storage temperature of -20°C (Figure 4.1). Moreover,
using these freezing conditions, no significant differences were found, for the chemical
and physical properties, between the samples subjected to different freezing methods,
slow or fast (Alvarenga et al., 2011).
4.1. Cheeses with Chlorella
In line with the previously presented food products, some studies on the rheology
characterization of cheese products with microalgae incorporation were performed. As
referred previously, considerable nutritional advantages can arise from the use of
microalgae biomass, besides colouring purposes, but some techno-functional and
structural modifications are also expected, that should be monitored through rheology
measurements.
This work was developed in collaboration with a large Portuguese cheese company
(Queijos Santiago), in lab to pilot scale.
In a first study (Batista et al., 2017a), Chlorella was added to cow’s milk to produce
fresh cheeses. The addition of the microalga resulted in an 11% increase in the milk-to-
cheese yield, rising from 330 g/L (control) to 368 g/L (5 g/L Chlorella). This should be
related to the microalgae high water absorption capacity which should also impart texture
and rheology modifications.
In terms of texture (Figure 4.2), both TPA and cutting tests revealed an increase in
firmness (and cohesiveness for TPA) with microalgae concentration up to 3 g/L.
Cheeses with higher Chlorella incorporation (4 and 5 g/L) showed lower texture
parameter values than the control. This could probably be related to the higher water
retention capacity of the algae, “softening” the cheese matrix texture.
The linear viscoelastic properties of the fresh cheeses were analysed by frequency
sweeps, and the resulting mechanical spectrum revealed well-structured gels (tan  ~ 0.2
for all frequency range). As observed for texture parameters, G’1Hz values increased with
alga incorporation up to 3 g/L, and further decreased for 4 and 5 g/L of Chlorella
addition (Figure 4.2).
Figure 4.2. Chlorella fresh cheeses firmness – cutting test (a) and mechanical spectra (b).
Afterwards, Chlorella was added on the initial stage of cured cheeses processing.
Cheese maturation was monitored in terms of colour, aw, texture, viscoelastic properties,
as well as chemical composition. As expected an increase in cutting force resistance
(firmness) was observed along maturation time (Figure 4.3). Cheeses with Chlorella
presented a final rind with firmness similar to the control while the texture of the interior
was softer and creamier, probably related to a higher water/whey retention.
Figure 4.3. Evolution of the firmness of the exterior rind (A) and of the interior (B) along Chlorella
cheeses maturation.
5. COOKIES
Dough-based products, such as cookies, are very complex and heterogeneous

systems, which are submitted to intense mechanical (e.g., kneading, extrusion) and
thermal processes (e.g., baking, drying). In a simplified manner, it can be said that these
products consist of suspensions of starch particles in concentrated biopolymer solutions
(e.g., gluten proteins), in which molecular to micrometer scale interactions determine the
global rheology properties, the baking process and the quality of the final product. The
main factor affecting their structure and texture is the moisture content and water
mobility, which is highly affected by the interaction with hydroxyl groups present in the
matrix (Hoseney et al., 1988) but the content of fat is also important.
The rheology of cookies, a solid material often within the glassy state, is
fundamentally based in texture measurements, in order to assess the fracture profile of
these crisp food products. There are mainly two types of texture analysis tests for this
type of product, to measure the hardness and resistance of cookies: i) simple penetration
with a small diameter cylindrical probe; ii) three point bend or “snap” test (Figure 5.1).
The results from these tests are often difficult to interpret with a high data variability,
associated with crispness and due to the heterogeneity and fracturability of this type of
samples. Therefore, a larger number of replications is often necessary. An acoustic
envelope detector can be used, coupled to the texturometer, to measure the acoustic
energy release during a physical test. This can allow the assessment of crispness through
the analysis of the sound spectrum (e.g., sound pulses, frequencies, maximum sound
peak, number of sound peaks, mean sound value) (Chen et al., 2005; Piteira, 2006).
Other physical parameters are also important regarding the texture quality of this type
of product. Water activity, aw, is an important physical parameter with impact on both the
texture and the shelf life of the cookies (Carter et al., 2015), since it conditions the
biopolymers hydration, as well as the temperature (T) and pressure (P) occurring during
glass transition. In fact, food polymers have an amorphous structure in which water acts
as a plasticizer (Slade & Levine, 1993), so the combined effects aw, T and P can cause the
biopolymers to change from a metastable glassy state to the rubbery state, characterized
by a general increase of the free volume, disorder and mobility of the polymer chains
(Ferry, 1980). This glass transition, strongly dependent on the chemical structure of the
biopolymers, brings about a sudden change in their thermal and mechanical properties.
Traditionally, it is identified by the discontinuous variation of the calorific capacity (Cp)
of amorphous materials, on a temperature sweep, which can be measured by differential
scanning calorimetry (DSC) (Rouilly et al., 2001).
The rheology characterization of the cookies dough can also give important
information in the research phases of this type of product development, providing insight
on the microstructural implications of formulation and/or processing variables on the
dough microstructure. In the processing of wheat-based food products, direct
relationships exist between the rheology properties of dough, dough structure and quality
of the finished product. Dough is the intermediate product between flour and cookies and
its rheological behaviour is of considerable importance in cookies manufacture as it
influences the machinability within processing and the quality of cookies (Manohar &
Rao, 2002; Piteira et al., 2006). Extensional flow behaviour (Piteira et al., 2006) and/or
linear viscoelastic oscillatory shear (Raymundo et al., 2014) tests can be used to analyse
this type of products, as will be discussed in the following subsections 5.1 and 5.2.
Figure 5.1. Acoustic envelope detector coupled to TA-XTplus texturometer

(www.stablemicrosystems.com).
Cookies are considered a convenient nutritiously dense snack food, widely consumed
by European citizens from all age groups. There is a tendency for research and innovation
in this market segment, which promotes the inclusion in cookies of healthy ingredients,
such as antioxidants, vitamins, minerals, proteins and fibres (e.g., Tumbas-Saponjak,
2016; Raymundo et al., 2014; Fradinho et al., 2015). In the present section, some
examples of case studies will be presented regarding rheology and texture implications of
adding different sources of fibres (e.g., psyllium), antioxidants and fatty acids
(microalgae biomass).
5.1. Cookies Enriched with Fibres
The health benefits of fibre consumption are well recognized. A diet low in dietary
fibre is associated with a spectrum of degenerative diseases, including diabetes, obesity,
coronary heart disease, bowl cancer and gallstones; it is a well-established fact that the
consumption of adequate amounts of dietary fibre reduces significantly the risk of these
diseases (Lairon et al., 2005). Fibre is known to be determinant of food structure, namely
the insoluble fibre has a strong impact on food texture (Dhingra et al., 2012). This was
the motivation to the development of novel cookies enriched with dietary fibre from
different origins, such as cereals (oat), fruits (orange) and legumes (pea) (Piteira, 2006;
Piteira et al., 2004; 2006). Cookies formulations and processing have been optimized
according to Response Surface Methodology (RSM). The contents of fibre and fat along
with cooking time were the independent variables, while texture, colour, moisture and aw
were the responses, completed with phase transition temperatures information. This study
demonstrated the potential use of dietary fibre as a functional ingredient at levels up to 6-

8% (Piteira et al., 2004).
The results from the rheology characterization of the cookies dough showed that
extensional uniaxial viscosity (Figure 5.2) is a more sensitive method to study the dough
rheology than the oscillatory tests, clearly showing differences when different types of
fibre are incorporated in cookies dough (Piteira et al., 2006). Furthermore, these results
were in agreement with the machinability studies, which showed that it was very difficult
to process these doughs at levels of fibre addition equal or above 8%for oat and orange
and 9–10% for pea, since higher viscosity levels lead to an increased difficulty in the
formation of the cellular structure in the final biscuits.
Figure 5.2. Flow (a) and rupture (b) of cookie dough after homogenous uniaxial extensional
deformation. Transient extensional viscosity curves of cookie doughs with orange fibers (c) (Piteira et
al., 2006).
More recently, our attention has been directed to the use of Psyllium extracted from
the husk of Plantago seeds as source of fibres (Raymundo et al., 2014; Fradinho et al.,
2015). Psyllium contains about 80% soluble fibre and is an excellent dietary source of
both soluble and insoluble fractions (Bijkerk et al., 2004).The effect of this fibre and
wheat flour content, on the linear viscoelastic behaviour (Figure 5.3) of the dough, was
studied through SAOS tests, where a significant increase in G’ values, at different
frequencies, with Psyllium concentration was observed (Raymundo et al., 2014).
Furthermore, the cookies formulation was optimized through RSM methodology, by
studying the effect of wheat flour (48-52%) and Psyllium (3-9%) contents on the
dimensional features (spread ratio, expansion factor), aw, colour (L*,a*,b*), texture
(firmness, adhesiveness, cohesiveness, springiness) and sensory properties (colour,
flavour, texture, overall acceptability) of dough and biscuits (Figure 5.4). An optimum
formulation was obtained by incorporating 6% Psyllium and 50% flour (Fradinho et al.,
2015).
Figure 5.3. Frequency sweeps of different Psyllium cookie dough formulations (A); Effect of Psyllium
concentration on cookie dough elastic modulus G’ (B) (P-Psyllium, F-Flour) (Raymundo et al., 2014).
Figure 5.4. Response surface scores of the effects of flour and Psyllium in different physical parameters
of the cookies (Fradinho et al., 2015).
5.2. Cookies Enriched with Microalgae Biomass
Some studies on the development of cookies with Chlorella vulgaris (Gouveia et al.,
2007) and Isochrysis galbana (Gouveia et al., 2008b) microalgae biomass were published
and more recently also with Arthrospira (Spirulina) platensis, Tetraselmis suecica and
Phaeodactylum tricornutum up to 6% (w/w) concentration (Batista et al., 2017a; 2017b).
For Chlorella vulgaris biscuits (Gouveia et al., 2007), an increase of added
microalgae biomass from 0.5 to 3.0% (w/w) reflected a significant increase on the
biscuits’ firmness, evidencing the positive effect of the alga in the biscuit structure,
reinforcing the short dough system (Figure 5.5). Similar results were observed for
Isochrysis biscuits (Gouveia et al., 2008b) which presented high texture stability with
time. The replacement of a small amount of flour by microalgae biomass resulted in the
inclusion of a complex biomaterial, rich in different proteins and polysaccharides. These
molecules have an important role on the water absorption process, which promotes the
increase of biscuit firmness, resulting in more compact structures (Batista et al., 2017a).
Figure 5.5. Firmness values of cookies with different concentrations of Chlorella vulgaris biomass
(Gouveia et al., 2007).
In a recent study (Batista et al., 2017b; 2017c), 6% Spirulina cookies presented

significantly higher firmness compared to the control (117% increase). Therefore it was
decided to investigate the rheology behaviour of these cookies dough. From the rheology
tests a linear increase in the wheat cookie dough’s viscoelastic moduli with Spirulina
concentration increase (R2=0.97) was observed. Moreover, at 6% Spirulina incorporation
it was possible to lower the flour content from 49 to 37% (w/w), while maintaining
viscoelastic properties similar to the control wheat cookie dough (Figure 5.6).
On the other hand, the addition of Spirulina to rice flour cookies clearly contributed
to a more elastic dough, with positive impact in the cookies sensory properties. A linear
increase in viscoelastic moduli with Spirulina concentration (R2=0.99) was also observed,
with a much pronounced structuring effect even at 2% Spirulina addition (63% increase
in G’1Hz, in relation to the control) (Batista et al., 2017b). The use of Spirulina biomass as
functional ingredient in bakery gluten-free products is thus promising, besides its great
functional level, with high impact on health, it adds a positive effect on the production
yield.
Figure 5.6. Mechanical spectra of wheat flour (F) and Spirulina (S) cookie doughs, with different
formulations. ■ 0%Sp (control, 49% WF), ♦ 2% Sp (47% WF), ▲ 4% Sp (45% WF), ● 6% Sp (43%
WF), ♦ 6% Sp (40% WF), ▲ 6% Sp (37% WF). G’, closed symbol, G’’ open symbol (Batista et al,
2017b).
Regarding the presented case studies, it can be concluded that, although cookies are
products with a relatively simple and standard processing conditions, the addition of
novel ingredients, such as fibres and antioxidants, can have a significant impact in the
final product texture and sensory properties. Therefore, along the novel food product
development process of these type of products, the rheology changes of the dough and
the consequent texture changes in the baked cookies, arising from formulation
adjustments should be evaluated. The addition of ingredients with higher water
absorption capacity, such as fibres and microalgae biomass, can be particularly
challenging regarding the interference with aw and flour hydration, but could be a useful
tool in the development of gluten-free cookies.
6. PASTA PRODUCTS
Pasta is a well-known staple food with origin in the traditional Italian cuisine, made
from an unleavened dough of durum wheat flour mixed with water and salt. Other
ingredients such as eggs, vegetables or algae can be also incorporated and the dough is
formed into sheets or various shapes, with a specific extrusion equipment. After this
mechanical process, pasta can then be cooked by boiling. If an additional drying
operation is added to the process, dried pasta can be obtained, with a longer shelf-life due
to the lower water activity.
The increasing demand for gluten-free products, related with a growing number of
people suffering from celiac disease, has led to the development of several alternatives to
the traditional pasta, produced from durum wheat flour. In gluten-free pasta, rice flour
and corn flour are predominantly used, and the gluten mimetic has been achieved using
various strategies, such as protein incorporation (Susanna & Prabhasankar, 2013), use of
hydrocolloids (Chauhan et al., 2017) and pre-gelatinization of starch (Li et al, 2014).
The rheology characterisation of pasta products can be focused on the pasta dough,
previously to the moulding phase, on the fresh pasta, before and after cooking, or even on
the rupture resistance of the dry pasta. Texture properties are the most widely used
characteristics to compare pasta products (Smewing, 1997), being strongly related to the
sensory attributes. Nevertheless, small amplitude oscillatory measurements can also be
used to evaluate the structure developed during the pasta formation.
Besides the pasta rheology characterization, the development of this type of products
also involves the evaluation of the quality technological attributes. They are a set of
empirical parameters, which allow to obtain important information about the cooking
process and are directly related to its consumption. The main properties are: i) Optimum
cooking time (min), considered as the time necessary to obtain complete starch
gelatinisation; ii) Swelling index; iii) Water absorption; iv) Cooking loss, described by
Tudorica et al. (2002).
6.1. Fresh Pasta Enriched with Microalgae Biomass
The use of pasta as an energy source is well known and is related to its complex
carbohydrate content, being also regarded as a product with low glycaemic index. These
nutritional advantages, coupled with the gastronomic versatility and an easy preparation,
as well as the appeal of pasta amongst consumers, have made this food product a
potential vehicle for nutraceuticals such as vitamins or polyunsaturated fatty acids
(Verardo et al., 2009).
Texture parameters, especially firmness, are important attributes for pasta cooking
quality. The consumer acceptance of pasta products is strongly dependent on their
texture, which results from its composition and cooking process, mainly the cooking
time. In the role of product development, the optimisation of texture parameters is a
critical issue and different analytical methods can be applied.
In 1989, the American Association of Cereal Chemists approved a standard method
(AACC 16-50, later updated to AACC 66-50) for measuring pasta firmness, which is
now widely used all over the world in research and quality control procedures. This
method is based on the use of a knife blade and assessment of the force required to cut
through pasta strands (Figure 6.1). From the texturogram (force versus time) obtained,
firmness is considered as the height of the peak.
Figure 6.1. Light Knife Blade measuring texture firmness of pasta prepared with different formulations,
based on the AACC 66-50, using a TA-XTplus Texture Analyser (Stable Micro Systems, UK) with a 5
kg load – A and respective texturogram (B).
Microalgae biomass are an important source of bioactive compounds with a huge

potential to be used as a food ingredient in pasta products, producing appealing, coloured
and bioactive solutions. Kadam & Prabhasankar (2010) revised the main aspects
associated with the incorporation of microalgae biomass in bread and pasta products.
Our research group evaluated the incorporation of Chlorella vulgaris (Cv) and
Spirulina maxima (Sm) biomass in pasta products (Fradique et al., 2010). These two
microalgae are the most popular in human food and their use is allowed by European
Food Safety Authority, positioning developed products in a higher Technology Readiness
Levels (TRL), close to the market. In relation to Chlorella vulgaris, two forms were used:
the natural produced microalge with a green colour (Cv green) and a sample that suffered
carotenogenesis, by nitrogen starvation, NaCl addition at high luminosity (Gouveia et al.,
1996), which presents an interesting orange colouration (Cv orange). The chemical
proximate composition of freeze-dried microalgae was considerably different for the
three samples studied (Batista et al., 2013) and this was a key factor for the different
behaviour as a food ingredient: Cv (green): 38.0% protein, 5.1% fat, 24.2% ash; Cv
(orange): 12.3% protein, 27.6% fat, 34.8% ash and Sm: 44.9% protein, 3.6% fat, 30.9%
ash. The incorporation of microalgae results in an increase of quality parameters, when
compared to the control sample without microalgae, followed by an increase in the raw
pasta firmness. For the three samples of microalgae studied, an increase in the biomass
concentration (0.5–2.0%) resulted in a general tendency to increase pasta firmness
(Figure 6.2), determined according to AACC method 66-50.01, using a TA-XTplus
Texture Analyser (Stable Micro Systems, UK).
Figure 6.2. Firmness raw (A) and cooked (B) pastas, enriched with microalgae biomass: Spirulina
maxima (Sm) and Chlorella vulgaris (Cv) green and orange. Control is the sample with no microalgae
addition (Fradique et al., 2010).
The final microalgae-enriched pastas, resulting from the optimisation of their texture,
exhibited attractive colours – orange and green products (Figure 6.3), due to the major
pigments existing in the algae: chlorophylls, phycocyanin and canthaxanthin. These
pastas have a high potential to be developed in an industrial scale, contributing to
increase the supply of healthy pastas.
A A B B C C
Figure 6.3. Pastas containing Chlorella vulgaris green (A) and orange (B) and Spirulina maxima (C)
biomass, after cooking (Fradique et al., 2010).
Although the number of microalgae species approved by EFSA is still reduced, the
study of species that are not yet approved but which have a high nutritional value and are
non-toxic, makes sense to prepare future applications, as it was presented for gels
enriched with different microalgae. Fradique et al. (2013) studied the possibility of enrich
fresh pasta with Isochrysis galbana and Diacronema vlkianum biomass as PUFA’s
source, following a similar methodology to Fradique et al. (2010) and obtaining similar
results. In addition, it was observed that fatty acid profile of pastas prepared with
Isochrysis galbana and Diacronema vlkianum biomass incorporation, presented a high
resistance to the thermal treatment applied during the cooking procedure. The increase of
the amount of the algae lead to a significant increase of EPA (Eicosapentaenoic Acid)
and DHA (Docosahexaenoic Acid), both in raw and cooked pastas.
6.2. Gluten-Free Pasta
Alternative methods have been used to access the texture properties of pasta
products, in order to minimise the experimental variations of the results. Stable Micro
Systems developed a specific probe for pasta – Spaghetti/Noodle Loop Tensile Rig
(Figure 6.4) to perform tension tests on loops of pasta or noodle sheet (Smewing, 1997).
Using this test it is possible to assess elasticity and breaking strength, giving an indication
the product behaviour during cooking.
The Spaghetti Loop Tensile Rig probe was used by our research group in the
development of gluten-free pasta using by-products from the rice industry. This industry
generates large amounts of by-products (husk- 20%, bran – 10% and broken rice-10%),
which can be used as feedstock for the gluten-free industries, originating added-value
products.
Gluten-free fresh pasta obtained with corn flour, rice flour (milling broken rice), egg-
white and bran was developed (Figure 6.5). In this process, the rice flour was submitted
to a pre-gelatinisation step (90ºC, 30 min), in order to increase the extensibility, a key
texture property for pasta.
Figure 6.4. Spaghetti/Noodle Loop Tensile Rig measuring extensibility of pasta, using a TA-XTplus
Texture Analyser (Stable Micro Systems, UK) with a 5 kg load (Smewing, 1997).
A A B B
Figure 6.5. Gluten-free pasta without rice bran (A) and with 2% (w/w) of this by-product (B), after
cooking during 3 min.
A significant (p<0.05) increase of pasta extensibility, resulting from the rice bran
addition was observed. On Figure 6.6 texturograms obtained with Spaghetti Loop Tensile
Rig are represented (crosshead speed of 1mm/s), for pasta formulations with rice, corn
and with bran addition. The extensibility, corresponding to the distance until the rupture
(mm), is higher for the samples with bran, which confirms the role of this by-product on
the reinforcement of the pasta structure, expressed as a higher resistance to an extensional
deformation. The impact of bran addition on the breaking strength, measured as the high
of the peak (N) is most noticeable on the corn pasta system.
Figure 6.6. Texturograms of cooked pastas prepared with rice and corn and with rice bran addition,
using a Spaghetti Loop Tensile Rig measuring in a TA-XTplus Texture Analyser (Stable Micro
Systems, UK) with a 5 kg load.
It is necessary to emphasize the importance of choosing the proper method for texture
evaluation, according to the sample characteristics and the type of development that is
intended to be made. The addition of rice bran implies a structural modification of the
dough, due to its high water absorption capacity and the high level of fibre content,
namely soluble fibre (Elleuch, 2011). The selected method for texture evaluation was
quite sensitive to optimise the final pasta composition, based on a physical characteristic
which is strongly related to the consumer acceptance of the pasta product.
One of the widely used processes to obtain gluten-free pastas involves a step of
starch pre-gelatinization. Optimization of this stage is fundamental to obtain a pasta with
the desired technological quality parameters and with a good sensorial acceptance. The
use of oscillatory tests to characterize gels formed during pre-gelatinization is an efficient
way of studying this phenomenon. On Figure 6.7, heating curves of rice flour solutions
with different concentrations are represented to compare the effect of temperature and
rice concentration on the efficiency of the pre-gelatinization process. Starch
gelatinization can be considered to occur when there is a marked increase in G’ values

upon temperature increase (Nunes et al., 2006a). It is clearly noticed that the higher the
concentration of the solution, the lower the temperature at which gelatinization occurs
and the higher the values of the elastic component.
8.00E+05
40% (w/w)
7.00E+05
50% (w/w)
6.00E+05 60% (w/w)
5.00E+05
G' (Pa)
4.00E+05
3.00E+05
2.00E+05
1.00E+05
0.00E+00
0 10 20 30 40 50 60 70 80 90
Temperature (ºC)
Figure 6.7. Heating curves of rice flour solutions with different concentrations.
The evaluation of the dough texture is a methodology which allows the development
of pasta products with properties that are in agreement with the consumer’s sensorial
patterns. By determining firmness and extensibility of the pasta product, it is possible to
evaluate differences in behaviour resulting from changes in process, cooking time or even
in the formulation. Nevertheless, the use of fundamental rheology tests also provided
important contributions to understand the phenomena associated with dough formation,
as for example the pre-gelatinization of the starch.
7. BREAD
A case study on cereal products is the development of bread with the incorporation of
brewers spent grain (BSG). Although BSG has no clearance yet to be used in food and
should only be used in animal feed formulation, its composition and the quantities that
are produced by the beer industry, justify the investment in developing a staple food
which can benefit from the addition of BSG.
The ratio of BSG produced per L of beer is around 20 kg of BSG for 100 L of beer.
In Portugal around 100,000 tons of BSG are produced each year corresponding to about
25,000 tons of dry matter of very good nutritional quality. The composition of BSG
(Muthusamy, 2014) is roughly 25% dry matter from which around 25% is partially
digested (by proteases) protein of good quality (mainly from barley); 20% are fibres,
some of them soluble; 12-15% is fat, 1.5% are minerals, 15% cellulose and 5% lignin and
vitamins include (ppm): biotin (0.1), choline (1800), folic acid (0.2), niacin
(44),pantothenic acid (8.5), riboflavin (1.5), thiamine (0.7) and pyridoxine (0.7). With 9
billion people forecasted for 2050, it would be a gross irrationality to ignore this material
as a food source.
One important drawback of this BSG is its high moisture which largely reduces its
shelf-life time. With an important quantity of reducing sugars from starch hydrolysis and
such nutritive richness it easily ferments with the development of intense negative odors
which turn it impossible to be used for food. The trick is to dry as soon as possible, but
with amounts of water of over 75% this is an expensive task. Of course engineering as
some alternatives like using continuous pressing to squeeze the most of the water out,
followed by tumble dryers against high speed hot air. Another alternative would be to
extrude (Stojceska et al., 2008) and further dry BSG for longer life time and safety. It
would be cheaper to transport and easy to store to be safe for use into food products.
As already mentioned, structure is the main limiting characteristic of these ingredient
incorporations, with impact on bread volume, crumb homogeneity and crust crispness -
Figure 7.1.
Figure 7.1. Breads with 0, 5 and 10% BSGf (brewer spent grain flour) addition.
Therefore, both fundamental rheology like dynamic small amplitude oscillatory

testing (SAOS) and empiric techniques like texture measurements as well as
extensography (extensograph Brabender) were applied to baking product development,
along with chemical quantifications. In Figure 7.2, the response surfaces for dough
maximum resistance (Rmax) measured in the Extensograph, therefore expressed in
empiric Brabender units (UB), as a function of BSG flour (BSGf) and SSL (sodium
stearyl laurate - emulsifier) addition to wheat dough and the Elongation in cm, resulting
from the same determinations, can be seen along with respective equations,
Response surface for dough resistance 45 min:
Rmax (BU) = 465.3 – 16.3 BSGf + 1.7 BSGf 2 + 255.0 SSL + 502.3 SSL2- 13.2 BSGf
SSL Adjusted R2 = 0.88
Response surface for dough extensibility 45 min
E (cm) = 17.7 – 0.5 BSGf + 0.002 BSGf 2 + 4.7 SSL -7.5 SSL2 – 0.0003 BSGf SSL
Adjusted R2 = 0.88
Figure 7.2. Response surfaces of maximum dough resistance (BU-Brabender Units) and Extensibility
(cm) for dough with different incorporations of BSGf (brewer spent grain flour) and SSL (emulsifier for
bread volume compensation).
Another interesting development, in a parallel study, was the addition of the

fermentation residue (FR), after bioethanol production from hominy feed, into baking
products (Pereira et al., 2014). Hominy feed is a by-product from the corn milling
industry with about 50% starch.
Additions of up to 7% of both BSG and FR resulted in increasing baking yield, since
water absorption increased almost linearly with BSG and FR addition, with the
production of softer breads that preserved freshness for longer, due to the high content in
fibre of these products. Furthermore the bread showed higher contents in protein,
antioxidants and vitamins, rich in Ca, K and Mg, as well as in 3 and 6 fatty acids,
being a differentiated product to be marketed in the foods for good health market. The
visual appearance was very good (Figure 7.3.) as well as aroma, taste and texture.
Figure 7.3. Dried Brewer Spent Grain (BSG) before milling; BSG flour and respective bread with 7%
of BSG flour addition.
Once again, rheology techniques were determinant for the success of product
development. This turn, the product developed was bread with an extra ingredient, to
introduce by-products into the stream production line, increasing food resources by
closing the production cycle as a contribution to sustainability in agreement with the
concept of circular economy.
ACKNOWLEDGMENTS
The authors acknowledge all partner institutions, researchers and students involved in
these projects, in particular, Cristiana Nunes, Mª Dolores Torres, Fátima Piteira, Patrícia
Fradinho, José Maria Franco, Luísa Gouveia, Nuno Bartolomeu Alvarenga, Susana
Ramos, Queijos Santiago, Novarroz, Mundi Globe Trading, Panidor and CampoFrio.
Financial support was granted by: Fundação para a Ciência e Tecnologia (FCT)
through the research unit UID/AGR/04129/2013 (LEAF), Projects “Vegetable proteins in
foods as an alternative to dairy desserts” (POCTI/AGR/38251/2001), “Pigments,
antioxidants and PUFA’s in microalgae based food products – functional implications”
(PTDC/AGR-ALI/65926/2006) and doctoral grant SFRH/BD/21388/2005; COMPETE
through projects: “Production of a Functional Ingredient from the ethanol fermentation
of hominy feed residue to be used on baking products from bread to pizza”-QREN 5581
ETANE FARE FUN 2009; “Technology to recover a by-product of the brewing industry,
BSG, for use in the baking industry” -QREN23141 DRECHE 2011; “Valorization of
bran and broken rice for the production of high value-added food products” -QREN
8749 ARROZ+ 2013; “Development of a snack based on cured meat derivative” -QREN
38593SI SNACK NOBRE 2015).
REFERENCES
Alvarenga, N., Silva, P., Garcia, J.R., Sousa, I. (2008). Estimation of Serpa cheese
ripening time using multiple linear regression (MLR) considering rheological,
physical and chemical data. Journal of Dairy Research, 75:233-239.
Alvarenga, N., Canada, J., Sousa, I. (2011). Effect of freezing on the rheological,
chemical and colour properties of Serpa cheese. Journal of Dairy Research, 78:80-
87.
Alvarenga, N.B., Ferro, S.P., Almodôvar, A.S., Canada, J., Sousa, I. (2013). Shelf life
extension of cheese: frozen storage. In Handbook of cheese in health: production,
nutrition and medical sciences, Eds. V.R. Preedy, R.R. Watson & V.B. Pate.
Wageningen Academic Publishers: Wageningen; pp. 87-102.
American Association of Cereal Chemists (1999). Semolina, Pasta, and Noodle AACC
International Method 66-50.01. P 1 – 3. Formerly 16-50 Pasta and Noodle Cooking
Quality—Firmness First approval November 1, 1989; Reapproval November 3, 1999.
Arozarena, I., Bertholo, H., Empis, J., Bunger, A., Sousa, I. (2001). Study of the total
replacement of egg by white lupin protein emulsifiers and xathan gum in yellow
cakes. European Food Research and Technology, 213:312-316.
Barnes, A., Hutton, J.F., Walters, K. (1989). An Introduction to Rheology. Rheology
Series 3. Elsevier: Amsterdam.
Batista, A.P.S. (2012). Natural Pigments and Microalgal Biomass in Colloidal Food
Systems: Rheological Characterization. European PhD Thesis. Universidad de
Huelva. Huelva, Spain. http://hdl.handle.net/10272/6472.
Batista, A.P., Gouveia, L., Nunes, M.C., Franco, J.M., Raymundo, A. (2008). Microalgae
biomass as a novel functional ingredient in mixed gel systems. In Gums and
Stabilisers for the Food Industry 14, Eds. P.A. Williams & G.O. Phillips. RSC
Publishing: Cambridge; pp. 487-494.
Batista, A.P., Nunes, M.C., Gouveia, L., Sousa, I., Raymundo, A., Cordobés, F.,
Guerrero, A., Franco, J.M. (2011). Microalgae biomass interaction in biopolymer
gelled systems. Food Hydrocolloids, 25:817-825.
Batista, A.P., Nunes, M.C., Fradinho, P., Gouveia, L., Sousa, I., Raymundo, A., Franco,
J.M. (2012). Novel foods with microalgal ingredients – Effect of gel setting
conditions on the linear viscoelasticity of Spirulina and Haematococcus gels. Journal
of Food Engineering, 110:182-189.
Batista, A.P., Gouveia, L., Bandarra, N.M., Franco, J.M., Raymundo, A. (2013).
Comparison of microalgal biomass profiles as novel functional ingredient for food
products. Algal Research, 2:164- 173.
Batista, A.P., Niccolai, A., Fradinho, P., Fragoso, S., Bursic, I., Rodolfi, L., Biondi, N.,
Tredici, M.R., Sousa, I., Raymundo, A. (2017a). Microalgae cookies: Sensory,
physical and chemical properties, antioxidant activity and in vitro digestibility. Algal
Research,26:161-171.
Batista, A.P., Bursic, I., Miranda, A., Fragoso, S., Fradinho, P., Raymundo, A., Sousa, I.
(2017b). Rheological characterization of Spirulina gluten-free cookie doughs. In The
multidisciplinary science of rheology: Towards a healthy and sustainable
development. Proceedings of IBEREO 2017 Meeting (6-8 Set). Valencia, Spain.
Batista, A.P., Ramos, S., Archer de Carvalho, B., Sousa, I., Raymundo, A. (2017a).
Developing innovative cheese products with Chlorella. In The multidisciplinary
science of rheology: Towards a healthy and sustainable development. Proceedings of
IBEREO 2017 Meeting (6-8 Set). Valencia, Spain.
Batista, I., Vidal, I., Lourenço, H.M., Sousa, I. (1999). Tenderisation of the dog cockle
(Glycymeris glycymeris) meat by polyphosphates. European Food Research and
Technology, 210:31-33.
Bijkerk, C.J., Muris, J.W.M., Knottnerus, J.A., Hoes, A.W., de Wit, N.J. (2004).
Systematic review: The role of different types of fiber in the treatment of irritable
bowel syndrome. Alimentary Pharmacology & Therapeutics, 19:245–251.
Blanshard, J.M.V., Mitchell J.R. (1988). Food Structure - Its creation and Evaluation.
Butterworth-Heinemann: Nothingham.
Bollinger, H. (2001). Food Marketing and Technology, 15:10-15.
Cabodevila, O., Hill, S., Armstrong, H., Sousa, I.M.N., Mitchell, J.R. (1994). Gelation
enhancement of soy protein isolate using the Maillard reaction and high
temperatures. Journal of Food Science, 59:872-875.
Carter, B.P., Galloway, M.T., Campbell, G.S., Carter, A.H. (2015). The critical water
activity from dynamic dew point isotherms as an indicator of crispness in low
moisture cookies. Journal of Food Measurement and Characterization, 9:463–470.
Chauhan, A., Saxena, D. C., Singh, S. (2017). Effect of hydrocolloids on microstructure,
texture and quality characteristics of gluten-free pasta. Journal of Food Measurement
and Characterization, 11:1188-1195.
Chen, J., Karlsson, C., Povey, M. (2005). Acoustic envelope detector for crispness
assessment of biscuits. Journal of Texture Studies, 36:139–156.
Clark, A.H., Ross-Murphy, S.B. (1987). Structural and mechanical properties of
biopolymer gels. Advances in Polymer Science, 83:57-192.
Clark, A.H. (1991). Structural and mechanical properties of biopolymer gels. In Food
Polymers, Gels and Colloids, Ed. E. Dickinson. Royal Society of Chemistry:
London; pp. 322–328.
Clark, D.C., Wilde, P.J., Wilson, D.R., Wustneck, R. (1992). The interaction of sucrose
esters with β-lactoglobulin and β-casein from bovine milk. Food Hydrocolloids, 6:
173-186.
Cunha, C.R., Viotto, W.H., Viotto, L.A. (2006). Use of low concentration factor
ultrafiltration retentates in reduced fat “Minas Frescal” cheese manufacture: Effect on
composition, proteolysis, viscoelastic properties and sensory acceptance.

International Dairy Journal, 16:215-224.
Dhingra, D., Michael, M., Rajput, H., Patil, R.T. (2012). Dietary fibre in foods: a review.
Journal of Food Science and Technology, 49:255–266.
Dias J., Alvarenga, N., Sousa, I. (2015). Effect of hydrocolloids on low-fat chocolate
fillings. Journal of Food Science and Technology, 52: 7209–7217.
Dickinson, E., Hong, S.T. (1995). Influence of water soluble non-ionic emulsifier on the
rheology of heat-set protein stabilized emulsions gels. Ibid, 43:25-60.
Dickinson, E. (1998). Proteins at interfaces and in emulsions: stability, rheology and
interactions. Journal of Chemical Society, Faraday Transactions, 94:1657-1669.
Dickinson, E. (2003). Hydrocolloids at interfaces and the influence in the properties of
dispersed systems. Food Hydrocolloids, 17:25-39.
Doublier, J.L., Launay, B., Cuvelier, G. (1992). Viscoelastic properties of food gels. In
Viscoelastic Properties of Foods, Eds. M.A. Rao & J.F. Steffe. Elsevier Applied
Science: Barking.
Doublier, J.L., Garnier, C., Renard, D., Sanchez, C. (2000). Protein-polysaccharide
interactions. Current Opinion in Colloid and Interface Science, 5:202-214.
Elleuch, M., Bedigian, D., Roiseux, O., Besbes, S., Blecker, C., Attia, H. (2011). Dietary
fibre and fibre-rich by-products of food processing: characterisation, technological
functionality and commercial applications: a review. Food Chemistry, 124:411-421.
Ferry, J.D. (1980). Viscoelastic Properties. John Wiley & Sons: New York; pp.280-291.
Fox, P.F., McSweeney, P.L.H. (2004). Cheese: An overview. In Cheese: Chemistry,
Physics and Microbiology. Eds. P.F. Fox, P.L.H. McSweeney, T.M. Cogan & T.P.
Guinee. Elsevier; vol. 1, pp. 1-18.
Fradinho, P., Nunes, M.C., Raymundo, A. (2015). Developing consumer acceptable
biscuits enriched with Psyllium fibre. Journal of Food Science and Technology,
52:4830–4840.
Fradique, M., Batista, A.P., Nunes, M.C., Gouveia, L., Bandarra, N., Raymundo, A.
(2010). Incorporation of Chlorella vulgaris and Spirulina maxima biomass in pasta
products. Part 1: Preparation and evaluation. Journal of the Science of Food and
Agriculture, 90:1656–1664.
Fradique, M., Batista, A.P., Nunes, M.C., Gouveia, L., Bandarra, N., Raymundo, A.
(2013). Isochrysis galbana and Diacronema vlkianum biomass incorporation in pasta
products as PUFA’s source. LWT - Food Science and Technology, 50:312 -319.
Franco, J.M., Guerrero, A., Gallegos, C. (1995). Rheology and processing of salad
dressing emulsions. Rheologica Acta, 34:513–524.
Franco, J. Raymundo, A.; Sousa, I., Gallegos, C. (1998). Influence of processing
variables on the rheological and textural properties of lupin protein-stabilized
emulsions. Journal of Agricultural and Food Chemistry, 46:3109-3115.
Gaspar, C., Sousa, I., Laureano, O. (1997). Estudio de la influencia de los iones y de la
fuerza iónica en el proceso de la gelificación. Informacion Tecnologica, 8:213- 217.
[Study of the influence of ions and ionic strength in the gelification process].
Gaspar, C., Sousa, I., Laureano, O. (1998). Production of reduced-calorie grape juice
jelly with gellan, xanthan and locust bean gums: sensory analysis and objective
texture. Z. Lebensm. Unters. Forsch., 206:169-174.
Gosal, W.S., Clark, A.H., Ross-Murphy, S.B. (2004). Fibrillar β-Iactoglobulin gels: part
2. Dynamic mechanical characterization of heat-set systems. Biomacromolecules,
5:2420-2429.
Gouveia, L., Veloso, V., Reis, A., Fernandes, H.L., Empis, J., Novais, J.M. (1996).
Evolution of pigment composition in Chlorella vulgaris. Bioresource Technology,
57:157–163.
Gouveia, L., Batista, A.P., Raymundo, A., Sousa, I., Empis, J. (2006). Chlorella vulgaris
and Haematococcus pluvialis biomass as colouring and antioxidant in food
emulsions. European Food Research and Technology, 222:362-367.
Gouveia, L., Batista, A.P., Miranda, A., Empis, J., Raymundo, A. (2007). Chlorella
vulgaris biomass used as colouring source in traditional butter cookies. Innovative
Food Science and Emerging Technologies, 8:433-436.
Gouveia, L.; Batista, A.P.; Sousa, I.; Raymundo, A. Bandarra, N. (2008a). Microalgae in
novel food products. In Food Chemistry Research Developments, Ed. K.N.
Papadopoulos. Nova Science Publishers Inc; pp. 75-111.
Gouveia, L., Coutinho, C., Mendonça, E., Batista, A.P., Sousa, I., Bandarra, N.M.,
Raymundo, A. (2008b). Functional biscuits with PUFA-ω3 from Isochrysis galbana.
Journal of the Science of Food and Agriculture, 88:891-896.
Graça, C., Raymundo, A., Sousa, I. (2016). Rheology changes in oil-in-water emulsions
stabilized by a complex system of animal and vegetable proteins induced by thermal
processing. LWT- Food Science and Technology, 74:263-270.
Guéguen, J. (2000). Pea proteins: new and promising protein ingredients. Industrial
Proteins, 8:6-8.
Haque, A., Morris, E.R. (1993). Thermogelation of methylcellulose. Part I: molecular
structures and processes. Carbohydrate Polymers, 22:161-173.
Hill, S.E., Ledward, D.A., Mitchell, J.R. (1998). Functional Properties of Food
Macromolecules. Aspen Publishers: Maryland.
Hoseney, R.C., Wade, P., Finley, J.W. (1988). Soft wheat products. In Wheat Chemistry
and Technology, Ed. Y. Pomeranz. American Association of Cereal Chemists
Monograph Series; pp. 407–456.
Kadam, S.U., Prabhasankar, P. (2010). Marine foods as functional ingredients in bakery
and pasta products. Food Research International, 43:1975–1980.
Klinkesorn, U., Sophanodora, P., Chinachoti, P., McClements, D.J. (2004). Stability and
rheology of corn oil-in-water emulsions containing maltodextrin. Food Research
International, 37:851-859.
Lairon, D., Arnault, N., Bertrais, S., Planells, R., Clero, E., Hercberg, S., Boutron-Ruault,
M.C. (2005). Dietary fiber intake and risk factors for cardiovascular disease in
French adults. American Journal of Clinical Nutrition, 82:1185-1194.
Li, M., Zhu, K., Guo, X., Brijs, K., Zhou, H. (2014). Natural Additives in Wheat-Based
Pasta and Noodle Products: Opportunities for Enhanced Nutritional and Functional
Properties. Comprehensive Reviews in Food Science and Food Safety, 13:347-357.
Line, V., Remondetto, G., Subirade, M. (2005). Cold gelation of -lactoglobulin oil-in-
water emulsions. Food Hydrocolloids, 19:269-278.
Lopes-da-Silva, J.A., Rao, M.A. (2007). Rheological behaviour of food gels. In Rheology
of Fluids and Semisolid Foods: Principles and Applications, Ed. M.A. Rao. Springer
Science.
Manohar, R.S., Rao, P.H. (2002). Interrelationship between rheological characteristic of
dough and quality of biscuits; use of elastic recovery of dough to predict biscuit
quality. Food Research International, 35:807-813.
Manoi, K., Rizvi, S. (2009). Emulsification mechanisms and characterizations of cold,
gel-like emulsions produced from texturized whey protein concentrate. Food
Hydrocolloids, 23:1837-1847.
Marco, C., Rosell, C.M. (2008). Breadmaking performance of protein enriched, gluten-
free breads. European Food Research and Technology, 227:1205-1213.
Mariotti, M., Lucisano, M., Pagani, M.A., Ng, P.K.W. (2009). The role of corn starch,
amaranth flour, pea isolate, and Psyllium flour on the rheological properties and the
ultrastructure of gluten-free doughs. Food Research International, 42:963-975.
McClements, D. J. (2015). Food emulsions: Principles, Practices, and Techniques (3rd
edition). CRC Press: Boca Raton, FL.
Moreira, R., Chenlo, F., Torres, M.D. (2012). Rheology of gluten-free doughs from
blends of chestnut and rice flours. Food and Bioprocess Technology, 6:1476-1485.
Muthusamy, N. (2014). Chemical composition of brewers spent grain – a review.
International Journal of Science, Environment and Technology, 3:2109-2112.
Nascimento, A.C., Pires, A.S., Gusmão, J.H., Sousa, I., Ferreira-Dias, S. (2006). The
effect of enzymatic transesterification on rheological properties of fat blends rich in
omega-3 polyunsaturated fatty acids. In Current Topics on Bioprocesses in Food
Industry, Eds. C. Larroche, A. Pandey & C.G. Dussap. Asiatech Publ. Inc: New
Delhi.
Nunes, M.C.H. (2006). Interacções Proteína/Polissacárido em Géis Proteicos de Origem
Vegetal. PhD Thesis. Instituto Superior de Agronomia, Universidade Técnica de
Lisboa. Lisboa, Portugal. [Protein/polysaccharide interactions in protein gels of
vegetable origin].
Nunes, M.C., Batista, P., Raymundo, A., Alves, M.M., Sousa, I. (2003). Vegetable
proteins and milk puddings. Colloids and Surfaces B: Biointerfaces, 31:21-29.
Nunes, M.C., Raymundo, A., Sousa, I. (2004). Effect of thermal treatment and
composition on the mechanical properties of pea/-carrageenan/starch desserts. In
Gums and Stabilisers for the Food Industry 12, Eds. P.A. Williams & G.O. Phillips.
Royal Society of Chemistry: Cambridge; pp. 54-64.
Nunes, M.C., Raymundo, A., Sousa, I. (2006a). Rheological behaviour and
microstructure of pea protein/-carrageenan/starch gels with different setting
conditions. Food Hydrocolloids, 20:106-113.
Nunes, M.C., Raymundo, A., Sousa, I. (2006b). Gelled vegetable desserts containing pea
protein, -carrageenan and starch. European Food Research and Technology,
222:622-628.
O’Kane, F.E., Vereijken, J.M., Gruppen, H., van Boekel, M.A.J.S. (2005). Gelation
behavior of protein isolates extracted from 5 cultivars of Pisum sativum. Journal of
Food Science, 70:C132-C137.
Park, Y.W. (2007). Rheological characteristics of goat and sheep milk. Small Ruminant
Research, 68:73-87.
Pereira, M., Malfeito-Ferreira, M., Macri, A., Belc, N., Sousa, I. (2014). Hominy feed
value for bioethanol and bakery functional ingredient. International Journal of Food
Processing Technology, 1:1-9.
Pires, C., Costa, S., Batista, A.P., Nunes, M.C., Raymundo, A., Batista, I. (2012).
Properties of protein powder prepared from Cape hake by-products. Journal of Food
Engineering. 108:268-275.
Piteira, M.F.C. (2006). Alimentos Funcionais – Incorporação de Fibras em Bolachas.
PhD Thesis. Instituto Superior de Agronomia, Universidade Técnica de Lisboa.
Lisboa, Portugal. [Functional Foods – Incorporation of fibers in cookies].
Piteira, M.F., Nunes, M.C., Raymundo, A., Sousa, I. (2004). Effect of principal
ingredients on quality of cookies with dietary fibre. In Gums and Stabilisers for the
Food Industry 12, Eds. P.A. Williams & G.O. Phillips. Royal Society of Chemistry:
Cambridge; pp. 475-483.
Piteira, M.F., Maia, J.M., Raymundo, A., Sousa, A. (2006). Extensional flow behaviour
of natural fibre-filled dough and its relationship with structure and properties.
Journal of Non-Newtonian Fluid Mechanics, 137:72-80.
Rao, M.A., Cooley, H.J. (1993). Dynamic rheological measurement of structure
development in high-methoxyl pectin/fructose gels. Journal of Food Science, 58:876-
879.
Raymundo, A., Franco, J., Gallegos, C., Empis, J.E., Sousa, I. (1998). Effect of thermal
denaturation of lupin protein on its emulsifying properties. Nahrung, 42:220-224.
Raymundo, A., Franco, J.M., Partal, P., Sousa, I., Gallegos, C. (1999). Effect of the
Lupin Protein/surfactant ratio on linear viscoelasticity properties of oil-in-water
emulsions. Journal of Surfactants and Detergents, 2:545-551.
Raymundo, A., Empis, J., Sousa, I., Gallegos, C., Franco, J.M. (2001). Influencia de la
relación proteína de altramuz/tensioactivo en las propiedades de flujo y texturales de
emulsiones aceite en agua. Grasas y Aceites, 52:3-4. [Effect of the lupin protein
surfactant ratio on the flow and textural properties of oil-in-water emulsions].
Raymundo, A., Franco, J.M., Empis, J., Sousa, I. (2002). Optimization of the
composition of low-fat oil-in-water emulsions stabilized by white lupin protein.
Journal of the American Oil Chemists Society, 79:783-790.
Raymundo, A., Gouveia, L., Batista, A.P., Empis, J.; Sousa, I. (2005). Fat mimetic
capacity of Chlorella vulgaris biomass in oil-in-water food emulsions stabilised by
pea protein. Food Research International, 38: 961-965.
Raymundo, A., Fradinho, P., Nunes, M.C. (2014). Effect of Psyllium fibre content on the
textural and rheological characteristics of biscuit and biscuit dough. Bioactive
Carbohydrates and Dietary Fibre, 3:96-105.
Riscardo, M.A., Franco, J.M., Gallegos, C. (2003). Influence of composition of
emulsifier blends on the rheological properties of salad dressing-type emulsions.
Food Science and Technology International, 9:53-63.
Rodríguez, A., Bunger, A., Castro, E., Sousa, I., Empis, J. (2003). Development and
optimization of soured goat’s cream butter. JAOCS, 80:987-992.
Ross-Murphy, S.B. (1991). Incipient behaviour of gelatin gels. Rheologica Acta, 30:401-
411.
Rouilly, A., Orliac, O., Silvestre, F. (2001). DSC study on the thermal properties of
sunflowers proteins according to their water content. Polymer, 42:10111-10117.
Schmitt, C., Sanchez, C., Desobry-Banon, S., Hardy, J. (1998). Structure and
technofunctional properties of protein-polysaccharide complexes: a review. Critical
Reviews in Food Science and Nutrition, 38:689-753.
Slade, L., Levine, H. (1993). The glassy state in applications for the food industry, with
an emphasis on cookie and cracker production. In The Glassy State in Foods, Eds.
J.M.V. Blanshard & P. Lillford. Nottingham University Press: Loughborough; pp.
333-371.
Sousa, I., Matias, E., Laureano, O. (1997). The texture of low calorie grape juice jelly. Z.
Lebenm. Unters. Forsch., 205:140-142.
Surh, J., Decker, E. A., McClements, D.J. (2006). Properties and stability of oil-in water
emulsions stabilized by fish gelatin. Food Hydrocolloids, 20:596-606.
Susanna, S., Prabhasankar P. (2013). A study on development of gluten free pasta and its
biochemical and immunological validation. LWT - Food Science and Technology,
50:613-621.
Smewing, J. (1997). Analyzing the texture of pasta for quality control. Cereal Foods
World, 42:1-4.
Stojceska, V., Ainsworth, P., Plunkett, A., Ibanoglu, S. (2008). The recycling of brewer’s
processing by-product into ready to eat snacks using extrusion technology. Journal of
Cereal Science, 47:469-479.
Tolstoguzov, V.B. (2003). Some thermodynamic considerations in food formulation.
Food Hydrocolloids, 17:1-23.
Tomé, A.S., Pires, C., Raymundo, A., Batista, I., Sousa, I. (2014). Protein gels and
emulsions from cape-hake by products and pea. Journal of the Science of Food and
Agriculture, 95: 289-298.
Torres, M.D., Raymundo, A., Sousa, I. (2013). Effect of sucrose, stevia and xylitol on
rheological properties of gels from blends of chestnut and rice flour. Carbohydrate
Polymers, 98:249-256.
Torres, M.D., Fradinho, P., Raymundo, A., Sousa, I. (2014a). Thermorheological and
textural behavior of gluten-free gels obtained from chestnut and rice flours. Food and
Bioprocess Technology, 7:1171-1182.
Torres, M.D., Raymundo, A., Sousa, I. (2014b). Influence of Na+, K+ and Ca2+ on
mechanical and structural properties of gels from chestnut and rice flours.
Carbohydrate Polymers, 102:30-37.
Trujillo-Cayado, L.A., Rodriguez, M.C.A., Muñoz, J., Raymundo, A., Sousa, I. (2016).
Development and rheological properties of ecological emulsions formulated with a
biosolvent and two microbial polysaccharides. Colloids and Surfaces B:
Biointerfaces. 141:53–58.
Tudorica, C.M., Kuri, V., Brennan, C.S. (2002). Nutritional and physicochemical
characteristics of dietary fiber enriched pasta. Journal of Agriculture and Food
Chemistry, 50:347–356.
Tumbas-Saponjak, V.T., Cetkovic, G., Canadanovic-Brunet, J., Pajin, B., Djilas, S.,
Petrovic, J., Loncarevic, I., Stajcic, S., Vulic, S. (2016). Sour cherry pomace extract
encapsulated in whey and soy proteins: incorporation in cookies. Food Chemistry,
207:27-33.
Van Vliet, T., Martin, A.H., Bos, M.A. (2002). Gelation and interfacial behaviour of
vegetable proteins. Current Opinion in Colloid and Interface Science, 7:462-468.
Vasconcelos, M.C.B.M., Nunes, F., García-Viguera, C., Benett, R.N., Rosa, E.A.S.,
Ferreira-Cardoso, J.V. (2010). Industrial processing effects on chestnut fruits
(Castanea sativa Mill.). 3. Minerals, free sugars, carotenoids and antioxidant
vitamins in chestnut fruits (Castanea sativa Mill.). International Journal of Food
Science and Technology, 45:496-505.
Verardo V., Ferioli F., Riciputi Y., Iafelice G., Marconi E. and Caboni MF. (2009).
Evaluation of lipid oxidation in spaghetti pasta enriched with long chain n-3
polyunsaturated fatty acids under different storage conditions. Food Chemistry,

114:472–477.
Verbeken, D., Thas, O., Dewettinck, K. (2004). Textural properties of gelled dairy
desserts containing kappa-carrageenan and starch. Food Hydrocolloids, 18:817-823.
Wu, S. (1989). Chain Structure and Entanglement. Journal of Polymer Science, 27:723-
741.
Xie, Y.R., Hettiarachchy, N.S. (1997). Xanthan gum effects on solubility and
emulsification properties of soy protein isolate. Journal of Food Science, 62:1101-
1104.
Zalazar, C.A., Zalazar, C.S., Bernal, S., Bertola, N., Bevilacqua, A., Zaritzky, N. (2002).
Effect of moisture level and fat replacer on physicochemical, rheological and sensory
properties of low fat soft cheeses. International Dairy Journal, 12:45-50.
Chapter 5
MISTIBILITY AND VISCOELASTIC PROPERTIES OF

POLY(STYRENE-CO-ACRYLONITRILE) BLENDS
Yuji Aoki*
Graduate School of Organic Materials Science, Yamagata University,
Yonezawa, Yamagata, Japan
ABSTRACT
This article gives a review of the miscibility and viscoelastic properties of

poly(styrene-co-acrylonitrile) (SAN) blends. SAN is miscible with poly(methyl
methacrylate) (PMMA), poly(-caprolactone) (PCL), poly(-methyl styrene)-co-
acrylonitrile (MSAN), poly[styrene-co-(maleic anhydride)] (SMA), poly[styrene-co-(N-
phenylmaleimide)] (SMI), and so on. The miscibility of SAN is mainly due to the
presence of unfavorable interaction between styrene (S) and acrylonitrile (AN)
comprising the copolymer. First, we explain the miscibility of SAN with PMMA, SMA,
and SMI. The phase behavior of MSAN/SAN blends is also explained in relation to slow
phase separation due to low mobility. The  parameters for monomer (i)-monomer (j)
interaction are evaluated from the miscibility map. Second, we report the viscoelastic
properties in the glass transition region. Time-temperature superposition principle can be
applied over the temperature range measured for miscible SMA/SAN blends, but cannot
be applied for immiscible blends although the principle can be applied below or above
the Tg of high Tg component. Temperature dependence of the shift factor aT can be
expressed in WLF equation denoted by aT = 8.86(T  Ts)/(101.6  T  Ts), when the Ts
is chosen to be fitted in the data. Third, we report the viscoelastic properties of SAN
blends in the terminal flow region. Miscible blends exhibit the terminal tails (G  2 and
G  ) and behaves as a viscous fluid having G  G at low frequencies. On the other
hand, immiscible blends do not show the terminal tails and have the long time relaxation
due to a inhomogeneous structure. Lastly, we discuss compositional dependences of the
*
Corresponding Author Email: y.aoki@yz.yamagata-u.ac.jp.
120 Yuji Aoki
zero-shear viscosity 0 and plateau modulus GN0 of miscible blends are discussed in
relation to entanglement couplings. It is emphasized that the viscoelastic properties of
miscible blends should be compared at an iso-free volume condition. Then, the
entanglement molecular weights in the blends are identical with those in the component
(co)polymers. The plateau modulus GN0 of miscible blends is identical with those in the
component (co)polymers and varies linearly with the blend composition.
Keywords: poly(styrene-co-acrylonitrile), miscibility, phase behavior, interaction

parameter, WLF equation, zero-shear viscosity, plateau modulus
1. INTRODUCTION
Miscibility of polymer blend systems has been investigated for 50 years. It is

generally accepted that the thermodynamic basis is an exothermic heat of mixing, since
entropic contributions are so small in high molecular weight polymer blend systems. It is
suggested that specific interactions, such as hydrogen bonding and n- complex
formation, are responsible for the exothermic heat of mixing [1, 2]. In the Flory-Hugins
formulation for polymer mixtures [3], the free energy of mixing G at a temperature T
for a blend of r1-mers and r2-mers can be written as
G/RT  [(1/r1) ln 1  (2/r2) ln 2]  blend12 (1)
where R is the gas constant, 1 and 2 are the volume fraction of r1-mer and r2-mer in the
blend, respectively, and blend is the net segmental interaction parameter between the two
polymers. For blends with high molecular weight, the combinatorial entropy term can be
neglected. Then, the miscibility depends simply on the sign of blend; if blend is negative,
the two polymers are miscible.
It has been demonstrated that many blend systems consisting of a homopolymer and
a copolymer, or two different copolymers, are miscible for a certain range of copolymer
composition even though the combination of their corresponding homopolymers are
immiscible. This would suggest that there are no effective specific interactions between
the components and that the driving force foe miscibility lies elsewhere. To explain the
above experimental facts, it was proposed that miscibility of copolymers is due to
repulsion between the two different monomer units comprising the copolymers [4-6], and
in a meanfield approach the overall Flory-Huggins interaction parameter between the two
polymers can simply expressed in terms of the respective segmental interaction
parameters. For the most general case of a blend of two statistical copolymers, (AxB1x)
and (CyD1y), where x and y denote the copolymer compositions in terms of volume
Mistibility and Viscoelastic Properties of Poly(Styrene-Co-Acrylonitrile) Blends 121
fraction, the blend expression involves six segmental interaction parameters ij for the
monomer(i)-monomer(j) interactions relevant to the system, and this is given by
blend = xyAC  (1  x)yBC  x(1 y)AD

 (1  x)(1  y)BD  x(1 x)AB  y(1  y)CD (2)
The first four terms on the right-hand side define additive interaction between
component monomers of the mixture of the two copolymers, whereas the remaining two
terms define the intramolecular forces between two different monomers comprising each
of the copolymers, as explained by ten Brinke et al. [5].
For a blend comprising a homopolymer (Ax, x  1) and a copolymer ((CyD1y), we
have
blend = yAC  (1  y)AD  y(1  y)CD (3)
Paul and Barlow [6] and Shiomi et al. [7] applied it to blends of two statistical
copolymers with a common monomer unit, (AxB1x) and (CyB1y). The blend expression
for the blend is given by a two-variable quadratic with three ij parameters:
blend = ABx2  (AC  BC  AB)xy  BCy2 (4)
For a blend of two copolymers that differ only in composition, (AxB1x) and (AyB1y),
the blend expression for the blend is given by
blend = (x  y)2AB (5)
Poly(styrene-co-acrylonitrile) (SAN) is mainly used for continuous phase of ABS

polymers. To improve heat resistance of ABS polymers, several approaches have been
taken. One approach is to prepare copolymers of -methylstyrene (-MS) and
acrylonitrile (AN), and the other is to prepare blends of SAN with other polymer. In any
case, miscibility and phase behavior of SAN blends are important factors. There have
been many studies of miscibility of SAN with homopolymers and copolymers. It is
reported that SAN is miscible with poly(methyl methacrylate) (PMMA) [8-22],
poly(-caprolactone) (PCL) [20, 23-27], poly[styrene-co-(maleic anhydride)] (SMA) [28-
37], poly[styrene-co-(N-phenylmaleimide)] (SMI) [32, 38], poly[(-methyl styrene)-co-
acrylonitrile] (MSAN) [39-42] and so on, when the acrylonitrile (AN) content of SAN is
within a specific range. The reason why SAN is miscible with many other (co)polymers
is mainly due to a strong repulsion between styrene (S) and acrylonitrile (AN) monomer
units. First, we explain the miscibility of SAN with PMMA, SMA, SMI, and MSAN. It is
122 Yuji Aoki
emphasized that 1) volume fraction should be used as copolymer composition to estimate

the interaction parameters from miscibility map and 2) longer time annealing is required
to decide phase diagrams
Until now, numerous experimental and theoretical studies have been performed to
investigate the rheological properties of miscible and immiscible polymer blends [43].
Prest and Porter [44] studied the viscoelastic properties of polystyrene (PS)/poly(2,6-
dimethyl phenylene oxide) (PPO) blends which is one of the most successful polymeric
materials. They reported that the relaxation processes of both molecular species exhibit
the same temperature dependence but are a function of composition. Wu [45-47] studied
the viscoelastic properties of miscible blends of PMMA with poly(vinylidene fluoride)
(PVDF), poly(ethylene oxide) (PEO), and SAN. He discussed the entanglement, friction,
and free volume between dissimilar chains in miscible polymer blends in connection with
the Flory-Huggins  parameter. He suggested that dissimilar chains are less likely to
entangle with each other than similar chains. For PMMA/PVDF blend, it results in a large
reduction in the zero-shear viscosity in the system. He also found that the free volume
tends to be linearly additive. Saito et al. [48] investigated the orientation relaxation of
dissimilar chains in the molten miscible blends PMMA/PVDF and
PMMA/poly(vinylidene fluoride-co-trifluoroethylene). They suggested that the dissimilar
polymers do not relax independently but cooperatively, and this behavior is induced by a
constraint due to the specific interactions between the dissimilar chains, e.g., weak
hydrogen bonding. Wu suggested that, as the interchain interaction becomes stronger, the
entanglement probability between dissimilar chains becomes smaller.
Han and Kim [49, 50] developed a molecular theory to predict the linear viscoelastic
properties of binary mixtures of miscible blends, using the concept of the tube model of
Doi and Edwards [51]. They assumed that each primitive chain reptates in a respective
tube but that molecular interactions between the two chemically dissimilar primitive
chains take place under the influence of an external potential, dependent upon the
interaction parameter of the constitutive components. Their theory predicts the
followings: (1) when the viscosity ratio 01/02 of the constitutive components is much
greater than 1, the log 0b versus blend composition curves show negative deviation from
linearity for   0 and positive deviations for   0. (2) When the 01/02 ratio is close to
1, the value of log 0b has a minimum at a certain blend composition for   0. Here, 0b,
01, and 02 indicate the zero-shear viscosity of the blend, component 1, and component
2, respectively. They found that the predictive behavior of log 0b versus blend
composition (positive deviation) is agreement with experimental data for PMMA/SAN
blends of nearly zero value of .
Another model for the viscosity of miscible polymer blends was reported by Haley
and Lodge [52]. The model combines a mixing rule for chain relaxation, based on the
double reputation concept as adopted by Tsenoglou [53], with a calculation of the
concentration dependence of the monomeric friction factors, based on the self-

concentration model of Lodge and McLeish [54].
Pathak et al. [55] reported the linear viscoelasticity of miscible PMMA/SAN blends
in details. They found that the blends are theromrheologically simple and gave specific
criteria for expecting thermorheological complexity, based on a simple compositional
model [56, 57].
There have been numerous studies of miscibility and viscoelasticity of polymer
blends. However, there are many contradictory conclusions in the studies. In this review,
we focus on SAN blends. We first explain the miscibility of the blends, which miscibility
is due to the presence of unfavorable interaction between the constituent monomers, not
to specific interchain interaction [32]. Here, we would like to emphasize that the volume
fraction should be used as copolymer composition for making miscibility maps and
longer time annealing is required to decide phase diagrams. We next explain the
viscoelastic properties of miscible and immiscible SAN blends [58, 59]. Then,
compositional dependence of the viscoelastic functions is discussed for these miscible
blends [30, 59-62]. Here, an importance of the free volume in the blends is emphasized.
2. MISCIBILITY OF SAN BLENDS
2.1. Miscibility Window of PMMA/SAN Blends
Miscibility and phase behavior of PMMA/SAN blends have been widely studied [8-
22]. It has been known that PMMA forms miscible blends with SAN over a limited
window of AN contents while PMMA is not miscible with homopolymers of styrene or
acrylonitrile. For PMMA/SAN blends, the interaction parameter blend can be expressed
by Eq. (3’).
blend = yMMA-S  (1 y)MMA-AN  y(1  y)S-AN (3’)
Miscibility is only to be expected if blend  0. In the case 0  MMA-S  MMA-AN  S-

AN, the repulsion between the different segments comprising the copolymer (S-AN) is
stronger than the repulsion between the homopolymer and copolymer segments (MMA-S
and MMA-AN). Therefore, the miscibility in a certain range of copolymer composition
does not originate from any specific interaction but arises from the special structure of
blend given by Eq. (3’). Using S-AN=0.84, MMA-S=0.03, and MMA-AN = 0.50 reported by
Nishimoto et al. [18], the compositional dependence of blend was calculated and
schematically shown in Figure 1. The miscibility is observed for SAN with AN content
between 9 and 33 wt %.
124 Yuji Aoki
Figure 1. Interaction parameter blend for PMMA/SAN blends as a function of concentration of SAN.
Figure 2. Phase diagram of PMMA/SAN blends.
PMMA/SAN blends exhibit phase separation on heating or lower critical

solution temperature (LCST) at relatively low temperatures. The phase diagram for
PMMA/SAN =60/40 blends by Suess et al. [15] is shown in Figure 2. A miscibility
window is observed in the temperature versus copolymer composition plane. The narrow
gap between the cloud point curve and the glass transition of PMMA/SAN blends
severely restricts the ability for melt processing to achieve single phase mixture.
2.2. Miscibility Maps of SMA/SAN and SMI/SAN Blends
Table 1 summarizes characteristics of copolymer samples used here [30, 32]. SMA-
15 is a commercially available polymer (Dylark 332, Arco Polymers Inc.) Other
(co)polymer samples were prepared by radical polymerization by our laboratory. Maleic
anhydride content (MA%) of SMA was determined by a coulometric titrator (Mitsubishi
Chemical Industry, Model CT-01). AN content (AN%) of SAN and N-phenylmaleimide
content (MI%) of SMI were determined by elementary analysis (Yanako CHN Coder).
Weight average molecular weight (Mw) was determined by low angle laser light
scattering (Chromatix, KMX-6). Polydispersity index Mw/Mn was determined by GPC
(Waters, 150 ALC/GPC). The elusion solvent was tetrahydrofuran.
The miscibility of SMA/SAN and SMI/SAN blends having various copolymer
compositions was determined using dynamic mechanical measurements. The
measurements were conducted by a viscoelastic spectrometer (Iwamoto Seisakusho Co.
Ltd.) at a heating rate of 1C/min and frequency of 20 Hz, and were started at 23C and
continued until the sample became too soft to be tested. Blend miscibility was monitored
using the criterion that a miscible blend exhibits one Tg whereas an immiscible blend
shows the Tg values for both components. [32]
Figure 3 shows the temperature T dependence curves of the dynamic storage modulus
E and loss modulus E for SMA-15/SAN blends having different SAN compositions.
Table 1. Molecular characteristics of PS, SAN, SMA, and SMI samples
Sample wt % of S mol % of S Mw/103 Mw/Mn /(g/cm3)

PS 100 100 260 2.0 1.052
AS-15 (SAN) 85.1 74.4 128 2.2 1.070
AS-20 (SAN) 80.5 67.8 89 1.7 1.073
AS-27 (SAN) 73.5 58.5 87 2.2 1.077
AS-30 (SAN) 71.3 55.9 80 2.2 1.079
AS-33 (SAN) 67.7 51.7 63 2.1 1.081
AS-40 (SAN) 61.3 44.7 58 1.9 1.084
AS-55 (SAN) 45.0 29.4   
SMA-15 (SMA) 84.7 83.9 185 2.0 1.098
M-9 (SMA) 69.8 68.5 171 2.1 1.174
P-90 (SMI) 65.3 75.8   1.125
P-91 (SMI) 58.1 69.8   1.150
P-93 (SMI) 44.0 56.6 270 2.0 1.196
126 Yuji Aoki
Figure 3. Dynamic storage and loss moduli, E and E, as a function of temperature for SMA-15 blends
with AS-15, AS-20, and AS-27.
These three SAN samples show almost the same temperature dependence curve. For
SMA-15/AS-15 blends, the E curve of each blend has single peak, indicating that AS-15
is miscible to SMA-15. For SMA-15/AS-20 blends, the E becomes broader, indicating
boundary between miscibility and immiscibility. For SMA15/ AS-27=50/50 blend, the E
has one peak and one shoulder, indicating that AS-27 is immiscible to SMA-15.
Figure 4a shows the T dependence curves of the E and E for the component
copolymers SMI (P-93) and SAN (AS-15), and their 50/50 (in weight) blend [32]. The
maximum of the E curve due to the glass transition was observed at about 120C and
220C, respectively, for AS-15 and P-93. The E decreases sharply with elevating T in the
glass transition region for the component copolymers. For the P-93/AS-15 blend, on the
other hand, the E curve exhibited two peaks and the E decreased by two steps in the
corresponding T region. The peak temperature in the E curve at low temperature was
found to agree with that of AS-15. The existence of two Tgs is an evidence of the
immiscibility bwtween P-93 and AS-15.
Figure 4bd shows the T dependence curves of the E and E for P-93 blends with
AS-20, AS-27, AS-33, and AS-40. Each blend exhibited a single and similar Tg. These
peaks in the E due to the glass transition existed at the intermediate temperature between
those of the respective components, although the peaks became to be somewhat
broadened. This fact is the evidence that these SMI/SAN blends are miscible.
Figure 4f shows the T dependence curves of the E and E for P-93/AS-55 blend. The
temperature dependence curve of the P-93/AS-55 blend was similar to that of P-93/AS-15
blend. This indicates that P-93 is immiscible with AS-55.
Based on the dynamic mechanical measurements, P-93 was found to be miscible with
SAN having AN content ranging from 20 to 40 wt%, although P-93 was immiscible with
SAN having below 20 wt % or above 40 wt%.
(a) (b)
(c) (d)
(e) (f)
Figure 4. Dynamic storage and loss moduli, E and E, as a function of temperature for P-93 blends
with AS-15 (a), AS-20 (b), AS-27 (c), AS-33 (d), AS-40 (e), and AS-55 (f). Open circles denote pure
copolymers, and close circles their blends [32].
Figure 5 shows the T dependence curves of the E and E for M-9 (SMA) blended
with AS-20, AS-30, and AS-40 (SAN) [32]. The M-9/AS-20=80/20 blend exhibited two
peaks in the E curve. The low temperature peak shifted to higher temperature and high
temperature peak shifted to lower temperature comparing that of the component
copolymers. The M-9/AS-20=60/40 blend showed very flat E curve, although the curve
did not have two peaks. The M-9/AS-20=40/60 and 20/80 blends exhibited one sharp
peak and a shoulder. The peaks shifted to high temperatures slightly and became broad.
These facts suggest that M-9 mixes partially with AS-20 and the Tg elevates. The
M-9/AS-20 blends are partially miscible and, in other words, a boundary between
miscible and immiscible region.
128 Yuji Aoki
Figure 5. Dynamic storage and loss moduli, E and E, as a function of temperature for M-9 blends with
AS-20 (a), AS-30 (b), and AS-40 (c) [32].
Figure 5b shows T dependence curves of the E and E for the M-9/AS-30 blends.
The E curve of M-9/AS-30 blend exhibits a single peak corresponding to the Tg. The
peaks are as sharp as those of the component copolymers. The peak temperature shifts to
higher temperature with increasing of M-9 content. This simple compositionally
dependent Tg is evidence of miscibility.
Figure 5c shows the T dependence curves of the E and E for the M-9 /AS-40
blends. The E curve of M-9/AS-40 blends has two peaks due to the Tg of each
component. This is evidence of the immisciblity of M-9/AS-40 blends. It was found that
M-9 is miscible with SAN having AN content ranging from about 20 to 35 wt %.
Figures 6 and 7, respectively, show the miscibility maps of SMA/SAN and SMI/SAN
blends. Here, the abscissa and ordinate represent the volume fraction of component
copolymers. The volume fraction was calculated from the specific volume of each
comonomer obtained by density measurements. Open circles denote “miscible,” “half-
closed circles” partially miscible, and “closed circles” immiscible blends. The boundaries
between miscible and immiscible domains are expressed by straight lines for both blend
systems.
Figure 6. Miscibility map for SMA/SAN blends, when the volume fraction is used as copolymer
composition [32]. The broken lines are calculated using AN-MA/AN-S = 0.04 and S-MA/AN-S = 1.89.
○: miscible, ●:immiscible.
Figure 7. Miscibility map for SMI/SAN blends, when the volume fraction is used as copolymer
composition [32]. The broken lines are calculated using AN-PMI/AN-S = 0.07 and S-PMI/AN-S = 0.30.
○: miscible, ●:immiscible.
130 Yuji Aoki
Shiomi et al. [7] reported the theoretical miscibility maps for blends of two random
copolymers having a common monomer. Our experimental miscibility maps shown in
Figures 6 and 7 can be used to estimate the segmental ij parameters for SMA/SAN and
SMI/SAN blends. Comparing the experimental data with theoretical ones by Shiomi
et al. the ij values were obtained as follows.
The boundaries between miscible and immiscible domains were expressed by two
straight lines intersecting at the original points, as described before. For SMA/SAN
blends, two straight lines are given by (x  1.8y)(x  1.05y), where x is AN volume
fraction and y is MA volume fraction. Compared with Eq. (4), we can obtain the next
relations
AN-MA/AN-S  S-MA/AN-S  1 = 2.85
S-MA/AN-S = 1.85  1.05 = 1.89 (6)
From these relations, AN-MA/AN-S was obtained to be 0.04. For SMI/SAN blends,
two straight lines are given by (x  0.90y)(x  0.33y) = 0, where x is AN volume fraction
and y is PMI volume fraction. We can get S-MI/AN-S = 0.30 and AN-MI/AN-S = 0.07.
These results confirm that the so-called repulsion effect can lead to exothermic mixing in
the absence of specific interactions.
To determine ij values, we adopted a solubility parameter approach. When there are
no specific interactions between the copolymers, ij is expressed as [63]
ij = (V/RT)(i  j)2 (7)
where i and j are the solubility parameters of the respective components and V is a
reference volume. V was taken as the geometric mean of the molar volumes of the
involved polymer repeat units, as used by Kressler et al. [13]. From the group molar
constant according to Hoy [64] and molar volume by density measurements, the
solubility parameters at 23C were obtained to be S = 9.1, AN = 12.0, and MA = 12.7
(cal/cm3)1/2. Inserting these values in Eq. (7), we got three  parameters, AN-S = 0.98, S-
MA = 1.72, and AN-MA = 0.05. The  parameter ratios, S-MA/AN-S = 1.76 and AN-MA/AN-S
= 0.05, are in excellent agreement with the values obtained from the miscibility map of
SMA/SAN blends. By inclusion of AN-S = 0.98 into the  parameter ratios, four
segmental interaction parameters were obtained and listed in Table 2. It is clear that the
observed miscibility maps can be explained by the strong repulsion between the
segmental units within the copolymers, without any attractive intermolecular interactions.
Table 2. Segmental interaction parameters
Segmental pair ij Segmental pair ij

AN-S 0.98 S-MA 1.85
AN-MA 0.04 S-MI 0.29
AN-MI 0.07  
Figure 8. Miscibility map for SMA/SAN blends, when the weight fraction is used as
copolymer composition.
Here, we would like to a comment on miscibility map. Figure 8 shows a miscibility

map of SMA/SAN blends. In this figure, the abscissa and ordinate are the copolymer
compositions in terms of weight fraction, not volume fraction. The miscibility region
include the diagonal line, which the compositions of SMA and SAN copolymers are
equal. From Eq. (4), we can get that blend = AC  0. This means that miscibility of SMA
and SAN is caused by specific interactions of AN and MA leading to an exothermic heat
of mixing, as reported by Paul and Barlow [6]. However, they used the copolymer
compositions in weight fraction. On the other hand, Kressler et al. [33] reported from
Fourier transform infrared spectroscopy that the miscibility of SMA/SAN blends is
caused chiefly by intramolecular interactions, and the intermolecular interactions can be
partly neglected. Heinen et al. [35] studied solid-state NMR study of miscibility of
SMA/SAN blends and found that the miscibility is associated mainly with a decrease of
intramolecular repulsion upon mixing, as opposed to a specific exothermic binary
interaction. We believe that the miscibility of SMA/SAN blends is due to mainly
intramolecular interaction between S and AN in SAN, and S and MA in SMA.
132 Yuji Aoki
2.3. Phase Behavior of MSAN/SAN Blends
Ikawa et al. [39] studied the miscibility of MSAN and SAN by NMR. Cowie et al.
[40] examined the miscibility by differential scanning calorimeter (DSC). Gan et al. [41]
also studied the DSC miscibility maps for various blends of MSAN and SAN and
calculated the segmental interaction energy densities or parameters. But, they did not
report the phase diagram of these blends. Especially, Gan et al. concluded that these
blends did not exhibit lower or upper critical solution temperature (LCST or UCST)
phase behavior.
Slow phase separation due to low mobility is exactly a phenomenon that blend
components are trapped into a non-equilibrium but apparent homogeneous state, when
blend samples are prepared from solvent casting other than freeze-drying. Indeed, some
literatures describe several examples of quite slow phase separation of blends [36, 65,
66]. Diffusion is very slow at a temperature just above Tg, so that the temperature range
where phase separation proceeds at a reasonable rate is even narrower.
In this subsection, the miscibility and phase behavior of MSAN/SAN blends are
reported [42]. AN content of MSAN used is 32 wt % and that of SAN ranges from 23 to
45 wt %. The miscibility and phase behavior were determined from Tg behavior of the
blends by DSC thermograms. The interaction parameters for MSAN-32/SAN systems
were calculated utilizing the FloryHuggins theory.
The molecular characteristics of MSAN and SANs used in this study are listed in
Table 3. Blend samples were prepared by casting from butanone solution. A desired
weight ratio of MSAN/SAN blends was dissolved in butanone at 5 wt % of total polymer.
The solution was poured into a flat-bottomed Petri dishes. Evaporation of solvent was
carried out slowly at room temperature. After the solvent had been evaporated at room
temperature, the cast film was dried in a vacuum oven for several days, and careful
stepwise heating from 70 to 130C prevented bubbling. Finally, subsequent annealing at
130C for 100 h was designed to ensure perfect removal of solvent. The well-annealed
specimen was cooled rapidly at room temperature. All the blends of MSAN-32 with
SANs prepared here were transparent at room temperature. It is difficult to learn much
about the phase behavior of their blends by visual observation, because two components
have very similar refractive indices, as seen in Figure 9.
The glass transition temperature Tg of various MSAN/SAN samples was measured
with a Perkin-Elmer DSC-7 using a heating rate of 10 K/min. A sample weight 715 mg
was used on the measurements. Tg was taken as the temperature corresponding to half of
the heat capacity change; that is, Tg is defined as the midpoint of the glass transition. The
error of Tg determination is 0.5C. Glass transitions were usually taken from a second
scan to ensure reproducible thermograms. The maximum temperature in the first scan
was kept below the annealing temperature of the sample. Annealing of various blends
was carried out at a desired temperature for about 100 h. After annealing, the sample was
cooled rapidly at room temperature, and another DSC scan was made to assess the
change in LCST behavior caused by annealing.
Table 3. Molecular characteristics of MSAN and SAN samples
Sample AN wt % (vol %) Mw/103 Mw/Mn

MSAN-32 31.2 (30.4) 63 2.1
SAN-23 23.2 (22.2) 122 2.1
SAN-26 25.3 (24.3) 95 2.8
SAN-30 28.6 (27.5) 78 2.3
SAN-33 31.9 (30.7) 62 2.2
SAN-37 36.4 (35.1) 85 1.9
SAN-40 39.1 (37.9) 63 2.0
SAN-45 43.7 (42.4) 70 1.9
Figure 9. Refractive index of SAN plotted against volume fraction of AN in SAN. That of MSAN-32 in
indicated by an arrow [42].
A typical measurement consisted of heating at 10 K/min to annealing temperature Ta,

keeping at that temperature for 5 min, and then cooling at 80 K/min to 328 K. This cycle
was repeated several times in succession. A fresh sample gave a thermogram with minor
134 Yuji Aoki
irregularities in the first run, but if it was held at a temperature 1020 K higher than the
Tg for 5 min or longer, all subsequent runs were completely reproducible within the
instrumental noise level.
Figure 10. DSC thermograms of blends of MSAN-32 and SAN with various AN contents [42].
The appearance of a single, composition dependent Tg of a blend is a generally

accepted criterion for blend miscibility. Figure 10 shows the DSC curves for various
MSAN/SAN (=50/50 in weight) blends annealed at 130C for 100 h. The MSAN-
32/SAN-30, SAN-33, SAN-37, and SAN-40 blends show only one glass transition. On
the other hand, the MSAN/SAN-23, SAN-26, and SAN-45 blends exhibit two glass
transitions. This result indicates that MSAN-32 is miscible with SAN-30, -33, -37, and -
40 but not with the other SAN samples of higher or lower AN content at 130C. A
sample that has undergone phase separation will contain in it some region richer in
MSAN and other region poorer in MSAN than the original mixing composition. As a
consequence, its DSC curve will show Tg’s higher or lower than Tg expected of the
component polymers. Here, we designate Tg(MSAN) for MSAN-32 rich phase and
Tg(SAN) for SAN rich phase. The Tg-composition curves for the various blends are
shown in Figure 11ag. The two Tg values for the immiscible blends (MSAN32/SAN-23
and MSAN-32/SAN-45) are close to those of MSAN-32 (Tg = 123C) and the respective
SAN sample. Tg’s stay relatively constant as shown in Figure 11a and g. On the other
hand, for a immiscible MSAN-32/SAN-26 blend Tg(SAN) is raised and Tg(MSAN)
lowered with increasing in MSAN content. This behavior indicates that MSAN-32/SAN-
26 blend is just at the boundary of miscibility and immiscibility.
Figure 11. Composition-dependent glass transition temperature Tg for MSAN-32/SAN blends [42].
136 Yuji Aoki
Blends of MSAN-32 with SANs containing 28.639.1 AN wt % showed one

composition-dependent Tg, indicating completely miscibility between MSAN-32 and
these SAN samples. The glass transition region becomes noticeably broader when the
two components are presented in comparable amounts. Such a phenomenon was seen for
many other miscible blends, and is probably due to the existence of concentration
fluctuation in the miscible blends. It is concluded that MSAN-32 is miscible with SAN
having AN contents between 28 and 40 wt % at 130C.
Miscible polymer blends tend to phase separation at elevated temperature. This
LCST behavior is characteristic for miscible blends. The visual indication method is
usually used to find LCST behavior. However, we cannot use the method because of the
similarity of refractive indices of MSAN and SAN. An alternative way to test LCST
behavior is to heat the blends to a temperature above the suspected cloud point, quench
them, and examine their Tg behavior. The appearance of a single Tg of a blend after
annealing is no LCST. When the temperature is above the cloud point, two Tg’s are found
in the DSC thermogram. The miscibleMSAN-32/SAN blends at 130C were examined
for the existence of LCST behavior. The phase diagrams of the blends were constructed
from the phase separation temperature obtained from DSC measurements.
Figure 12ac shows the DSC traces for the MSAN-32/SAN-30 = 50/50 samples
annealed at 145, 150, and 160C for various hours. We can observe that the phase
separation took place at longer annealing time and the phase separation time became
shorter at higher temperature. Figure 13ad shows Tg versus composition curves for the
blend annealed at various temperature for 100 h. Tg varies almost linearly with MSAN-
32 content for the samples annealed below 140C, and two Tg’s are found for the samples
annealed at 145 and 150C. The lower Tg value is 4C higher than that of SAN-30 (Tg =
112C), indicating the presence of MSAN-32 in the SAN-30 rich phase, but the upper Tg
value is close to that of MSAN-32. These results show that MSAN32/SAN-30 blend has
a phase separation temperature between 140 and 145C. Miscible blends of MSAN-32
with SAN-37 and SAN-40 also showed LCST behavior. The other miscible blend
(MSAN-32/SAN-33) remained single phase up to 250C. Phase separation temperature
was not observed for MSAN-32/SAN-33 blend. For the blends, the phase separation
temperature may exist above their decomposition temperature.
The phase diagrams were constructed from the phase separation temperatures
obtained from the DSC measurements and are shown in Figure 14ad. Open circles
denote “one phase” and closed circles “two phases.” This is the first observation of the
presence of LCST behavior for MSAN/SAN blends. Gan et al. [41] reported that
MSAN/SAN blends did not exhibit LCST phase behavior. Their samples were dried in a
vacuum oven at about 150C for 2 days. Figure 14 indicates that we cannot observe
LCST behavior for MSAN-32/SAN-30, SAN-37, and SAN-40 blends, after annealing
above 150C.
Figure 12. DSC thermograms of MSAN-32/SAN-30 = 50/50 blends annealed at 145, 150, and 160C
for various hours [42].
138 Yuji Aoki
Figure 13. Composition-dependent glass transition temperature Tg for MSAN-32/SAN-30 blends

annealed at 130, 140, 145, and 150C for 100 h [42].
Figure 15 shows the miscibility map for MSAN-32/SAN=50/50 blends. Open and
closed circles, respectively, refer to miscible and immiscible blends. Larger circles denote
the results of our work at 130C and small circles those of Ikawa et al. [39]. Our data are
in good agreement with those by Ikawa et al.
We explain the miscibility map according to the FloryHuggins theory (Eq. (1)). For
MSAN/SAN blends of two random copolymer containing a common monomer unit,
SxAN1x)r1 and (MSyAN1y)r2, blend is given by
blend = S-ANx2  (S-MS  MS-AN  MS-AN)xy  MS-ANy2 (4’)
S-AN, MS-AN, and S-MS are the segmental  parameters between styrene and
acrylonitrile, -methylstyrene and acrylonitrile, and styrene and -methylstyrene
monomers, respectively. When blend is smaller than blend, crit =1/2(1/r11/2 + 1/r21/2)2, the
two polymers are miscible each other. Even if all the segmental  parameters are positive,
the blend can be negative in a limited range of copolymer compositions.
Figure 14. Phase diagrams of blends of MSAN-32 and SAN with various AN contents. Open circles
denote “one phase” and closed circles “two phases” [42].
We adopted a solubility parameter approach to determine ij, as reported for

SMI/SAN and SMA/SAN blends and could get for the parameters S-AN =0.747, MS-AN =
0.736, and M-MS = 0.001 at 130C. The straight lines in Figure 14 were calculated from
Eqs. (1) and (8) using ij set and the constant value of blend, crit (= 0.002). The observed
miscibility map is in good agreement with that predicted using the solubility parameter
approach. It is clear that the miscibility of MSAN/SAN blends can be explained by the
strong repulsion between the segmental units within the copolymers. As S-AN and MS-AN
become small at elevating temperature, the blend can become higher than blend, crit. Then,
140 Yuji Aoki
phase separation on heating could be caused by free volume contribution to mixing as

well as complex variations in segmental interactions with temperature.
This study is a first observation of LCST behavior for MSAN/SAN blends. Previous
work could not observe LCST behavior, because the blend samples were prepared at
higher temperature than the LCST. To decide the phase diagrams of MSAN/SAN blends,
very long isothermal annealing are required just above the Tg.
Figure 15. Miscibility map for blends of MSAN and SAN [42]: Open and closed circles denote our
results and small circles the results by Ikawa et al. [39].
3. EFFECTS OF MISCIBILITY ON THE VISCOELASTIC PROPERTIES
3.1. Glass Transition Region
In subsection 2.2., we reported that miscibility of polymer blends can be judged from
the temperature dispersion curves of E and E or tan . Temperature dispersion
measurement is a simple and conventional method. The measurements were used to
judge the miscibility for many polymer blends. However, it is impossible to judge the
miscibility in some cases, as described for M-9/AS-20 blend. In this subsection, we
describe the frequency  dependent curves of miscible and immiscible SMA/SAN blends
in the glass transition region.
Figure 16 shows the master curves of E and E plotted against normalized frequency
/max for miscible M-9/AS-30 blend. Here, max is the frequency which E is a
maximum. It was found that this blend obeys time-temperature superposition over
temperature range measured. At high , the E values did not depend on the blend
composition. The E peaks became broader for intermediate blend composition. On the
other hand, the curves of M-9/AS-20 and AS-40 blends could not be superposed at all
temperatures measured, but could be superposed below or above 140C. Figures 17 and
18, respectively, show the master curves of E and E for M-9/AS-20 blends below and
above 140C. This indicates that M-9/AS-20 blend is immiscible. E values did not
depend on the blend composition at high , as same as for M-9/AS-30 blends. However,
the decrease in E decreasing  becomes very weak, especially for M-9/AS-20 = 60/40
blend. This blend consists of two phases and the M-9 phase cannot relax below 140C.
(We did not show the master curves for M-9/AS-40 blends, because the master curves
were similar to those of M-9/AS-20 blends).
Figure 16. Master curves of storage and loss tensile moduli, E and E, plotted against normalized
frequency, /max, for M-9/AS-30 blends with various compositions.
142 Yuji Aoki
frequency, /max, for M-9/AS-20 blends with various compositions at temperatures below 140C.
frequency, /max, for M-9/AS-20 blends with various compositions at temperatures above 140C.
Figures 19 and 20 show the T dependence of the shift factor aT for M-9/AS-30 and
M-9/AS-20 blends. When the reference temperatures Ts were chosen at the temperatures
shown in these figures, all the curves could be expressed in the WLF equation [67]
denoted by Eq. (8);
log aT = c1(T  Ts)/(c2 + (T  Ts) (8)
where Ts is the reference temperature and c1=8.86, and c2=101.6. (Ts  50) corresponds to
the glass transition temperature Tg of the blends. The free volume fraction f of a polymer
at T is given by
f = f0 + ((T  Ts) (9)
where f0 is the free volume fraction at Ts and  the temperature coefficient of the free
volume fraction. According to the Doolittle equation [68] which describes the
relationship between the viscosity of a system and the temperature dependence of the free
volume fraction, f0 is B/(2.303c1) and  is f0/c2, where B is a constant of the order of 1.
The results in Figures 19 and 20 indicate that both f0 and  are constant at each Ts.
Figure 19. Shift factor, aT,, plotted against a distance from the reference temperature, T  Ts, for M-
9/AS-30 blends. The dotted curve indicates the WLF equation calculated using the reference
temperatures, Ts, shown in the figure.
144 Yuji Aoki
Figure 20. Shift factor, aT, plotted against a distance from the reference temperature, T  Ts, for
M-9/AS-20 blends (left panel, below 140C and right panel, above 140C). The dotted curve represents
the WLF equation calculated using the reference temperatures, Ts, shown in the figure.
Figure 21 shows the peak temperature of loss modulus T(Emax) and (Ts  50) plotted
against blend composition for M-9/AS-20, AS-30, and AS-40 blends. For miscible M-
9/AS-30 blend, one characteristic temperature corresponding to Tg was found. The
temperature increased linearly with M-9 composition. On the other hand, for M-9/AS-20
and AS-40 blends two characteristic temperatures were observed. Low Ts of the blends
was almost the same to those of SANs and were constant irrespective of blend
composition. This fact indicates that AS-20 and AS-40 do not dissolve into M-9 at all.
High Ts of M-9/AS-20 blend shifts to lower temperature with increasing AS-20 content
indicating that some portion of AS-20 dissolve into M-9. This blend is located in a
boundary of miscible and immiscible state. High Ts of M-9/AS-40 is constant irrespective
of AS-40 content for M-9/AS-40 blend, indicating that M-9/AS-40 blend is perfectly
immiscible.
The master curves of the E and E for miscible M-9/AS-30 blends were already
shown in Figure 16. In this figure, the axis of abscissa is a normalized frequency /max
(max is  at E peak). The free volume fraction f of the blends at max was found to
be constant (f = 0.006) irrespective of blend composition, as shown in Figure 22. In other
words, the peak position of E master curves appears at the same  when the curves are
compared at a constant free volume fraction. These results show that the average
relaxation time of the blend does not change. However, the distribution broadens for the
blends, as shown in Figure 16. It may be explained by a local concentration fluctuation in
the blends.
Figure 21. Peak temperature of loss modulus, T(Emax), and (Ts  50) plotted against blend composition
for M-9/AS-20, AS-30, and AS-40 blends.
Figure 22. Free volume fraction f plotted against blend composition for M-9/AS-30 blend.
146 Yuji Aoki
The master curves of the E and E below 140C for immiscible M-9/AS-20 blends
are shown in Figure 17. M-9/AS-20=0/10 exhibits those of an amorphous polymer. With
increasing M-9 content, the master curves become flat and flat, and show very weak 
dependence for 60/40 blend. This weak  dependence suggests that M-9 phase having
high Tg becomes a continuous phase and no relaxation are observed in this temperature
region. Above 140C, SAN chains already relax and SMA chains are the origin of
elasticity. The E values of the blends approach to that of M-9 increasing M-9 content.
It was found that time-temperature superposition principle can be applied to miscible
SMA/SAN blends over the entire temperature range measured. At a constant free volume
fraction, the blends exhibit the maximum value in E at the same  irrespective of blend
composition, although the peaks broaden. On the other hand, the principle can be applied
for immiscible blends in a temperature range, below or above Tg of high Tg component.
Below the Tg, the high Tg component behave like a rigid filler. Above Tg of low Tg
component, the low Tg component behave like a viscous liquid. Accordingly, the
principle can be applied below and above the intermediate temperature between two Tg.
3.2 Terminal Region
In this subsection, the effects of miscibility on the viscoelastic properties of

SMA/SAN and PMMA/SAN blends in the terminal region are reported [30, 61].
Figure 23. Master curves of storage and loss shear moduli, G’ and G’’, for SMA-15/AS-15 blend at
200ºC [30].
Figure 24. Master curves of storage and loss shear moduli, G’ and G’’, for SMA-15/AS-20 blend at
200ºC [30].
Figure 25. Master curves of storage and loss shear moduli, G’ and G’’, for SMA-15/AS-27
blend at 200ºC.
148 Yuji Aoki
Figures 23-25 show the master curves of the G and G for miscible SMA-15/AS-15
blend, and immiscible SMA-15/AS-20 and SMA-15/AS-27 blends, respectively, as a
function of aT. The miscible SMA-15/AS-15 blend shows the terminal tails (G  2
and G  ) and behaves as a viscous liquid having G  G at low  irrespective of
blend composition. On the other hands, for the immiscible SMA-15/AS-20 and SMA-
15/AS-27 blends a long time relaxation process is clearly observed through the G. This
long time relaxation is due to the inhomogeneous structure of immiscible blends.
Next, the effects of  parameter on the viscoelastic properties are explained for
PMMA/SAN blends [59]. Molecular characteristics of the samples used here are shown
in Table 4. Figures 26 and 27 show the master curves of PMMA-7/SAN = 50/50 blends
as a function of aT at 200C. All the blends show the terminal zone at low frequencies,
the plateau zone at intermediate frequencies, and entry the glass transition zone at higher
frequencies. At low , PMMA-7/SAN-10, SAN-15, SAN-18, SAN-20, SAN-23, and
SAN-27 blends exhibit a Newtonian flow; the dynamic viscosity   G/ is
independent of frequency and G is proportional to (aT)2. These blends are miscible and
exhibit almost the same master curves. It should not be surprising to observe such a
behavior, because of the similar viscoelastic properties of the component SANs. On the
other hand, PMMA-7/SAN-0, -30, -33, -37, and -40 blends show a long time relaxation
and thus are immiscible blends. This relaxation should arise from an inhomogeneous
structure. However, these immiscible PMMA-7/SAN blends were also
themorheologically simple.
Table 4. Molecular characteristics of PMMA and SAN samples
Sample AN(wt %) Mw/103 Mw/Mn

PMMA-7  72 1.7
SAN-0 (PS) 0 113 2.2
SAN-10 9.6 151 1.9
SAN-15 14.7 138 2.5
SAN-18 18.4 136 2.1
SAN-20 20.1 145 2.0
SAN-23 23.2 131 2.3
SAN-27 27.0 84 2.4
SAN-30 28.6 78 2.1
SAN-33 31.8 56 2.4
SAN-37 36.5 74 2.0
SAN-40 38.7 63 2.1
Figure 26. Master curves of storage and loss moduli, G (top panel) and G (bottom panel) at 200C
for PMMA-7/SAN-0, SAN-10, SAN-15, SAN-18, SAN-20, and SAN-23 blends [59].
Figure 28 show log G versus log G plots for various PMMA-7/SAN = 50/50 blends.
It was found that the plots give temperature independent correlation for both miscible and
immiscible blends, as already shown by Han and Chuang [69]. It was also seen that for
the miscible blends (open symbols) such plots yield correlations which become
independent of AN wt % of SAN with slope of 2. On the contrary, immiscible blends
(closed symbols) the slopes becomes slower at low G and G region. These results
indicate that the G versus G plot is a very useful tool for determining polymer-polymer
miscibility.
The effects of shear flow on the phase behavior of PMMA/SAN blend were reported
by several authors [70-73]. The shear effect is an important and interesting phenomenon.
However, we did not evaluate the flow effects on the miscibility of PMMA/SAN blends,
because we thought that the effects of AN content of SAN on the miscibility are larger
than the flow effect.
150 Yuji Aoki
Figure 27. Master curves of storage and loss moduli, G (top panel) and G (bottom panel) at 200C for
PMMA-7/SAN-23, SAN-27, SAN-30, SAN-33, SAN-37, and SAN-40 blends [59].
Figure 28. Log G versus log G plot for various PMMA-7/SAN blends [59].
Figure 29 shows plot of the plateau modulus GN0 against AN wt % in SAN for
various PMMA/SAN blends. Here, the plateau modulus GN0 was determined as the
storage modulus G at the frequency where tan  is at a minimum from the master curves.
The GN0 value of SAN was found to be a linear function of AN wt % in SAN. The solid
line in the figure denotes a straight line calculated by assuming the additivity of the
plateau modulus of component copolymers. Most interesting result seen in Figure 29 is
that the GN0 shows a linear function of AN wt % in SAN irrespective of polymer
miscibilitr.
Figure 29. Plot of plateau modulus GN0 against AN wt % in SAN for various PMMA-7/SAN
blends [59].
The entanglement molecular weight in immiscible blends should have “apparent”

values, because the above method to determine the plateau modulus is meaningless for
immiscible blends.
Dynamic viscoelastic properties of blends of PMMA and SAN with various AN
contents have been investigated. It was found that PMMA/SAN blends are miscible when
the AN content of SAN ranges from 10 to 27 wt % and are immiscible when the AN
content in SAN is less than 10 wt % or greater than 30 wt %. Melt rheology of the blends
was characterized by the plots of G against G. Log G against G plots exhibited a
straight line of slope 2 for the miscible blends, but did not show a straight line for the
immiscible blends. These facts suggest that G against G plots can be used for
determination of rheological miscibility of polymer blend systems. GN0 determined as the
G at the frequency where tan  is at a minimum, varied linearly with the AN content of
SAN irrespective of blend miscibility. This result indicates that the additivity rule holds
well for the entanglement molecular weights in miscible PMMA/SAN blends. However,
the entanglement molecular weight in immiscible blends should have “apparent” values,
152 Yuji Aoki
because the above method to determine the plateau modulus is meaningless for
immiscible blends.
4. CHARACTERISTIC FEATURES OF THE VISCOELASTIC PROPERTIES

FOR MISCIBLE SAN BLENDS
In this section, characteristic features of the viscoelastic properties for miscible

SMI/SAN, PMMA/SAN, and PCL/SAN blends are reported [60-62]. Polymer
characteristics of the samples used here are summarized in Tables 5 and 6. The storage
modulus G and loss modulus G for these SAN blends were measured as a function of
angular frequency . Measurements were taken at different temperatures to investigate
the temperature T dependence of the viscoelastic properties. The -dependent curves of
G and G at various temperatures can be superposed into a master curve by shifting the
curves along  axis by a factor aT. Compositional dependence of the shift factor, zero-
shear viscosity, and plateau modulus for miscible SAN blends are discussed.
Table 5. Molecular characteristics of SMI and SAN samples
Sample S wt % Mw/103 Mw/Mn

P-30 (SMI) 57.7 163 1.9
AS-2 (SAN) 73.7 78 2.8
AS-3 (SAN) 74.9 118 2.8
AS-4 (SAN) 74.3 173 2.7
Table 6. Molecular characteristics of PMMA and SAN samples
Sample Mw/103 Mw/Mn AN wt % st ht it

PMMA-7 72 1.7  57.0 38.1 4.9
PMMA-60 593 3.2  49.2 40.3 10.5
SAN-18 116 2.1 18.4   
SAN-27 84 2.4 27.0   
4.1. Shift Factor
In the left panel of Figure 30, the shift factors aT are plotted against T for P-30
(SMI)/AS-3(SAN) blends. The reference temperature Ts is 200C for all samples.
T dependence of aT becomes steep with an increase in P-30 content. In the right panel of
Figure 30, these aT data are replotted against temperature difference (T  Ts) and are in
good agreement with the WLF equation [63] with constant coefficients c1 (= 8.86) and
c2 (= 101.6) shown with the dotted line in Figure 30(b), when Ts values are chosen to be
136, 145, 155, 166, 181, and 198C for P-30/SAN-3 = 0/100, 20/80, 40/60, 60/40, 80/20,
and 100/0, respectively. Provided that aT data are taken over a sufficiently wide
temperature range, this procedure determines Ts within 1 C. The Ts of the blend was
found to be a linear function of SMI mole fraction.
Figure 30. Left panel: Temperature dependence of the shift factor aT for P-30/AS-3 blends at the
reference temperature of 200 C. Right panel: shift factor, aT, plotted against a distance from the
reference temperature, T  Ts, for P-30/AS-3 blends. The dotted curve indicates the WLF equation
calculated using the reference temperatures, Ts, shown in the figure [60].
Previously, we showed the master curves of G and G of SMA-15/SAN-15 blends in

Figure 23. The T dependence of aT of the blends was studied. It was found that these aT
data of the blends are found to be excellently described by the WLF equation. The Ts of
the blend was found to be a linear function of SMA mole fraction.
The compositional dependence of Ts for P-30/AS-3 and SMA-15/SAN-15 blends was
found to be a linear function of SMI and SMA mole fraction. These findings indicate that
the additivity of free volume of the component copolymers is applicable for miscible
SMI/SAN and SMA/SAN blends. In other words, SAN (low Tg component) is a diluent
for SMI or SMA (high Tg component). More interestingly, Tg of the SMA/SAN and
SMI/SAN blends is determined by the mole fraction of MA or MI.
Next, we show another example of the composition dependence of Ts, shown in
Figure 31 for PMMA-7/SAN-18 blend. Ts of the blend was found to be a linear function
of the weight fraction x of PMMA. Accordingly, Tsblend is given by
154 Yuji Aoki
Tsblend = (1  x)TsSAN + xTsPMMA (10)
where, TsSAN is the Ts of SAN-18 and TsPMMA that of PMMA-7. Eq. (10) can be
understood in terms of the free volume concept and the additivity of the free volume of
the component polymers. The free volume fraction f of a polymer at a temperature Ts is
given by f  f0 + (T  Ts) as shown in Eq. (9). The result shown in Figure 31 indicates
that f0 and  are constant at each Ts irrespective of blend composition. Then, free volume
fraction of the blends fblend is given by
fblend = f0 + (T  Tsblend) (11)
Putting Eq. (10) into Eq. (11)
fblend = (1  x)fSAN + xfPMMA (12)
where fSAN and fPMMA are the free volume fraction of SAN and PMMA at a constant
temperature.
Pathak et al. [55] studied the temperature dependence of the shift factor aT for
PMMA/SAN blends. They evaluated WLF coefficients by plotting log aT versus
(T  Tg)/(T  T) using the Vogel temperature T as a free parameter to linearize the plot
according to Plazek [74] and reported this procedure to be the most objective and
reliable. We measured the DSC curves of PMMA-7/SAN-18 blends. Figure 32 shows the
composition dependence of Tg(1) and Tg(2) of the blends, where Tg(1) is the onset
temperature and Tg(2) the midpoint temperature of the glass transition. As expected, both
glass transition temperatures are a linear function of PMMA weight fraction. Difference
between Tg(1) and Tg(2) is about 20C and depends slightly on the blend composition.
The accuracy of the midpoint temperature Tg(2) is much better than that of the
measurement of the onset temperature Tg(1). Reproducibility in Tg(2) results in 0.5C
error, but the error in the measurement of Tg(1) is much larger, resulting in larger
than 2C. We suppose that Tg(1) is a value near to “real” Tg determined on dilatometry.
The Vogel temperature T obtained from the above method depends strongly on the Tg
value used, although they used Tg as the midpoint of DSC curves. Therefore, we like to
utilize Ts as an adjustable parameter.
PMMA-7/SAN-27 and PMMA-60/SAN-18 blends were also therrmorheologically
simple. The reasons may be due to dynamic symmetry of PMMA and SAN, as already
explained by Pathak et al. [55]. The temperature dependence of aT for these blends is
essentially same as that for the PMMA-7/SAN-18 blends, although we do not show them
here.
Figure 31. Compositional dependence of the reference temperature Ts for PMMA-7/SAN-18

blend [61].
Figure 32. Compositional dependence of the onset temperature Tg(1) and the midpoint temperature
Tg(2)of the glass transition for PMMA-7/SAN-18 blend [61].
A miscible poly(-caprolactone) (PCL)/SAN blend were also studied [62]. The

frequency dependence curves of G and G at various temperatures were superposed onto
the master curves at 160C. This blend has very large Tg difference of about 150C
between the component polymers. The temperature dependence of the shift factor aT
became very steep with an increase in SAN content, when the reference temperature was
160C for all samples. However, when the reference temperatures Ts were chosen to be
156 Yuji Aoki
140, 92, 61, 25, and 0C for PCL/SAN = 0/100, 20/80, 40/60, 60/40, and 80/20,
respectively (for pure PCL we could not precisely determine the Ts from the temperature
dependence of aT), the aT could be expressed by the WLF equation Eq. (8). It is said that
the WLF equation is applicable to temperatures that are approximately 100C above the
Tg. However, our previous studies show that the WLF equation is applicable to wider
temperature range for SAN and miscible SAN blends. Provided that data are taken over a
sufficiently wide temperature range, this procedure determines Ts to within 1C.
Figure 33 shows the compositional dependence of Ts for the PCL/SAN blend. Chiu
and Smith [23] reported that the compositional dependence of the Tg for PCL/SAN
blends represents the Gordon-Taylor equation [75]:
Tsblend = TsPCL + K(TsSAN  Tsblend)wSAN/wPCL (13)
where wPCL and wSAN are the weight fractions, TsPCL and TsSAN are the Ts of pure SAN and
PCL, respectively, and Tsblend is the blend Ts. For a miscible system, plot of Tsblend versus
(TsSAN  Tsblend)wSAN/wPCL will yield a straight line with a slope of the empirical constant K
and an ordinate intercept of TsPCL. The data were fitted by a least-square regression
technique, and the slope of the straight line was found to be about 0.6. The Ts of PCL
obtained as the ordinate intercept was 250 K. The broken line in Figure 33 indicates the
Gordon-Taylor equation in the case of K = 0.6 and TsPCL =
250 K. Chiu and Smith’s data [23] are essentially equivalent to our results. PCL (low Tg
component) is considered as a diluent for SAN (high Tg component).
Figure 33. Compositional dependence of the reference temperature Ts for PCL/SAN blend [62].
In this subsection, we discussed the shift factor of miscible SAN blends. It was found
that the aT of the blends are well described by the WLF equation irrespective of the large
difference of Tg values, when the reference temperature was chosen about (Tg + 50C).
This result demonstrates that the free volume of miscible SAN blends is determined by
only the Tg of the component (co)polymers.
For miscible polymer blends, several mixing rules for the glass transition temperature
Tg, for example, Fox equation, Gordon-Tailor equation, and so on, have been proposed.
We found that the compositional dependence of Tg is a linear function of mole fraction
for SMA/SAN and SMI/SAN blends. Molecular weight of MI monomer is very different
from that of MA monomer; MI = 173 and MA = 98. But, the Tg values of the blends are
decided by mole fraction of MA or MI. This means that Tg for SMA/SAN and SMI/SAN
blends is a function of only the number of MA or MI. On the other hand, the
compositional dependence of Tg is a linear function of weight fraction for PMMA/SAN
blend. For PCL/SAN blend, that is expressed by Gordon-Tailor equation. The mixing
rule of Ts for SAN blends may depend on the blended polymer type.
Lodge and McLeish [54] reported that in a miscible blend the local environment of a
monomer of type A will, on average, be rich in A compared to the bulk composition 
due to chain connectivity. As a result, the local dynamics of the two chains may exhibit
difference on temperature and overall composition. As low Tg polymers are generally
more flexible, the associated “self-concentration” s is larger, the local dynamics in the
mixture may be quite similar to the pure material. An effective Tgeff for each component
can be estimated from the composition-dependent bulk average Tg as Tg(eff). This
approach can reconcile the failure of time-temperature superposition. We would like to
emphasize that time-temperature superposition holds well for miscible SAN blends.
Then, we can guess that “self-concentration” effect is weak for miscible SAN blends.
4.2. Zero-Shear Viscosity
In subsection 3.2., we showed the master curves of G and G for SMA-15/SAN-15

blend and the component copolymers (Figure 23) [30]. At high frequencies, the G and
G exhibit very weak  dependence and are independent of blend ratio. At low
frequencies, the blend shows the terminal tails (G  2 and G  ). Therefore, the
characteristic parameters in the terminal zone, zero-shear viscosity 0 and steady state
compliance Je0 can be calculated from these data.
Figure 34 shows 0 and Je0 as function of blend composition for SMA-15/SAN-15
blends. At a constant temperature of 200C, 0 increases with increasing SMA content,
because of the increase in Tg of the blend. However, when f = 0.0672, the 0 values are
158 Yuji Aoki
almost constant independent of blend composition. Je0 also shows hardly any dependence
on blend composition.
The zero-shear viscosity of polymer melts at equal free volume fraction is given by
0 = 0(Mc)( Mw/Mc)3.43.8 Mw  Mc (14)
where Mc is a critical molecular weight. If 0(Mc) of SAN-15 is equal to that of SMA-15

and 0  Mw3.5, we obtain the next relation,
0SAN-15/0SMA-15 = (MwSAN-15/McSAN-15)/(MwSMA-15/McSMA-15)3.5 (15)
Putting the experimental values into Eq. (15), we obtain McSMA-15/McSAN-15 = 1.3. McSMA-15
is larger than McSAN-15, as the C-C bonds of MA unit in SMA chains cannot rotate freely.
Above relation indicates that the number of entanglement couplings in SMA-15 chains is
almost identical to that in SAN-15 chains, although Mw of SAN-15 is lower than that of
SMA-15.
The number of entanglement couplings in SMA-15 chains is nearly equal to that in
SAN-15 chains. Thus, the number of entanglement coupling per chain in the blends is
almost equal to that in the component copolymers. This may possibly explain why both
0 and Je0 are independent of the blend composition.
Figure 34. Compositional dependence of zero-shear viscosity 0 and steady state compliance Je0 for
SMA-15/SAN-15 blend [30].
Next, we studied the dynamic viscoelastic properties of miscible blends of SMI and
three kinds of SAN with the same AN content but different molecular weights [60]. The
composition and molecular weight used in this study are already shown in Table 5.
Figure 35 shows the master curves of G and G for P-30/SAN-3 blend and the
component copolymers, respectively, as a function of aT. At low , G is proportional
to aT for each sample. Therefore, the zero-shear viscosity 0 of the blends was
calculated from these data.
Figure 35. Master curves of storage and loss moduli, G (top panel) and G (bottom panel) at 200C for
P-30/AS-3 blend [60].
Figure 36 shows the compositional dependence of 0 at 200, 220, and 245C. For P-
30/AS-2 blend, 0 increases concavely with increasing P-30 content. For P-30/AS-3
blend, the 0 value has a minimum at about 20 wt % P-30 and then increases with
increasing P-30 content. For P-30/AS-4 blend, the compositional dependence of 0 is
very small because the difference between 0 of P-30 and that of SAN-4 is small.
160 Yuji Aoki
Figure 36. Compositional dependence of zero-shear viscosity 0 at constant temperatures of 200, 220,
and 245C for P-30/AS-2, AS-3, and AS-4 blends [60].
Figure 37. Compositional dependence of zero-shear viscosity 0 at a constant free volume fraction of
f = 0.00672 for P-30/AS-2, AS-3, and AS-4 blends [60].
The existence of the minima found for P-30/AS-3 blend seems to support Wu’s
explanation that dissimilar chains are less likely to entangle with each other than similar
chains [45-47]. However, as shown in Figure 36, compositional dependences of 0 are
varied by measuring temperature. This fact implies that 0 behavior is affected by a free
volume in the blends, as shown for SMA-15/SAN-15 blend. Compositional dependence
of 0 at a constant free volume fraction of 0.00672 is plotted in Figure 37. The 0
decreases convexly with increasing of P-30 content. This result reminds us of the
viscoelastic properties of the blends of homologous polymers with different molecular
weight.
The molecular weight dependence of 0 of SAN at a constant free volume fraction

was expressed by 0 = K(Mw/Me)3.8. Using the entanglement molecular weight Me of
SAN, Me(SAN) = 11,000, evaluated by Wu [46, 47], the number of entanglement,
nw = Mw/Me, was estimated to be 7.1, 10.7, and 15.7 for AS-2, AS-3, and AS-4,
respectively. Assuming that constant K of SMI is the same as that of SAN, nw of P-30 is
5.3. When the entanglement molecular weights in the blends are identical with those in
the component copolymers, nw in the blends, nwblend, can be expressed by
nwblend = w1nw1 + w2 nw2 (16)
where wj and nwj, respectively, are the weight fraction and the number of entanglement of
pure component j. Using Eq. (16), 0 was calculated and plotted by solid lines in Figure
37. The calculated curves are in excellent agreement with the experimental values. These
results suggest that the dissimilar polymers relax independently and entanglement
molecular weights in the blend are the same as those in each component copolymer. Wu
[45-47] calculated the Mws for the blends by Mw = w1Mw1 + w2Mw2. But, we believe that
the above relation is meaningless and Eq. (16) should be used instead of the additivity of
the molecular weights.
The zero-shear viscosity of miscible SMI/SAN blends exhibits a behavior similar to
that of a blend consisting of homologous polymers with different molecular weights. It is
emphasized that the rheological behavior of miscible blends should be compared at an
iso-free volume condition. Then, it is found that the entanglement molecular weights in
the miscible blends are identical with those in the component copolymers.
Figure 38 shows the master curves of G and G for PMMA-7/SAN-18 blends and
the component polymers. The zero-shear viscosity 0 can be calculated from these data at
low . The terminal relaxation times of PMMA-7 and SAN-18 are too close to resolve
cleanly.
Figure 39(a), (b), respectively, shows the compositional dependence of 0 at 200C
and f = 0.049. Both 0 versus PMMA content curves deviate positively from additivity
for both polymers. These experimental results quantitatively agree with the results by Wu
[47] and by Han and Kim [49, 50]. Wu proposed that 0 of miscible blends is expressed
by Eq. (17) [46].
log 0 = 1 log 01 + 2 log 02 + 12 log  (17)
where 0j and j are the zero-shear viscosity and the volume fraction of pure component j,
and  is a constant for a blend system. Wu [47] reported that the zero-shear viscosity of
PMMA/SAN blends at 180C showed a positive deviation from linearity, and he obtained
162 Yuji Aoki
log  = 0 comparing the 0 at the same molecular weight. He stated that log  = 0 is not
expected to be generally applicable to other miscible blends. Our data exhibited positive
deviation from linearity, which are different from Wu’s data, and fit well to log  = 0.3,
as shown in Figure 39(b). We can interpret our data as follows.
for PMMA-7/SAN-18 blend [59].
We proposed that the rheological behavior of miscible blends should be compared at

an iso-free volume condition. When the entanglement molecular weight Me in a blend is
identical with that in the component polymer, the number of entanglements nwblend can be
expressed by Eq. (16). Using this equation, 0 values of PMMA-7/SAN-18 were
calculated. The calculated values are very similar to the solid line (log  = 0.3) in Figure
39(b) and are in good agreement with experimental ones. These findings are convincing
evidence that the zero-shear viscosity of PMMA/SAN blends is also similar to that of a
blend consisting of homologous polymers with different molecular weight.
Figure 39. Compositional dependence of zero-shear viscosity 0 at 200C (top panel) and at a constant
free volume fraction of f = 0.049 (bottom panel) for PMMA-7/SAN-18 blend [61].
Parts (a) and (b) of Figure 40 show the master curves of G and G for the PCL/SAN
blends as a function of . At low frequencies, the blend shows the terminal tails (G  2
and G  ). The zero-shear viscosity 0 was calculated from these data.
164 Yuji Aoki
for PCL/SAN blend [62].
Figure 41(a) shows the compositional dependence of 0 at constant T. The 0 values

increase concavely upward with an increase in weight fraction of SAN, as found by Han
and Yang [76]. However, compositional dependences of 0 were found to vary with
temperature. This fact implies that the 0 behavior is affected by a free volume within the
blends. Figure 41(b) shows the compositional dependence of 0 at a constant free volume
fraction of f = 0.049 (T = Ts), where the 0 of pure PCL was estimated from Ts = 250 K.
0 increases concavely downward with an increase in weight fraction of SAN.
Figure 41. Compositional dependence of zero-shear viscosity 0 at 160C (top panel) and at a constant
free volume fraction of f = 0.049 (bottom panel) for PCL/SAN blend [62].
Using Eq. (16) and 0 = const.(nwblend)3.8, where nwPCL = 3.3 and nwSAN = 7.1, the 0
values of the PCL/SAN blends were calculated according to the method of previous
studies. The broken line in Figure 41(b) indicates the calculated 0 values. It is found that
the calculated values are in good agreement with experimental values. When compared at
an iso-free volume condition, the compositional dependence curve is the same as that of a
blend consisting of two homologous polymers with different molecular weights. Then, it
is concluded that the entanglement coupling of the PCL/SAN blends are identical to those
in the component polymers.
It was found that the zero-shear viscosity change of miscible PCL/SAN blends at
constant temperature consists of two terms: the free volume change and the change of
number of entanglement couplings. The free volume change can be estimated from the
temperature dependence of the shift factors aT. Because the time-temperature
superposition principle is applicable over the entire investigated temperature range for the
166 Yuji Aoki
PCL/SAN blends, the temperature dependence of aT could be expressed by the WLF

equation. The change of number of entanglement couplings can be estimated from the
compositional dependence of the zero-shear viscosity at an iso-free volume condition. It
is concluded that the relaxation behavior of the PCL/SAN blends is similar to that of a
blend consisting of homologous polymers. It is emphasized that the zero-shear viscosity
of miscible blends should be compared at an iso-free volume condition to discuss the
change in number of entanglement couplings.
4.3. Plateau Modulus
Lastly, we discuss the plateau modulus GN0 of miscible SAN blends [61]. Figures 42
and 43, respectively, show the master curves of G and tan  for PMMA-7/SAN-18 and
PMMA-60/SAN-18 at 200C. GN0 was determined as the storage modulus G at the
frequency where tan  is at a minimum. Parts (a) and (b) of Figure 44 show the plateau
modulus versus volume fraction 1 of PMMA for PMMA-7/SAN-18 (a) and PMMA-
60/SAN-18 (b) blends. The GN0 shows a linear function of 1 for PMMA-7/SAN-18 and
PMMA-7/SAN-27 blends (not shown here) but shows a negative deviation from linearity
for PMMA-60/SAN-18 blends.
Table 7. Entanglement molecular weight Me and number of entanglement

ne for PMMA-7, PMMA-60, SAN-18, and SAN-27
polymer Mw/103 Me/103 ne

PMMA-7 72 7.98 9.0
PMMA-60 593 6.84 87
SAN-18 116 12.8 9.1
SAN-27 84 11.5 7.3
Wu [45-47] proposed that GN0 of miscible blends is expressed by Eq. (18),
GN0 = 1 GN10 + 2 GN20 + 12(e  1)(GN10 + GN20) (18)
where j is the volume fraction and GNj10 the plateau modulus of component j, assuming
binary interaction contacts with constant probability of entanglement for each type of
contact. The e is a constant for a given system, given by
e = [(12)1/2/Me12]/[(1/2)(1/Me1 + 2/Me2)] (19)

8 5
PMMA-7/SAN-18, 200oC
6 3
4 1
2 -1
-3 0 3 6
log (aT/rad s-1 )
Figure 42. Master curves of storage modulus G and loss tangent tan  at 200C for PMMA-7/SAN-18
blend [61].
8 5
PMMA-60/SAN-18, 200oC
6 3
4 
2 -1
-3 0 3 6
log (aT/rad s-1 )
Figure 43. Master curves of storage modulus G and loss tangent tan  at 200C for PMMA-60/SAN
blend [61].
The Me12 is the entanglement molecular weight of a hypothetical pure component

having a density (12)1/2 and the same entanglement probability as that between
dissimilar chains in the real blend. Figure 44(a) indicate that e = 1.0 for PMMA-7/SAN-
18 and PMMA-7/SAN-27 blends (not shown). It means that entanglement molecular
weights in these blends are same as those in each component polymer independent of 
168 Yuji Aoki
parameter. On the other hand, e of PMMA-60/SAN-18 blend is in agreement with Wu’s

result of 0.775, as shown in Figure 44(b).
Figure 44. Compositional dependence of plateau modulus G N0 of PMMA-7/SAN-18 (top panel) and
PMMA-60/SAN-18 (bottom panel) blends [61].
Let’s consider the reason for the discrepancy of e for PMMA/SAN blends. First, we
calculated the entanglement molecular weight Me between entanglement points for SAN-
18, SAN-27, PMMA-7, and PMMA-60 by
Me = RT/GN0 (20)
where  is the density and R the gas constant. Table 7 shows Me and the number of
entanglement ne for SAN-18, SAN-27, PMMA-7, and PMMA-60. Next, we compare the
master curves shown in Figures 42 and 43. Master curves of G and tan  of PMMA-
7/SAN-18 blends are very similar for each blend composition, because the number of
entanglement coupling of each polymer is almost the same. The tan  values at minimum
were 0.360.41, almost independent of the composition. On the other hand, tan  curves
of PMMA-60/SAN-18 blends are skewed, and the tan  values at minimum were
0.130.41, dependent on the composition. Such behavior is similar to that of a blend
consisting of homologous polymers with different molecular weight. An extreme case is
a blend of two homologous polymers with narrow molecular weight distribution. In this
case, we cannot determine the plateau modulus from the storage modulus at the
frequency where tan  is a minimum. For PMMA-60/SAN-18 blends, the difference in
the number of entanglement coupling is very large; about 10 times. Therefore, the above
method to determine the plateau modulus would not be applicable to PMMA-60/SAN-18
blends. Wu and Beckerbauer [77] reported that GN0 determined by the above method is
invariant with respect to molecular weight and molecular weight distribution and is equal
to the GN0 value determined by the integration of the G curve. However, Mw/Mn of their
samples is within about 4, although the ratio of ne for PMMA-60/SAN-18 blends is about
10. Our data indicate that the plateau modulus of PMMA/SAN blends is a linear function
of blend composition.
Figure 45 shows the master curves of G and G for MSAN-32/SAN-33 blends and
the component copolymers at 170C. GN0 was determined as the storage modulus G at
the frequency where tan  is at a minimum. Figure 46 shows the plateau modulus versus
weight fraction of MSAN-32. The GN0 shows a linear function of the fraction, as same as
for PMMA-7/SAN-18 blend which molecular weight asymmetry of the component
copolymers is not large. Therefore, this result supports that GN0 values, in other words,
entanglement molecular weights in miscible SAN blends, are identical with those in the
component (co)polymers.
Wu [45-47] proposed that specific interchain interactions, responsible for miscibility
of dissimilar polymers, tend locally to align the chain segments for association and thus
stiffen the chains, resulting in reduced entanglement between dissimilar chains. This local
ordering can be regarded as a kind of nematic interaction. Watanabe et al. [78] and Doi
and Watanabe [79] formulated the dynamics of Rouse chains that interact with each other
by nematic potential. They concluded that the rheological properties on the long time
scale are very insensitive to the nematic interaction, an interaction between the segments
of the chains and matrix. This conclusion indicates that interchain specific interactions
responsible for the miscibility do not influence the rheological properties on the long time
scale such as the plateau modulus and zero-shear viscosity of miscible blends and
supports our experimental results.
170 Yuji Aoki
Figure 45. Master curves of storage and loss moduli, G and G at 170C for MSAN-32/SAN-33 blends
having various compositions.
Figure 46. Compositional dependence of plateau modulus GN0 of MSAN-32/SAN-33 blend.

Recently, Watanabe et al. [80, 81] studied the mixing rule of the plateau modulus GN0
for cis-polyisoprene (PI)/poly(p-tert-butylestyrene) (PtBS) blends using in details. They
found that GN0is well described by a linear mixing rule of the entanglement length with
the weighting factor being equated to the number fraction of Kuhn segments of the
components, not by the reciprocal mixing rule utilizing the component volume fraction of
entanglement as the weighting factor. Their finding is different from our result. We can
speculate that this difference is due to the difference of polymer blend species; 1) the G
and G data of PI/PtBS blends do no obey the time-temperature superposition, although
those of our SAN blends obey, 2) for the PI/PtBS blends, the component chains have a
huge chemical difference, although our SAN blends have a similar chemical difference.
We determined GN0 from G at the angular frequency  where tan  has the minimum
value. This method is sufficiently accurate only when the molecular weight distribution is
not broad, as Liu et al. [82] reported evaluation of different methods for the determination
of GN0. Our SAN blend samples used here have similar molecular weight and its
distribution. We believe that our conclusion of the mixing rule is correct for our SAN
blends explained here.
CONCLUSION
Miscibility and viscoelastic properties of SAN blends are reported. SAN is miscible
with PMMA, PCL, MSAN, SMA, and SMI. The miscibility of SAN is mainly due to the
presence of unfavorable interaction between styrene and acrylonitrile comprising the
copolymer. The  parameters for monomer-monomer interaction were evaluated from the
miscibility map. Two points are emphasized here; one is that the volume fraction must be
used as copolymer composition, and another is that some blend systems, for example
MSAN/SAN blends, exhibit a very slow phase separation because of low mobility.
Miscible blends exhibit the terminal tails at low , although immiscible blends do not
show the terminal tails and have a long time relaxation due to an inhomogeneous
structure. Shift factors are well described by the WLF equation with two constant
coefficients 8.86 and 101.6 for all the SAN blends studied here. Compositional
dependences of 0 and GN0 for miscible blends are discussed in relation to entanglement
couplings. The 0 values should be compared at an iso-free volume condition, and then,
the entanglement molecular weight in the blends are identical with those in the
component polymers. The GN0 is also identical with those in the component polymers and
varies linearly with the blend composition. It is concluded that miscible SAN blends
behave as a mixture of homologous polymers, although the glass transition temperature
Tg varies with the chemical species.
172 Yuji Aoki
REFERENCES
[1] Paul DR, Newman S (1978) Polymer Blends; Academic, New York.
[2] Olabisi O, Robeson LE, Shaw MT (1979) Polymer-Polymer Miscibility; Academic,
New York.
[3] Scott RL (1949) J. Chem. Phys. 17:279.
[4] Kambour RP, Blender JT, Bopp RC (1983) Macromolecules 16:753.
[5] ten Brinke G, Karasz FE, MacKnight WJ (1983) Macromolecules 16:1827.
[6] Paul DR, Barlow JW (1984) Polymer 25:487.
[7] Shiomi T, Karasz FE, MacKnight WJ (1986) Macromolecules 19:2274.
[8] Stein VDJ, Jung RH, Illers KH, Hendus H (1974) Angew. Makromol. Chem. 36:89.
[9] Naito K, Johnson GE, Allara DL, Kwei TK (1978) Macromolecules 11:1260.
[10] McBrierty VJ, Douglass DC, Kwei TK (1978) Macromolecules 11:1265.
[11] Jelenic J, Kirste RG, Schmitt BJ, Schmitt-Strecker S (1979) Makromol. Chem.
180:2057.
[12] Amrani F, Hung JM, Morawetz H (1980) Macromolecules 13:649.
[13] Kressler J, Kammer HW, Klostermann K (1986) Polym. Bull. 15:113.
[14] Piglowski J, Kressler J. Kammer HW (1986) Polym. Bull. 16:493.
[15] Suess M, Kressler J, Kammer HW (1987) Polymer 28:957.
[16] Fowler ME, Barlow JW, Paul DR (1987) Polymer 28:1177.
[17] Cowie JMG, Lath D (1988) Makromol. Chem., Macromol. Symp. 16:103.
[18] Nishimoto M, Keskkula H, Paul DR (1989) Polymer 30:1279.
[19] Frezzotti D, Ravanetti GP (1994) J. Therm. Anal. 41:1237.
[20] Higashida N, Kressler J, Inoue T (1995) Polymer 36:2761.
[21] Rudolf B, Ougizawa T, Shimomai K, Higashida N, Inoue T (1996) Macromol.
Symp. 112:31.
[22] Hsu W-N (1999) J. Appl. Polym. Sci. 74:2894.
[23] Chiu S, Smith TG (1984) J. Appl. Polym. Sci. 29:1781.
[24] Chiu S, Smith TG (1984) J. Appl. Polym. Sci. 29:1794.
[25] Janarthanan V, Kressler J, Karasz FE, MacKnight WJ (1993) J. Polym. Sci., Part B,
Polym Phys. 31:1013.
[26] Li W, Prud’homme RE (1994) Polymer 35:3260.
[27] Svobona P, Kressler J, Ougizawa T, Inoue T, Ozutsumi K (1997) Macromolecules
30:1973.
[28] Hall WJ, Kruse RL, Mendelson RA, Trementozzi QA (1983) ACS Symp. Ser.
229:49.
[29] Kato T, Kobayashi N, Takahashi A (1983) Kobunshi Ronbunshu 40:661. (in
Japanese)
[30] Aoki Y (1984) Polym. J. 16:431.
[31] Mendelson RA (1985) J. Polym. Sci., Polym. Phys. Ed. 23:1975.
[32] Aoki Y (1988) Macromolecules 21:1277.

[33] Kressler J, Kammer HW, Schmidt-Naake G, Herzog K (1988) Polymer 29:686.
[34] Maccarinelli D, Schreiber HP, Scholtens BJR, Reid VMC (1995) J. Adhesion
55:43.
[35] Heinen W, Wenzel CB, Rosenmoller CH, Mulder FM, Boender GJ, Lugtenburg J,
de Groot HJM, van Duin M, Klumperman B (1998) Macromolecules 31:7404.
[36] Maruta J, Ougizawa T, Inoue T (1988) Polymer 29:2056.
[37] Kim JH, Barlow JW, Paul DR (1989) J. Polym. Sci., Polym. Phys. Ed. 27:223.
[38] Jar P-YB, Kuboki T, Takahashi K, Shinmura T (1998) J. Appl. Polym. Sci. 69:513.
[39] Ikawa K, Tanaka S, Ueno K (1991) Polym. Prepr. Jpn 40(9):3497.
[40] Cowie JMG, Elexpuru EM, McEwen IJ (1992) Polymer 33:1993.
[41] Gan PP, Paul DR, Padwa AR (1994) Polymer 35:3351.
[43] Utracki LA (1990) Polymer Alloys and Blends, Thermodynamics and Rheology,
Carl Hanser Verlag.
[44] Prest WM, Porter RS (1972) J. Polym. Sci., A-2 10:1639.
[45] Wu S (1987) J. Polym. Sci., Polym. Phys. Ed. 25:557.
[46] Wu S (1987) J. Polym. Sci., Polym. Phys. Ed. 25:2511.
[47] Wu S (1987) Polymer 28:1144.
[48] Saito H, Takahashi M, Inoue T (1988) J. Polym. Sci., Polym. Phys. Ed. 26:1761.
[49] Han CD, Kim JK (1989) Macromolecules 22:1914.
[50] Han CD, Kim JK (1989) Macromolecules 22:4249.
[51] Doi M, Edwards SF (1986) The Theory of Polymer Dynamics, Claredon Press,
Oxford.
[52] Haley JC, Lodge TP (2004) J. Rheol. 48:463.
[53] Tsenoglou C (1989) New Trends Phys. Phys. Chem. Polym. [Proc. Int. Symp.] 375.
[54] Lodge TP, McLeish TCB (2000) Macromolecules 33:5278.
[55] Pathak JA, Colby RH, Kamath SY, Kumar SK, Stadler R (1998) Macromolecules
31:8988.
[56] Zetsche A, Fisher EW (1994) Acta Polym. 45:168.
[57] Kumar SK, Coby RH, Anastasiadis SH, Fytas G (1996) J. Chem. Phys. 105:3777.
[58] Aoki Y, unpublished work.
[59] Aoki Y (2008) J. Appl. Polym. Sci. 108:2206.
[61] Aoki Y, Tanaka T (1999) Macromolecules 32:8560.
[62] Aoki Y, Arendt O (2001) J. Appl. Polym. Sci. 82:2037.
[63] Hildebrand JH, Scott RL (1950) The Solubility of Nonrelectrolytes, 3rd ed,
Reinhold, New York.
[64] Hoy KL (1970) J. Paint Technol. 42:76.
[65] Nishimoto T, Keskkula H, Paul DR (1991) Polymer 32:272.
174 Yuji Aoki
[66] Kwak KD, Okada M, Nose T (1991) Polymer 32:864.

[67] Williams ML, Landel RF, Ferry JD (1955) J. Am. Chem. Soc. 77:4701.
[68] Doolittle AK (1951) J. Appl. Phys. 22:1471.
[69] Han CD, Chuang HK (1985) J. Appl. Polym. Sci. 30:4431.
[70] Lyngaae-Jorgensen J, Sondegaard K (1987) Polym. Eng. Sci. 27:344.
[71] Lyngaae-Jorgensen J, Sondegaard K (1987) Polym. Eng. Sci. 27:351.
[72] Madbouly SM, Ougizawa T, Inoue T (1999) Macromolecules 32:5631.
[73] Hong Z, Shaw MT, Weiss RA (2000) Polymer 41:5895.
[74] Plazek DJ (1982) J. Polym. Sci., Polym. Phys. Ed. 20:729.
[75] Gordon M, Taylor JS (1952) J. Appl. Chem. 2:493.
[76] Han CD, Yang H-H (1987) J. Appl. Polym. Sci. 33:1199.
[77] Wu S, Beckerbauer R (1992) Polymer 33:509.
[78] Watanabe H, Kotaka T, Tirrel M (1991) Macromolecules 24:201.
[79] Doi M, Watanabe H (1991) Macromolecules 24:740.
[80] Watanabe H, Chen Q, Kawasaki Y, Matsumiya Y, Inoue T, Urakawa O (2011)
Macromolecules 44:1570.
[81] Matsumiya Y, Rakkapao N, Watanabe H (2015) Macromolecules 48:7889.
[82] Liu C, He J van Ruymbeke E, Keunings R, Bailly C (2006) Polymer 47:4461.
Chapter 6
RHEOLOGY OF HONEY
Clara Tovar1, María Shantel Rodríguez-Flores2,

Olga Escuredo2 and María del Carmen Seijo2
1
Department of Applied Physics, University of Vigo,
Ourense, Spain
2
Department of Plant Biology and Soil Sciences,
University of Vigo, Ourense, Spain
ABSTRACT
Honey is a complex fluid with viscoelastic properties, predominantly similar to those

of liquids, but with certain elastic components. Properties of honey are influenced by
composition, temperature, sugar crystals and principally moisture content. In most
research works, honey is reported to be a Newtonian liquid, but the presence of natural
colloids contributed to its time dependant behaviour (thixotropy). For the present study,
two kinds of measurements have been made. First, small amplitude oscillatory shear
(SAOS) tests, to determine the viscoelastic moduli in the linear viscoelastic (LVE) range,
when storage (G’) and loss (G’’) moduli are independent on the shear strain/stress
applied. SAOS tests allow evaluate the time-dependence of G’ and G’’, providing the
mechanical properties of honey at rest. Second, steady-state shear tests, which give the
relationship between the shear viscosity (η) and the shear rate (). The obtained flow
curves showed honey changes in response to continuous deformation. In the samples
analysed, the power law model indicated a weak shear-thinning behaviour. In addition,
the time-dependence of flow behaviour using a step test with three intervals has been
assayed: reference, high-shear interval and regeneration interval to test thixotropy. These
experimental proofs have been completed applying upward ramp, hold time, and
downward ramp in order to determine the hysteresis area and verify the possible
thixotropy.
Moreover, the viscoelastic behaviour has been analysed by transient tests which
provide creep and creep-recovery curves, using stress values (0) (during creep stage), in
the LVE range. These tests serve to compare the viscosity values obtained from flow
176 Clara Tovar, María Shantal Rodríguez-Flores, Olga Escuredo et al.
curves, and at the same time to obtain a percentage of elasticity after removing load at the
rest phase.
Temperature (T) has the most significant role on the viscosity of honey, its
temperature dependence was examined applying a constant shear rate (50 s -1) and an
upward T-ramp, at a constant heating rate (1°C/min) from 20 to 50°C. In all studied
samples, the viscosity values followed an Arrhenius relationship and the activation
energy (Ea) was obtained.
Keywords: Rheology, honey, water content, thixotropy, botanical origin
INTRODUCTION
Honey is the sweet natural substance produced by the honeybee A. mellifera. It has a
vegetable origin, being both the nectar and the honeydew the main raw materials
originated from the phloem of the superior plants. This origin determines their
composition in sugars, as well as other nutrients such as amino acids, electrolytes, plant
hormones, phenolic acids, vitamins, organic acids, etc. (Crane, 1975; Le Hir et al., 2008;
Won-Kyong et al., 2010). These sweetened secretions are mixed with saliva and other
glandular secretions from honeybees that contain small amounts of lipids, vitamins and
enzymatic proteins that will play a key role in the transformations that the nectar or the
honeydew will suffer in order to become honey. The main components are fructose and
glucose, however, other components such as oligosaccharides, polysaccharides, enzymes,
amino acids, organic acids, minerals, flavonoids, vitamins, pollen grains, waxes and other
phytochemicals were present. The complex chemical composition, color, aroma and
flavour of honey depend mainly on its botanical origin and also other external factors as
geographical region and weather conditions (Sanz et al., 2004; Guler et al., 2007;
Alvarez-Suárez et al., 2010; Escuredo et al., 2013; Nayik et al., 2014).
The botanical origin of the honey is the main criterion used for its typification, since
it determines its organoleptic characteristics and composition. According to this, honey
can be classified as multifloral or unifloral honey if the raw material origin is nectar.
Multifloral honey comes from several plants but any of them predominating enough to
condition their sensorial characteristics. Instead unifloral honeys, have specific sensorial
characteristics, and come mainly from the nectar of specific plant species. On the other
hand, the honeydew honey comes from secretions produced in plants outside the
nectaries, generally due to the action of aphids or other pests, but also from excretions of
green parts of plants, as is the case of Quercus pyrenaica acorns (Rodríguez-Flores et al.,
2015).
To classify a unifloral honey, sensorial, physical-chemical, and pollen requirements
are necessary. In spite of this, there is no honey that is 100% unifloral, since for the
production of honey, there are several plants that participate nuancing its characteristics.
Rheology of Honey 177
Therefore, the honeys with the same botanical origin, range in their both pollen and
physicochemical composition, and consequently may have different physical behaviors.
This can be reflected in the honey analysis from the rheological point of view.
Rheology is an useful tool to analyse the physical and structural response of honeys
(Smanalieva & Senge, 2009; Wei et al., 2010). For fluid samples, the principal
rheological properties are the flow curves even in complex systems with several
continuous phases (Torres et al., 2015), and the viscoelastic response if there is
crystallization. Both of them are influenced by sugar and moisture contents, temperature,
as well as the presence of crystals and natural colloids (Bhandari et al., 1999; Oroian
et al., 2011; Dobre et al., 2012).
The rheological study of honey provides information for a quality control and for
making a regulation that still does not exist, so, different honey samples produced in
Northwest Spain were studied regarding their rheological response and the relationship
with physicochemical and pollen profile.
METHODS
Honey Samples
To study the rheological behaviour of honey, six samples produced in Northwest

Spain were selected. The samples were produced in different geographical and botanical
areas and were extracted from the combs by centrifugation. All of them are unheated
honeys. The standard physicochemical and palynological analysis were performed in
order to know their quality, the main components and the botanical origin.
Palynological Analysis
The principal plants used for honeybees to obtain honeys were studied through the
pollen spectra samples. 10 g of honey were dissolved in bidistilled water and centrifuged
at 4500 rpm (3373 g) for 10 min. The obtained sediment was redissolved and centrifuged
for another 5 min. The final volume of the sediment was used to prepare the slide. The
pollen spectra were determined by counting and identifying a minimum of 800 pollen
grains using a Nikon Optiphot II microscope (400x and 1000x, when needed). The results
were expressed as the percentage of the pollen type over the total pollen counted and
identified constituting the pollen spectra of the sample.
Water Content, pH, Conductivity Electrical and Color
Water content was measured with a digital refractometer (ABBE URA-2WAJ-325;

Auxilab S. L., Navarra, Spain). The refractive index values, at 20ºC, were converted to
moisture contents. Measurements of pH and electrical conductivity (EC) were performed
directly on a solution of honey (5 g) dissolved in bi-distilled water (25 mL) with a
Multiparameter pH, mV, COND, TDS, Temperature digital benchtop instrument with
microprocessor, resolution ± 0.1/0.01 pH/0.01 µS. Finally, the results were expressed
as mS cm-1. Color measurements were made with the digital instrument from Hanna
(HANNA C 221 Honey Color Analyzer, Rhode Island, USA). Results were expressed
using the Pfund scale in mm.
Sugar Composition Analysis
For analyzing sugars composition, an ion Dionex ICS-3000 chromatography system

(Sunnyvale, California, EEUU) was used. The separation of sugars was performed with a
Carbo Pac PA1 column (3 X 250 mm) and a pulsed amperometric detector (PAD) with a
gradient of two mobile phases (A and B). Phase A involved ultrapure water, while phase
B involved 200 mM NaOH (HPLC grade, Merck). The sugar content of honey solutions
(10 mg L-1) was calculated using the calibration curves of standard solution for each pure
sugar (Sigma, Aldrich). The identified sugars were fructose, glucose. Sugars were
quantified by comparison of the peak area obtained with those of standard sugars. The
CHROMELEON Chromatography Management System was used for acquisition of all
the chromatograms. The results for each sugar were expressed as g Kg-1 of honey.
The experimentation of each analysis was performed in duplicate.
Rheological Measurements
Both small amplitude oscillation stress (SAOS) and steady shear rheological
measurements were performed using a controlled rate rheometer RS600, (Haake Thermo
Electron, karlsruhe, GmbH) with computer control. A cone/ plate (C35/2) sensor system
was used, with a diameter of 35 mm and 2° angle, a gap between the cone and plate was
fixed at 104 µm. Samples were allowed to rest for 10 min before analysis to ensure both
thermal and mechanical equilibrium at the time of measurement. Samples were covered
with a trap-solvent system to avoid evaporation during measurements. Temperature was
kept at 20 ± 0.1°C controlled by a Peltier element in plate.
Rotational Tests: Flow Curves
To analyse the flow behaviour under isothermal conditions, samples were subjected
to controlled shear rate () with a test program with three intervals: 1) upward  ramp in
600s from  = 0.1 to 500 s-1; 2) peak hold, high  = 500 s-1 for 300 s; 3) downward  ramp
in 600 s from  = 500 to 0.1 s-1. The possible time dependence was assessed using a step
test with three intervals: 1) Reference interval  = 5 s-1 for 60 s to obtain the “viscosity at
rest” and to be compared with the values in the third test interval; 2) high-shear interval
500 s-1 for 30 s in order to break the internal structure; 3) regeneration interval:  = 5 s-1
for 600 s to evaluate structural regeneration and consequently the possible thixotropic
response (Mezger, 2006).
Temperature Dependence of Flow Behaviour
The viscosity was examined as a function of temperature using a rotational tests at

constant  = 50 s-1. A thermal profile was programmed at constant upward ramp of
1°C/min from 20 to 50°C to compare the thermal dependence of apparent viscosity (η) of
different honeys in terms of the relative fluidity considering the Arrhenius relation
(Figura & Teixeira, 2007).
Creep and Recovery Tests
An instantaneous small stress σ0 (1.5 Pa) was applied for 60 s to each sample in the
creep tests and the resulting change in strain over time was monitored. When the stress
was released, some recovery could be observed for 300s. The creep-recovery results are
described in terms of the shear compliance (J) function (Eq. 1) to compare the
deformability of the different honeys on larger time scales (Steffe, 1996).
 (t )
J (t )  (1)
0
Dynamic Mechanical Analysis
Using this kind of oscillatory tests, both frequency (1 Hz) and stress amplitude
(1.5 Pa) were kept constant. In addition, the measuring temperature was constant (20ºC).
The complex viscosity (η*), storage (elastic) modulus (G’), loss (viscous) modulus (G’’),
and loss factor (tan δ), were determined with increasing time up to 1800 s. Data were
obtained in such a way as to ensure that the resulting deformation maintains the internal
structure of honeys.
Statistical Analyses
At least five independent batches were tested for each experiment and data are
presented as averages. Statistical analysis was carried out using Microsoft Excel
software. Trends were considered significant when means of compared sets differed at
p < 0.05 (Student’s t-test).
Physicochemical Properties of Honeys, Colour and Botanical Origin
Honey samples were produced mainly from plants as Castanea, Eucalyptus, Erica, or
Rubus being the principal pollen in the pollen spectra of samples (Table 1). The
physicochemical parameters exhibited some differences. Water content varied between
16.8% (sample B) and 19.2% (sample D). Colour corresponds to light amber (68 mm) of
sample F to dark brown (samples A and B). Also pH varied from 3.7 to 4.1, electrical
conductivity are considered medium high being from 548 µS/cm of samples D to 850
µS/cm of sample A. Regarding sugar content, the main reduced sugars were fructose and
glucose constituting more than 60% of sugar in honeys. Sample A had the sum of
fructose and glucose 74%, while the samples E and F had 67%. Fructose was almost the
best represented sugar varying from 47.5% (sample A) to 39.9% (samples E and F).
Glucose content varied from 26.3% (sample D) to 29.6% (sample B). Relations between
reduced sugars and water content are also shown in Table 1 as they are important to the
crystallization of honey (Escuredo et al., 2014).
Steady Shear Flow
Rheological testing to determine flow curves show how the material changes in
response to continuous deformation, i.e., how it may be processed or shaped into a useful
product. Flow behaviour provides some physical properties of a complex fluid as a
function of shear rate or frequency of deformation (Larson, 1999). It is combined with
flow resistance (internal frictions) among molecules and particles, for that reason the test
was made under isothermal conditions, in the present study at 20°C.
Figure 1 shows the flow curves of different honeys, in terms of the fluid response
when a steady shearing flow was achieved in both upward and downward ramps during a
suitable time interval. So, samples exert a force on a surface in the direction parallel to
the flow:  has been reported vs . Power law (eq. 2) was found to be the best fit for the
flow curves based on the estimated experimental errors (Table 2).
   ·  
·
(2)
 
where  is the consistency coefficient (Pa sn), and n is the flow behaviour index
(dimensionless). In general, all samples showed a weak shear-thinning flow (n < 1). So,
the flow curves exhibited a convex profile in which the tangential slope is decreasing
with increasing shearing (Larson, 1999). For that reason, the viscosity (η) slightly
decreased during the upward ramp and increased during the downward ramp, so, it is
called apparent viscosity (Steffe, 1996).
In general,  varied in both up (u) and down (d) ramps (u > d), this difference
indicates that both (up and down) curves do not overlap suggesting a thixotropic
response. The highest values of  and the greater differences between u and d were
found in sample B (Table 2).  reflects the strength of the attractive forces among
permanent dipoles principally in sugar and water molecules and could be related with the
lowest water and higher glucose contents, giving the highest G/W ratio, and a high Erica
pollen content. Moreover, the significant difference between u and d (A and B)
suggests a partial loss of the intermolecular conectivity in a complex fluid by shearing
(Larson, 1999). This result is consistent with the lowest values of both n u and nd
exponents (Table 2), indicating some weakening in binding forces by shearing (Figura &
Teixeira, 2007) and a more shear-sensitive fluid.
Conversely, for sample D with the highest water content (lowest G/W ratio), it was
found similar and lower values of both u and d together with the highest nu and nd
exponents (Table 2). This result suggests the main role of the water molecules in
structure of honeys appearing as a “coupling agent,” that induces some intermolecular
adhesion by hydrogen bonding between sugars, producing a reversible flow (D).
Flow behavior index (n) of the other samples were ranged between 0.91 and 0.97
which were in close agreement with literature values for honey samples (Ahmed et al.,
2007).
Table 1. Palynological and physicochemical characterization of honey samples
Sample Water Color pH EC Glucose Fructose G/W F/G Botanical origin

(%) (mm) (µS/cm) (%) (%)
A 17.0 116 4.0 850 26.5 47.5 1.6 1.8 Castanea (35.4%), Rubus (35.4%), Erica (18.9%), Other (10.3%)
B 16.8 141 4.3 745 29.6 40.8 1.8 1.4 Erica (43.5%), Castanea (18.2%), Cytisus (16.9%), Rubus (5.1%),
Other (16.3%)
C 18.0 90 4.0 773 28.3 41.4 1.6 1.5 Castanea (72.8%), Rubus (17.5%), Erica (5.0%), Other (3.8%)
D 19.2 107 3.7 548 26.3 45.0 1.4 1.7 Erica (49.2%), Castanea (24.3%), Rubus (8.8%),
Eucalyptus (4.7%), Other (4.6%)
E 18.0 78 4.1 667 27.1 39.9 1.5 1.5 Eucalyptus (65.0%), Castanea (26.1%), Rubus (3.6%),
Other (14.9%)
F 17.6 68 3.9 676 27.1 39.9 1.5 1.5 Eucalyptus (66.1%), Castanea (6.2%), Cytisus (6.2%),
Rubus (3.3%), Other (18.2%)
EC: electrical conductivity, G/W: glucose/water, F/G: fructose/glucose, G+F: glucose + fructose.
C down C up F down F up E down E up
D down D up A down A up B down B up

10000
9000
8000
7000
Shear stress (Pa)
6000
5000
4000
3000
2000
1000
0
0 100 200 300 400 500
Shear rate (1/s)
Figure 1. Comparison of flow curves for honeys A, B, C, D, E, F with up-down ramps at 20°C.
Table 2. Power law parameters Ku, nu and Kd, nd from equation 2 for different
honeys at 20°C
Sample Ku (Pa·sn) nu r2Eq.1 Kd (Pa·sn) nd r2Eq.2

A 39.0 ± 3.1Ad 0.901 ± 0.007Ab 0.995 32.71 ± 2.9 Bd 0.929 ± 0.007Ba 0.995
B 47.6 ± 4.8Ae 0.887 ± 0.006Aa 0.996 36.4 ± 4.4Bd 0.926 ± 0.007Ba 0.995
C 19.5 ± 1.6Ab 0.942 ± 0.003Ac 0.999 18.4 ± 1.5Ab 0.954 ± 0.004Bb 0.998
D 10.3 ± 1.0Aa 0.970 ± 0.002Ae 0.999 11.0 ± 1.1Aa 0.964 ± 0.003Bc 0.999
E 18.9 ± 1.3Ab 0.957 ± 0.004Ad 0.998 21.1 ± 1.6Ab 0.934 ± 0.003Ba 0.999
F 31.8 ± 1.6Ac 0.914 ± 0.006Ab 0.996 28.2 ± 1.4Bc 0.933 ± 0.005Ba 0.997
Values are given as mean ± standard deviation.
a-e: different small letters in the same column indicate significant differences (p < 0.05) among honeys.
A-B: different capital letters indicate significant differences (p < 0.05) for up-down ramps for each sample.
Time Dependence of Shear Flow (Thixotropy)
Thixotropy is a physical phenomenon which shows the time dependence of the fluid
flow. It is associated to the reduction of the structural strength by shearing, and further
structural regeneration when load was removed during the last period of rest (Sahin &
Sumnu, 2006). Naturally, in many fluids when are subjected to a constant, , η decreases
(become thinner) with time, but, flow could induce some changes in fluid structure that
are erased only after hours of quiescence. So, thixotropy is a kind of viscoelasticity with
long relaxation time (Larson, 1999). For that reason, the thixotropic flow should be
analysed considering both decomposition and regeneration stages over the fluid structure.
In the present study a three-step test has been applied, so, it is possible to determine
the proportion between the instantaneous η value at the end of the third step (period of
rest, η at 390 s), and the corresponding η value at the end of the first step (η at 60 s)
(Figure 2). So, for samples with “true thixotropic flow,” when the strong shear rate
stopped, the original structure is restored and the regeneration of η was practically 100%.
If there is some irreversible structural breakdown, the time dependence of flow behaviour
is called “apparent thixotropy” (Figura & Teixeira, 2007), which is equivalent to partially
reversible thixotropy (Sahin & Sumnu, 2006). This behaviour is asscociated to some
reduction in the percentage of the molecular regeneration. Specifically, honeys A, C, D,
E, F exhibited “true thixotropy” because the thixotropic behaviour was associated with
almost complete structural regeneration: 96, 98, 100, 100, and 98% respectively. The
only exception was the sample B, whose percentage (81%) was significantly lower than
the other samples, indicating a partially reversible thixotropic flow (apparent thixotropy),
that could be related with its higher G/W ratio which involves a higher level of
intermolecular organization (sugar-sugar and sugar-water) molecules which is naturally
more dificult to be completely restaured after the last rest step.
D F E B A C
45
40
35
Apparent viscosity (Pa·s)
30
25
20
15
10
0
0 50 100 150 200 250 300 350 400
t (s)
Figure 2. Comparison of viscosity curves in different thixotropic responses for samples A, B, C, D, E, F

at 20°C.
Creep Tests
When a constant load is applied to honeys, the resulting strain is measured as a

function of time. There is a possibility of some recovery of the material when the stress is
released and honey tries to return to its original state. These data are often used to
characterize the long-term behaviour of samples under controlled shear stress conditions
(Steffe, 1996).
Figure 3 shows different creep curves in terms of the creep compliance J(t) (eq. 1)
data for different samples up to 30 s. Sample D was not shown due to the practicaly nule
values of J(t). Creep J(t) data showed a constant slope, indicating that steady state was
reached in a few seconds. The slope is proportional to the inverse of shear viscosity (1/η)
as was indicated in equation 5, starting from equation 3:
d (t ) 1
 ·dt (3)
0 
J
1 t
 dJ   · dt
0
0
(4)
1
J (t )  ·t (5)

η is one of the most important rheological parameters that provides a material

constant at fixed temperature, is defined as the resistance to flow of a fluid (Sahin &
Sumnu, 2006). Samples B and A had the highest values of η (Table 3), which were
similar than the respective consistency coefficients at 20°C (analysed above). Conversely,
honeys C and E had similar and the lowest η values (Table 3), which is consistent with
the highest moisture content (Table 1). At the initial times (7-10 s) of the creep step, the
steady state was verified when constant shear rate values were reached. Moreover, the
software of the Rheometer (Rheowin 4.6) provides the instantaneous values of η(t) (Pa·s)
together with J(t) (1/Pa) data. So, multiplying both magnitudes may be obtained a
characteristic time of material (τ) using the equation 6:
   (t )·J (t ) (6)
When the steady state was reached, τ reached a constant value for all samples
(τ = 30 s), appart from sample B (τ = 34 s). Higher values of τ could be related to the
presence of a large fraction of bigger molecules (colloids, proteins and or sugars) in
honey B (Witczak et al., 2011; Al-Mahasneh et al., 2014).
E B A F C
1,80
1,44
J (1/Pa) 1,08
0,72
0,36
0,00
0 50 100 150 200 250 300
t (s)
Figure 3. Creep and creep-recovery curves for samples A, B, C, E, F at 20°C.
Table 3. Values of shear viscosity from equation 5 for different honeys at 20°C
Sample η (Pa·s)
A 36.5 ± 3.6c
B 43.1 ± 3.0c
C 17.1 ± 1.7a
D nf
E 19.8 ± 1.0a
F 25.5 ± 1.5b
Values are given as mean ± standard deviation.
a-c: different small letters indicate significant differences (p < 0.05) among honeys.
nf: not found.
When removing the stress, a creep recovery (reformation curve) was recorded up to
300 s (Figure 3). This rest phase produced a permanent deformation value Jr(t), which is
characteristic of the liquid-like behaviour, that is larger in a less viscous material (C).
Conversely, for the most viscous sample (B) Jr(t) reached the lowest values (Figure 3) in
line with the highest flow resistance. In general creep-recovery tests characterized honeys
as purely viscous materials showing practically no recovery during reformation stage.
There were two exceptions (samples E and B), specifically for sample E, with a high
content of Eucalyptus and low content of Castanea in relation to sample F, was
calculated a percentage of elasticity (14%) according to equation 7:
J 30  J 300
Elasticity (%)  ( )·100 (7)
J 30
A similar (qualitative) effect was found in sample B which had a lower partial
recovery (6%).
Time-Dependent Behaviour at Constant Dynamic Mechanical Analysis

(DMA)
DMA tests were used to examine the time dependence (at constant frequency) of
viscoelastic materials without chemical changes during the measurement at small stress.
So, it has been made the DMA test applying 0 = 1.5 Pa during t = 1800 s at ω = 6.28
rad/s, to simulate the viscoelastic behaviour at rest, under the more stable physical
conditions to preserve the internal order in honeys. Data were summarized in Table 4
where are shown the storage (elastic) modulus (G’), loss (viscous) modulus (G’’),
complex viscosity (η*) and loss factor (tan ). These data indicate that for all samples
G’’ > G’ during the all time, so honeys exhibited a liquid-like response, thus tan  >> 1.
Moreover, η* had similar values (Table 4) than those of shear viscosity (creep tests)
(Table 3).
Table 4. Viscoelastic parameters obtained from dynamic mechanical analysis

(DMA) at 6.28 rad·s-1 and 20°C
Sample G' (Pa) G'' (Pa) η* (Pa·s) tan δ

A 3.58 ± 0.06c 224.31 ± 0.78e 35.71 ± 0.12e 63 ± 2b
B 6.05 ± 0.25d 248.62 ± 0.59f 39.59 ± 0.09f 41 ± 2a
D 1.50 ± 0.11b 65.80 ± 0.32a 10.48 ± 0.05a 44 ± 3a
C 1.41 ± 0.12ab 109.05 ± 0.40b 17.36 ± 0.06b 78 ± 7c
E 1.31 ± 0.07a 113.05 ± 0.15c 17.99 ± 0.02c 87 ± 5c
F 3.76 ± 0.21c 164.34 ± 0.14d 26.17 ± 0.07d 44 ± 3a
Values are given as mean ± standard (average of instantaneous values during testing time.
a-f: different small letters in the same column indicate significant differences (p < 0.05) among honeys.
By other hand, G’’ represents the viscous behaviour of any material, it gives the
deformation energy spent in the viscoelastic measurements (Mezger, 2006). It was
observed that sample B had the highest η* and G’’ values, showing a greater irreversible
deformation. That means that the relative motion between particles produces higher
frictional forces, thus, more deformation energy is dissipated. This fact is consistent with
the apparent thixotropy analysed above, that indicated only a partial molecular
regeneration for a rest period after strong shearing. Conversely, honey D had the lowest
η* and G’’ values (Table 4) indicating a greater fluidity, consequence of the highest
moisture content that act as lubricating system, improving the mobility between particles.
Temperature Dependence of the Rheological Properties
In the present section, shear viscosity is examined as a function of temperature under

a rotational motion at fixed shear rate, applying an upward ramp from 20 to 50°C at
constant heating rate (1°C/min). Honeys were thermo-rheologically simple materials, i.e.,
materials which maintained the physical state during heating (Mezger, 2006). Thermal
response produced a notable decrease of η with increasing temperature according to
Arrhenius relation (equation 8).
EA
  A·e RT
(8)
where: Ea = activation energy (J·mol-1); R = gas constant (8.314 J·mol-1·K-1); T =

temperature (K). A = constant (Pa·s).
95
82
69
Ea (kJ/mol)
56
43
30
A B C D E F
Honeys
Figure 4. Values of flow activation energy from Arrhenius curve (eq. 8) for samples A, B, C, D, E, F.
Molecules in honey are closely spaced with cohesive forces, principally by dipolar
interactions and hydrogen bonds (Pardkarand & Irudayaraj, 2001). These forces decrease
with increasing temperature and flow becomes freer (Sahin & Sumnu, 2006).
Specifically, Ea is a characteristic material parameter which gives the energy needed for
the molecules to exceed the potential barrier to move against the internal flow resistance
(Mezger, 2006). It can be noticed that sample C and D had the lowest values of Ea
(Figure 4) which could be related with the highest moisture content, more intense for
sample D, which possess the lowest G/W ratio thus, this sample had the highest density
of the intermolecular hydrogen bonds in water-water interactions whose lifetime
exponentially diminishes with increasing temperature (Juszczak & Fortuna, 2006). This
fact could explain the lowest potential barrier of sample B comparing with the other
honeys (Figure 4).
CONCLUSION
The proportion of glucose/water in honeys of different botanical origins was a

relevant property to justify their rheological responses. When glucose/water ratio is high,
consistency and the degree of conectivity are also high, becoming samples more shear-
dependent with lower recovery ability. These less hydrated honeys show an apparent
thixotropy in line with the high complex viscosity and loss modulus, which reflect a
greater energy dissipated in a irreversible deformation. Water in honeys has an important
plasticizer role that improves the recovery ability of the structure after shearing at
constant temperature. Thermal profiles of more hydrated honeys give lower activation
energies due to their greater fluidity.
REFERENCES
Ahmed, J., Prabhu, S. T., Raghavan, G. S. V. & Ngadi, M. (2007). Physico-chemical,

rheological, calorimetric and dielectric behavior of selected Indian Honey. Journal of
Food Engineering, 79, 1207-1213.
Al‐Mahasneh, M. A., Rababah, T. M., Amer, M., & Al‐Omoush, M. (2014). Modeling
physical and rheological behavior of minimally processed wild flowers honey.
Journal of Food Processing and Preservation, 38(1), 21-30.
Alvarez-Suarez, J. M., Tulipani, S., Romandini, S., Bertoli, E., & Battino, M. (2010).
Contribution of honey in nutrition and human health: a review. Mediterranean
Journal of Nutrition and Metabolism, 3(1), 15-23.
Bhandari, B., D'Arcy, B. & Kelly, C. (1999). Rheology and crystallization kinetics of
honey: present status. International Journal of Food Properties, 2(3), 217-226.
Crane, E. (1975). Honey. A comprehensive survey. International Bee Research
Association (IBRA). Ed. Heinemann. London. U.K., 608 pp.
Dobre, I., Georgescu, L. A., Alexe, P., Escuredo, O., & Seijo, M. C. (2012). Rheological
behavior of different honey types from Romania. Food Research International,
49(1), 126-132.
Escuredo, O., Míguez, M., Fernández-González, M., & Seijo, M. C. (2013). Nutritional
value and antioxidant activity of honeys produced in a European Atlantic area. Food
Chemistry, 138(2), 851-856.
Escuredo, O., Dobre, I., Fernández-González, M. & Seijo, M. C. (2014). Contribution of
botanical origin and sugar composition of honeys on the crystallization phenomenon.
Food Chemistry, 149, 84-90.
Figura, L. O. & Teixeira, A. A. (2007). Food Physics. Springer, New York.
Guler, A., Bakan, A., Nisbet, C. & Yavuz, O. (2007). Determination of important
biochemical properties of honey to discriminate pure and adulterated honey with
sucrose (Saccharum officinarum L.) syrup. Food Chemistry, 105(3), 1119-1125.
Juszczak, L., & Fortuna, T. (2006). Rheology of selected Polish honeys. Journal of Food
Engineering, 75(1), 43-49.
Larson, R. G. (1999). The structure and rheology of complex fluids. Oxford University
Press, New York, 232 pp.
Le Hir, R., Beneteau, J., Bellini, C., Vilaine, F. & Dinant, S. (2008). Gene expression
profiling: keys for investigating phloem functions. Trends in Plant Science, 13(6),
273-280.
Mezger, T. G. (2006). The rheology handbook. (2nd ed.). Vicentz Network, Hannover,
29 pp.
Nayik, G. A., Bhat, F. M., Muzaffar, K., Gull, A., Wani, S. A., Shah, T. R. & Nanda, V.
(2014). Honey: Its History and Religious Significance: A Review. Universal Journal
of Pharmacy, 3(01), 5-8.
Oroian, M., Amariei, S., Escriche, I., & Gutt, G. (2013). Rheological aspects of Spanish
honeys. Food and Bioprocess Technology, 6(1), 228-241.
Paradkar, M. M., & Irudayaraj, J. (2002). Discrimination and classification of beet and
cane inverts in honey by FT-Raman spectroscopy. Food Chemistry, 76(2), 231-239.
Rodríguez-Flores, M. S., Escuredo, O. & Seijo, M. C. (2015). Assessment of
physicochemical and antioxidant characteristics of Quercus pyrenaica honeydew
honeys. Food chemistry, 166, 101-106.
Sahin, S. & Sumnu, S. G. (2006). Physical properties of foods. Springer. New York.
47 pp.
Sanz, M. L., González, M., De Lorenzo, C., Sanz, J. & Martínez‐ Castro, I. (2004).
Carbohydrate composition and physicochemical properties of artisanal honeys from
Madrid (Spain): occurrence of Echium sp honey. Journal of the Science of Food and
Agriculture, 84(12), 1577-1584.
Smanalieva, J., & Senge, B. (2009). Analytical and rheological investigations into
selected unifloral German honey. European Food Research and Technology, 229(1),
107-113.
Steffe, J. F. (1996). Rheological methods in food process engineering. (2nd ed.). Freeman
Press, East Lansing. 295 pp.
Torres, M. D., Hallmark, B. & Wilson, D. I. (2015). Determination of the shear and
extensional rheology of bubbly liquids with a shear-thinning continuous phase.
Rheologica Acta, 54(6), 461-478.
Wei, Z., Wang, J., & Wang, Y. (2010). Classification of unifloral honeys from different
floral origins and geographical origins based on rheometer. Journal of Food
Engineering, 96(3), 469-479.
Witczak, M., Juszczak, L., Gałkowska, D. (2011). Non-Newtonian behaviour of heather
honey. Journal of Food Engineering 104, 532-537.
Won-Kyong, C., Chen, X. Y., Rim, Y., Chu, H., Kim, S., Kim, S. W., Park, Z. Y. & Kim,
J. Y. (2010). Proteome study of the phloem sap of pumpkin using multidimensional
protein identification technology. Journal of Plant Physiology, 167(10), 771-778.
Chapter 7
UNUSUAL NONLINEAR RHEOLOGICAL BEHAVIOR

OF BRANCHED POLYMERS: MULTIPLE STRESS
OVERSHOOTS IN SHEAR STARTUP -
EXPERIMENTS AND THEORY
Federico Sebastiano Grasso1, Salvatore Coppola1,

Fabio Bacchelli1,*, Giovanni Ianniruberto2 and Giuseppe Marrucci2
1
Versalis SpA, Ravenna, Italy
2
Dipartimento di Ingegneria chimica, dei Materiali e della Produzione industriale,
Università di Napoli “Federico II”, Napoli, Italy
ABSTRACT
Rheological properties of polymers for elastomer applications are strongly dependent

on their molecular structure. Most commercially available elastomers are entangled
polydisperse, branched polymers with an unknown molecular structure. From the
experimental point of view, rheological measurements can be carried out in the linear and
nonlinear viscoelastic regimes. Both types of measurements can be very fruitful in
characterizing polymeric materials and, if combined, they can disclose peculiar physical
properties providing a deep insight into molecular structure. That is because the two
regimes deal with very different aspects of molecular dynamics. As in the literature the
linear viscoelastic regime is widely explored, in this chapter we choose to focus on
nonlinear rheological measurements for a number of elastomers, i.e., Polybutadiene
(PBR), styrene-butadiene copolymers (SBR) and nitrile butadiene (NBR) rubber, which
differ significantly in branching content.
*
Corresponding Author Email: fabio.bacchelli@versalis.eni.com.
194 Federico Sebastiano Grasso, Salvatore Coppola, Fabio Bacchelli et al.
For such materials, single point data (e.g., Melt Flow Index or Mooney Torque) are
unable to describe the rheological features of polymers with complex branching pattern.
A significant example is the case of emulsion-polymerized elastomers, characterized by
broad Molar Mass Distribution and branched architecture, both depending on the
conversion degree and temperature attained in the reactors.
Furthermore, polymers with a highly branched molecular structure can display
multiple overshoots during shear startup experiments at constant shear rate. Such
overshoots can be related to complex molecular dynamics, such as chain orientation,
chain stretching, and branchpoint withdrawal, the extent of the latter being strongly
dependent on the previous deformation history.
It is well known that shear startup data collected by means of a Mooney-type
viscometer can display stress overshoots, particularly prominent for randomly branched
materials. Noticeably, similar results were recently obtained also for model branched
polymers tested with a strain controlled rheometer equipped with a cone-partitioned plate
geometry. The extremely slow attainment of a steady state, and the high degree of
nonlinearity are related to a significant branching content of the high molecular weight
fraction of the material. Although scientifically intriguing, as well as relevant for polymer
processing, such phenomena have been seldom investigated in the rheological literature.
Only recently, this phenomenology has been the subject of careful experimental works, as
well as of attempts of theoretical understanding.
The aim of this work is to present some original results together with a short review
of some of the landmark experimental results and theoretical understanding gained in the
last years on this topic.
Keywords: shear stress growth, stress overshoots, recovery, elastomers, branching, linear
and nonlinear viscoelasiticity, Mooney
INTRODUCTION
It is well known that the dynamics of entangled polymeric liquids is very complex,
both at equilibrium and under non-equilibrium conditions, due to the fact that the
polymer molecules are constrained within “tubes” formed by the surrounding chains [Doi
and Edwards, 1986]. Several molecular mechanisms are believed to be responsible for
such a complex behavior: reptation out of the tube, tube-length fluctuations, tube renewal
through constraint release, chain stretch due to friction, etc. One aspect of such
complexity shows up in the start-up of shearing flows in the nonlinear range, where
overshoots in the shear stress σ and, often, in the first normal stress difference are
observed, before achieving the steady state. A single maximum in shear stress for
monodisperse linear polymers was first rationalized by Doi and Edwards [Doi and
Edwards, 1986] as due to tube orientation, occurring when the shear rate exceeds the
reciprocal of the disengagement time to exit the tube (for linear chains, approximately the
reptation time). In the basic Doi-Edwards theory, where chain stretch is ignored, the
maximum in shear stress occurs at a fixed value of the deformation (≈2.3) for all values
Unusual Nonlinear Rheological Behavior of Branched Polymers 195
of the shear rate (exceeding the threshold value), and no maximum in the first normal
stress difference is predicted.
On the other hand, chain stretch becomes important when the shear rate approaches
the reciprocal of the retraction time of the chain in the tube, expected to be of the order of
the Rouse time [Doi and Edwards, 1986]. It was then shown [Pearson et al., 1991] that
chain stretch displaces the location of the maximum to larger values of strain, and that a
maximum appears in the first normal stress difference as well. It is also readily shown
that, by further increasing the shear stress, the time at which these maxima occur
asymptotically approaches the retraction time, a concept emphasized by Osaki and
coworkers [Inoue et al., 2002].
For bimodal homopolymer blends, two maxima were sometimes observed [Liu et al.,
1983; Osaki et al., 2000]. Osaki and coworkers studied in detail this unusual effect by
properly tailoring the relative amount and the relaxation times of the two components of
the blend. The double maximum is then understood as follows. The maximum occurring
at a lower strain is contributed by the majority component of the blend having the smaller
molar mass, and it typically falls close to the expected value of 2.3, being due to tube
orientation. The contribution to stress of the minority component of the blend remains
negligible at that strain value. The stress due to the minority component having a larger
molar mass prevails at higher strains, when the longer chains become stretched, thus
providing a second maximum at a value of time close to the retraction time of the long
chains. Recently Graham [2002] and Islam [2006] also predicted the double maximum
for binary blends by using tube models.
Previous observations report the onset of multiple overshoots in monomodal
polymers developed for the rubber industry. When a high polymer is sheared under
standard conditions in a Mooney plastometer [Mooney, 1934, 1936], the strain on the
shearing mechanism is measured and can be plotted against time. It has been the custom
for the rubber industry to depend on the Mooney value, i.e., the measured torque, after
four minutes shearing (ASTM D412). ASTM D3346 reports a test method using the
shearing disk viscometer to obtain an indication of the processability of emulsion styrene-
butadiene rubbers (eSBR) through the analysis of the first and second overshoot. This
procedure was used to separate those polymers that are easy to process from those that
are difficult to process within a group of polymers of the same type. Such empirical test
was found to be suitable for comparing a production lot with a standard of known
processability characteristics. The difference between Mooney torques at two specified
shearing times will rank emulsion SBR (eSBR) polymers that differ appreciably in this
property according to their processability.
It was suggested by Fensom [1956], after testing various rubbers in an early work,
that the entire Mooney viscometer curve provides crucial information about the polymer
architecture and molar mass distribution. It was observed that, in addition to the primary
peak, a secondary rise may also appear in certain Mooney curves. Cross-linking in
structure, as indicated by high gel content, seemed to be correlated with secondary rise in
shearing force shown during a long shear start-up test.
According to early works, it is worth noting that the shear-startup runs of LDPE
exhibit the classical overshoot predicted in Doi–Edwards theory without showing double
peaks, although a hint of a second peak is perhaps present in the branched high-density-
polyethylene (HDPE) reported in recent data of Das et al. [2014], and in the LDPE data
reported by Bacchelli [2008] who used a Mooney-type viscometer for reaching large
values of shear rates and deformations. Using Mooney-type viscometers, double peaks in
shear-startup data of commercial elastomers were also observed by Einhorn [1963], by
Nakajima and Harrell [1982] and by Bacchelli [2008] himself. More recently, Kasehagen
and Macosko [1998] reported unusual nonlinear effect in shear stress growth experiments
for branched model polybutadienes using a traditional cone-plate geometry.
Multiple stress-overshoots were further interpreted in terms of the withdrawal idea of
the pompom model by Marrucci et al. [2008]. Very recently multiple overshoots were
again found in nearly monodisperse comblike branched polystyrene (PS) melts (and
similarly interpreted) by Snijkers et al. [2013], who used a state-of-the-art, cone-
partitioned-plate rheometer.
New data of shear startup on branched styrene-butadiene random (SBR) copolymers
have been recently reported by Coppola et al. [2014], where the novelty consists in
repeating the startup run after different rest times at zero stress. In the cited work by
Coppola et al., the aim was one of exploring the “damage” introduced by the first run, as
well as the subsequent recovery, if any, upon waiting increasingly long times. Differently
from a linear sample, branched melts showed multiple peaks during the first run, as
previously reported by Bacchelli [2008]. The repeated runs show an intriguing novel
feature with respect to the case of linear polymers, namely, the first peak goes up
initially, instead of down. The second peak goes down and seemingly recovers only after
an extremely long time, longer than the largest relaxation time practically accessible to
linear viscoelasticity, the latter not reaching the terminal behaviour. All such features of
nonlinear viscoelasticity of highly branched polymers were then interpreted by Coppola
et al. [2014] using a simple theory inspired by the well-known pompom model.
In the present chapter, we review the intriguing experimental results on SBR and the
theoretical approach to their interpretation provided by Coppola et al. [2014], and we also
enlarge the horizon of this investigation presenting original, previously unpublished, data
on other classes of polymers, like Polybutadiene (PBR) and emulsion NBR.
METHODS
Linear Viscoelasticity
For characterizing SBR samples, the experimental protocol was based on

creep/recovery measurements and frequency sweeps at different temperatures, as better
detailed in the following.
The samples were press-molded at 90°C for 20 minutes to obtain disk shaped
specimens of thickness 1-2 mm. The specimens were then allowed to cool down to room
temperature for about 20 minutes. Temperature and time for the molding stage were
chosen in such a way as to avoid thermo-oxidative reactions in the material. Small
Amplitude Oscillatory Shear (SAOS) measurements were conducted on a strain-
controlled rheometer (ARES 2KFRTN1) equipped with parallel plate fixtures (8 mm – 25
mm diameter), with a temperature control tolerance better than 0.1°C (achieved via a
convection oven and a liquid nitrogen Dewar), under nitrogen environment to reduce the
risk of degradation. The SAOS tests were performed in the temperature range -40°C to
110°C. Strain sweep experiments were conducted to determine the linear viscoelastic
region for the frequency sweeps.
Creep/recovery tests were performed at 110°C with a MCR501 instrument (Anton
Paar-Physica), equipped with parallel plate fixtures (25 mm diameter). Temperature was
controlled by means of a nitrogen-fed convection oven (CTD600). Each SBR sample has
been tested at three different stress levels in order to be sure that the results were in the
linear range. The stress levels were in the range 50 - 300 Pa.
Before performing the final frequency sweep or creep/recovery test, several
frequency sweep tests in the range 0.01 - 100 rad/s were performed to ensure
achievement of thermal equilibrium and relaxation of residual stresses due to squeezing
of the sample between the parallel plates. The same tests also assured absence of
progressive degradation.
In the specific case of eNBR samples, due to the high normal force showed by the
polymer sample when squeezing it, the linear viscoelastic characterization had to be
performed by means of a closed chamber dynamic-mechanical rheometer, RPA 2000
(Alpha Technologies) equipped with a serrated biconical chamber. The temperature
chosen for testing the NBR samples was 130°C and the applied strain was 10%. A delay
time of 40 minutes at 130°C before starting the test was adopted, in order to allow the
relaxation of stresses produced by the squeezing flow.
NonLinear Viscoelasticity
Overcoming well-known experimental problems encountered with conventional

cone-plate geometry, such as edge fracture and wall slip, has been the subject, in the last
decades, of a number of works that pointed out the need of developing alternative
geometries to the conventional cone-plate one. In the last years, it has been widely
accepted the idea of adopting a cone-partitioned-plate geometry to measure the torque far
away from the sample rim, in order to delay the negative effect of edge fracture. The
cone-partitioned-plate geometry was first developed by Pollet [1955] and has been
subsequently implemented and developed in different versions by a number of authors
[Meissner et al., 1989; Schweizer, 2002, 2003; Ravindranath and Wang, 2008; Schweizer
et al., 2008; Li and Wang, 2010; Snijkers and Vlassopoulos, 2011]. Such geometry can
be designed and built to be mounted on commercial rheometers with high torque and
normal force sensitivity.
However, for the rubbery high-molar-mass polymers studied here, instruments like
ARES could not be used because of the limited normal force range, insufficient to
squeeze down the sample.
The problem of measuring the nonlinear shear response of high-molar-mass rubbery
polymers had been faced long ago by Mooney [1934], who developed a shearing disk
apparatus (later called Mooney viscometer), in which a flat grooved disk rotates in the
rubber in a flat serrated cavity, so that there is a rotational shear flow between opposing
flat surfaces, a sort of cylindrical Couette flow at the periphery, and some complex flow
pattern in between those regions. The main advantage of such closed chamber rheometer
is that it allows performing, in a very simple way, rheological experiments up to large
shear rates without the problem of edge fracture. It was indeed strongly felt at that time
the need to reach sufficiently large shear rates, at least of order 1–10 s-1, in order to get
shear rates similar to those encountered during rubber processing operations like mixing,
milling, and extruding. Multispeed Mooney viscometers equipped with the standard disk
rotor were investigated by several authors. A first accurate description of torque-time
curves is reported by Nakajima and Harrell [1982], who pointed out the problem
associated with the shear rate distribution within the material in the cavity between the
flat rotor and the stator.
Many years before, in order to obtain a more uniform flow field, Piper and Scott
[1945] had already developed a modification of the Mooney viscometer in which a
biconical rotor with rounded edge was adopted. Similar biconical rotors were used in a
subsequent work by Galvin and Whorlow [1975] on the shear stress growth of HDPE,
LDPE, ethylene-vinyl acetate copolymer, while Best and Rosen [1968] adopted a
biconical rotor with sharp edge in order to determine the flow curves of a natural rubber,
and of a SBR, in a wide range of temperatures and shear rates. More recently, Montes et
al. [1988] also used a variable speed Mooney viscometer equipped with a biconical rotor,
emphasizing the role played by several operating variables (chamber pressure, grooves,
etc.) in avoiding slip at the wall.
All the nonlinear viscoelastic experimental results presented in this chapter have been
obtained with a variable speed Mooney viscometer MV2000 VS (Alpha Technologies
Inc.).
In order to elucidate the possible role of the uniformity of flow field, for SBR and
PBR, we show results obtained with a Mooney biconical rotor design, similar to that of
This choice is justified by the reduced interference for such design between the rotor edge
and the chamber wall with respect to rotors with rounded or flat edge.
It is important to underline that the observed phenomenology of interest in this work
(i.e., long stress growth transient, multiple overshoots) is not related to nonuniformity of
the flow field as it has been observed even when a uniform flow field is applied
[Bacchelli, 2008; Coppola et al., 2014]. The shear startup experiments can therefore
provide similar qualitative information even if a nonuniform shear flow field is applied,
like that occurring between two parallel plates rotating with respect to each other. For this
reason, the original results of shear startup on eNBR presented in this chapter have been
obtained using a grooved standard “large” flat rotor, designed according to ASTM
D1646. Even if a nonuniform shear flow field is obtained with this geometry, we adopted
it because the densely-grooved surface of the flat rotor allows avoiding issues about wall
slip instabilities that may occur with some non oil-extended, highly elastic samples like
eNBR.
For the latter material results obtained with the flat rotor, Nakajima & Harrell [1982]
approach has been adopted to convert the torque dial reading (ML) in shear stress
(expressed in Pascal units):
0.382  104  ML  3  n 
R 
 
 n 
 2 
  (3  n)hL 
n

4  1   
n
2 n

   
 2 R n 1 1   R   
n
   Re   
   
For sake of simplicity, based on LVE information, we adopted n=0.3 for all the NBR
samples. The specific choice of such value may slightly vary for each eNBR sample but
we decided to retain a single value for all samples as the main differences between the
experimental results is qualitative rather than quantitative.
MULTIPLE STRESS OVERSHOOTS DURING SHEAR STARTUP
In Figure 1 original shear startup tests are shown for three different classes of
materials: high cis polybutadiene (PBR), TDAE oil-extended random styrene-butadiene
copolymer obtained from an emulsion polymerization (OE eSBR) and emulsion
polymerized nitrilic rubber (eNBR) with 33% ACN content. Test temperature and test
time were set at 100°C and 180 mins, respectively. Samples were preliminarily
conditioned at the test temperature for 30 minutes in order to significantly reduce residual
stresses due to the squeezing. For eNBR, experimental conditions were slightly different
(test temperature = 130°C and 60 min delay time before test) because of eNBR higher
Mooney Torque in comparison with PBR and OE eSBR. All the tests were run on a
Variable Speed Mooney viscometer at standard rotational speed (2 rpm) and at 0.1 rpm
for comparison. That is because decreasing the rotor speed implies applying lower shear
rate values, and lower shear rates can show in most cases stronger discrepancies between
different samples.
For a given rotor speed, profiles of the curves have different shapes for each class of
materials. All the materials show a first major peak occurring within 2÷20 strain unit
range. At higher strain values, PBR curves progressively decrease, whereas both
emulsion-polymerized polymers show multiple peaks. In all cases, the transient response
is not “usual”, in the sense that there is not a single overshoot at a low strain followed by
a clear steady-state value.
At 2 rpm rotor speed, shear startup curves are very similar for both PBR and OE
eSBR in the region of the first major peak. At 2 rpm, eNBR shear startup curve is
different in the region of the first major peak. Not only in terms of peak shape, but also
because the first major peak occurs at a slightly higher strain.
When decreasing rotor speed down to 0.1 rpm, differences in the curves become
more prominent in the linear response region (i.e., at small strain values), confirming that
materials differ significantly in terms of molecular structures. At higher strain values,
eNBR shows a prominent second peak due to the presence of complex molecular
structures with significant branching content.
Interestingly, despite the prolonged tests, none of the considered materials reached
the final steady-state value. Figure 1 therefore demonstrates that an “unusual” stress
response characterized by very long transients and possibly multiple overshoots is a
universal behavior displayed by polymers possessing different chemistry. Such long
transients are typically displayed by polymers with significant Long Chain Branching
content and broad relaxation time spectrum, as better illustrated later in this chapter.
It is important to emphasize that similar, unusual rheological responses have been
observed not only for commercial polymers [Bacchelli, 2008] but also for well-defined
model polymers [Snijkers et al., 2013; Ianniruberto and Marrucci, 2014].
In Figure 2 we show original results of a more systematic investigation made on

PBR. Shear startup secondary peaks can be observed for PBR possessing a rheologically
significant long chain branching (LCB) content. It is interesting to note that the first peak
progressively shifts to the right as the shear rate increases while the second essentially
stays at a constant strain or even slightly shifts towards lower strains.
In Figure 3 we show the stress response to shear startup of different eNBR samples.
We considered materials with an acrylonitrile (ACN) content ranging from 28% to 39%.
Furthermore, we also considered a single material featuring two main chain populations
with different ACN values, but so that the average ACN content is 33%. Despite the
different acrylonitrile content, all the considered grades have similar Mooney Torque at
standard conditions, i.e., ML(1+4) at 2 rpm and T = 100°C.
Figure 1. Shear startup curves for three different material classes: high cis PBR, OE eSBR and eNBR.
All tests were carried out with standard Mooney flat rotor (large type). Two different rotor speed
were applied 0.1 and 2 rpm; we calculated an equivalent shear rate value at steady-state of 0.08 and
1.57 s-1, respectively.
The shear startup test was applied by using large standard flat rotor and test
temperature was set at 130°C. Tests were prolonged to 180 min at slow speed rotor (0.1
rpm) to assess materials behavior at high strain values. In particular, the performed test
reveals that the lower the content of acrylonitrile, the more prominent the peak at the
highest strain values.
Interestingly, in Figure 3 we note that none of the eNBR polymers shows an evident
peak at 2 strain units. Only the polymer sample with 33% average content of ACN
reveals a slight change in the slope of the curve at approximately 2 strain unit. In fact, the
very first visible overshoot occurs at about 20 strain units.
Figure 2. Shear startup for PBR material featuring long chain branching. Test temperature was set at
45°C. A large rotor type with grooves and biconical profile was used to obtain a uniform flow field. A
wide number of rotor speed values was applied in order to achieve shear rate values ranging from 0.005
to 7.5 s-1.
Figure 3. Shear startup of eNBR with different content of acrylonitrile but similar Mooney Torque at
standard test conditions. One of the polymers shows two different chain populations with distinguished
acrylonitrile content. Still the average ACN content is 33%. In this figure test conditions are set to ML
(60 + 180), T = 130°C, 0.1 RPM. The performed test reveals that the lower the content of acrylonitrile,
the more prominent the peak at the highest strain values. No overshoot is visible at 2 strain unit
deformation, only the polymer with 33% average ACN content reveals a slight slope change in the
startup curve.
An higher ACN content is apparently related to a lower branching content produced

during the polymerization process. This is suggested by the comparison between Figure 3
and Figure 4. The polymer samples showing a more evident second overshoot (Figure 3)
are the ones showing also higher elasticity (higher G’ and lower tan, Figure 4) and
broader relaxation time spectrum.
Figure 4. A (above) and B (below). Elastic moduli and damping factor (tan) at T = 130°C for the same
four eNBR samples (see text) of Figure 3.
In Figure 5, stress response to shear startup experiments carried out by Coppola et al.
[2014] for three SBR samples, is displayed. The main features of the three SBR samples
used by Coppola et al. [2014] are reported in Table 1. All three samples are oil extended
and the amount of oil is reported in Table 1. The Linear SBR sample is a linear
polydisperse polymer sample obtained by an anionic polymerization process in solution.
The Branched SBR is a randomly branched sample obtained by an anionic
polymerization process in solution. The eSBR sample is a highly branched, very
polydisperse, polymer sample obtained by emulsion polymerization. Differently from the
eSBR sample, the branching content of Branched SBR sample can be designed on
purpose.
Table 1. The main features of the SBR samples
Mw Mw/Mn Oil Tg
SBR Polymer
(kg/mol) - % (°C)
Linear sSBR 500 2 27.3 -32.5
Branched sSBR 600 2.5 27.3 -24.0
Highly Branched eSBR >500 >4 27.3 -51.0
The transient response observed (Figure 5) is characterized by the presence of

multiple overshoots for both branched samples and the transient is much longer than that
for the linear one. For the Branched SBR, after the first overshoot, another maximum is
observed at a strain of the order of 300 units. For the Highly Branched eSBR, three stress
maxima are observable, positioned at very different strains: the first “usual” stress
overshoot appears at about 2 strain units, as expected; a second overshoot is observed at
about 20 strain units, and a third one at about 500 strain units. It is important to underline
that such results are perfectly reproducible. Repeating tests on different specimens of the
same sample provides identical transient curves, as it was shown by Coppola et al.
[2014]. As it was extensively discussed by Coppola et al. [2014], the strain at which the
overshoots occur can be used as a fingerprint of the kind of polymer architecture. More
complex branching patterns typically imply the presence of overshoots located at very
large strains (of order 100 strain units or larger). The height of such overshoots is related
to the relative importance of the response associated to branched structures at the chosen
shear rate. Branched SBR sample and Highly Branched eSBR sample both should
possess a complex branching pattern. Of course, it is clear that Highly Branched eSBR
sample possesses a more complex structure compared to Branched SBR as its response is
characterized by a higher number of overshoots. These considerations are consistent with
Linear Viscoelastic (LVE) characterization even if apparently the nonlinear transient
response shown in Figure 5 is even richer of structural information. Figure 6A and Figure
6B report the master curves of dynamic elastic moduli and damping factor (tan) reduced
to 60°C, for the SBR samples of Table 1, as they were obtained by Coppola et al. [2014].
It is noteworthy that the frequency range extends over as many as 14 decades. Highly
Branched eSBR shows a slightly higher plateau modulus, as a consequence of a smaller
entanglement molecular weight Me, and a frequency dependence of G’ and G’’ similar to
one another in a very wide frequency range (from 10-5 to 1 rad/s). Highly Branched eSBR
also shows different moduli in the glass transition zone, due to the different chemistry.
Linear SBR and Branched SBR show instead a similar trend in the high frequency range
(10 to 109 rad/s), while differing in the terminal region. Indeed Branched SBR shows a
very wide, and incomplete transition from rubbery to terminal behavior and, similarly to
the Highly Branched SBR, there is a wide frequency range (from 10-5 to 102 rad/s) where
the elastic and viscous moduli are nearly parallel. This behavior is sometimes called “gel-
like” and is typical of polydisperse branched high molar mass polymers. In particular, it
is frequently observed for diene polymers obtained from emulsion radical
polymerization. Consistently with the nonlinear rheological response, LVE
characterization shows higher elasticity at low frequency for Highly Branched eSBR if
compared to Branched SBR.
Based on the reviewed results, also in the case of SBR the peculiar transient response
to shear startup appears to be a characteristic fingerprint of the polymer architecture.
However, it appear sensible to ask if similar information on the polymer structure could
have been obtained by other simpler and well known tests such as the Mooney Stress
Relaxation. The Mooney Stress Relaxation is indeed a quick test that provides
information about the nonlinear stress relaxation after shear flow. In Figure 7, original
Mooney Stress Relaxation test results performed on the Highly Branched eSBR sample
are shown. Results are shown of stress relaxation after shear startup of different duration
performed with a Mooney Viscometer. It is interesting to note that with increasing the
time of shear flow, significantly different stress relaxation curves are obtained. The stress
relaxation appears to be accelerated when increasing the duration of the previously
applied shear flow. The conclusion that such tests can discriminate between different
architectures is however incorrect. This is demonstrated in Figure 8 where Mooney Stress
Relaxation results are shown for the Branched SBR sample of Table 1. For this sample
we have observed in Figure 5 a significant qualitative difference from Linear SBR. In
Figure 8 we observe instead a stress relaxation curve for the Branched SBR sample which
is quite insensitive to the duration of the previously applied shear flow. This behavior is
qualitatively identical to that typically shown by linear samples.
Based on these observations, it is clear that just an analysis of the transient response
to shear startup provides unique information about the structure – property relationship
for branched polymers.
Figure 5. Shear startup experiments at T = 60°C with Mooney VS equipped with biconical rotor on
SBR samples listed in Table 1. Results obtained by Coppola et al. [2014].
Figure 6. A (above) and B (below). Elastic moduli (A) and damping factor, tan (B) at T = 60°C for the
same three SBR samples of Figure 5.
As already indicated by the Mooney Stress Relaxation tests, high molar mass,
branched polymers are extremely sensitive to the previous deformation history. Such
feature can also be pointed out by performing experiments, in which, on the same
specimen, shearing tests are repeated one after the other. Two kinds of experiments are
shown in Figure 9 and Figure 10, respectively.
Figure 7. Mooney Stress Relaxation results for Highly Branched eSBR sample (see Table 1).
Figure 8. Mooney Stress Relaxation results for Branched SBR sample (see Table 1).
In Figure 9, it is possible to observe the results obtained when performing a second

shear startup experiment a short time after the first one. Four different durations have
been adopted for the first run. The delay time between the first and second run is rather
short, 1 min. Interestingly, if in the first run the second peak was not completely
overcome then it would appear a second peak also in the second run (see lowest couple of
curves in Figure 9). Otherwise, in the second run there is no clear evidence of a second
peak and just the first peak at low strain is observable. This suggests that the recovery
process of the second peak is much slower than that of the first peak.
Figure 9. Results obtained repeating a shear startup experiment after a first shear startup run of different
duration. Each couple of curves (1rst and 2nd run) has been artificially shifted in the vertical direction
for sake of clarity. A delay (resting) time of 1 min was adopted between the first and second run. Test
conditions: 5 rpm, flat standard large rotor, T = 100°C.
In Figure 10, experimental results are shown of different shear startup repetitions
made on the same specimen for two oil-extended eSBR samples, referred to as I and II in
the picture. For each test, a delay time of 30 min at test temperature (100°C) was adopted
before starting the shearing stage (0.075 s-1 for 3 hours, flat standard large rotor).
Between two subsequent tests on the same specimen, samples rested at room temperature
(23°C), in stress free condition, for 1 month. By assuming the validity of Time-
Temperature Superposition (TTS), resting time (23°C) could be rescaled to the equivalent
time value (in hours) at test temperature (100°C). Stress curves were therfore represented
in Figure 10, one after the other, on the same time scale. Also this experiment, like the
previous one displayed in Figure 9, reveals that the transient response at each run is
strongly influenced by the previous deformation history. This effect becomes more and
more visible with increasing the number of shearing test repetitions. Noticeably, after the
fourth shearing test, steady-state is not yet reached. In all cases, a first overshoot was
always observed at each repetition although its height was changing with the number of
repetitions. The shape of the transient response following the first overshoot was strongly
changing with the number of repetitions.
The phenomenology emerging from the experiments described in Figure 9 and Figure
10 calls for a deeper and more systematic investigation on the recovery process of the
original transient response after a shear startup experiment. This investigation is the
subject of the following section.
Figure 10. Shear startup repetitions made with Mooney VS (standard flat rotor) at T = 100°C and 0.075
s-1 on the same specimen for two different oil extended eSBR samples, named I and II (see text).
RECOVERY: EXPERIMENTAL RESULTS
In the case of branched samples, if the shear flow is interrupted and the test repeated
after different rest times, different stress growth patterns are observed. In Figure 11,
Figure 12 and Figure 13, the stress growth results at the shear rate of 0.25 s -1 are shown
for the three SBR samples listed in Table 1. This rate has been chosen as the largest shear
rate at which also the Linear SBR still obeys the Cox-Merz rule. All samples were
sheared in a first run for 1 hour at 60°C, then the same test was repeated after a rest time
of 30 min at the same temperature (60°C) with no torque applied. Next, the samples were
cooled down to room temperature, and carefully removed from the rheometer. Even if the
structural recovery process was expected to be extremely slow at room temperature, it
was nevertheless preferred to do so in order to avoid any risk of inducing
degradation/crosslinking reactions in the materials if kept at higher temperatures for a
long time. After the rest time at room temperature, the sample was introduced again in
the rheometer, and the test started after a delay time of 30 minutes for temperature
equilibration, as also done in the first run.
For the Linear SBR, tests following the room temperature recovery were not made
since the second run after 30 min rest time at 60°C shows that the structural recovery is
already completed (Figure 11). On the contrary, the branched samples show a very rich
phenomenology in their extremely slow recovery process. For them, different time effects
are found for the first overshoot, and for the subsequent ones. Contrary to what observed
in previous works on linear polymers [Stratton and Butcher, 1973; Dealy and Tsang,
1981; Robertson et al., 2004; Wang et al., 2012; Ianniruberto and Marrucci, 2014], for the
Branched SBR the first stress overshoot in the second run is higher than in the first run on
the virgin material (see Figure 12). The height of the first stress overshoot decreases in
subsequent runs with increasing rest time, presumably approaching the original height at
very long rest times (not observed). The shape of the second overshoot at large strains
significantly deviates from that of the first run. Apparently, an even longer time is needed
to fully recover.
Figure 11. Shear startup of Linear SBR sample (see Table 1). The 2nd run refers to a test repetition on
the same sheared specimen after a 30min rest time at 60°C.
Figure 12. Shear startup of Branched SBR sample (see Table 1). The 2nd run refers to a test repetition
on the same sheared specimen after a 30min rest time at 60°C. Further test runs were applied on
sheared samples after different rest times at 25°C. More precisely, after 20, 60 and 180 days for 3 rd, 4th
and 5th runs, respectively.
Figure 13. Shear startup of Highly Branched eSBR sample (see Table 1). The 2 nd run refers to a test
repetition on the same sheared specimen after a 30min rest time at 60°C. Further test runs were applied
on sheared samples after different rest times at 25°C. More precisely, after 30 and 60 days for 3 rd and
4th runs, respectively.
The recovery process of the Highly Branched SBR is even slower and more complex,
although roughly similar to that of the Branched SBR (see Figure 13). Indeed, also in this
case, when the test is repeated a short time after the interruption, the first stress overshoot
is higher than in the first run. After longer recovery times, the first stress overshoot falls
slightly below the original height. For the Highly Branched SBR, the recovery of the
second overshoot appears even more difficult to observe.
The slow recovery processes found here, much slower than in the case of linear
polymers, are consistent with those observed for LDPE [Rokudai, 1979; Leblans and
Bastiaansen, 1989], and theoretically interpreted by Borrigaud et al. [2003], who also
used the pompom model. All these authors recognized that the flow-induced change in
rheological properties is not due to degradation, but results from conformational changes
of the branched molecule relaxing back to the equilibrium conformation over an
exceptionally long time.
In conclusion, the multiple overshoots (peaks) displayed in Figure 2, Figure 3, Figure
5, Figure 12 and Figure 13, and the structural recovery process during the rest time
revealed in Figure 12 and Figure 13, all together indicate an extremely complex
viscoelastic behavior for which we proposed a theoretical approach [Coppola et al.,
2014]. Our approach is briefly recalled in the following section and is also implemented
on the new data presented in this chapter for eNBR polymer samples.
MULTIPLE STRESS OVERSHOOTS AND RECOVERY:

THEORETICAL FRAMEWORK
Starting from the classical theory of Doi and Edwards [1986] for monodisperse linear
polymers, the appearance of a shear stress maximum (overshoot) in shear startup
experiments is well understood. A maximum is predicted to occur at a shear deformation
𝛾 ≈ 2, independently of the shear rate 𝛾̇ , as long as 𝛾̇ exceeds the reciprocal orientational
relaxation time 𝜏𝑑 of the entangled polymer. The maximum is due to the fact that the xy-
component of the orientation tensor (x and y being the shear and gradient directions,
respectively) starts from zero at equilibrium due to randomness, grows up with increasing
deformation during shear startup, but then declines towards zero due to molecular
alignment to the x-direction (the y component of the unit vectors describing orientation of
the molecular segment between consecutive entanglements going down to zero).
Doi-Edwards original theory did not include chain stretch. The latter arises when 𝛾̇
becomes larger than, not only of the reciprocal orientational time 𝜏𝑑 , but also of the
reciprocal Rouse time 𝜏𝑅 (𝜏𝑅 < 𝜏𝑑 ). As shown by Pearson et al. [1991], chain stretch
displaces the location of the maximum to larger values of the shear deformation 𝛾. This is
due to the fact that chain stretch (which increases stress levels) starts to decline only
when chains become significantly oriented in the x direction, because stretch is due to
frictional forces linked to the y component of the orientation. In particular, Marrucci et al.
[2008] have shown that, in the limit of large values of 𝛾̇ , the location of the maximum
obeys the simple relationship 𝛾 ≈ 𝛾̇ 𝜏𝑅 ⁄2.
In the case of binary homopolymer blends (a bimodal molar mass distribution), a
double shear stress peak may appear in shear startup. Indeed, one may observe a first
peak around 𝛾 ≈ 2 due to the short polymer that merely orient, and the second peak at a
higher value of 𝛾 due to the longer polymer that is both oriented and stretched. However,
such a double peak is observed only if the short polymer is the major component of the
blend, because, otherwise, the response of the long polymer prevails, and only the single
high 𝛾 peak can be observed.
To conclude on linear polymers, we mention what is expected to happen when, after
the steady state has been reached in a first startup run, and the sample is maintained at
rest for some time 𝑡𝑟𝑒𝑠𝑡 , a second startup is switched on at the same rate. In general, the
response of the second run is different from that of the first, unless the rest time is longer
than the largest relaxation time of the sample. Indeed, for shorter rest times, some
orientation has still survived, and therefore the stress observed will be lower. In
particular, for the case of binary blends, if the rest time is shorter than the orientation time
of the longer component, the second peak may well disappear.
Branched polymers (with the exclusion of star-like polymers) somehow resemble
bimodal blends, insofar as arms typically behave like the short polymer component, and
merely orient, while the backbones between branchpoints are prone to stretch (similarly
to the long component). An important difference should however be noted. Indeed, the
stretch in the backbone of a branched polymer is upper limited by a withdrawal process
of the arms within the entanglements of the backbone, a process first described for a
pom-pom polymer by McLeish and Larson [1998]. As a consequence of the withdrawal
mechanism, the second peak is located at a nearly fixed value of the shear deformation,
independently of the shear rate. Here again we expect that in a repeated startup the stress
be lower, and in particular the second peak may disappear.
For the general case of a complex branched structure, the model proposed by
Coppola et al. [2014] is summarized as follows. To begin with, the linear viscoelastic
response of the polymer needs to be determined in the form of a finite set of orientational
relaxation times 𝜏𝑑,𝑖 , and associated moduli 𝑔𝑖 , so that the shear stress in the linear range
is given by a sum over the Maxwell modes:
𝑔𝑖 𝑡
𝜎(𝑡) = ∑𝑖 ∫ 𝑑𝑡′[𝛾(𝑡) − 𝛾(𝑡′)]𝑒𝑥𝑝[−(𝑡 − 𝑡′)⁄𝜏𝑑,𝑖 ].
𝜏𝑑,𝑖 −∞
In the nonlinear range, the fast modes (i.e., those with the smaller values of 𝜏𝑑,𝑖 ) are
associated to short external branches or to short linear polymers, and are expected to
orient only (no stretch). Conversely, the slower modes (largest values of 𝜏𝑑,𝑖 ) will at least
contain a fraction 𝑥𝑖 of stretchable material (internal backbones of the branched structure,
or long arms and linear polymers). Therefore, the shear stress is written as
𝜎(𝑡) = 6 ∑𝑖 𝑔𝑖 𝑆𝑖 (1 − 𝑥𝑖 + 𝑥𝑖 𝜆2𝑖 ),
where 𝑆𝑖 is the xy component of the i-th mode orientation tensor given by
1 𝑡
𝑆𝑖 (𝑡) = ∫ 𝑑𝑡′𝑄[𝛾(𝑡) − 𝛾(𝑡′)]𝑒𝑥𝑝[−(𝑡 − 𝑡′)⁄𝜏𝑑,𝑖 ],
𝜏𝑑,𝑖 −∞
with 𝑄(𝛾) = 𝛾⁄(4 + 𝛾 2 + √4 + 𝛾 2 ) , and where the chain stretch ratio 𝜆𝑖 obeys the
differential equation:
𝑑𝜆𝑖 𝜆 −1 2 𝜆 −1
= 𝛾̇ 𝑆𝑖 [𝜆𝑖 − 𝜆̂𝑖 (̂𝑖 ) ] − 𝑖 .
𝑑𝑡 𝜆𝑖 −1 𝜏𝑠,𝑖
In the last equation, 𝜏𝑠,𝑖 is the backbone stretch relaxation time (𝜏𝑠,𝑖 < 𝜏𝑑,𝑖 ) of the i-th
mode, and 𝜆̂𝑖 is the parameter that determines the withdrawal process. The 𝜆̂ parameter is
expected to be much larger than unity, with a minimum value 𝜆̂ = 2 corresponding to the
withdrawal of only two arms. Although the stretch equation explicitly refers to branched
polymers, it should be noted that, by artificially setting 𝜆̂ = 0, the equation reduces to the
stretch equation for a linear polymer, or for the arm of a branched structure. For such a
case, the stretch time 𝜏𝑠 coincides with the Rouse time 𝜏𝑅 .
The nonlinear equation for 𝜎 properly reduces to the linear one for small
deformations or slow flows. Indeed, in such a limit, all 𝜆𝑖 ’s attain the equilibrium value
𝜆𝑖 = 1, and 𝑄(𝛾) reduces to 𝛾⁄6.
The parameters 𝑥𝑖 , 𝜏𝑠,𝑖 , 𝜆̂𝑖 of the nonlinear model are fitting parameters. In many
cases, even the highest orientational relaxation times 𝜏𝑑,𝑖 cannot be obtained from the
linear response, although they play a crucial role in the nonlinear one, especially in
repeated startup runs. The nonlinear model described above was shown to qualitatively
reproduce the experimental results on SBR elastomers, both in terms of double peaks in
shear startup, and in the slow recovery of the second peak during repeated startup runs
after different rest times [Coppola et al., 2014].
As a further example, we will use the simple model described above to simulate the
shear startup behavior of the eNBR elastomers with increasing ACN content reported
above in Figure 3. The first peak in those curves is attributed to linear polymers (and/or
external arms of the branched structures), while the second peak at very large strain
values is attributed to internal segments of hyperbranched structures. For the sake of
simplicity, we will here attempt at describing the curves in Figure 3 with only two
stretching “species” (the model parameter 𝑥 being set to unity for both).
Each species is responsible for one of the two maxima. The first species (linear
polymer or external arms) shows a maximum at 𝛾 values significantly larger than the
purely orientational location 𝛾 ≈ 2, hence requiring that the polymer is also stretched. As
mentioned before, the equation for 𝜆 will therefore be used by setting 𝜆̂ = 0, and we
expect that the maximum be located at 𝛾 ≈ 𝛾̇ 𝜏𝑅 ⁄2. The second species (backbones of
hyperbranched polymer) shows a maximum located at 𝛾 ≈ 200, therefore requiring large
values of 𝜆̂.
The curves reported below in Figure 14 were obtained with the parameter values of
Table 2, each line referring to one of the curves in Figure 3 above, i.e., to a different
ACN content. Some of the parameters were kept constant for all curves. In particular, the
high values of 𝜏𝑑 imply that the orientation of both species does not relax during startup.
Similarly the large value of 𝜏𝑠 for the hyperbranched species implies that also the internal
backbone stretch has no time to relax. The crucial parameters are 𝜏𝑅 for the linear
species, and 𝜆̂ for the branched one. It should also be noted that by increasing ACN
content 𝜆̂ decreases, implying that the branched structure becomes progressively less
complex.
The predicted curves of Figure 14 do not match completely the experimental ones in
Figure 3. This comes to no surprise, however, in view of the crudeness of the model, and
of the fact that only two species were assumed to be present. In particular, when
compared to the experimental curves, predictions drop down too steeply both moving
leftwards of the first maximum, and rightwards of the second one. Presumably, this is due
to the fact that we are not accounting for minor amounts of species, “smaller” and
“larger”, respectively, of the only two here considered.
Notwithstanding the good semi-quantitative agreement between the model
predictions and the experimental double-peak startup curves, both here and in the
previous work [Ianniruberto and Marrucci, 2014; Coppola et al., 2014], the model should
not be taken too literally, and further analysis of the complex nonlinear behavior of
branched structures is certainly required.
Table 2. Model parameters
Linear polymer Hyperbranched polymer

𝒈 (𝐤𝐏𝐚) 𝝉𝒅 (𝐬) 𝝉𝑹 (𝐬) 𝒈 (𝐤𝐏𝐚) 𝝉𝒅 (𝐬) 𝝉𝒔 (𝐬) 𝝀̂
5 4
10 2500 230 1.5 10 10 100
10 2500 200 1.5 105 104 85
8 2500 220 1.2 105 104 75
5 4
8 2500 210 1.2 10 10 45
Figure 14. Shear stress model predictions during shear startup at 𝛾̇ = 0.0785 𝑠 −1 for the eNBR
elastomers with different ACN contents, to be compared with the data in Figure 3. Top-to-bottom
curves were obtained with the parameters of Table 2 (top-to-bottom lines).
CONCLUSION
In this work, we presented recent (published and unpublished) experimental data on

shear startup of branched polymers. We showed that for such materials a stress response
to shear startup, characterized by long transient and possibly multiple overshoots, can be
observed. Moreover, repeating the shear startup on the same specimen does not provide
the same results of the first run unless an extremely long resting time is adopted. In some
cases the full recovery is not practically observable. The recovery route of the original
response is also unusual, showing in many cases a higher first peak in the second run,
contrary to the most common observations available in the literature for linear polymers.
For such complex phenomenology, a theoretical approach had been previously proposed
by us, which proved to be successful in providing a possible qualitative explanation for
the main features of the nonlinear rheological response of SBR polymers with complex
branching patterns.
Furthermore, herein the same model was applied to eNBR samples with different
ACN content. Although these polymers have a similar Mooney viscosity (measured in
standard test conditions), their behavior in shear startup is significantly different,
revealing different macromolecular architectures. The model, in a very simple version,
qualitatively matches the experimental results, confirming that multiple overshoots in
shear startup can be explained by considering branched architectures.
In this chapter, we demonstrated that the proposed theoretical approach enhances the
available tools of “analytical rheology” (i.e., of the structural characterization by means
of rheological linear and nonlinear measurements) of commercial branched polymers.
The investigation presented in this chapter is also relevant for processability evaluations,
as the sensitivity of a polymer to previous deformation history is of remarkable
importance in processing applications.
REFERENCES
Bacchelli, F. Kautsch. Gummi Kunstst. 2008, 61, 188-191.

Best, D. M.; Rosen, S. L. Polym. Eng. Sci. 1968, 8, 116-125.
Bourrigaud, S.; Marin, G.; Poitou, A. Macromolecules 2003, 36, 1388-1394.
Chang, T. ACS Macro Lett. 2013, 2, 601–604.
Coppola, S.; Bacchelli, F.; Marrucci, G.; Ianniruberto, G. J. Rheol. 2014, 58, 1877-1901.
Das, C.; Read, D. J.; Auhl, D.; Kapnistos, M.; den Doelder, J.; Vittorias, I.; McLeish, T.
C. B. J. Rheol. 2014, 58, 737–757.
Dealy, J. M.; Tsang, W. K. W. J. Appl. Polym. Sci. 1981, 26, 1149-1158.
Doi, M.; Edwards, S. F. The Theory of Polymer Dynamics; Oxford University Press:
Oxford, UK, 1986.
Einhorn, S. C. Rubber World 1963, 48, 40–45.
Fensom, D. S. Rubber Chem. Technol. 1956, 29, 269-277.
Galvin, P. T.; Whorlow, R. W. J. Appl. Polym. Sci. 1975, 19, 567–583.
Graham, R. S. PhD Thesis; The University of Leeds: Leeds, UK, 2002.
Ianniruberto, G.; Marrucci, G. ACS Macro Letters 2014, 3, 552-555.

Inoue, T.; Yamashita, Y.; Osaki, K. Macromolecules 2002, 35, 1770–1775.
Islam, M. T. Rheol. Acta 2006, 45, 1003–1009.
Kasehagen, L. J.; Macosko, C. W. J. Rheol. 1998, 42, 1303-1327.
Leblans, P. J. R.; Bastiaansen, C. Macromolecules 1989, 22, 3312-3317.
Li, X.; Wang, S. Q. Rheologica Acta 2010, 49, 985-991.
Liu, T. Y.; Soong, D. S.; Williams, M. C. J. Rheol. 1983, 27, 7–35.
McLeish, T. C. B.; Larson, R. G. J. Rheol. 1998, 42, 81–110.
Marrucci, G.; Ianniruberto, G.; Bacchelli, F.; Coppola, S. Prog. Theor. Phys. Suppl.
2008, 175, 1–9.
Meissner, J.; Garbella, R. W.; Hostettler, J. J. Rheol. 1989, 33, 843-864.
Montes, S.; White, J. L.; Nakajima, N.; Weissert, F. C.; Min, K. Rubber Chemistry and
Technology 1988, 61, 698-716.
Mooney, M. Industrial and Engineering Chemistry Analytical Edition 1934, 6, 147–151.
Mooney, M. J. App. Physics 1936, 7, 413–420.
Nakajima, N.; Harrell, E. R. Rubber Chem. Technol. 1982, 55, 1426–1436.
Osaki, K.; Inoue, T.; Isomura, T. J. Polym. Sci.: Polym. Phys. 2000, 38, 2043–2050.
Pearson, D.; Herbolzheimer, H.; Grizzuti, N.; Marrucci, G. J. Polym. Sci.: Polym. Phys.
1991, 29, 1589–1597.
Piper, G. H.; Scott, J. R. J. Sci. Instr. 1945, 22, 206-210.
Pollett, W. F. O. Br. J. Appl. Phys. 1955, 6, 199-206.
Ravindranath, S.; Wang, S. Q. J. Rheol. 2008, 52, 957-980.
Robertson, C. G.; Warren, S.; Plazek, D. J.; Roland, C. M. Macromolecules 2004, 37,
10018–10022.
Rokudai, M. J. Appl. Polym. Sci. 1979, 23, 463–471.
Schweizer, T. Rheologica Acta 2002, 41, 337–344.
Schweizer, T. J. Rheol. 2003, 47, 1071–1085.
Schweizer, T.; Hostettler, J.; Mettler, F. Rheologica Acta 2008, 47, 943–957.
Snijkers, F.; Vlassopoulos, D. J. Rheol. 2011, 55, 1167–1186.
Snijkers, F.; Vlassopoulos, D.; Ianniruberto, G.; Marrucci, G.; Lee, H.; Yang, J.; Chang,
T. ACS Macro Letters 2013, 2, 601–604.
Stratton, R. A.; Butcher, A. F. J. J. Polym. Sci.: Polym. Phys. Ed. 1973, 11, 1747–1758.
Wang, Y.; Li, X.; Zhu, X.; Wang, S. Q. Macromolecules 2012, 45, 2514–2521.
Chapter 8
VISCOSITY AND VISCOELASTICITY OF BABY FOODS
María Dolores Alvarez, Wenceslao Canet and Beatriz Herranz

Department of Characterization, Quality and Safety, Institute of Food Science,
Technology and Nutrition (ICTAN-CSIC), Ciudad Universitaria, Madrid, Spain
ABSTRACT
Traditionally, infants have been supplemented by ready-to-eat semisolid formulated

foods. Composition of these baby foods is being continually modified and adapted with
the aim of producing healthier foods. However, there are no recent rheological studies of
baby foods. A comprehensive study on the flow and the viscoelastic behavior of four
current commercial types of Spanish vegetable-based infant purees was carried out. The
effect of a wide range of temperatures, which are encountered during processing and
storage of baby foods, on their rheological properties is documented. The flow behavior
of the vegetable-based infant purees was analyzed at different temperatures (5-65°C)
giving particular attention to their time-dependent properties in a wide shear rate range
(5-200 s–1). The power law model parameters describing the flow behavior of samples
depended on the kind of infant puree, its water content and the measurement temperature.
Arrhenius model was used to explain the temperature effect on the apparent viscosity at
50 s–1. The infant purees exhibited thixotropic behavior for all the temperatures tested.
For the same temperature, differences in hysteresis loop magnitudes were observed
among the purees, being more noticeable at lower temperatures. Two models were used
to describe the time-dependent behavior, namely Weltman model, and second-order
structural kinetic model. For all the infant purees, the initial shear stress and the extent of
thixotropy increased and decreased significantly with increases in the shear rate and the
temperature. The breakdown rate of the puree associations also accelerated at higher
shear rates, but no trend was observed with the temperature. In turn, the viscoelastic
properties of the four vegetable-based infant purees were evaluated in the temperature
range of 5 to 80°C. The samples behaved like weak gels, with the exception of rice and
chicken puree at 35, 50 and 65°C, which behaved like a macromolecular solution. At 5-
65°C, the storage modulus (G') and the loss modulus (G") could be well described by a
power function of the frequency (R2 > 0.92), and the dependency of their parameters
220 María Dolores Alvarez, Wenceslao Canet and Beatriz Herranz
with the temperature was modeled by the Arrhenius’s model and quadratic functions. The
approach named ‘weak gel model’ was also applied to the baby foods, and both G' and
G" values decreased with an increase in the temperature between 5 and 50°C. Master
curves were obtained applying the time-temperature superposition (TTS) principle, and
the horizontal shift factor was sensitive to the choice of the viscoelastic property being
selected for the shifting procedure. At 5-50°C, the infant purees behaved like
thermorheologically complex materials. The rheological measurements were well
supported by the particle size distributions. Controlling the temperature, the moisture
content and the particle size distribution may be employed to optimize the applicability,
stability, and sensory of baby foods. The results provide a useful fingerprint of the
rheological behavior of these products. The generated data and findings can be
extrapolated to other infant systems for developing new products for convenience to
newborn consumers and manufacturers.
Keywords: baby foods, flow behavior, thixotropy, Arrhenius model, viscoelasticity,

time-temperature superposition, particle size, microstructure
INTRODUCTION
The demand for weaning foods has risen significantly with an increase in the number
of working mothers and the demand for convenience foods [1]. For this reason, the
infants are being supplemented by ready-to-eat formulated foods. The processing
industries are successfully manufacturing and marketing weaning foods. Numerous
weaning food preparations are marketed to satisfy different age groups and nutritional
needs, and the majority of these foods are semisolid in nature that makes them easy to be
handled and fed to infants [2]. The foods include various forms of vegetables, rice or
pasta, and meats from a wide variety of sources such as beef, chicken, turkey, pork and
lamb [3]. At present, the Spanish formulations manufactured by a prominent and well-
recognized manufacturer (Nestlé) of baby foods in Europe contain different ingredients,
and commonly include corn starch, olive oil and salt appropriately mixed with different
vegetables and meat products (ranging between 11% and 12% w/w).
The formulation of a food product specifies the molecules that go into it, and the
processing of these molecules leads to the formation of structures. The ingredients are
themselves an assembly of molecules with associated structures (and textures) which can
be modified, if these ingredients are combined with others and/or further processing is
employed [4].
Strained baby foods are obtained by processes that include both heat treatment and
tissue destructuration and have been described as suspensions of soft particles in a
viscous serum or gel [5-7]. Viscosity can arise from viscous deformation of the liquid
phase (long-range hydrodynamics or lubrication forces between particles), or direct
sliding friction between the particles [8]. These suspensions are in general highly
polydisperse in their particle size, shape and in their chemical composition. In addition,
Viscosity and Viscoelasticity of Baby Foods 221
some carbohydrates, especially starches, and proteins undergo changes/modifications

during the processing resulting in a viscous dispersion, forming solutions or gels
depending on the temperature and the concentration [2]. Generally, the fluids with
suspended particles have a certain shear-sensitive structure, and most vegetable purees
that are used as intermediate products are expected to undergo further low (mixing
operations) to high (pumping) shear stresses [9]. It is therefore essential to understand
their behavior in such conditions for the process engineering calculations as well as for
the product development and the quality control.
The composition of the food product along with its structure determines the rheology
of the food product [10]. The rheological studies contribute to our knowledge of the
molecular structure or distribution of the molecular components of foods, as well as
enabling us to predict the structural changes of the food systems during the
manufacturing processes [11]. When a material is sheared at a constant shear rate, the
viscosity of a thixotropic material will decrease over a period of time, implying a
progressive breakdown of the structure [12]. In addition, the characterization of the time-
dependent rheological properties of food systems is important to establish relationships
between the structure and the flow, and to correlate the physical parameters with the
sensory evaluation [13]. Most food products are of a complex rheological nature, and
their viscosity does not depend only on the shear rate but also on the times that shear is
applied.
Steady shear viscometry, covering the shear rate range 0.1–100 s–1, generally
indicated the presence of yield stress and gave a good fit of data to the Herschel– Bulkley
model in sweet potato puree infant food [2]. However, little information is available on
the applicability of the time-dependent flow models to vegetable or fruit puree-based
baby foods. The Weltman model was recommended for banana puree [5], which was also
better that the Hahn model in describing the thixotropic behavior of four different infant
foods (vegetable, meat, fish and dessert with six fruits) [14].
On the other hand, food products can exhibit viscous or elastic behavior, or a
combination of the two, which are generally recognized as viscoelastic properties. The
method most commonly used to study the viscoelastic property of foods is oscillatory
viscometry, which provides more details about the rheological behavior than
conventional rheometric parameters such as the flow-behavior index or the consistency
coefficient [1]. G' (the storage modulus) is a measure of the deformation energy stored by
the sample during the shear process and represents the elastic behavior of a material [15].
G" (the loss modulus) is a measure of the deformation energy used up by the sample
during the shear process and so lost by the sample; it represents the viscous behavior of
the material.
Wide temperature ranges are encountered during the processing and the storage of
pureed foods, and therefore, the effect of the temperature on rheological properties needs
to be documented [16]. The viscoelastic properties of sweet potato puree infant food have
been evaluated in the range of 5–80°C [2]. The dynamic viscoelastic behavior of three
commercial vegetable puree based baby foods (pea, corn and wax bean) [1] and of three
meat-based strained baby foods [10] have been investigated in the ranges of 20–80°C and
of 5–80°C, respectively. The rheological behavior of three strained pureed baby foods
(apple, apricot and banana) was also studied under small amplitude oscillatory
measurements in the range of 5–80°C [17].
A growing concern for safer foods has led research into increased infant food
formulation and process monitoring procedures [2]. The emphasis of the baby food
industry has been mostly on the nutritional content of such foods, with various
combinations to provide infants with the adequate energy [10]. The improvements in the
weaning foods have been achieved by the incorporation of protein-rich components from
meat and fish and by reducing, to some extent, the flatulence and the nutritional stress
factors of the legumes. Currently, the baby food manufacturers of the products
characterized in this chapter recommend on their labels that protein intake in infants
should be lower than in adults, and that an excess of proteins is entirely inadvisable.
Nevertheless, in spite of the fact that the composition of baby foods, along with all
the other foodstuffs, is being continually modified and adapted with the aim of producing
healthier foods, there are no recent rheological studies of current commercial vegetable
and fruit puree-based baby foods. The flow characterization of baby foods is very
important with respect to the eating characteristics of the product [17]. The puree should
be thick enough to stay in the spoon, but not so thick as to make swallowing difficult
[18]. This fact becomes especially important because consumption of baby foods is
directed at a population who cannot express their opinion [14], thus impeding the
acceptance or rejection of the products on the part of the baby consumers. A sound
knowledge of the viscoelastic properties of foods is important in the design and
prediction of product stability [19], and they are very sensitive to the chemical
composition and the physical structure [20]. Consequently, a complete rheological
characterization of baby foods entails both viscosity and viscoelastic measurements.
The first objective of this chapter was to characterize either the flow behavior, with
particular attention being given to their time-dependent properties under the effect of
different shearing conditions (5, 25, 50, 100 and 200 s–1), or the viscoelastic behavior of
four current commercial samples of Spanish vegetable-based infant purees (all of them
formulated together with meat products), studying the effect of the temperature (5, 20,
35, 50, 65, and 80°C) on their rheological properties. The suitability of the time-
temperature superposition (TTS) principle and the nature of thermorheological behavior
to the infant purees were investigated furthermore. A second objective was to evaluate to
what extent the steady and dynamic rheological properties of these baby foods are related
to their microstructural and particle size features.
MATERIALS AND METHODS
Infant Food Samples
Batches of four types of vegetable puree-based strained baby foods in 250-g retail
jars were purchased from a local market. These baby foods were named by the
manufacturer (Nestlé España, S.A. Esplugues de Llobregat, Barcelona, Spain) as: rice
and chicken, selected vegetables with bass, vegetables and beef, and vegetables and
chicken purees. The product specifications based on a 100 g serving size provided by the
manufacturer are presented in Table 1. The selected vegetable with bass puree is intended
for an infant population aged 8 months and over, whereas the other three purees are
prepared for an infant populations of over 6 months. Each batch consisted of 50 jars of
each vegetable puree, which were randomly purchased to obtain a more accurate
representation of the product. After purchasing, the jars were stored in a dry, cool place
with a minimal light environment and refrigerated immediately after opening. Each
sample opened was discarded within 3 h.
Table 1. Composition and ingredient information of the vegetable

puree-based baby foods*
Composition Puree (g 100 g–1)

Rice and chicken Selected vegetables Vegetables and Vegetables and
with bass beef chicken
Protein 3.3 3.0 3.4 3.1
Carbohydrate 8.8 7.4 6.5 7.3
of which sugar 2.4 2.5 1.2 1.0
Fat 2.5 2.3 2.1 2.2
of which 0.5 0.3 1.0 0.4
saturated
Fiber 0.6 1.5 1.3 1.2
Sodium 0.042 0.066 0.055 0.065
Ingredients Cooking water, Cooking water, Cooking water, Cooking water,
reconstituted reconstituted potato, beef (12%), potato, chicken
skimmed milk skimmed milk carrot, peas, (11%), carrot,
powder, potato, powder, vegetables tomato, onion, peas, tomato,
chicken (12%), (58%) (carrot, potato, green bean, leek, onion, green
carrot, rice (4%), onion, peas, green corn starch, olive bean, leek, corn
corn starch, olive oil beans and leek), bass oil (1.3%) and salt. starch, olive oil
(1.5%), lactose and (12%) corn starch, (1.3%) and salt.
salt. olive oil (2%),
lactose and salt.
*Provided by the manufacturer.
The total soluble solid (TSS) content g 100 g–1 (w/w) and the pH of the samples were
determined using a hand Atago (Itabashi-ku, Tokyo, Japan) dbx-30 refractometer and a
Schott CG pH meter (Model 842; Schott-Geräte GmbH, Mainz, Germany), respectively.
The expressible water for each sample was measured using centrifugal force according to
Eliasson and Kim [21]. The dry matter content (%) was also determined by drying the
samples in a Philips microwave oven (model M-718, 700 W) with output power at 70%.
Rheological Properties of the Baby Foods
A Bohlin CVR 50 controlled stress rheometer (Bohlin Instruments Ltd., Cirencester,

UK) was used to conduct the steady shear and small-amplitude oscillatory shear (SAOS)
experiments using a plate–plate sensor system (PP40, 40 mm) with 1-mm gap in all the
infant purees. The time-dependency issue is well known in the baby food industry for
products that use starch as a thickener. Therefore, the influence of the sample loading on
the time-dependent changes may influence the results. As an unavoidable syneresis was
detected in most cases as the jars were opened, all the products were thoroughly stirred
and allowed to rest for 10 min. Then, the sample was carefully spooned out with
minimum structural damage before it was loaded into the rheometer. Besides, the samples
were allowed to relax for 5 min after the measuring system reached the testing position.
This ensured thermal equilibrium at the time of the measurement. A solvent trap was
used in order to prevent moisture losses. The temperature control was achieved with a
Peltier Plate system (-40 to +180°C; Bohlin Instruments). The sample temperature was
ramped between 5 and 80°C with incremental steps of 15°C. However, after data
analysis, for all the infant purees, values at 80°C were discarded because of the anomalies
observed in the rheological measurements (sample desiccation).
Steady Shear Rheological Measurements

The following curves were obtained:
1. Flow curves with a controlled shear stress (CSS) in the range of 2.99-200 Pa,
during a 10 min period at temperatures of 5, 20, 35, 50, and 65°C. The
experimental data were described by the power law model (Eq. 1) on the one
hand and the Herschel-Bulkley model (Eq. 2) on the other hand [16] using the
Bohlin software CVO 120 (v. 06.40):
  K  n
(1)
   0  K 
n
(2)
where σ is the shear stress (Pa),  is the shear rate (s–1), K is the consistency index (Pa sn),
n is the flow behavior index (dimensionless) and σ0 is the yield stress (Pa). The effect of
the temperature on the apparent viscosity was described by the Arrhenius relationship:
 a   A exp( Ea / RT ) (3)
where ηa is the apparent viscosity at a specific shear rate (50 s–1), η∞A is the frequency
factor, Ea is the activation energy (J mol–1), R is the gas constant (J mol–1K–1), and T is the
temperature (K).
2. Flow curves with a controlled shear rate (CSR) at temperatures of 5, 20, 35, 50,
and 65°C; the hysteresis loop was obtained by registering the shear stress at shear
rates from 0.1 to 100 s–1 in 5 min and down in 5 min. The areas under the
upstream data points (Aup) and under the downstream data points (Adown) as well
as the hysteresis area (Aup - Adown) were obtained using Bohlin software CVO 120
(v. 06.40). The percentage of relative hysteresis area [21, 22] was calculated by:
Ar  ( Aup  Adown ) / Aup 100

(4)
Both the upstream and downstream data points were also fitted to the power law
model (Eq. 1).
3. Curves showing the time-dependent behavior of the shear stress in the time range
of 0-40 min at 5, 20, 35, 50, and 65°C with a constant shear rate of 5, 25, 50, 100,
and 200 s–1. The experimental data were fitted to the Weltman model (Eq. 5) and
to the structural kinetic model (SKM) of Nguyen et al. [23] (Eq. 6), considering
second-order kinetics with n = 2 [9, 12, 22].
  A  B ln t (5)
Here, σ is the shear stress at the time t (Pa), A is the initial shear stress (Pa), B is the
time coefficient of the thixotropic breakdown (Pa), and t is the time of shearing (s).
1 n
 (   e) 
   (n  1)kt  1 (6)
 ( 0   e) 
In Eq. (6), η0 is the initial apparent viscosity at t = 0 (structured state), ηe is the

equilibrium apparent viscosity as t → ∞ (non-structured state), k = k(  is the rate
constant, and n is the order of the structure breakdown reaction. Details and assumptions
of this model can be found in [12].
Six and three replicates were performed for the flow behavior determination (1) and
the time dependence characterization [2, 3], respectively. The extent of the fit for the
applied models was evaluated by determining the correlation coefficient (R2) between the
experimental and predicted shear stress values.
SAOS Measurements
To determine the linear viscoelastic (LVE) region, stress amplitude sweeps were run
at a constant oscillatory frequency (ω) of 1 rad s–1 by varying the shear stress (σ) of the
input signal from 3 to 200 Pa. For each sample, the LVE region was established at 20°C.
In preliminary experiments, the effects of four temperatures: 5, 20, 35, and 50°C on the
LVE domain were investigated, and not difference appeared in the domains obtained at
these four temperatures. Frequency sweeps oscillatory tests were performed at variable ω
over the range 0.1-100 rad s–1, keeping the amplitude stress at a constant value (3 Pa)
within the LVE region. Changes in storage modulus (G', Pa), loss modulus (G", Pa),
complex modulus (G*, Pa), and phase angle (δ, °) were recorded. Six replicates were
performed for the frequency sweeps. To determine the influence of the temperature on
structural changes in the different baby foods, temperature sweep tests (time cure) were
additionally carried out at the heating rate of 2°C min–1. The values of the dynamic
properties were measured at a constant amplitude stress of 3 Pa, using ω = 1 rad s–1, and
at various temperatures from 5 to 90°C. Three replicates were performed for the
temperature sweeps.
Analysis of SAOS Data

Usually the mechanical spectra can be expressed from the angular frequency
dependence of G' and G". Given that the appearance of the data on logarithmic
coordinates was nearly linear, a power law model was used to characterize the ω
dependence of these moduli as follows (Eqs. (7) and (8)):
G'  G'0  n '

(7)
G"  G"0  n"

(8)
where, G'0 (Pa) and G"0 (Pa) are the storage and viscous moduli at 1 rad s–1 respectively,
and the exponents n' and n" (both dimensionless) denote the influence degree of the ω on
both moduli. G'0 is the energy stored and recovered per cycle of sinusoidal shear
deformation [24]. Its increase indicates rigidity of the sample associated with the
formation of an elastic gel structure [25]. G"0 is the energy dissipated or lost as heat per
cycle of sinusoidal strain, and indicates the extension of the viscous element in the
sample [24].
Gabriele et al. [26] introduced the concept of power law relaxation modulus to
describe the rheological behavior of dough, jam, and yoghurt. The weak gel model
parameter z, is the “coordination number”, which is the number of flow units interacting
with each other to give the observed flow response, and the flow regime is characterized
as follows (Eq. (9)):
|G*| = ((G'(ω))2 + (G" (ω))2)1/2 = Aω1/z (9)
where, A is a constant which can be interpreted as the ‘interaction strength’ between the
flow rheological units; these are basic conformational structures which are connected by
small threads and made up of flow units with a certain degree of coordination [24, 26].
The three-dimensional structure characterizing a gel is described in terms of A, which is
related to the overall stiffness or resistance to the deformation within the LVE region at
angular frequency of 1 rad s–1, while z can be used as a convenient measure for the
strength of interaction of a gel. In turn, 1/z was taken as index of the viscoelastic nature
of the material (a measure of the physical cross-links in a protein network) [24, 25].
Furthermore, the effect of the temperature (5-50°C) on G'0, G"0, n' and n" was
studied on the vegetable-based infant purees using the Arrhenius equation (Eq. (10); R is
the gas constant (J mol–1K–1) and T is the absolute temperature (K)), where each
parameter (A) is modeled by a pre-exponential factor (A0) and the activation energy (Ea):
A  A0 exp( Ea / RT ) (10)
The time-temperature superposition (TTS) principle was applied to investigate the

temperature dependence of G' and G" at various temperatures. It is known that TTS is a
procedure whereby modulus data taken as a function of frequency at various temperatures
are superposed to incorporate the temperature dependence into the time scale [27, 28].
The resulting response covers an expands the frequency range, which can be very useful
for approximating the behavior of a material at frequencies or times inaccessible with the
instrument used to obtain original data, as indicated by different authors [29-32]. The
application of the superposition method for the given system allows for a preliminary
assessing of the system thermal stability (i.e., the lack of phase transitions or chemical
reactions) [33]. The amount of shifting along the horizontal (x-axis) in a typical TTS plot
required to align the individual experimental data into the master curve that can be
generally described using the Williams-Landel-Ferry (WLF) model [34] (Eq. (11)):
 C1 (T  Tr )
log aT 
C2  (T  Tr ) (11)
where aT is the shift factor, C1 (dimensionless) and C2 (K) are the model constants and Tr
is the reference temperature, conveniently chosen.
The aT is defined as follows [27]:
aT = t(T)/t(Tr) = ω(T)/ω(Tr) (12)
where t(T) is the time required to reach a particular response at temperature T and t(Tr) is
the time required to reach the same response at Tr, and ω is the oscillatory frequency. In
this study, the corresponding shift factors obtained at Tr of 35°C were calculated with the
Bohlin Rheometer TTS Software (vs.06.40).
Particle Size Distribution
The particle size distribution analysis was performed with a Laser Diffraction Particle
Size Analyzer (Mastersizer 2000 Malvern Instruments, Worcestershire, UK), connected
to a cell for liquid measurements (Hydro 2000S mixing, Malvern Instruments,
Worcestershire, UK) with distilled water as dispersant. The refractive index used was
1.52. Feed rate and air pressure were adjusted to keep the sample passing through the
measurement unit in the obscuration range 7 – 9%. The statistics of the distribution are
calculated from the results using the derived diameters, D [m,n] according to Eq. (13)
[35]:
1 / m n
 m3 
Dm, n    V d
i i 
 n3 

 V i d i 
(13)
According to this formula, D [3,2] (μm) represents the surface weighed mean, also
known as the surface area moment mean or Sauter mean, while D [4,3] (μm) represents
the volume weighed mean or mass moment mean diameter. Further parameters obtained
by the measurement were d (0.1) (μm), as the size of particle below which 10% of the
sample lies; d (0.5) (μm), the size in microns at which 50% of the sample is smaller and
50% is larger. This value is also known as the mass median diameter or the median of the
volume distribution. d (0.9) (μm) is the size of the particle below which 90% of the
sample lies. The span (S) was expressed as S = [d (0.9) - d (0.1)/ d (0.5)]. The
calculations were done with the software provided with the equipment (Mastersizer 2000,
version 5.60). The measurements were done in triplicate, and the results are expressed as
the mean (n=3) ± standard deviation.
Scanning Electron Microscopy (SEM)
The vegetable-based infant puree microstructure was examined by SEM using a

Hitachi model S-2.100 microscope (Hitachi, Ltd., Tokyo, Japan). Samples were air-dried,
then mounted and sputter-coated with Au (200 Ǻ approx.) in an SPI diode sputtering
system metallizer. Photomicrographs were taken with a digital system Scanvision 1.2 of
Röntgenanalysen-Technik (800 × 1200 pixel, Rontec, GmbH, Berlin, Germany).
Statistical Analysis
A two-way ANOVA was used to evaluate the effect of the infant puree type, the
measurement temperature and the interactions on the rheological properties derived from
the time independent flow behavior and the hysteresis loops. Moreover, to establish the
effect of the measurement temperature on each kind of infant puree separately, as well as
the effect of the kind of infant puree at each measurement temperature, two one-way
ANOVA were performed. A three-way ANOVA with interactions was applied to
evaluate how the infant puree type, the measurement temperature and the shear rate affect
the parameters derived from the shear stress decay of the products. Additionally, to
establish the effect of the measurement temperature at each shear rate in each vegetable-
based infant puree, as well as the effect of the shear rate on each measurement
temperature, two one-factor ANOVA were applied to the data.
In turn, a two-way ANOVA was used to evaluate the effect of the infant puree type,
the measurement temperature and the interaction on the parameters of the power law
models for G', G" and G* (G'0, G"0, n', n", A, and z). The minimum significant
differences were calculated by Fisher’s least significant difference (LSD) tests at
significance level 0.05. Furthermore, in order to get a preliminary view of the relationship
between variables, Pearson product-moment correlations were determined with

significance levels based on the Student’s distribution. The statistical analyses were
performed using the SPSS Statistics 19.0 software (SPSS, Inc., Chicago, IL, USA).
Physicochemical Parameters of Baby Foods
The Spanish formulations contain different ingredients (Table 1), and commonly
include corn starch, olive oil and salt appropriately mixed with other different vegetables
and meat products (ranging between 11-12% w/w). The TSS content of each of rice and
chicken, selected vegetable with bass, vegetables and beef, and vegetables and chicken
purees was 11.2, 11.1, 9.2 and 9.5 g 100 g–1 (w/w), respectively. The respective values
for the pH were 4.62, 5.88, 5.60 and 5.84, for the expressible water (sineresis)
percentages 42.4%, 29.6%, 37.4% and 18.7%, and for the dry matter content 82.8%,
84.1%, 85% and 84.8%. Therefore, the rice and chicken puree had the highest TSS,
moisture content and syneresis, but the lowest pH.
These baby foods are rich in carbohydrates (especially rice and chicken puree) with
protein and fat contents varying within a narrow range from 3 to 3.4 g 100 g –1 and from
2.1 to 2.5 g 100 g–1, respectively (Table 1). In turn, the energy values ranged between 256
and 303 kJ 100 g–1. The protein and fat contents were higher and lower, respectively, than
those reported by Ahmed and Ramaswamy [1, 10] for vegetable (pea, corn and wax bean)
and meat-based (chicken, lamb and beef) strained baby foods. This can be explained by
the fact that all the vegetable puree-based baby foods presented in this chapter also
contained meat products as mentioned above, although the energy values of the samples
were quite similar to those reported by the same authors for pea, corn and wax bean
purees [1].
The rheological properties of the vegetable-based infant purees can be determined by
the size of solid phase particles [36]. It is thus clear from literature that by controlling the
particle size distribution it is possible to influence the food processing optimization [7].
The rheological properties of these infant foods are also influenced by their dry matter
content, as well as by the current addition of olive oil fractions and thickeners. All the
infant purees presented here contain corn starch as thickener, although the amount
incorporated is not specified on labels (Table 1). On the other hand, the rheological
behavior of a lot of suspensions depends on the measurement temperature, the shear rate
and the time of shearing [5].
Steady Shear Rheological Measurements of Baby Foods
In this chapter, the steady shear rheological measurements have been separated into
the time-independent flow behavior and the time-dependent experiments.
Time-Independent Flow Behavior

Examples of the flow curves obtained with controlled shear stress (CSS) for rice and
chicken puree together with the power law model fits obtained at the different
temperatures are presented in the Figure 1a, whilst the Figure 1b shows the shear stress
curves and the Hershel-Bulkley model fits at 35°C for all the four vegetable-based infant
purees. The reference to the classification of the time-independent flow behavior of fluid
foods shows that all the curves exhibited a non-Newtonian shear-thinning behavior,
which is characterized by a curvature downwards on the shear rate axis. The apparent
viscosity decreases markedly with increasing the shear rate and with an increase in the
temperature (Figure 1a). Other baby foods were also reported as having non-Newtonian
flow behavior [2, 10, 14]. The rice and chicken puree exhibited the lowest shear stress
values over all the shear rate range applied, whilst the vegetable and chicken puree had
the highest ones (Figure 1b). The power law and the Hershel-Bulkley models were used
to describe flow curves of the vegetable-based infant purees. The parameters of the
rheological models employed to describe the experimental data and the values of the
apparent viscosity at 50 s–1 are summarized in the Tables 2 and 3.
Based on the R2 correlation coefficient values, the Hershel-Bulkley model exhibited a
better fit than the power law model, though, some flow curves failed to adequately fit the
former model. In addition, a negative yield stress was found in some cases. The Hershel-
Bulkley model was found to be suitable for describing the flow of different meat-based
infant foods after eliminating time dependency [14], as well as the flow of fruit and sweet
potato puree-based baby foods [2, 17].
However, the yield stress value that is estimated from the flow curves depends on the
estimation equation [37, 38]; therefore in this case yield stress should be interpreted as a
model parameter rather than as the actual yield stress value which is a physical property
of the fluid itself. On the other hand, the consistency coefficient had higher values for the
Herschel-Bulkley model than for the power law model (Table 2), while the flow behavior
index values were lower for the Herschel-Bulkley model than for the power law model.
This was probably due to that a non-linear regression technique was used to estimate the
parameters (σ0, K and n) of the Herschel-Bulkley model, and the fact that the power law
model disregards the yield stress [39]. Conversely, it is known that the consistency
coefficient of a product should be higher when taking out the yield stress. Linear
regression techniques also can be used for the determination of the parameters of the
Herschel-Bulkley model (when the magnitude of the yield stress is known from an
independent experiment). Reasonably, the resulting straight line’s intercept will be lower
than that obtained from the power law model. The results confirm that the estimated
values of the yield stress should be used only when experimentally determined values are
not available [16, 37]. Nevertheless, the yield stress values obtained in this study for rice
and chicken puree are similar to the data reported by such authors as Ahmed and
Ramaswamy [2] for sweet potato puree based baby food, but these authors recorded
much lower values for the Herschel-Bulkley consistency coefficient at the same
temperatures. In other commercial vegetable infant puree, Alonso et al. [14] also obtained
lower values for the Herschel-Bulkley consistency coefficient but higher flow behavior
index and yield stress values than those obtained in this study. The latter authors recorded
flow after eliminating the time dependence by shearing the samples at 57.6 s –1 for 20
min; this fact probably led to a drop in the value of the consistency coefficient and an
increase in the flow behavior index value.
It should be noted that the standard deviation (SD) for the consistency coefficient in
the Hershel-Bulkley model was higher than that in the power law model for the following
purees: selected vegetables with bass, vegetables and beef, and vegetables and chicken
(Table 2). Nevertheless, although the power law model does not take into account the
yield stress, the model fitted the data quite well (R2 ≥ 0.99), and for rice and chicken
puree, the degree of fit was higher than that obtained for the rest of the purees (R2 =
1.00). The inclusion of olive oil in the formulation of these baby foods in percentages
ranging from 1.3 to 2 (Table 1) could account for the fact that the flow behavior of these
products may be better described by the simpler power law model. The addition of extra
virgin olive oil (either alone or blended) reduced the apparent viscosity and the
pseudoplasticity of mashed potatoes, thereby producing softer systems, indicating that the
olive oil behaves as a soft filler [40]. Therefore, power law model was selected for the
subsequent studies.
The results of the two one-way variance analysis showed that the kind of baby food
and the measurement temperature, as well as the interaction between these two factors
significantly affected the time-independent rheological parameters (Table 2). The
consistency coefficient and the yield stress values, together with the apparent viscosity at
50 s–1 were the lowest for the rice and chicken puree with the highest moisture content
(17.2%) and syneresis (42.4%). These same parameters were higher for the vegetable and
chicken puree with lower water content (15.2%) and the highest ability to immobilize
water (syneresis 18.7%). All the vegetable-based infant puree samples had n values of
less than one which confirms that they exhibited non-Newtonian shear thinning flow
behavior. The shear-thinning behavior is classically encountered in heterogeneous
systems containing a dispersed phase. At rest, the particles are linked together by weak
forces. When hydrodynamic forces during shear are sufficiently high, the interparticle
linkages are broken, resulting in a structural unit size reduction that, in turn, offers lower
resistance to flow during shear [9]. The most pseudoplastic properties corresponded to
the vegetable and chicken puree with the lowest flow behavior indexes from both
rheological models, which correspond with the highest shear stress values obtained for
this puree (Figure 1b). The highest σ0 value for this baby food indicates that the
interparticular forces were greater than those of the other purees [14].
(a) Rice and Chicken Baby Food

200
5 °C 20 °C 35 °C 50 °C 65 °C Power law model
150
Shear stress (Pa)
100
50
0
0 20 40 60 80 100
Shear rate (s-1)
(b)
Rice and Chicken Baby Food at 35 °C
75 Selected Vegetables with Bass Baby Food at 35 °C
Vegetables and Beef Baby Food at 35 °C
Vegetables and Chicken Baby Food at 35 °C
Herschel-Bulkley model
Shear stress (Pa)
50
25
0
0 20 40 60 80 100
Shear rate (s-1)
Figure 1. Flow curves obtained with controlled shear stress (CSS mode) of the vegetable-based infant
purees measured by increasing the shear rate (forward measurements). (a) Rice and chicken puree at
different temperatures fitted to the power law model; (b) Vegetable-based infant purees at 35°C fitted to
the Herschel-Bulkley model.
Table 2. Rheological parameters of the power law and Herschel-Bulkley models describing the flow curves of the vegetable-based
infant purees and the apparent viscosity at 50 s–1
Vegetable-based Temperature Power law model Herschel-Bulkley model ηa at 50 s–1 Change in ηa

infant puree type (°C) n
K (Pa s ) n R 2 σ0 (Pa) K (Pa s )n n R 2 (Pa s) per °C (%)
Rice and 5 13.52 ± 0.64Ad 0.45 ± 0.01A,Ba 1.00 2.83 ± 0.54Ad 16.85 ± 0.96Ad 0.40 ± 0.01Ba 1.00 1.55 ± 0.03Ad
chicken 20 10.77 ± 0.98Bd 0.44 ± 0.00Ba 1.00 1.85 ± 0.21Bc 11.97 ± 0.22Bd 0.41 ± 0.00Ba 1.00 1.17 ± 0.01Bd 1.64
C B B C B
35 7.71 ± 0.06 d 0.44 ± 0.00 a 1.00 1.64 ± 0.40 c 9.25 ± 0.48 c 0.41 ± 0.01 a 1.00 0.89 ± 0.01Cc 1.60
D A B D B
50 5.93 ± 0.05 c 0.45 ± 0.00 a 1.00 1.54 ± 0.03 c 7.36 ± 0.06 c 0.41 ± 0.01 a 1.00 0.70 ± 0.01Dc 1.42
65 5.04 ± 0.13Dd 0.41 ± 0.00Ca 1.00 0.24 ± 0.11Cb 4.71 ± 0.09Ec 0.43 ± 0.01Aa 1.00 0.51 ± 0.01Ed 2.00
Selected 5 29.91 ± 0.46Ac 0.31 ± 0.01A-Cb 0.99 8.55 ± 0.08Ac 42.71 ± 0.64Ac 0.23 ± 0.00Cb 1.00 1.94 ± 0.01Ac
B A, B A,B B
vegetables 20 23.07 ± 1.18 c 0.32 ± 0.02 b 0.99 6.50 ± 1.78 b 32.44 ± 1.25 c 0.25 ± 0.00 b 1.00 1.57 ± 0.02Bc 1.27
B,C
with bass 35 18.64 ± 0.52Cc 0.36 ± 0.02Ab 0.99 6.24 ± 0.94Bb 28.68 ± 2.69Bb 0.24 ± 0.01B,Cb 1.00 1.30 ± 0.03Cb 1.15
50 16.90 ± 1.95Cb 0.29 ± 0.04B,Cb 1.00 3.24 ± 0.22Cb 22.82 ± 2.56Cb 0.26 ± 0.01Bb 1.00 1.07 ± 0.01Db 1.18
D C C D A
65 14.31 ± 0.57 c 0.26 ± 0.01 c 1.00 2.08 ± 0.53 a 13.56 ± 1.79 b 0.32 ± 0.01 b 1.00 0.84 ± 0.02Ec 1.43
A A A A D
Vegetables 5 36.01 ± 0.81 b 0.30 ± 0.01 c 0.99 13.35 ± 0.18 b 54.84 ± 0.93 b 0.21 ± 0.00 c 0.99 2.22 ± 0.02Ab
and beef 20 26.92 ± 0.35Bb 0.30 ± 0.01Ab 0.99 9.77 ± 0.16Ba 39.78 ± 0.64Bb 0.22 ± 0.00Cc 0.99 1.69 ± 0.01Bb 1.59
35 22.44 ± 0.13Cb 0.27 ± 0.00Bc 0.99 4.05 ± 0.17C 26.94 ± 0.25Cb 0.24 ± 0.00Bb 1.00 1.33 ± 0.04Cb 1.42
D A C,D D A
50 16.68 ± 0.82 b 0.29 ± 0.02 b 1.00 3.03 ± 1.02 b 19.98 ± 1.21 b 0.26 ± 0.01 b 1.00 1.07 ± 0.03Db 1.30
65 12.80 ± 0.43Eb 0.29 ± 0.01A,Bb 1.00 2.37 ± 0.08Da 18.49 ± 2.43Da 0.25 ± 0.00Bc 1.00 0.78 ± 0.04Eb 1.81
Vegetables 5 58.65 ± 0.45Aa 0.27 ± 0.00A, Bd 0.99 14.26 ± 0.19Aa 76.09 ± 2.62Aa 0.21 ± 0.01Bc 1.00 3.29 ± 0.02Aa
and chicken 20 B
52.48 ± 0.61 a 0.26 ± 0.01 c B B
0.99 10.93 ± 0.53 a 64.36 ± 3.96 a B B
0.21 ± 0.00 d 0.99 2.90 ± 0.01Ba 0.79
C A,B C C B
35 41.48 ± 0.43 a 0.26 ± 0.00 c 0.99 7.94 ± 1.28 a 52.20 ± 2.04 a 0.22 ± 0.01 c 0.99 2.23 ± 0.02Ca 1.54
50 31.05 ± 2.23Da 0.27 ± 0.01A,Bb 0.99 4.86 ± 0.40Da 39.20 ± 0.94Da 0.22 ± 0.01Bc 0.99 1.74 ± 0.10Da 1.46
65 22.50 ± 0.46Ea 0.28 ± 0.00Ab,c 1.00 2.40 ± 0.17Ea 16.69 ± 2.86Ea,b 0.29 ± 0.04Ab,c 1.00 1.28 ± 0.00Ea 1.76
Values are given as mean (n = 6) ± standard deviation (SD).A-E For the same column and for the same vegetable-based infant puree type and parameter, means without
the same letter are significantly different (P < 0.05) according to the LSD multiple range test.a-d For the same column and for the same temperature and parameter,
means without the same letter are significantly different (P < 0.05) according to the LSD multiple range test.
Without doubt, the dry matter content of the vegetable-based infant purees had an
effect on their rheological properties: the lower it was, the lower the yield stress and the
consistency coefficients were. The opposite was true with respect to the flow behavior;
the lower the dry matter content, the higher the flow behavior index values. The rice and
chicken puree, with the highest carbohydrate content (8.8 g 100 g–1), was clearly less
viscous because of the plasticizing and diluting effect of the water [41].
However, as the vegetables and beef puree, with the lowest amount of water (15%),
had intermediate rheological properties, it would appear that these properties do not
depend only on the water content. The rheological property values may result from other
factors such as the content and the degree of protein hydration that are responsible for the
pseudo-gel structure formations in many foods, as well as the carbohydrate content [1, 2,
10]. Alonso et al. [14] suggested that the flow resistance in meat and fish infant foods is
mainly due to the protein gel formation, produced by a molecular interaction between the
protein molecules through hydrogen bonding, ionic bonding, disulfide bonding and
hydrophobic association, although the authors did not give any data about product
composition. In this study, the rice and chicken and vegetables and chicken purees had
the lowest and highest viscous properties respectively, in spite of both containing meat
proteins from the same specie. Furthermore, the chicken content of rice and chicken
puree (12%) is higher than that for vegetables and chicken (11%). In turn, the vegetables
and beef puree has the highest total protein content (3.4 g 100 g–1), but also the smallest
amount of carbohydrates (6.5 g 100 g–1) (Table 1). Definitely, the total protein content of
the vegetable-based infant foods would appear to be too low and similar to justify that the
protein content alone is responsible for the different rheological behavior observed
between the samples. On the other hand, because the pseudoplasticity is related to the
average size of the particle of the dispersed phase [42], it can be presumed that the rice
and chicken puree had the smallest particles, while the vegetables and chicken puree had
the largest ones, so much so that this fact could be appreciated with the naked eye, due to
dissimilar finishing procedures. Moreover, an increase in particle content generally leads
to higher viscous properties of pureed fruits and vegetables [6].
The temperature significantly affects the viscosity of the vegetable-based infant
purees which decreases when the temperature increases. In this study, the two
consistency coefficients, the yield stress and the apparent viscosity decreased when the
temperature was increased (Table 2). In the case of the rice and chicken puree, the
consistency coefficient values from the power law model fell from 13.52 at 5°C to 5.04
Pa sn at 65°C, and the apparent viscosity from 1.55 at 5°C to 0.51 Pa s at 65°C. A similar
tendency was observed in the other baby foods. The effect of the temperature on the
apparent viscosity is commonly described by the Arrhenius relationship (Eq. 3). The
applicability of the Arrhenius model to the apparent viscosity at 50 s–1 for the vegetable-
based infant purees is shown in the Figure 2.
2.0
Rice and Chicken Baby Food (y = 1715.9x - 5.71; R^2 = 0.99)
Selected Vegetables with Bass Baby Food; y = 1279.8x - 3.92; R^2 = 0.99
Vegetables and Beef Baby Food; y = 1589.1x - 4.90; R^2 = 0.99
Vegetables and Chicken Baby Food; y = 1492.8x - 4.10; R^2 = 0.97
1.5 Arrhenius model
ln (a at 50 s-1, Pa s)
1.0
0.5
0.0
-0.5
-1.0
0.0028 0.0030 0.0032 0.0034 0.0036 0.0038
(1/T, K)
Figure 2. Plots of the apparent viscosity (CSS mode) versus the temperature for the vegetable-based
infant purees.
The flow activation energy values (Ea) for rice and chicken, selected vegetables with
bass, vegetables and beef and vegetables and chicken purees, were 14.27, 10.64, 13.21
and 12.41 kJ mol–1, respectively. Therefore, the vegetables and chicken puree had the
worst fit to the Arrhenius model. The flow activation energy is a measure of the changes
in viscosity with increasing the temperature and for Newtonian fluids ranges from 14.4 kJ
mol–1 for water to more than 60 kJ mol–1 for sugar solutions and concentrated juices.
Non-Newtonian fluids, including purees, have much lower flow activation energy values
[43]. Consequently, the rice and chicken puree was the most sensitive to the temperature
changes. The differences in the Ea values between samples are the result of many factors
connected with the sample composition. The change in the viscosity by unit of
temperature, i.e., per °C gave values between 0.79 and 2.0% for the vegetable-based
infant purees. Roopa and Bhattacharya [44] obtained values of this parameter ranging
between 1.5 and 2.2% for aqueous alginate dispersions. The apparent viscosity value of
the rice and chicken puree with the highest Ea value decreased over 3-fold as the
temperature increased, while the apparent viscosity value of the selected vegetables with
bass puree, with the lowest Ea value, decreased over 2.3-fold as the temperature
increased. The higher sensitivity of the rice and chicken puree could be ascribed to their
higher water content, as well as to an additional gelatinization of the rice starch present in
this product. The lower Ea values of the selected vegetables with bass puree could be due
to the presence of more fiber (1.5%) binding water (Table 1). The changes in the
viscosity by unit of temperature also increased with the temperature for all the infant
foods (Table 2), which means that the binding of ingredients with the water decreases at
elevated temperatures.
Time-Dependent Flow Behavior

Two methods were employed: hysteresis loop and shear stress decay. As an example,
the Figure 3 shows the flow curves obtained with controlled shear rate (CSR) at different
temperatures (5–65°C) for the selected vegetables with bass and vegetables and chicken
purees. The presence of a hysteresis area between the curves representing the shear stress
versus shear rate data for increasing and decreasing shear rates indicates that the flow of
the samples was time-dependent [22]. Therefore, the presence of hysteresis loops in the
Figure 3 indicates a thixotropic behavior of the vegetable-based infant purees. As a result,
the loop area designates the energy required to break down the structure that is not
recovered during the experimentation period [44], and depended on the kind of infant
puree and measurement temperature.
The hysteresis loop areas for all the baby foods studied at the different temperatures
are shown in the Table 3. The two-way analysis of variance showed that the kind of puree
and the temperature, as well as their interaction significantly affected the hysteresis loop
area, which was larger in the vegetables and chicken puree and smaller in the rice and
chicken puree, implying more and less damage to the structures, respectively. For the
same temperature, differences in the form (Figure 3) and magnitude of the hysteresis loop
(Table 3) were observed among the vegetable-based infant purees, these differences
being more noticeable at 5°C. However, the significance of the interaction when
considering the kind of puree sample and the temperature as factors, and with the
corresponding F value (Fint) being 94.76 (P = 0.000), indicated that the effect of
temperature on this area was different depending on the type of infant puree.
As expected, the loop area was significantly smaller at the higher temperatures for all
the samples. As a result, the hysteresis loop area became smaller as the temperature
increased from 5 to 65°C for each infant puree. This probably indicates that the
temperature effect on the molecular structure damages the puree suspension. However,
while the areas obtained at 65°C were less than half the values at 5°C for the rice and
chicken puree, these areas at 65°C were less than a quarter of the values at 5°C for the
rest of the vegetables-based infant purees (Table 3). Apparently, these results disagree
with the time-independent flow activation energy values obtained for the different
samples, which indicated that the rice and chicken puree is more affected by the
temperature variations. Moreover, at 65°C, there were no significant differences between
the hysteresis area values corresponding to the rice and chicken puree and the selected
vegetables with bass one. This phenomenon results from the fact that increasing the
temperature reduces the thixotropy, and therefore the differences between the samples
become less significant when the temperature increases. However, a comparison of the
straight loop areas between different viscous systems may not render valid conclusions
on the extension of the time-dependent structural breakdown [22]. In this chapter, the rice
and chicken and vegetables and chicken purees showed not only the smallest and largest
loop areas respectively, but also in the same order the lowest and highest resistance to
flow.
Figure 3. Flow curves obtained with controlled shear rate (CSR mode) measured by increasing (forward
measurements) and decreasing the shear rate (backward measurements) at temperatures of 5, 20, 35, 50
and 65°C. (a) Selected vegetables with bass puree; (b) Vegetables and chicken puree.
An alternative approach may be the use of the relative thixotropic area under the
ascending curve, defined as the ratio of the “absolute” hysteresis area, which provides a
more accurate study of the rate of the internal structural breakdown when comparing
systems with different viscosities [21]. Considering the relative thixotropic areas, a
significant kind of baby food-temperature interaction (Fint = 6.69, P = 0.000) was also
found. At the higher temperature (65°C), for the selected vegetables with bass, vegetables
and beef and vegetables and chicken purees, the relative areas were smaller but the
decrease with respect to the values at 5°C was lower than that for the directly registered
loop areas (Table 3). Conversely, for the rice and chicken puree, the relative thixotropic
areas increased as the temperature increased. This increase indicates that in the rice and
chicken puree the areas under the upstream data points decrease with the temperature
more significantly than their corresponding hysteresis loops, thus supporting the results
obtained for the time-independent flow behavior. Furthermore, the differences between
the samples were not the same when the relative areas were compared. Assuming that a
hysteresis loop area is an index of the energy needed to destroy the structure responsible
for the flow time dependence, the experimental data showed that at 5, 20 and 35°C the
vegetables and chicken puree was the one needing the highest energy to break down such
structure, while if the relative areas are considered the highest energy values at the same
temperatures were shown by the vegetables and beef puree. The power law model was
also used to describe the shear stress and shear rate data obtained at the CSR mode. The
model exhibited a better fit when the descending curves were considered.
The Table 3 shows the fitting parameters of the power law model for the
experimental data of the vegetable-based infant purees at the different temperatures
measured in the backward direction. The fitting of the experimental data was satisfactory
with R2 ≥ 0.99 for all the samples. Therefore, the application of the CSR mode
considering the downward curves leads to better fits of the experimental data to the
power law model than the CSS mode considering the ascending ones (Table 2). As
expected, the consistency coefficients had lower values for the down curves in the CSR
mode (Table 3) than for the up ones using the CSS approach, while the values of the flow
behavior index were greater for the downstream data points. In each puree type, the
consistency coefficient, K, significantly decreased with the temperature. Unlike the
consistency coefficient, n increased significantly with increasing the temperature, the
only exception being the vegetables and beef puree.
The time-dependent flow properties of the vegetables-based infant purees were also
measured by shearing the samples at constant shear rates of 5, 25, 50, 100 and 200 s –1
over a temperature range of 5-65°C. The rate and extent of the shear stress decay varied
among the baby foods, the temperatures and the shear rates (Figure 4). The observed
time-dependent flow behavior of the vegetable-based infant purees was modeled using
both an empirical equation − the Weltman model [45] − and the modified structural
kinetics approach proposed by Nguyen et al. [23]. The decreases in the shear stress were
more pronounced for the higher shear rates (Figure 4a) and lower temperatures (Figure
4b). A similar effect was reported for stirred yoghurt by Ramaswamy and Basak [46] and
O’Donnell and Butler [47] and for buttermilk by Butler and McNulty [48]. Their results
showed that after 60 min of shearing at various constant shear rates, yoghurt and
buttermilk continued to show structural decay.
Table 3. Experimental hysteresis area, relative thixotropic area values and

rheological parameters of the power law model describing the flow downward
curves of the vegetable-based infant purees
Vegetable- Temperature Hysteresis area Relative Power law model

based infant (°C) (Pa s-1) hysteresis area (Backward measurements)
puree type (%) K (Pa sn) n R2
Rice and 5 1178.80 ± 27.70Ad 42.40 ± 0.74Dd 10.99 ± 0.02Ad 0.54 ± 0.00Ba 1.00
B C
chicken 20 973.40 ± 26.80 c 46.30 ± 0.41 c 7.41 ± 0.21Bd 0.57 ± 0.01Ba 1.00
35 800.63 ± 6.86Cc 49.22 ± 0.70Bb,c 5.30 ± 0.16Cd 0.58 ± 0.01A,Ba 1.00
D
50 648.89 ± 11.51 b 50.94 ± 0.28 a 4.15 ± 0.58 d 0.59 ± 0.04A,Bb 1.00
A, B D
E
65 504.03 ± 4.14 b 52.29 ± 0.36Aa 2.45 ± 0.23Ec 0.65 ± 0.02Aa 1.00
A A
Selected 5 2015.25 ± 28.35 b 46.74 ± 0.15 c 24.01 ± 0.24 c 0.40 ± 0.00Bb 1.00
A
B A
vegetables 20 1346.15 ± 32.75 b 42.92 ± 1.12 c 18.86 ± 0.04Bc 0.40 ± 0.00Bb 1.00
with bass 35 1093.50 ± 10.50Cb 42.07 ± 0.93Ac 15.02 ± 0.73Cc 0.40 ± 0.01Bb 1.00
D A
50 865.30 ± 26.70 b 42.18 ± 1.02 b 12.74 ± 0.05Cb 0.39 ± 0.00Bb 1.00
E
65 501.17 ± 24.18 b 41.83 ± 6.52 a,b 4.98 ± 1.86Db,c 0.49 ± 0.02Ab 1.00
A
Vegetables 5 1748.11 ± 36.20 c 65.10 ± 1.04Aa

A 18.38 ± 0.81Ab 0.37 ± 0.00Cc 0.99
and beef 20 1333.46 ± 131.37Bb 62.72 ± 3.01Aa 13.44 ± 0.15Bb 0.39 ± 0.00Bb 1.00
B A
35 1087.50 ± 89.01 b 59.47 ± 4.26 a 10.21 ± 0.06Cb 0.40 ± 0.00Ab 1.00
C B
50 783.69 ± 27.70 b 49.83 ± 0.82 a 7.89 ± 0.10Dc 0.40 ± 0.00Ab 1.00
D C
65 327.25 ± 31.19 c 34.87 ± 2.43 b 6.84 ± 0.01 a,b 0.37 ± 0.00Cc 1.00
D
Vegetables 5 3811.30 ± 17.00Aa 58.20 ± 0.02Ab 41.84 ± 0.32Aa 0.33 ± 0.00Cd 0.99
and chicken 20 2853.00 ± 4.40Ba 55.53 ± 0.01A,Bb 31.97 ± 0.37Ba 0.34 ± 0.00B,Cc 1.00
35 2274.00 ± 26.30Ca 54.60 ± 0.98A-Ca,b 25.53 ± 0.64Ca 0.34 ± 0.00Bc 1.00
50 1430.95 ± 29.25Da 52.49 ± 1.02B,Ca 15.26 ± 0.49Da 0.37 ± 0.00Ab 1.00
65 661.27 ± 25.54Ea 51.55 ± 1.99Ca 8.97 ± 0.60Ea 0.36 ± 0.00Ac 1.00
Mean values (n = 3) ± standard deviation (SD). A-E For the same column and for the same vegetable-based
infant puree type and parameter, means without the same letter are significantly different (P < 0.05)
according to the LSD multiple range test. a-d For the same column and for the same temperature and
parameter, means without the same letter are significantly different (P < 0.05) according to the LSD
multiple range test.
The Weltman model (Eq. 5) indicates the logarithmic decrease in the shear stress
with the shearing time. The Weltman model regression parameters A and B are presented
to compare the stress decay behavior of the vegetable-based infant foods (Tables 4-7).
The three-way analysis of variance showed that the kind of baby food, the measurement
temperature and the shear rate significantly affected the A and B values, although the
three binary interactions between factors also had a significant effect on both parameters.
In most cases the experimental data fitted well to the Weltman model, but in the case of
the stress decay the fit to the Weltman model was worse when measured at 5 s–1 (R2 ≤
0.90), and sometimes even at 25 s–1 as in the case of the rice and chicken puree at 20 and
50°C (Table 4). This result is ascribed to a tiny build-up at low shear rates. Also, the
stress decay in the selected vegetables with bass puree fitted rather poorly to the Weltman
model at 65°C (Table 5). This fact can be explained as a result of an increased
intermolecular interaction because of desiccation occurring towards the end of the 40 min
shearing period at high temperature, as explained below.
(a) Rice and Chicken Baby Food at 35 °C

150
125 5 s-1 25 s-1 50 s-1 100 s-1 200 s-1 Weltman model
Shear stress (Pa)
100
75
50
25
0
0 500 1000 1500 2000 2500
Time (s)
(b) Vegetables and Beef Baby Food Sheared at 100 s-1
150
5 °C 20 °C 35 °C 50 °C 65 °C Weltman model
125
Shear stress (Pa)
100
75
50
25
0
0 500 1000 1500 2000 2500
Time (s)
Figure 4. Experimental and predicted shear stress values as a function of the shearing time. (a) Rice and
chicken puree at 35°C at different shear rates (5-200 s–1). (b) Vegetables and beef puree sheared at 100
s–1 at different temperatures (5-65°C).
Table 4. Parameters from the Weltman and second-order structural kinetic models
for the rice and chicken baby food at various temperatures and shear rates
Temperature Shear Weltman model Second-order structural kinetic model

(°C) rate A (Pa) -B (Pa) R 2 k × 102 (s–1) η0/ηe R2
–1
(s )
5 5 49.29 ± 0.01Ae 2.33 ± 0.02Ae 0.90 0.21 ± 0.01A,Be 1.59 ± 0.03C, Db 0.98
A A
25 104.93 ± 4.00 d 7.05 ± 0.04 d 0.97 0.35 ± 0.01Bd 2.14 ± 0.04Aa 0.99
50 145.78 ± 2.00Ac 10.40 ± 0.40Ac 0.98 0.42 ± 0.02Ac 2.27 ± 0.11Ba 0.99
A A A B
100 192.83 ± 1.10 b 14.44 ± 0.03 b 0.99 0.49 ± 0.01 b 2.36 ± 0.13 a 0.99
200 261.59 ± 5.00Aa 19.93 ± 1.03Aa 0.99 0.58 ± 0.02Aa 2.38 ± 0.01Ba 0.99
20 5 28.44 ± 1.40Be 1.89 ± 0.10Be 0.82 0.15 ± 0.03Bd 1.38 ± 0.07Dc 0.95
25 69.20 ± 1.10Bd 3.90 ± 0.10Bd 0.85 0.25 ± 0.01C c 1.78 ± 0.11Bb 0.99
50 95.50 ± 0.30Bc 6.32 ± 0.30Bc 0.95 0.33 ± 0.01Bb 2.07 ± 0.07Ba,b 0.99
100 139.61 ± 1.20Bb 9.93 ± 0.02Bb 0.98 0.41 ± 0.01Ba 2.25 ± 0.12Ba 0.99
B B C
200 191.18 ± 1.18 a 13.93 ± 0.03 a 0.99 0.46 ± 0.02 a 2.36 ± 0.03Ba 0.99
35 5 27.56 ± 2.12Be 1.48 ± 0.02Cd 0.95 0.30 ± 0.03Ad 1.72 ± 0.12Cb 0.98
25 58.20 ± 1.20Cd 3.94 ± 0.03Bc, d 0.99 0.42 ± 0.02Ac 2.16 ± 0.02Aa,b 0.98
50 74.00 ± 3.50Cc 5.32 ± 0.98Bb, c 0.99 0.47 ± 0.01Ab,c 2.23 ± 0.23Ba 0.98
100 106.17 ± 2.03 b 7.75 ± 1.20 a,b 0.99 0.51 ± 0.00Ab
C B,C 2.32 ± 0.11Ba 0.98
200 136.04 ± 4.04Ca 9.75 ± 0.10Ca 0.99 0.58 ± 0.01A,Ba 2.35 ± 0.03Ba 0.98
50 5 17.39 ± 0.81Ce 1.15 ± 0.02Dd 0.89 0.15 ± 0.02Bd 2.25 ± 0.01Bc 0.97
D C C A
25 38.12 ± 0.12 d 2.46 ± 0.40 c, d 0.89 0.20 ± 0.02 c,d 2.26 ± 0.11 c 0.98
50 54.93 ± 2.03Dc 4.47 ± 0.41Bb,c 0.90 0.25 ± 0.02Cb,c 2.97 ± 0.02Ab 0.99
100 76.15 ± 1.02Db 6.16 ± 1.20C,Db 0.95 0.28 ± 0.01Ca,b 3.03 ± 0.10Ab 0.99
200 110.75 ± 5.22Da 9.70 ± 0.20Ca 0.96 0.33 ± 0.01Da 3.43 ± 0.13Aa 0.99
65 5 13.57 ± 1.12Ce 1.17 ± 0.01Dc 0.90 0.27 ± 0.03Ac 3.31 ± 0.11Aa 0.98
E C A,B A
25 28.89 ± 2.10 d 1.78 ± 0.12 b,c 0.98 0.39 ± 0.01 b 2.08 ± 0.08 b 0.97
50 39.16 ± 1.03Ec 2.33 ± 0.23Ca-c 0.99 0.44 ± 0.02Ab 1.95 ± 0.01Bb,c 0.98
100 54.64 ± 1.32Eb 3.47 ± 0.22Da,b 0.98 0.40 ± 0.02Ba,b 1.97 ± 0.12Bb,c 0.99
200 72.58 ± 0.40Ea 3.77 ± 1.10Da 0.98 0.51 ± 0.03B, Ca 1.70 ± 0.02Cc 0.96
Mean values (n = 3) ± standard deviation (SD). A-E For the same column and for the same shear rate and
multiple range test. a-d For the same column and for the same temperature and parameter, means
without the same letter are significantly different (P < 0.05) according to the LSD multiple range test.
In all the vegetables-based infant purees, the parameter A representing the shear
stress needed for the structure to start degrading, increased significantly as the shear rate
increased but decreased significantly as the temperature increased. Similar results were
reported by Choi and Yoo [5] for banana baby food. In turn, the parameter B indicating
the quantity of structure degraded during shearing (time coefficient of thixotropic
breakdown) also increased as the shear rate increased, except for the vegetables and
chicken puree at 5 and 50°C at intermediate shear rates (Table 7). A negative value of B
measures how fast the shear stress drops from the initial value to the final equilibrium
(steady-state) value [11]. Analogously, the B parameter values decreased when the
measurement temperature increased, but again with the exception of the vegetables and
chicken puree at the higher temperatures over the shear rate range of 50-200 s–1.
Therefore, the baby foods became somewhat less thixotropic at higher temperatures. The
rice and chicken and vegetables and chicken purees showed the lowest and highest initial
stress values at all the temperatures (Tables 4 and 7), which were less than one-third at
65°C than at 5°C. It is interesting to note that the A and B values were higher for the
selected vegetables with bass puree (Table 5) than for the vegetables and beef one (Table
6). Additionally, the B value was higher for the vegetables and chicken puree than for the
rest of the baby foods at all the temperatures, indicating its higher rate of structural
breakdown by shearing and degree of thixotropy. According to Aguilar et al. [36], the
Weltman model parameter values depend on the size distribution of the solid particles
suspended in the continuous phase.
The structural kinetics model postulates that the change in the time-dependent flow
properties is associated with the shear-induced breakdown of the internal fluid structure
and that the rate of this breakdown during shear depends on the kinetics of the structured
state→non-structured state process [12, 22]. The rate constant, k can be considered as a
measure of the rate of structure breakdown, i.e., the degree of thixotropy. On the other
hand, the ratio of the initial to equilibrium viscosity, η0/ηe, can be considered as a relative
measurement of the amount of structure breakdown, or in other words as a relative
measure of the extent of the thixotropy [49]. The initial viscosity (t = 4 s), η0, was
obtained from the viscosity versus time data. As an example, the Figure 5 shows the
effect of the temperature on the thixotropic behavior of the selected vegetables with bass
and vegetables and chicken purees at different shear rates. When a material is sheared at a
constant shear rate, the viscosity of a thixotropic material will decrease over a period of
time, implying a progressive breakdown of structure [12]. According to Van Hecke et al.
[9] as the particles in carrot puree should be considered as non-deformable owing to their
cellulose content and high pulp volume fraction, the observed thixotropic effect could be
attributed to a large particle size distribution. The smaller particles could act as lubricant
for the flow of the larger particles [7], thereby reducing the viscosity as the shear
continues. The apparent viscosity decreased significantly with the time at all the shearing
levels and for all the temperatures, particularly in the initial stages of shear (Figure 5).
This implies that the configuration of the polysaccharides and proteins in the suspensions
is changed by the shearing force. Toward the end of the 40 min shearing period, the
viscosity tended towards a plateau, especially at the higher temperatures. Analogously,
the viscosity tends to decay towards an equilibrium viscosity more rapidly at high shear
rates, which is lower than that at low shear rates.
Table 5. Parameters from the Weltman and second-order structural

kinetic models for the selected vegetables with bass baby food
at various temperatures and shear rates

(°C) rate A (Pa) -B (Pa) R2 k × 102 (s–1) η0/ηe R2
(s–1)
5 5 95.25 ± 1.10Ae 5.44 ± 0.33Ad 0.97 0.36 ± 0.03Ae 1.87 ± 0.11Ab 0.97
25 144.78 ± 3.10Ad 10.48 ± 0.48Ac 0.99 0.52 ± 0.01Ad 2.23 ± 0.13Aa,b 0.99
50 174.77 ± 14.02Ac 12.67 ± 1.10Ab,c 0.99 0.61 ± 0.01Ac 2.36 ± 0.03Aa 0.99
100 205.88 ± 1.11Ab 14.49 ± 0.12Ab 0.99 0.72 ± 0.02Cb 2.32 ± 0.11Aa 0.99
200 249.42 ± 4.22Aa 17.43 ± 1.12Aa 1.00 0.86 ± 0.02B,Ca 2.43 ± 0.12Aa 0.99
20 5 68.17 ± 1.12Bd 2.99 ± 0.01Bd 0.94 0.23 ± 0.01Ae 1.64 ± 0.10A,Bb 0.90
25 113.25 ± 13.04Bc 7.75 ± 0.21Bc 0.99 0.40 ± 0.01Bd 2.18 ± 0.02Aa 0.99
50 128.68 ± 4.10Bc 8.94 ± 0.11Bc 0.98 0.47 ± 0.02Cc 2.19 ± 0.04Ba 0.99
100 167.47 ± 1.47Bb 12.08 ± 1.01Bb 1.00 0.57 ± 0.01Db 2.26 ± 0.12A,Ba 0.99
200 201.44 ± 2.11Ba 14.23 ± 0.12Ba 1.00 0.77 ± 0.01C,Da 2.39 ± 0.02Aa 0.99
35 5 48.97 ± 1.02Ce 1.68 ± 0.11Cd 0.88 0.27 ± 0.01Ad 1.38 ± 0.10Bc 0.98
25 83.47 ± 2.40Cd 5.28 ± 0.51Cc 0.96 0.40 ± 0.02Bc,d 1.94 ± 0.03A,Bb 0.99
50 106.73 ± 3.02B,Cc 6.97 ± 0.02Cc 0.98 0.46 ± 0.01Cb,c 2.03 ± 0.03Ca,b 0.99
100 133.61 ± 5.10Cb 9.16 ± 0.16Cb 0.99 0.58 ± 0.01Da,b 2.23 ± 0.12A,Ba,b 0.99
200 164.80 ± 2.10Ca 11.64 ± 1.02Ca 1.00 0.73 ± 0.03Da 2.26 ± 0.11Aa 0.99
50 5 40.70 ± 1.20De 1.58 ± 0.11Cd 0.91 0.31 ± 0.03Ae 1.46 ± 0.07Bc 0.96
25 67.14 ± 1.01C,Dd 3.46 ± 0.46Dc 0.99 0.55 ± 0.02Ad 1.69 ± 0.11Bb,c 0.99
50 83.40 ± 2.20C, Dc 4.55 ± 0.12Db,c 0.99 0.66 ± 0.01Ac 1.77 ± 0.03Da,b 0.99
100 101.49 ± 0.40Db 6.02 ± 1.02Da,b 0.99 0.81 ± 0.01Bb 1.96 ± 0.01B,Ca 0.99
200 128.06 ± 2.03Da 7.77 ± 0.22Da 0.99 0.90 ± 0.02Ba 1.98 ± 0.02Ba 0.99
65 5 27.68 ± 1.20Ee 0.93 ± 0.03De 0.81 Not applicable
25 45.40 ± 3.10Dd 1.38 ± 0.11Ed 0.88 Not applicable
50 61.30 ± 1.80Dc 2.93 ± 0.10Dc 0.94 Not applicable
100 73.60 ± 3.10Eb 3.84 ± 0.20Db 0.90 1.02 ± 0.03Ab 1.82 ± 0.02Cb 0.97
200 92.01 ± 2.00Ea 5.16 ± 0.10Ea 0.90 1.69 ± 0.05Aa 2.53 ± 0.04Aa 0.99
parameter, means without the same letter are significantly different (P < 0.05) according to the LSD multiple
range test. a-d For the same column and for the same temperature and parameter, means without the same
letter are significantly different (P < 0.05) according to the LSD multiple range test.
The equilibrium viscosity, ηe, was determined as outlined by Butler and McNulty
[48] and O’Donnell and Butler [47], whereby the latter portions of the viscosity curves
were fitted to an exponential decay curve of the form:
   e  ae bt (14)
where ηe was determined by choosing a trial value of ηe and calculating the line of best fit
between log(η – ηe) and t. The value of ηe that gave the best correlation was selected. For
the shear rate range investigated, the R2 values were all greater than 0.99. In this chapter,
the time at which the first viscosity measurement was taken (t = 4 s) was considered to be
the zero time (t0). Analogously, the k values were determined using the first 25 min of the
data. In practice, this time period covers most of the conditions of flow encountered in
the food industry [48]. Subsequently, to determine the applicability of the second-order
SKM to the viscosity data of the vegetable-based baby foods, [(η0 – ηe)/(η – ηe)] -1 was
plotted vs. time for each puree, shear rate and temperature. Some examples are shown in
the Figure 6. The closeness of the linear fit validates the assumption of the second-order
rate equation given by Eq. (6), at least during the first 25 min of the shearing time. In this
plot (Figure 6), the k values are the slopes of the lines. Note that with the large time
values, a variation between the experimental results and the straight line approximation is
to be expected as η approaches ηe when t → ∞ [9, 48].
for the vegetables and beef baby food at various temperatures and shear rates

(°C) rate (s–1) A (Pa) -B (Pa) R2 k × 102 (s–1) η0/ηe R2
5 5 60.33 ± 0.03Ae 2.49 ± 0.11Ac 0.95 A
0.35 ± 0.03 e A
1.53 ± 0.12 b 0.93
25 100.93 ± 2.02Ad 6.36 ± 0.36Ab 0.99 0.56 ± 0.01A,Bd 1.96 ± 0.03Aa 0.99
50 121.07 ± 3.02Ac 7.82 ± 0.10Bb 1.00 0.61 ± 0.04Ac 2.04 ± 0.04Ca 0.99
100 153.26 ± 4.02Ab 10.03 ± 1.03A,Ba 1.00 0.72 ± 0.02Ab 2.09 ± 0.05Ca 0.98
200 188.44 ± 3.86Aa 11.59 ± 0.30Aa 1.00 0.95 ± 0.03Aa 2.13 ± 0.05Ba 0.98
20 5 53.16 ± 1.05Be 2.05 ± 0.05Bc 0.88 0.28 ± 0.04Ad 1.46 ± 0.10Ac 0.97
25 92.72 ± 2.12Bd 6.43 ± 0.33Ab 0.96 A,B A
0.40 ± 0.01 c,d 2.15 ± 0.05 b 0.99
50 114.60 ± 3.80Ac 8.32 ± 0.18Ab 0.98 0.48 ± 0.01Bb,c 2.30 ± 0.03Ab 0.99
100 144.06 ± 1.02Bb 11.14 ± 1.10Aa 1.00 0.59 ± 0.02Ba,b 2.56 ± 0.01Aa 0.99
200 168.61 ± 1.99Ba 12.12 ± 0.12Aa 0.99 0.73 ± 0.03Ba 2.68 ± 0.06Aa 0.99
35 5 40.49 ± 1.21Ce 1.57 ± 0.20Ce 0.88 0.26 ± 0.02Ae 1.44 ± 0.11Ac 0.99
25 68.79 ± 2.20Cd 4.50 ± 0.15Bd 0.96 0.37 ± 0.01Bd 2.09 ± 0.06Ab 0.99
50 83.51 ± 1.14Bc 5.82 ± 0.12Cc 0.98 0.48 ± 0.01Bc 2.16 ± 0.03Bb 0.99
100 105.29 ± 1.09Cb 7.60 ± 0.10Bb 1.00 B
0.57 ± 0.01 b 2.27 ± 0.02Ba,b 0.99
200 126.20 ± 4.01Ca 9.06 ± 0.06Ba 0.99 0.73 ± 0.02Ba 2.45 ± 0.04A, Ba 0.99
50 5 28.00 ± 1.02De 1.11 ± 0.10De 0.96 A
0.42 ± 0.06 d 1.51 ± 0.10Ac 0.91
25 48.68 ± 2.31Dd 2.48 ± 0.40Cd 0.99 0.59 ± 0.04Ac 1.68 ± 0.10Bb,c 0.99
50 64.19 ± 1.10Cc 3.94 ± 0.13Dc 0.99 0.64 ± 0.01Ab,c 1.89 ± 0.01Da,b 0.99
100 78.06 ± 2.03Db 4.86 ± 0.14Cb 0.99 0.73 ± 0.02Ab 1.95 ± 0.00Da,b 0.99
200 98.28 ± 1.27Da 6.45 ± 0.32Ca 0.98 A
0.97 ± 0.01 a 2.14 ± 0.14Ba 0.99
65 5 21.50 ± 1.40Ee 0.72 ± 0.02Dd 0.84 Not applicable
25 34.18 ± 1.02Ed 1.82 ± 0.12Cc 0.96 Not applicable
50 40.19 ± 1.01Dc 2.12 ± 0.01Eb 0.97 0.50 ± 0.02Bb 1.91 ± 0.04Da 0.99
100 51.78 ± 0.11Eb 3.01 ± 0.00Ca 0.99 A
0.67 ± 0.01 a D
1.93 ± 0.03 a 0.99
200 63.67 ± 0.33Ea 3.16 ± 0.12Da 0.92 Not applicable
parameter, means without the same letter are significantly different (P < 0.05) according to the LSD multiple
range test. a-d For the same column and for the same temperature and parameter, means without the same
letter are significantly different (P < 0.05) according to the LSD multiple range test.
for the vegetables and chicken baby food at various temperatures and shear rates

(°C) rate (s–1) A (Pa) -B (Pa) R 2 k × 102 (s–1) η0/ηe R2
A A A A
5 5 126.06 ± 1.16 d 6.54 ± 0.27 c 0.96 0.37 ± 0.05 c 1.72 ± 0.02 c 0.97
25 244.27 ± 4.30Ac 19.95 ± 1.03Ab 0.99 0.56 ± 0.03B,Cb 2.62 ± 0.06Ba,b 0.99
50 249.45 ± 3.10Ac 19.17 ± 0.11Ab 0.99 0.61 ± 0.01Cb 2.65 ± 0.01Ba 0.99
A A B A
100 272.69 ± 1.81 b 19.60 ± 0.40 b 0.99 0.69 ± 0.08 a,b 2.67 ± 0.11 a 0.99
200 351.88 ± 1.11Aa 25.77 ± 0.22Aa 1.00 0.80 ± 0.04 Ca 2.33 ± 0.13Cb 0.99
20 5 110.03 ± 2.02Bc 5.65 ± 0.33Bc 0.92 0.28 ± 0.05Ad 1.69 ± 0.06A,Bd 0.99
25 175.60 ± 2.10Bb 14.91 ± 0.30Bb 0.98 0.53 ± 0.01C,Dc 2.93 ± 0.02Ab 0.99
50 189.28 ± 3.20Bb 15.16 ± 1.01Bb 0.99 0.68 ± 0.01Bb 2.63 ± 0.06Bc 0.99
B B B A
100 207.35 ± 2.70 a 16.18 ± 0.18 b 0.97 0.75 ± 0.02 b 2.73 ± 0.05 b,c 0.99
200 217.99 ± 7.10Ba 18.38 ± 0.29Ba 0.97 0.98 ± 0.02B a 3.27 ± 0.12Aa 0.99
35 5 73.75 ± 1.21Ce 3.25 ± 0.20Cd 0.95 0.33 ± 0.07Ad 1.57 ± 0.00Bd 0.97
25 117.94 ± 5.03Cd 8.35 ± 0.30Cc 0.97 0.43 ± 0.02Dc,d 2.27 ± 0.04Cc 0.99
C C D C
50 134.23 ± 0.23 c 9.62 ± 0.12 c 0.99 0.51 ± 0.03 b,c 2.30 ± 0.06 c 0.99
100 153.58 ± 2.11Cb 11.80 ± 0.02Cb 0.98 0.61 ± 0.05Ba,b 2.63 ± 0.04Aa 0.99
200 200.54 ± 2.02Ca 13.79 ± 1.10Ca 0.99 0.69 ± 0.01Ca 2.45 ± 0.03B,Cb 0.97
50 5 45.99 ± 2.09Dc 2.47 ± 0.12C, Db 0.92 0.29 ± 0.02Ae 1.64 ± 0.01A,Bc 0.98
D D
25 75.56 ± 1.20 b 6.47 ± 0.21 a 1.00 0.66 ± 0.02 d 2.92 ± 0.05Aa
B 0.99
E D
50 77.71 ± 2.20 b 6.38 ± 0.02 a 0.99 0.86 ± 0.01Ac 2.93 ± 0.05Aa 0.99
100 82.78 ± 0.20Eb 6.39 ± 0.30Da 0.99 1.08 ± 0.06Ab 2.87 ± 0.01Aa 0.99
200 92.17 ± 3.17Ea 6.82 ± 0.11Ea 0.98 1.76 ± 0.04Aa 2.72 ± 0.02Bb 0.98
65 5 41.83 ± 1.11De 2.33 ± 0.10Dd 0.99 Not applicable
25 63.50 ± 1.20Ed 4.34 ± 0.11Ec 0.97 1.10 ± 0.05A 2.16 ± 0.10C 1.00
D D
50 88.51 ± 1.30 c 6.84 ± 0.12 b 0.99 Not applicable
100 97.64 ± 0.34Db 7.14 ± 0.04Db 0.99 Not applicable
200 120.37 ± 2.20Da 9.91 ± 0.99Da 0.98 Not applicable
A-E
Mean values (n = 3) ± standard deviation (SD). For the same column and for the same shear rate and
multiple range test. a-d For the same column and for the same temperature and parameter, means without
the same letter are significantly different (P < 0.05) according to the LSD multiple range test.
For the different vegetables-based infant purees, the values of k and η0/ηe as a
function of the applied shear rate and the temperature are reported in Tables 4-7. Both
rate constant (k) and structural breakdown (η0/ηe) were significantly (P<0.05) affected by
all the main factors (kind of puree, temperature and shear rate) and all the interactions of
these factors.
For the rice and chicken infant puree k increased significantly as the shear rate
increased at all the temperatures investigated (Table 4). This means that the breakdown
rate of the structure subjected to external shear stress increased with increasing the
applied shear stress, and exponential power law models were also found to describe this
dependence of k on the shear rate at each temperature studied. The effect of the shear rate
on the k values was the same for the rest of the baby foods studied, increasing
significantly and exponentially with the shear rate at temperatures ranging between 5 and
50°C (Tables 5-7).
(a) Selected Vegetables with Bass Baby Food Sheared at 100 s-1
2.5 5 °C 20 °C 35 °C 50 °C 65 °C Structural kinetic model
2.0
Apparent viscosity (Pa s)
1.5
1.0
0.5
0.0
0 500 1000 1500 2000 2500
Time (s)
(b) Vegetables and Chicken Baby Food Sheared at 25 s-1
10
5 °C 20 °C 35 °C 50 °C 65 °C Structural kinetic model
8
Apparent viscosity (Pa s)
0
0 500 1000 1500 2000 2500
Time (s)
Figure 5. Experimental and predicted apparent viscosity values as a function of the shearing time at
different temperatures (5-65°C). (a) Selected vegetables with bass puree at a constant shear rate of 100
s–1; (b) Vegetables and chicken puree at a constant shear rate of 25 s –1.
(a) Rice and Chicken Baby Food at 50 °C

6
5 s-1 25 s-1 50 s-1 100 s-1 200 s-1
4
[(0 - e)/( - e)] -1
0
0 200 400 600 800 1000 1200 1400 1600
Time (s)
(b) Vegetables and Chicken Baby Food Sheared at 100 s-1
20
5 °C 20 °C 35 °C 50 °C 65 °C
15
[(0 - e)/( - e)] -1
10
0
0 200 400 600 800 1000 1200 1400 1600
Time (s)
Figure 6. Second-order structural kinetic model fits, Eq. (6), for two vegetable-based baby foods. (a)
Rice and chicken puree at 50°C and different shear rates; (b) Vegetables and chicken puree at a
constant shear rate of 100 s–1 and different temperatures.
Previous studies have also shown that for stirred yogurt, k could be related to the
shear rate using power law models [47]. In turn, the extent of the hysteresis increased
significantly as the shear rate increased in the temperature range of 5 to 50°C for the rice
and chicken puree, although the η0/ηe ratio decreased significantly with the shear rate at
65°C (Table 4). Nguyen et al. [23] also found that η0/ηe decreased with increasing the
shear rate. For selected vegetables with bass and vegetables and beef purees, the value of
η0/ηe also increased significantly with the shear rate in the same range of 5 to 50°C
(Tables 5, 6). Finally, for the vegetables and chicken puree, η0/ηe again tended to increase
with the shear rate, but in this case the highest values did not always correspond with the
highest shear rate.
It should be stated that at 65°C, for the selected vegetables with bass, vegetables and
beef, and vegetables and chicken purees, the shear stress decay data could not be fitted to
the second-order SKM for most shear rates as the viscosity increased towards the end of
the 40 min shearing period. This fact is ascribed to an increased extent of structuring due
to a water loss suffered by the samples under the experimental conditions used in this
study, which impeded the estimation of the value of ηe in these cases. However, this
effect was not detected during shearing, in the rice and chicken puree, probably due to the
fact that this product has higher moisture content. In contrast, meat protein denaturation
[10] and starch gelatinization [1, 2] occurring in the range of 65 to 80°C were found to be
the primary factors responsible for the unexpected rheological behavior of three of the
different baby foods in this temperature zone, which was confirmed by thermal transition
behavior. At each shear rate and infant puree studied, both the decay rate constant (k) and
the extent of the thixotropy did not have any trend with temperature between 5 and 65°C
(Tables 4-7). Nguyen et al. [23] reported that k values for starch pastes increased or
decreased with the temperature depending on the type of starch and on the applied shear
rate. For the products analyzed in this chapter, these characteristics will depend not only
on the starch content of the samples but also on the elaboration process conditions,
mainly on the severity of the cooking process and the degree of grinding.
Examination of the suitability of the time-dependent rheological models, as assessed
by the determination coefficients (Tables 4-7), indicated that except in the above-
mentioned cases, both were suitable for characterizing the time-dependent rheological
behavior of the vegetable-based infant purees. In all the baby foods studied, both the
Weltman parameters A and B, showed a clear tendency to increase and decrease in line
with the shear rate and the temperature, respectively. Therefore, the Weltman model
could be considered better than the second-order SKM for characterizing these products
under the conditions studied, besides being less tedious.
On the other hand, given that the Weltman parameters represent different measures of
the shear stress, one would expect to find a logarithmic relationship behavior between
them and the shear rate [46]. In the present chapter, with only five shear rates employed
for testing in the range of 5 to 200 s–1, exponential power law models were found to
describe the dependence of A and B on the shear rate for the rice and chicken puree, all
yielding high R2 (Table 8). In the case of the selected vegetables with bass and vegetables
and beef baby foods (Tables 9 and 10 respectively), power law models were also the most
adequate to characterize the variations in the A values as a function of the  .
Table 8. Power law, structural kinetic and Arrhenius models for the
rice and chicken baby food
Temperature (°C) Power law model R2 Second-order structural kinetic model R2

5 A = 24.16   1.00 k= 1.39×10-3   0.99
0.99
B = 0.985  
20 A = 12.62   1.00 k = 9.20×10-4   0.99

0.99
B = 0.732  
35 A = 13.88   1.00 k = 2.31×10-3   0.99
0.99
B = 0.686  
50 A = 7.71   1.00 k = 1.04×10-3   0.99
0.99
B = 0.427  
65 A = 6.57   1.00 k = 2.12×10-3   0.97

0.97
B = 0.653  
Arrehnius At 5 s–1, A/    Ea = 16.02 kJ mol–1 R2 = 0.95

model
At 25 s–1, A/    Ea = 16.52 kJ mol–1 R2 = 0.99
At 50 s–1, A/    Ea = 16.58 kJ mol–1 R2 = 0.99
At 100 s–1, A/    Ea = 16.25 kJ mol–1 R 2 = 0.99
At 200 s–1, A/    Ea = 16.16 kJ mol–1 R2 = 0.97
A = initial shear stress (Pa) in the Weltman model; B = time coefficient of thixotropic breakdown (Pa) in the
Weltman model; k = rate constant in the structural kinetic model;  = the shear rate (s–1); Ea = the
activation energy (kJ mol–1).
However, the logarithmic models were found to describe better the dependence of B
on the shear rate at some of the temperatures used. In the case of the vegetables and
chicken puree (Table 11), both the logarithmic and power models were found to be
satisfactory, dependent on the temperature to describe the variations in A as a function of
the   although these models had lower R2 than those fitted for A in the rest of the baby
foods. Moreover, a logarithmic model described the dependence of B on the shear rate
only at 35°C (Table 11).
Table 9. Power law and logarithmic, structural kinetic and

Arrhenius models for the selected vegetables
with bass baby food
Temperature (°C) Power law and logarithmic R2 Second-order structural kinetic R2

model model
5 A = 62.74   1.00 k = 2.45×10-3   1.00
1.00
B = 3.18ln(  
20 A = 42.77   1.00 k = 1.39×10-3   0.99
0.99
B = 3.03ln(  
35 A = 28.38   0.99 k = 1.72×10-3   0.99
0.99
B = 2.66ln(  
50 A = 24.78   1.00 k = 2.02×10-3   0.98
1.00
B = 0.818  
65 A = 16.26   1.00 Not applicable
0.95
B = 0.379  
Arrehnius At 5 s–1, A/    Ea = 15.55 kJ mol–1 R2 = 0.99

model
At 25 s–1, A/    Ea = 14.73 kJ mol–1 R2 = 0.98
At 50 s–1, A/    Ea = 13.14 kJ mol–1 R2 = 0.99
At 100 s–1, A/    Ea = 13.24 kJ mol–1 R2 = 0.98
At 200 s–1, A/    Ea = 12.66 kJ mol–1 R2 = 0.98
Weltman model; k = rate constant in the structural kinetic model;  = the shear rate (s–1); Ea = the
activation energy (kJ mol–1).
In order to use the Arrhenius approach to describe the effect of the temperature on the
flow behavior of the vegetable-based infant purees, the apparent viscosities were
evaluated, in this case, as A/  ratios for the different shear rates, and regressed against
the reciprocal absolute temperature. The activation energy (Ea) values show the
temperature sensitivity of the inherent sample viscosity, and given that the A values are
estimates of the shear stress prior to the structure decay, they can be expected to match
the Ea of the flow [46].
The associated Ea values of the stress/viscosity decay are also summarized in the
Tables 8-11, and for all the infant purees these values were slightly higher that the Ea
values reported above for the same products while obtaining the shear stress-shear rate
data under the upward ramp shearing sequences (CSS mode). The fits of the A/  ratios to
the Arrhenius model had lower R2 values and were more dependent on the shear rate in
the case of the vegetables and beef and vegetables and chicken baby foods (Tables 10 and
11).
Table 10. Power law and logarithmic, structural kinetic and

Arrhenius models for the vegetables and beef baby food
Temperature (°C) Power law and logarithmic model R2 Second-order structural kinetic model R2
5 A = 36.87    1.00 k = 2.32×10-3   0.98
1.00
B = 2.48ln(  
20 A = 32.65   1.00 k = 1.79×10-3   0.99
0.99
B = 2.83ln(  
35 A = 24.90   1.00 k = 1.60×10-3   0.99
1.00
B = 2.04ln(  
50 A = 16.29   1.00 k = 2.93×10-3   0.97
0.99
B = 0.528  
65 A = 13.26   1.00 Not applicable

0.98
B = 0.69ln(  
Arrehnius model At 5 s–1, A/    Ea = 13.97 kJ mol–1 R2 = 0.96
At 25 s–1, A/    Ea = 14.46 kJ mol–1 R2 = 0.94
At 50 s–1, A/    Ea = 14.29 kJ mol–1 R2 = 0.91
At 100 s–1, A/    Ea = 14.30 kJ mol–1 R2 = 0.93
At 200 s–1, A/    Ea = 13.94 kJ mol–1 R2 = 0.94
Weltman model; k = rate constant in the structural kinetic model;  = the shear rate (s–1); Ea = the activation
energy (kJ mol–1).
Table 11. Power law and logarithmic, structural kinetic and Arrhenius models for
the vegetables and chicken baby food
Temperature (°C) Power law and logarithmic model R2 Second-order structural kinetic model R2
5 A = 55.54ln(   0.94 k = 2.74×10-3   0.99
20 A = 29.42ln(   0.97 k = 1.72×10-3   0.98
35 A = 49.01   0.99 k= 2.32×10-3   0.99
B = 2.82ln(   1.00
50 A = 11.91ln(   0.94 k = 1.38×10-3   0.99

65 A = 26.20   0.99 Not applicable
Arrehnius At 5 s–1, A/    Ea = 16.01 kJ mol–1 R2 = 0.95
model
At 25 s–1, A/    Ea = 18.46 kJ mol–1 R2 = 0.99
At 50 s–1, A/    Ea = 15.55 kJ mol–1 R2 = 0.92
At 100 s–1, A/    Ea = 15.59 kJ mol–1 R2 = 0.90
At 200 s–1, A/    Ea = 15.83 kJ mol–1 R2 = 0.84
Weltman model; k = rate constant in the structural kinetic model;  = the shear rate (s–1); Ea = the activation
energy (kJ mol–1).
Oscillatory Shear Measurements of Baby Foods
From the stress sweep tests performed at several temperatures, a value of 3 Pa was
found to guarantee that all the rheological tests were performed within the LVE region.
The Figures 7-9 show the ω dependence of G' (storage modulus) and G" (loss modulus)
of the vegetable-based baby foods at 5-80°C. In the complete temperature range studied,
for the selected vegetables with bass, vegetables and beef and vegetables and chicken
purees, the values of G' were higher than those of G" at all the values of ω employed.
Thus, these products can be classified as weak gels [1, 10, 17]. For example, at 20°C
(Figure 7b), it was found that G' exceeded G", by approximately a factor of two for the
selected vegetables with bass and of three for the vegetables and beef and vegetables and
chicken purees. Under small deformation, the weak gels resemble the strong gels in their
mechanical behaviors, but as the deformation increases, the three-dimensional networks
undergo a progressive breakdown into smaller clusters [29]. Such behavior has been
earlier reported for these vegetable puree-based baby foods [1], sweet potato puree infant
food [2], meat-based strained baby foods [10], fruit puree based baby foods [17] and
salmon-based baby foods [3].
Analogously, at 5, 20 and 80°C, the rice and chicken puree behaved like a weak gel.
Nevertheless, at 20°C (Figure 7b), the G' exceeded G" by just about a factor of one and a
half for the rice and chicken product. Moreover, for the rice and chicken baby food at 35,
50 and 65°C, an intermediate step between typical behaviors of the concentrated
solutions and the gelled systems was observed (Figures 8a, 8b and 9a), indicating less
intense particle interactions at these temperatures, which contributed to weaken the less
elastic network of this type of puree. At a low ω, the G" is higher than G' and both
moduli vary sharply with the ω. This behavior is said ‘liquid-like’. As the ω increases, a
cross-over of the curves is experienced because the G' increases faster than G". Beyond
this intercept, the G" is lower than G' and the response of the material is said ‘solid-like’.
Such behavior is typical of a macromolecular solution with topological entanglements
[33]. In a weak gel, the network is held together by linkages which are stronger and more
permanent than the entanglements, but which are nonetheless much more susceptible to
disruption than those in the strong gel systems (or indeed in the entanglement case) [50].
Among the baby foods studied, it was found that the vegetables and chicken puree
exhibited a greater elasticity with a higher magnitude of G' at 5, 20, 35, 50 and 65°C,
while the rice and chicken puree was the least. This reflects an increased resistance to the
shear forces in the vegetables and chicken puree, which was more rigid than the rest.
Analogously, at 5, 50 and 65°C, the vegetables and chicken puree showed more viscous
properties with a higher magnitude of G" than the vegetables and beef, selected
vegetables with bass and rice and chicken purees (in that order) (Figures 7a, 8b, 9a). In
contrast, at 80°C, the vegetables and beef puree showed the highest elastic and viscous
properties (Figure 9b), reflecting an atypical structural reinforcement probably due to its
lower moisture content together a strong loss of the liquid phase itself occurring at this
temperature. The ratio of G" to G', represented by the resulting tangent of phase angle
(δ), gives a relative measure of the energy lost versus the energy stored in the cyclic
deformation [1]. A δ value of 90° indicates that the material is fully viscous, while an
elastic material is characterized by a δ value approaching 0°. For example, at 35°C, a δ of
30-54.2° and of 15.3 to 24° was observed for the rice and chicken and vegetables and
chicken purees, respectively. This result supports their higher and lower viscous nature,
respectively. The selected vegetables with bass and vegetables and beef purees showed
an intermediate moderately similar pattern with the δ values ranging from 27.8 to 31.1°
and from 20.4 to 30.9°, respectively.
(a)
1000
100
G' and G" (Pa)
Rice and chicken puree at 5 °C (G')

Rice and chicken puree at 5 °C (G")
10
Selected vegetables with bass puree at 5 °C (G')
Selected vegetables with bass puree at 5 °C (G")
Vegetables and beef puree at 5 °C (G')
Vegetables and beef puree at 5 °C (G")
Vegetables and chicken puree at 5 °C (G')
Vegetables and chicken puree at 5 °C (G")
1
0.1 1 10 100
 (rad s-1)
(b)
1000
100
G' and G" (Pa)

10 Rice and chicken puree at 20 °C (G")
1
0.1 1 10 100
 (rad s-1)
Figure 7. Effect of the vegetable-based baby food type on the mechanical spectra (G', storage modulus;
G", loss modulus; ω, frequency). (a) At 5°C; (b) At 20°C. Points represent mean values of six replicates
± standard deviation (SD).
(a)
1000
100
G' and G" (Pa)
10 Rice and chicken puree at 35 °C (G')

1
0.1 1 10 100
 (rad s-1)
(b)
1000
100
G' and G" (Pa)
10 Rice and chicken puree at 50 °C (G')

1
0.1 1 10 100
 (rad s-1)
G", loss modulus; ω, frequency). (a) At 35°C; (b) At 50°C. Points represent mean values of six
replicates ± standard deviation (SD).
The effects of the temperature on the G' and G" values of all the vegetable-based
infant purees are also shown in the Figures 7-9. In the rice and chicken puree, both G' and
G" values decreased systematically with an increase in the temperature up to 65°C while
a significant increase was noticed at 80°C. This result was also confirmed by the
temperature sweep tests (Figure 10a) where a sharp increase of both dynamic moduli and
a reduction of the phase angle are evidenced commencing in a temperature of about 73 ±
3°C. It is a well-known phenomenon that increasing the temperature would decrease the
viscosity of many fluid foods due to an increase in the kinetic energy [51]. In turn, for the
selected vegetables with bass, vegetable and beef and vegetables and chicken baby foods,
both the G' and G" values decreased between 5 and 50°C with a turn toward higher
values at 65°C and increasing further as the temperature was elevated to 80°C. Moreover,
the Figures 10b and 11 supported the earlier viscoelastic behavior of these three baby
foods at selected temperatures, where the temperature sweep tests evidenced a sharp
increase of both G' and G" at and above 61, 57 and 65 ± 3°C for the selected vegetables
with bass (Figure 10b), vegetable and beef (Figure 11a) and vegetables and chicken
(Figure 11b) baby foods, respectively. The δ values were especially variable in the high
temperature zone. In various starches containing different levels of cassia gum [52], with
further increase in the temperature, the G' and G" values increased more steeply and
reached maxima, owing to the gelatinization with the formation of a network of the
swollen starch granules.
(a)
1000
100
G' and G" (Pa)
10
1
0.1 1 10 100
 (rad s-1)
(b)
100000
10000
G' and G" (Pa)
1000
100

10 Selected vegetables with bass puree at 80 °C (G")
1
0.1 1 10 100
 (rad s-1)
G", loss modulus; ω, frequency). (a) At 65°C; (b) At 80°C. Points represent mean values of six
replicates ± standard deviation (SD).
The changes of both G' and G" values versus the ω followed patterns similar with the
temperature to that observed by Ahmed and Ramaswamy [10] in meat-based strained
baby foods. The sudden change in both storage and loss moduli values at and above 65°C
was attributed by these authors to a possible denaturation of the proteins present in
chicken, lamb and beef baby foods as supported by differential scanning calorimetry
(DSC). According to Campo-Deaño et al. [15], in suwari gels, a gel strengthening from
50 to 65°C involves the denaturation of actin and can be ascribed both an increase in the
number of cross-linking between protein aggregates or strands, and to a deposition of the
additional denatured proteins in the existing protein networks, thus reinforcing the gel
matrix. In contrast, in a sweet potato puree infant food, a deviation in the rheological
behavior at and above 65°C was attributed to the gelatinization and possible amylose-
lipid complex formation of the sweet potato starch as confirmed by the two distinct
thermal peak transitions observed (at 57 and 94.5°C) [2].
These same authors also observed that both pea and wax bean puree infant foods
exhibited an increase in both G' and G" values at 50°C [1], somewhat coinciding with the
initiation of the starch gelatinization, while the corn puree infant food exhibited a
stronger elastic characteristics with the starch gelatinization occurring at 80°C. The types
of polymorphs present in these formulations were believed to be a reason for these
different behaviors. Analogously, in dough, the temperature sweep tests at 1 Hz
evidenced a sharp increase of both G' and G" values at 50°C that was ascribed to the
starch gelatinization [26], while in cherry jam, the network coordination decreased when
increasing the temperature up to about 60°C, where a slight increase of the connectivity
occurred, also ascribed to the starch gelatinization.
Indeed, the baby food formulations investigated in this chapter contained a mixture of
different starches and proteins (Table 1). The starches undergo different order-disorder
phase transitions in the presence of an excess water, which are normally associated with
the gelatinization process [1]. The thermal processes are well established and used by the
food industry to supply foods with a high score in nutritive and organoleptic attributes
[53]. Positively, the nutritional quality of the product, as well as other quality attributes
like microbiological and sensory quality, are essential factors in the baby food industry,
and therefore, in addition to the traditional cooking, a final sterilizing process is
furthermore applied to the commercial baby foods investigated in this chapter. Surely,
under such treatments, both starch gelatinization and protein denaturation phenomena
occurred, but they are irreversible processes [54], and consequently, these structural
changes are unexpected to occur again following an additional thermal treatment. Indeed,
the thermal behavior of the vegetable-based infant purees was assessed by DSC, and
there was no evidence of thermal transitions in the tested infant purees in the range of 5
to 90°C (data are not shown), corroborating the irreversibility of both starch
gelatinization and protein denaturation phenomena. As a result, in this case, the
abnormalities observed in the rheological measurements of the baby foods at and above
65°C, are undoubtedly associated with increased intermolecular interaction because of

loss of solvent taking place in the tested samples at the higher temperatures under the test
conditions used in this chapter, and with a consequent increase in the volume fractions of
these suspensions.
(a)
Rice and Chicken Baby Food
100000 80
G' (Pa)
G" (Pa)
10000 Phase angle (°)
60
1000
G' and G" (Pa)
100 40
 (°)
10
20
0.1 0
0 10 20 30 40 50 60 70 80 90 100
Temperature (°C)
(b)
Selected Vegetables with Bass Baby Food
100000 80
G' (Pa)
G" (Pa)
60
G' and G" (Pa)
1000
 (°)
100 40
10
20
0.1 0
0 10 20 30 40 50 60 70 80 90 100
Temperature (°C)
Figure 10. Temperature sweeps of the vegetable-based baby foods heated from 5 up to 90°C (G',
storage modulus; G", loss modulus; δ, phase angle). (a) Rice and chicken baby food; (b) Selected
vegetables with bass baby food. Points represent individual values of one of three replicates.
(a)
Vegetables and Beef Baby Food
100000 80
G' (Pa)
G" (Pa)
60
1000
G' and G" (Pa)
100 40
 (°)
10
20
0.1 0
0 10 20 30 40 50 60 70 80 90 100
Temperature (°C)
(b)
Vegetables and Chicken Baby Food
100000 80
G' (Pa)
G" (Pa)
60
1000
G' and G" (Pa)
100 40
 (°)
10
20
0.1 0
0 10 20 30 40 50 60 70 80 90 100
Temperature (°C)
Figure 11. Temperature sweeps of the vegetable-based baby foods heated from 5 up to 90°C (G',
storage modulus; G", loss modulus; δ, phase angle). (a) Vegetables and beef baby food; (b) Vegetables
and chicken baby food. Points represent individual values of one of three replicates.
On the other hand, the ω dependence of the G' and G" values can provide valuable
information about the structure of a gel [55]. Given that the appearance of the data on
logarithmic coordinates was nearly linear (Figures 12 and 13), it was possible to model
the storage and loss moduli as a power function of the oscillatory ω (Eqs. (7) and (8)).
In turn, the Table 12 shows the power law parameters for both G' and G" moduli as a
function of the temperature. The R2 values were always higher than 0.92 and 0.94 at the
ranges of 5 to 65°C and of 5 to 50°C, respectively. At 5-50°C, for the selected vegetables
with bass, vegetables and beef and vegetables and chicken purees, the correlation
coefficients of regressions were higher for G'0 than for G"0, indicating that the behavior
of G" was less linear than that of G". A comparison of these parameters showed that, in
general, G'0 and G"0 were greater in the vegetables and chicken puree than in the rest,
which is associated with a denser structure accounting for the higher rigidity of this baby
food.
According to Carreau et al. [56], the higher storage modulus value indicates either a
strong particle-particle interaction or a network type structure in a stabilized form or both.
The rice and chicken puree was weaker than the other baby foods because it had the
lowest G'0 and G"0 values and it was more dependent on the ω. These higher n' and n"
values are partially ascribed to the fact that the rice and chicken puree had the highest
moisture content and syneresis, whilst the vegetables and chicken puree, with
predominantly lower n’ and n” values, had the lowest expressible water percentage.
Analogously, the moisture content also negatively influenced all the viscoelastic
parameters of Spanish honeys [28]. In turn, Herranz et al. [55] observed a negative
correlation of n' with the water binding capacity, confirming that when a network gel
approximates to a true gel (low n'), the gel functionality is better and hence the water
binding capacity noticeably increases. According to Chen et al. [57], a well-structured
protein matrix with a large number of rheological units cross-linked by small filamentous
threads produces a fine, uniform structure with numerous small pores, which would
probably result in more absorptive capacity and better water retention.
The Table 12 shows that G'0 was 30.12 (Pa sn') at 5°C decreasing to 6.17 (Pa sn’) at
65°C, but sharply increasing to 1407.19 (Pa sn') at 80°C for the rice and chicken baby
food. The same temperature effect was observed for the rest of the vegetable-based baby
foods, although in the case of the selected vegetables with bass (Table 12) and vegetables
and beef baby foods (Table 13), higher significant G'0 values at 65°C would indicate
higher degrees of elasticity when compared with the G'0 values at 50°C. Therefore, in the
four vegetable-based baby foods investigated, the G'0 values decreased significantly with
an ncrease in the temperature up to 50°C. The results for G'0 are in accordance with those
obtained by Ahmed and Ramaswamy [10] when measuring the dynamic rheological
parameters over the same temperature range of 5-50°C for commercial meat-based
strained baby foods. It is evident from the data shown in the Tables 12 and 13 that the
G"0 values followed similar trends when compared with G'0 values for all the types of
samples. Moreover, the selected vegetables with bass, vegetables and beef and vegetables
and chicken purees exhibited a more solid-like characteristics in the temperature range of
5-80°C, because the associated G'0 values for G' were higher that the G"0 values for G".
In contrast, the rice and chicken puree with the highest water content exhibited more
liquid-like characteristics in the temperature range of 35-65°C. Analogously, the

magnitudes of G'0 and G"0 also decreased with the increase in the water content in
different honeys [27, 28]. On the other hand, the difference (G'0 – G"0) was used as a new
measure of the gel strength in surimi [58].
8
5 °C (G') 5 °C (G") 20 °C (G') 20 °C (G")
35 °C (G') 35 °C (G") 50 °C (G') 50 °C (G")
6
log G', log G" (Pa)
0
-3 -2 -1 0 1 2 3 4 5
log  (rad s-1)
(b)
Selected Vegetables with Bass
8
5 °C (G') 5 °C (G") 20 °C (G') 20 °C (G")
35 °C (G') 35 °C (G") 50 °C (G') 50 °C (G")
6
log G', log G" (Pa)
0
-3 -2 -1 0 1 2 3 4 5
log  (rad s-1)
Figure 12. Effect of the temperature ranging between 5 and 50°C on the mechanical spectra of the
vegetable-based baby foods (G', storage modulus; G", loss modulus; ω, frequency). (a) Rice and
chicken baby food; (b) Selected vegetables with bass baby food. Points represent individual values of
one of six replicates.
(a) Vegetables and Beef Baby Food
8
5 °C (G') 5 °C (G") 20 °C (G') 20 °C (G")
35 °C (G') 35 °C (G") 50 °C (G') 50 °C (G")
6
log G', log G" (Pa)
0
-3 -2 -1 0 1 2 3 4 5
log  (rad s-1)
(b)
Vegetables and Chicken Baby Food
8
5 °C (G') 5 °C (G") 20 °C (G') 20 °C (G")
35 °C (G') 35 °C (G") 50 °C (G') 50 °C (G")
6
log G', log G" (Pa)
0
-3 -2 -1 0 1 2 3 4 5
log  (rad s-1)
Figure 13. Effect of the temperature ranging between 5 and 50°C on the mechanical spectra of the
vegetable-based baby foods (G', storage modulus; G", loss modulus; ω, frequency). (a) Vegetables and
beef baby food; (b) Vegetables and chicken baby food. Points represent individual values of one of six
replicates.
In addition, the magnitudes of the straight lines were small (Tables 12 and 13), and at
5-50°C the n" values were higher than the n' values for the selected vegetables with bass,
vegetables and beef and vegetables and chicken baby foods, showing that G" was more
frequency-dependent than G' in this temperature range. A good gel is known to exhibit
the same proportional change in G' and G" with the ω over a wide range; in other words,
n' and n" must be identical [58]. Consequently, this type of behavior is also associated
with a weak gel formation [17]. Rosalina and Bhattacharya [29] reported similar behavior
during the dynamic rheological measurements of starch gels. In contrast, for these three
baby foods at 65 and 80°C, G' was more frequency-dependent than G", which it is
ascribed to loss of liquid phase because of the earlier mentioned desiccation of the
samples occurring in both temperatures. In other baby foods, Ahmed and Ramaswamy [2,
10] obtained lower values of either n' or n" than those obtained in this chapter. The
inclusion of olive oil in the formulation of these baby foods in percentages ranging from
1.3 to 2 (Table 1) could account for the fact of these vegetable-based infant purees were
weaker gels than those studied previously by other authors because olive oil behaves as a
soft filler [59]. However, the n" values obtained for the vegetables and chicken puree at
20-80°C are similar to the data reported by other authors such as Ahmed and
Ramaswamy [1] for wax bean puree infant food, and Rosalina and Bhattacharya [29] for
four types of starch gels, although in both cases G' showed lower dependence on the ω (n'
< 0.1).
The material functions of a food system in the LVE regime can be well described by
only two parameters (A and z) [26]. This approach, named the weak gel model, was also
applied to the baby foods presented in this chapter. The strength of the network, A, and
the coordination number, z, were computed using the Eq. (9) and are also reported in the
Tables 12 and 13. In the range of 35-65°C, the rice and chicken baby food behaved as a
macromolecular solution as evidenced by the presence of the moduli crossover (Figures
8a, 8b and 9a), thus indicating the impossibility to find a realistic value of the z parameter
when using the definition of critical gel reported in the Eq. (9). For that reason, the weak
gel model was not applied for the characterization of the rice and chicken puree in this
temperature range (Table 12). For the selected vegetables with bass, vegetable and beef
and vegetables and chicken baby foods, it can seen that the coordination number was
almost constant up to 35°C, whilst the A value decreased significantly with increasing the
temperature up to 50°C. This was also confirmed by the temperature sweep tests shown
in the Figures 10b, 11a and 11b; both moduli decreased continuously and the G' and G"
curves appeared slightly parallel between 5 and 50°C. This means that the temperature
has no effects on the extension of the network between 5 and 35°C for these baby foods
as already evidenced by the almost constant value of the z parameter. On the contrary, a
reduction of the network strength between 5 and 50°C was observed by the decrease of
both G' and G" moduli For frozen and refrigerated cheeses, both values of z and A
decreased as the temperature increased from 10 to 50°C [32]. It was found that at each
temperature studied, the vegetable and beef and vegetables and chicken baby foods had a
similar but significantly higher value of z, reflecting a higher number of interactions, but
the interactions were stronger for the vegetables and chicken baby food because the A
value, at each temperature, was significantly higher than that one exhibited by the other
three studied baby foods. It is worth noting that the parameters so determined indicate
that the network coordination decreases when increasing the temperature up to about
60°C, where a slight increase of the connectivity occurs, probably because of the solvent
evaporation.
There are just few studies which have modeled the power law parameters of the
dynamic viscoelastic properties of foods as function of the temperature. Ahmed and
Ramaswamy [1] described that the temperature effect was not systematic in the
evaluation of G'0, G"0, n' and n" of some vegetable-puree based baby foods between 20
and 80°C. In developed in laboratory salmon-based baby foods, Ramamoorthi et al. [3]
modeled the temperature dependency of G'0 and G"0 in a constant ω of 10 rad s–1 using
the Arrhenius’s model. Analogously, in two differently processed squid surimis, Campo-
Deaño et al. [60] modeled the temperature dependency between 10 and 45°C of the
complex viscosity (η*) in a constant ω of 0.628 rad s–1 using an Arrhenius equation. For
practical purposes, then, the Ea is used as a material–specific parameter to quantify the
decrease of the η* when the temperature is raised.
The Ea analysis has been applied to liquid foods [32]. From the molecular point of
view, according to the fluids flow theory, the Ea can be interpreted as the energy required
by a molecule to jump from one equilibrium position to another position in the liquid
medium [32]. It can therefore be used as a measurement of a characteristic thermal
resistance to the breakdown by other food gels [61].
In this case, the range of temperature considered for the four baby foods was 5-50°C.
In this range, the thermal effect disrupted the junction zones stabilized by thermolabile
interactions, such us hydrogen bonds or electrostatic interactions as evidenced earlier. As
it can be seen (Table 14), the Arrhenius’s model has well-described the experimental
values of G'0 and G"0, mainly for the vegetables and chicken baby food with R2 higher
than 0.97. However, it has not suited well for the values of n' and n". Overall, the Ea
values were higher for the rice and chicken baby food than for the rest of the baby foods.
For the rice and chicken puree, consequently, took longer to soften with heat, producing
an increase in the Ea value. This may be due to its higher viscous nature, i.e., “liquid-like”
behavior at 35 and 50°C (Figure 10). The selected vegetables with bass and vegetables
and beef baby foods had intermediate Ea values for G'0 and G"0, and the vegetables and
chicken product recorded the lower Ea values for both G'0 and G"0 (13.49 kJ mol–1 and
10.89 kJ mol–1, respectively).
According to Tunick [61], higher values of the Ea of cheeses can be related to more
resistance to break down of its body. In this chapter, the lower values of the Ea for the
vegetables and chicken baby food could indicate that its structure breaks down more
easily than that of the rice and chicken one. This would reflect a negative correlation
between the Ea and the gel strength because increased thermal resistance to gel rupture
means decreased elasticity, and hence a decrease in the A value. This result was
consistent with that observed by Campo-Deaño et al. [60] in two differently processed
squid surimis. Also, the lower value of the Ea for frozen cheeses indicated that its
structure breaks down more rapidly than that for refrigerated cheeses [32].
Table 12. Power law parameters of Eqs. (7), (8) and (9) from the frequency sweep tests, at 5-80°C, for the rice and chicken and
selected vegetables with bass baby foods
Vegetable- Tª G' G" G*

based baby (°C) (G'0 ± SD) n' R2 (G"0 ± SD) n" R2 (A ± SD) z R2
food type (Pa sn') (Pa sn") (Pa s1/z)
Rice and 5 30.12±0.70Ad 0.381±0.012Ca 0.998 23.47±0.90Ad 0.376±0.010Ca 0.993 37.52±1.75Ad 2.65±0.07Bc 0.997
A D B C A
chicken 20 29.41±0.53 d 0.336±0.007 a 0.987 19.82±0.28 d 0.365±0.005 a 0.992 35.51±0.35 d 2.90±0.03Ac 0.990
35 10.25±0.44Bd 0.483±0.014Ba 0.988 11.01±0.19Cc 0.414±0.001Ba 0.991 Not applicable Not applicable -
50 8.60±0.28Bc 0.526±0.030A, Ba 0.995 10.11±0.20Cc 0.441±0.011Aa 0.995 Not applicable Not applicable -
C A D A -
65 6.17±0.99 c 0.559±0.025 a 0.994 8.34±1.03 b 0.440±0.019 a 0.989 Not applicable Not applicable
80 1407.19±86.07 0.420±0.021 0.921 654.24±40.93 0.433±0.030 0.936 1553.56±94.62 2.37±0.12 0.926

Selected 5 240.26±14.21Ac 0.176±0.006Db 0.998 106.92±1.30Ac 0.206±0.007Cb 0.959 263.04±13.71Ac 5.52±0.12Bb 0.996
vegetables 20 212.68±2.32Bc 0.160±0.003Eb 0.990 91.57±1.17Bb 0.210±0.005B, Cc 0.981 231.59±1.65Bc 5.91±0.07Ab 0.989
with bass 35 148.53±4.56Cc 0.190±0.005Cb 0.993 76.75±1.63Cb 0.200±0.002Cd 0.971 167.41±4.72Cc 5.22±0.11Cb 0.995
D B D A, B D D
50 97.22±9.17 b 0.207±0.004 b 0.988 52.54±1.16 b 0.222±0.004 b 0.977 110.67±8.60 b 4.75±0.10 b 0.989
65 144.10±14.89Cb 0.280±0.007Ab 0.921 73.47±4.01Ca 0.230±0.009Ab 0.993 163.24±14.91Ca 3.75±0.09Eb 0.947
80 2222.61±280.35 0.336±0.010 0.868 846.94±111.15 0.279±0.012 0.881 2394.01±299.53 3.05±0.09 0.880
Mean values (n = 6) ± standard deviation (SD). A-E For the same column and for the same vegetable-based infant puree type and parameter, means without the same
letter are significantly different (P < 0.05) according to the LSD multiple range test.a-d For the same column and for the same temperature and parameter, means
without the same letter are significantly different (P < 0.05) according to the LSD multiple range test. Values at 80°C were discarded for statistical analyses.
Table 13. Power law parameters of Eqs. (7), (8) and (9) from the frequency sweep tests, at 5-80°C, for the vegetables
and beef and vegetables and chicken baby foods
Vegetable-based Tª G' G" G*

baby food type (°C) (G'0 ± SD) n' R2 (G"0 ± SD) n" R2 (A ± SD) z R2
(Pa sn') (Pa sn") (Pa s1/z)
Vegetables 5 329.37±5.08Ab 0.154±0.004Cc 0.997 119.15±4.26Ab 0.200±0.009Cb 0.940 350.70±6.16Ab 6.23±0.17Aa 0.993
B C B A B
and beef 20 281.78±9.38 b 0.143±0.005 c 0.992 94.29±0.58 b 0.225±0.004 b 0.977 297.69±9.06 b 6.54±0.19Aa 0.990
35 210.94±5.05Cb 0.137±0.003Cc 0.988 80.53±0.22Ca 0.215±0.004A, Bb 0.974 226.22±4.74Cb 6.69±0.10Aa 0.988
50 105.79±7.91Eb 0.200±0.011Bb 0.986 57.54±1.37Db 0.206±0.006B, Cb 0.976 120.57±7.58Eb 4.99±0.27Bb 0.988
D A D B, C D C
65 139.80±22.73 b 0.237±0.019 c 0.922 66.49±10.57 a 0.204±0.002 c 0.991 155.94±25.74 a 4.38±0.28 a 0.947
80 4637.77±2255.46 0.309±0.011 0.910 1668.58±746.75 0.224±0.025 0.665 4992.49±2434.27 3.32±0.09 0.916
Vegetables 5 453.38±23.87Aa 0.143±0.005Cc 0.998 133.81±4.55Aa 0.207±0.010Ab 0.957 473.16±24.13Aa 6.70±0.27Aa 0.995
and chicken 20 313.27±9.93Ba 0.144±0.006Cc 0.994 100.73±3.21Ba 0.218±0.007Ab, c 0.967 329.53±10.26Ba 6.54±0.23A, Ba 0.993
35 253.76±11.41Ca 0.143±0.005Cc 0.992 77.16±0.34Cb 0.219±0.006Ab 0.965 265.62±11.05Ca 6.62±0.19Aa 0.990
D B C A D B
50 198.87±23.84 a 0.162±0.004 c 0.989 71.14±5.81 a 0.206±0.007 b 0.965 211.51±24.34 a 5.97±0.17 a 0.997
65 199.77±31.73Da 0.220±0.012Ac 0.950 83.52±9.61Ca 0.187±0.005Bc 0.985 217.65±33.24C, Da 4.70±0.25Ca 0.967
80 3700.75±2119.65 0.284±0.028 0.897 1133.46±563.38 0.241±0.024 0.803 3788.45±2072.07 3.60±0.32 0.901
Mean values (n = 6) ± standard deviation (SD). A-E For the same column and for the same vegetable-based infant puree type and parameter, means without the same
letter are significantly different (P < 0.05) according to the LSD multiple range test.a-d For the same column and for the same temperature and parameter, means
without the same letter are significantly different (P < 0.05) according to the LSD multiple range test. Values at 80°C were discarded for statistical analyses.
Table 14. Values for the parameters of the power law model for storage (G') and loss (G") moduli, as a function of the
oscillatory frequency (ω): Arrhenius and quadratic models for describing the variation with the temperature for the vegetable-
based baby foods
Vegetable-based Parameters of the power Quadratic: G'0 = a· T2 + bT + c

baby food type law model for G' and G"
A0 Ea (kJ mol–1) R2 a(C–2) b(C–1) c R2
Rice and G'0 (Pasn') 0.001 23.87 0.853 -10–3 -0.50 34.45 0.844
chicken G"0 (Pa sn") 0.03 15.51 0.919 3.1·10–3 -0.49 26.52 0.934
n' Not applicable 10–4 -1.5·10–3 0.37 0.812
n" Not applicable 4·10–5 -7·10–4 0.37 0.908
Selected vegetables G'0 (Pa sn') 0.37 15.17 0.930 -2.64·10–2 -1.84 252.57 0.990
with bass G"0 (Pa sn") 0.83 11.37 0.930 -9.8·10–3 -0.64 109.89 0.997
n' Not applicable 4·10–5 -1.2·10–3 0.18 0.855
n" Not applicable Not applicable
Vegetables G'0 (Pa sn') 0.15 18.16 0.874 -6.4·10–2 -1.43 337.55 1
and beef G"0 (Pa sn") 0.81 11.59 0.969 2.1·10–3 -1.44 125.27 0.990
n' Not applicable 8·10–5 -3.6·10–3 0.17 0.916
n" Not applicable -4·10–5 2.1·10–3 0.19 0.818
Vegetables G'0 (Pa sn') 1.29 13.49 0.992 9.47·10–2 -10.69 500.69 0.992
and chicken G"0 (Pa sn") 1.17 10.89 0.970 3.01·10–2 -3.06 148.79 0.999
n' Not applicable 2·10–5 -7·10–4 0.15 0.908
n" Not applicable -3·10–5 1.5·10–3 0.20 0.994
Analogously, in cheese, the rheological analysis showed that the fresh cheese had the
lowest Ea, and the highest G' and G" values at 50°C [61] and, in addition, the protein and
moisture contents were highly correlated with the Ea. Yoo [27] and Oroian et al. [28] also
observed that in Korean and Spanish honeys the Ea values were a function of the moisture
content-temperature. Certainly, the dry matter content of the vegetable-based infant
purees had an effect on their activation energy: the lower it was, the higher the Ea value
was. It could be therefore concluded that the Ea value does depend on the water content
of the baby foods. However, the susceptibility of the vegetable-based infant purees to the
changes in the temperature may also result from other factors, such as the content and
hydration degree of the protein substances that are responsible for the formation of the
pseudogel structure of the baby foods. Another factor could be the content and
gelatinization degree of starch. Ramamoorthi et al. [3] reported similar Ea values for both
G' and G" moduli than those observed for the rice and chicken baby food when
measuring the dynamic rheological parameters over a temperature range of 25-55°C for
commercial chicken gravy samples. The authors just cited reported similar Ea values for
G" for different thermally treated salmon baby foods than those observed for all the
vegetable-based infant baby foods investigated in this chapter, whilst the Ea values for G'
were lower ranging from 3.16 to 6.29 kJ mol–1. In turn, the Ea values, between 10 and
45°C, calculated from the temperature sweeps for two differently processed squid surimis
[61], very similar to those obtained for G" in the case of the selected vegetables with
bass, vegetables and beef and vegetables and chicken baby foods.
Then, a quadratic (second-degree polynomial function) function was also used for
modeling the G'0, G"0, n' and n" values (Table 14), with one exception. The quadratic
function did not suit well the temperature dependency of n" for the selected vegetables
with bass baby food. However, for the rice and chicken, vegetables and beef and
vegetables and chicken baby foods, this function described the temperature dependency
of both G'0 and G"0 values better than the Arrhenius’s model, with values of R2 higher for
the quadratic function. Augusto et al. [62] modeled the values of n' and n" as a second-
order polynomial function in relation to the temperature in peach juices with fibers (0-
40°C), while the values of G'0 and G"0 were modeled using the Arrhenius equation. In
siriguela (Spondias purpurea L.) pulp, G'0 and G"0 showed a sigmoidal decay behavior
with the temperature ranging between 0 and 80°C [63]. According to the authors just
cited, it is important to highlight that the property behavior in relation to the temperature
is function not only of the product itself but also of the studied temperature range.
TTS Method
The Figures 14 and 15 show the changes in the ln (G' and G") versus ln (ω) for the
vegetable-based baby foods at 5-50°C. Because the frequency sweep tests allow the
viscoelastic properties to be determined on various timescales, their results are said to
constitute the fingerprints of the gel viscoelasticity [58]. The application of the TTS
principle requires a data collection at multiple frequencies under isothermal conditions

and the repetition of these measurements at several temperatures [64]. The viscoelastic
properties collected at one temperature and (and at multiple frequencies) may be shifted
along the frequency axix [34].
(a) Rice and Chicken Baby Foods

7
6
5 °C (G') 20 °C (G') 35 °C (G') 50 °C (G')
4
log G' (Pa)
0
1 2 3 4 5 6
log G" (Pa)
(b) Selected Vegetables with Bass Baby Food

7
5 °C (G') 20 °C (G') 35 °C (G') 50 °C (G')

6
4
log G' (Pa)
0
3 4 5 6
log G" (Pa)
Figure 14. Modified Cole-Cole plot for the vegetable-based baby foods. (G', storage modulus; G", loss
modulus). (a) Rice and chicken baby food; (b) Selected vegetables with bass baby food. Points
represent individual values of one of six replicates.
The modified Cole-Cole plot was used to decide on the temperature dependency of
the isotherms (Figures 16 and 17). The technique suppresses the effect of the temperature
by plotting the log G' vs. log G" [31]. The lack of overlap of the isotherms to form a
single straight line, and the marked shift between the isortherms would indicate a
temperature dependency and the need for vertical shifts to form a reference master curve
[34]. Nevertheless, while relationships like Arrhenius, Williams-Landel-Ferry (WLF),
exist to define the aT variation, no such empirical relationships have been reported in the
literature for the vertical shifts factors bT [65].

7
5 °C (G') 20 °C (G') 35 °C (G') 50 °C (G')
4
log G' (Pa)
0
3 4 5 6
log G" (Pa)
(b) Vegetables and Chicken Baby Food
7
5 °C (G') 20 °C (G') 35 °C (G') 50 °C (G')
4
log G' (Pa)
0
3 4 5 6
log G" (Pa)
Figure 15. Modified Cole-Cole plot for the vegetable-based baby foods. (G', storage modulus; G", loss
modulus). (a) Vegetables and beef baby food; (b) Vegetables and chicken baby food. Points represent
individual values of one of six replicates.

8
7 5 °C (G') 20 °C (G') 35 °C (G') 50 °C (G')
5 °C (G") 20 °C (G") 35 °C (G") 50 °C (G")

6
log G', log G" (Pa)
5
0
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9
log aT (rad s-1)
(b) Selected Vegetables with Bass Baby Food

8
5C (G') 20C (G') 35C (G') 50C (G')
7 5C (G") 20C (G") 35C (G") 50C (G")
5
log G', log G" (Pa)
0
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9
log aT (rad s-1)
Figure 16. Frequency scanning master curves of the vegetable-based baby foods at 5-50°C (reference
temperature, TR = 35°C), considering loss modulus (G") for the rice and chicken baby food (a), and
storage modulus (G') for the selected vegetables with bass baby food (b). Points represent individual
values of one of six replicates.
Nevertheless, neglecting G" and considering only G' to construct the master curves to
approximate the relaxation modulus are often considered as a valid approach given the
much smaller contribution of the loss modulus [34, 65]. Therefore, for the selected
vegetables with bass, vegetables and beef and vegetables and chicken baby foods, the
continuous master curves of G' and G" were obtained using horizontal shifting alone,
when G' is considered and G" is neglected, i.e., the log aT values obtained for G' were
used to superpose both moduli (G' and G") (Figures 16b, 17a and 17b).

8
5C (G') 20C (G') 35C (G') 50C (G')
7 5C (G") 20C (G") 35C (G") 50C (G")
5
log G', log G" (Pa)
0
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9
log aT (rad s-1)
(b) Vegetables and Chicken Baby Food

8
5C (G') 20C (G') 35C (G') 50C (G')
7
5C (G") 20C (G") 35C (G") 50C (G")
5
log G', log G" (Pa)
0
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9
log aT (rad s-1)
Figure 17. Frequency scanning master curves of the vegetable-based baby foods at 5-50°C (reference
temperature, TR = 35°C), considering the storage modulus (G') for the vegetables and beef baby food (a)
and the vegetables with chicken baby food (b). Points represent individual values of one of six
replicates.
As expected, while good overlap is observed for the storage modulus (G'), very minimal
or no overlap is observed for the loss modulus (G") at the same points. For the selected
vegetables with bass, vegetables and beef, and vegetables and chicken baby foods, the
superposition of G' modulus appears satisfactory from a visual perspective. Only for the
vegetables and beef baby food, a slight deviation was observed for the sample heated at
50°C for this particular replicate (Figure 17a). Besides, it should be noted that the overlap
of the storage modulus was obtained without any need for the vertical shift factors. Some
authors suggested that because of the small magnitude of the vertical shift, its effect
could not be detected, whereas in contrast Oroian et al. [28] found that in honey samples,
the vertical shift factor had a great magnitude and so, it could not be neglected. An
important point in developing the master curves is the possibility of extending the
viscoelastic properties over a larger scale of time or temperature, otherwise inaccessible
in the laboratory tests. Note that for the G' modulus, data collected over nearly six
decades of frequency (Figures 12 and 13) were transformed to cover nearly 14 orders of
magnitude (Figures 16b, 17a and 17 b). The use of the TTS principle allows the creation
of a master curve spanning several decades of time/frequency at a single reference
temperature by using data measured at several temperatures but for a much shorter
duration [65]. In turn, for the rice and chicken baby food, minimal overlap was observed
when the log aT values obtained for G' modulus were used to superpose both G' and G"
moduli (data are not shown). Yoo [27] and Oroian et al. [28] found that the G" values of
honey samples at various temperatures could be superposed into a master curve using a
reduced frequency, indicating that the TTS principle could be used successfully for the
honeys in a liquid-like state. Therefore, the TTS principle was also applied to bring the
G" values, for the rice and chicken baby food at 5-50°C, together into a master curve
(Figure 16a), and certainly, a good overlap of the G" modulus isotherms was observed
when the log aT values obtained for the loss storage were used to superpose both moduli.
It can be seen that a continuous master curve for the G" modulus spanning nine decades
of time is obtained for the rice and chicken baby food in a more like-liquid state.
The Figure 18 presents the values of the shift factors (aT) as a function of the
temperature for all the four baby foods. The aT value reflects the dependency of the
molecular mobility coefficient on the temperature rearrangement rates [64]. As
mentioned above, for the rice and chicken baby food, the aT values correspond to the TTS
master curve for the G" modulus, while for the selected vegetables with bass, vegetables
and beef, and vegetables and chicken baby foods, the aT values correspond to the TTS
master curves for the G' modulus. The calculated values of the aT coefficient are not close
to unity, and therefore, the scaling has a significant effect on extending the frequency
window [66]. The values of aT monotonically decreased with the temperature. Similar
results were observed in other studies [32, 64]. At 5 and 20°C, the aT values increased
with a lower water content, a major number of interactions and also with the stronger
interactions (higher values of the A and z factors from the weak gel model).
log aT
Rice and Chicken (G")
Selected Vegetables with Bass (G')

-2 Vegetables and Beef (G')
Vegetables and Chicken (G')
-4
0 10 20 30 40 50 60
Temperature (°C)
Figure 18. Horizontal shift factors (aT) for the vegetable-based baby foods as a function of the
temperature (for reference temperature, TR = 35°C).
Particle Size Analysis
The rheological properties of the vegetable-based baby food can also be determined
by the size of the solid phase particles as shown by Aguilar et al. [36]. The particle size
distributions and the statistics from the same, for the examined vegetable-based baby
foods, are shown in the Figures 19 and 20. The rice and chicken baby food showed a
well-defined bimodal distribution, with a first population of particles defined in few
microns ranging from 0.05 to 0.2 μm, and a main population with much more higher
polydispersity ranging from 1 to 750 μm (Figure 19). Regarding the selected vegetables
with bass, vegetables and beef, and vegetables and chicken baby foods, an only highly
polydispersed population was observed with the particles at above 1 to exceeding 1,000
μm, moreover, approaching to the main population observed for the rice and chicken
baby food. As a consequence, the dispersion index (S) of each of the rice and chicken,
selected vegetable with bass, vegetables and beef, and vegetables and chicken baby foods
was 5.91 ± 0.27, 2.58 ± 0.02, 2.50 ± 0.01 and 2.42 ± 0.01, respectively. This result would
indicate that the different components in all the baby foods were superimposed on a well-
structured matrix made up of swollen starch granules coming from different sources
(potato, carrot, corn, etc.) and protein. Similar particle size distributions were observed
for carrot, broccoli and tomato suspensions [8].
8
Rice and Chicken Baby Food
Selected Vegetables with Bass Baby Food
Vegetables and Beef Baby Food
% volume 6 Vegetables and Chicken Baby Food
0
0.01 0.1 1 10 100 1000 10000
particle size (µm)
Figure 19. Particle size distributions for the vegetables-based baby foods.
In any case, there were remarkable and significant differences between the particle
sizes of the four baby foods (Figure 20). The rice and chicken puree consists of smallest
particles, which is exhibited in their lowest values of all particle sizing parameters. Even,
the d (0.1) and D [3,2] values obtained for this baby food (0.07 ± 0.00 μm and 0.29 ±
0.01 μm, respectively) were so low compared with the values for the other purees that can
not be appreciated on the Figure 20. In systems with a broader particle size distribution,
the volume based particle size parameter D [4,3] will be highly influenced by the large
particles, whereas the presence of small particles will be reflected in a lower D [3,2]
value [8]. In turn, the vegetables and chicken baby food exhibited the highest D [3,2] and
d (0.1) values, while the vegetables and beef baby food exhibited the highest D [4,3], d
(0.5) and d (0.9) values. The selected vegetable with bass puree had particle sizes
intermediate. Summarizing, most of the particles had a diameter ranging between 0.07
and 371.53 μm, 39.15 and 437.54 μm, 34.06 and 547.97 μm, and 42.34 and 477.72 μm,
for the rice and chicken, selected vegetable with bass, vegetables and beef, and
vegetables and chicken baby foods, respectively. Therefore, the highest difference
between lowest and highest particle diameters was shown by the vegetables and beef
baby food. The differences observed in the particles sizes of the four baby foods would
be mainly due to a greater or lesser degree of homogenization suffered by each product.
A significant decrease was also observed in both D [3,2] and D [4,3] values with the
degree of homogenization for carrot, tomato, apple and potato pulp suspensions [67]. In
this sense, the first population observed in the rice and chicken baby food could be
explained by the small oil droplets aggregates separated from the starch and protein
matrix because of a stronger shear suffered by this product. According to Servais et al.
[7], a viscosity reduction and better mixing can be obtained using bimodal particle size
distribution, therefore improving the shear, time and the temperature stability of the final
products. Positively, the rice and chicken baby food, with a broader particle size
distribution, showed a less structured system, much less viscoelastic than the rest as
explained by the G' and G" values found in their mechanical spectra, and that were lower
than in the other baby foods. Such result emphasizes that the particle distribution is an
important factor influencing the rheology of these baby foods.
Rice and Chicken Baby Food Selected Vegetables with Bass Baby Food
600 Vegetables and Beef Baby Food Vegetables and Chicken Baby Food
500
400
Particle size (μm)
300
200
100
0
d (0.1) d (0.5) d (0.9) D [3,2] D [4,3]
Figure 20. Statistics from the particle size distributions for the vegetables-based baby foods.
To get a comprehensive overview of the reciprocity of the physical properties of the

baby foods, relationships between the oscillatory rheological properties measured at
20°C, the dry matter, the syneresis and the particle sizing parameters were established as
shown in Table 15. Predominantly, the correlations were significant at 1%, and therefore,
only significant correlations at 5% are marked with an asterisk. The highest negative
correlation between the oscillatory rheological properties was found between the phase
angle (δ) and the strength of the network (A). With respect to the oscillatory rheological
properties, the dry matter content presented the highest negative correlation with the δ,
and the syneresis the highest negative correlation with “interaction strength” between the
flow rheological units (z), while the correlation between the moisture content and the
syneresis of the products was only significant at 5%. The moisture content significantly
affected all the characteristic parameters of the particle size distributions. The highest
correlation between the particle size parameters was observed between D [4,3] and d
(0.5) (r = 0.996), while in food powders and their mixtures, Benković et al. [68] found
the highest significant correlation between D [3,2] and d (0.1) (r = 0.97). In turn, the
surface-based diameter, D [3,2], was the sizing parameter that correlated best with the
phase angle and the strength of the network, whereas the volume-based diameter, D [4,3],
correlated best with the dry matter. Therefore, besides the particle size, the moisture
content strongly influenced the oscillatory rheological properties of the vegetable-based
baby foods.
Table 15. Matrix of correlations between selected rheological

parameters measured at 20°C, dry matter, syneresis and particle size
parameters for the vegetable-based baby foods
δ A z Dry Syneresis D [4,3] D [3,2] d (0.1) d (0.5) d (0.9) S

matter
δ 1 -0.996 -0.981 -0.971 0.690 -0.967 -0.996 -0.929 -0.975 -0.879 0.950
A - 1 0.986 0.953 -0.727 0.952 0.998 0.945 0.964 0.849 -0.958
z - - 1 0.938 -0.667 0.955 0.978 0.963 0.973 0.851 -0.983
Dry matter - - - 1 -0.560* 0.978 0.956 0.859 0.976 0.932 -0.906
Syneresis - - - - 1 -0.489 * -0.739 -0.785 -0.529* -0.270* 0.681
D [4,3] - - - - - 1 0.949 0.856 0.996 0.966 -0.918
D [3,2] - - - - - - 1 0.942 0.960 0.845 -0.953
d (0.1) - - - - - - - 1 0.893 0.696 -0.983
d (0.5) - - - - - - - - 1 0.944 -0.947
d (0.9) - - - - - - - - - 1 -0.793
S - - - - - - - - - - 1
*Significant correlations at P ≤ 0.05 level.
Microstructure Examination
The Figures 21-24 show the morphological differences between the baby foods
studied. The microstructure of the baby foods consists of a network made up of mainly
different starch type components released from the granules and protein, together with fat
and fiber from different sources. In the rice and chicken baby food (Figure 21), it would
appear that finer grinding produced a more homogeneous, regular microstructure, with
amylose and amylopectin released from the gelatinized starch granules forming a
reticular network, closely resembling an emulsion and with less appreciable protein
structures. Although the starch granule integrity is expected to be lost during the
manufacturing process (cooking), some gelatinized starch granules which were not
broken down are still recognizable. In this chapter, the finer the particle sizes were the
lower the flow rheological properties, as observed in the case of the rice and chicken
puree.
Figure 21. Micrographs of rice and chicken baby food.
The presence of meat protein structures was more apparent in the micrographs
corresponding to the other three vegetable-based foods foods studied (Figures 22-24),
because they had all been more coarsely ground, especially in the case of the vegetables
and chicken baby food. In this product, the protein network structure was more
continuous and dense (Figure 24), and much less dispersed in the starch component
matrix. This transition to a more discontinuous phase containing structures which were
coarser, more irregular, compact and aggregated is reflected by a very significant increase
in the flow rheological properties in comparison to the rest of the baby food systems. The
microstructures of the selected vegetables with bass (Figure 22) and vegetables and beef
(Figure 23) baby foods were intermediate between those of the rice and chicken and
vegetables and chicken baby foods, a fact that was also reflected in intermediate
rheological properties. In the selected vegetables with bass baby food, the higher fiber
content resulted in a more compact microstructure with a more porous appearance. Some
oil droplets can be observed immersed in the network. The bass protein structures
remained separate from, and hence not integrated in the starch matrix, though in some
cases they had occupied empty cavities (Figure 22).
Figure 22. Micrographs of selected vegetables with bass baby food.
In the vegetables and beef baby food, the meat protein appeared to be integrated in
the amylose and amylopectin matrix, exhibiting a kind of mesh with connections between
the structural elements. Some published reports have indicated that during heating, the
starch exhibits little affinity for the myofibrillar proteins, so that instead of interacting
with the protein matrix, it acts as passive filler [69]. This fact could explain the lower
time dependent rheological properties obtained for this baby food in comparison to the
puree containing fish. On the other hand, all the baby foods contain potato and carrot.
According to Bengtsson and Tornberg [67], carrot and potato pulp suspensions consist of
large cell clusters and aggregates, respectively, which are degraded into smaller cell
clusters when homogenized. However, a high content of insoluble pectin in carrot and
potato pulp keeps the cells closely associated with each other. Both the vegetables and
beef and vegetables and chicken baby foods also contain tomato, but the tomato
suspensions were found to consist of large, single cells and cell fragments, which were
easily degraded by homogenization.
Figure 23. Micrographs of vegetables and beef baby food.

Figure 24. Micrographs of vegetables and chicken baby food.
CONCLUSION
It was found that the vegetable-based infant baby foods exhibited a shear thinning
and thixotropic behavior from 5 up to 65°C. Data of the baby foods studied fitted the
power law model well, mainly in the case of the rice and chicken baby food with a higher
moisture content and degree of homogenization. In addition, in all the infant baby foods,
the degree of fit was higher after partially eliminating the flow time dependence. The
hysteresis loop area was larger in the vegetables and chicken baby food but smaller in the
rice and chicken baby food, implying more and less damage to the structures,
respectively. On the other hand, all the vegetable-based baby foods became less
thixotropic at higher temperatures. Both initial shear stress and the time coefficient of
thixotropic breakdown (parameters A and B in the Weltman model, respectively),
increased and decreased, respectively, at high shear rates and temperatures. In other
words, increasing the shear rate leads to an increase in the degree of thixotropy. In the
vegetables and chicken baby food, higher values for the B parameter indicated its higher
rate of structural breakdown by shearing. This fact is ascribed to the lower degree of
homogenization observed in this baby food. The rate and extent of viscosity decay were
also higher at the higher shear rate, but the amount of structural breakdown of the infant
purees showed no tendency to either increase or decrease with the temperature.
For the selected vegetables with bass, vegetables and beef, and vegetables and
chicken baby foods at 5, 20, 35, 50, 65 and 80°C, the dynamic shear data revealed a
solid-like behavior, with G' dominating over G" and both viscoelastic properties showing
a slight dependence on the imposed frequency. In contrast, at 35, 50 and 65°C, the rice
and chicken baby food showed a liquid-like behavior at low frequency, with spectra
closer to those exhibited by a concentrated macromolecular solution and both dynamic
moduli strongly dependent upon frequency. Conversely, the Ea of G'0 and G"0 for the rice
and chicken bay food was higher than for the other three baby foods, suggesting more
resistance to breakdown of this puree. The rheological data also showed apparent
abnormalities at and above 65°C because of the loss of solvent occurring at the higher
temperatures. However, in all the vegetables-based baby foods, both G' and G" moduli, as
well as the parameters derived from the weak gel model decreased systematically with an
increase in the temperature between 5 and 50°C. In addition, the A values for the
vegetables and chicken baby food were higher, evidencing that this infant food presented
higher strength than the rest at 5-80°C.
When the applicability of the TTS principle was analyzed, the horizontal shift factor
was to found vary with the choice of the viscoelastic properties (G' or G") selected for
shifting. For the selected vegetables with bass, vegetables with beef, and vegetables and
chicken baby foods, the overlap of G' isotherms was larger as compared with when G"
was considered, while the opposite was true for the rice and chicken baby food. Thus,
Ferry’s criteria are violated, and in the strict sense, the TTS method is not applicable to
the vegetable-based infant purees under study. However, it is shown that if only G' is
considered, a master curve spanning 14 decades of time is obtained using horizontal
shifting alone for those baby foods behaved like weak gels. These observations, coupled
with the results obtained from the particle size distributions and correlations, seem to
suggest that moving from monomodal to bimodal distributions had a significant effect on
the oscillatory properties of these products, with the rice and chicken baby food
presenting higher maximum packing fraction of the particles and reduced both dry matter
content and viscosity.
In view of the results, the factors that could best explain the differences in the
rheological and microstructural properties of the vegetable-based baby foods are: the
degree of grinding undergone by each baby food, the meat or fish particle size and the
degree of homogenization of the product, as well as the origin of the various proteins
(chicken, bass or beef). Results also evidenced a strong dependence of the baby food
viscosity on the water content. An improved understanding of the rheological behavior
and the particles of the baby foods prepared mainly with vegetable material could aid in
the structuring of these products without adding other stabilizing agents.
ACKNOWLEDGMENTS
The authors wish to thank the Spanish Ministry of Economy, Industry and
Competitiveness for financial support (AGL2011-28569).
REFERENCES
[1] Ahmed, J., & Ramaswamy, H. S. (2006a). Viscoelastic and thermal characteristics
of vegetable-puree based baby foods. Journal of Food Process Engineering, 29,
219–233.
[2] Ahmed, J., & Ramaswamy, H. S. (2006b). Viscoelastic properties of sweet potato
puree infant food. Journal of Food Engineering, 74, 376–382.
[3] Ramamoorthi, L., Lee, Y., & Brewer, S. (2009). Effect of food matrix and heat
treatment on the rheological properties of salmon-based baby food. Journal of Food
Engineering, 95, 432–437.
[4] Borwankar, R. P. (1992). Food texture and rheology: a tutorial review. Journal of
Food Engineering, 16, 1-16.
[5] Choi, Y. H., & Yoo, B. (2004). Characterization of time-dependent flow properties
of food suspensions. International Journal of Food Science and Technology, 39,
801-805.
[6] Colin-Henrion, M., Cuvelier, G., & Renard, C. M. G. C. (2007). Texture of pureed
fruit and vegetable foods. Stewart Postharvest Review, 5, 1-14.
[7] Servais, C., Jones, R., & Roberts, L. (2002). The influence of particle size
distribution on the processing of food. Journal of Food Engineering, 51, 201–208.
[8] Lopez-Sanchez, P., Chapara, V., Schumm, S., & Farr, R. (2012). Shear elastic
deformation and particle packing in plant cell dispersions. Food Biophysics, 7, 1–
14.
[9] Van Hecke, E., Nguyen, P. U., Clausse, D., & Lanoisellé, J. L. (2012). Flow
behaviour of carrot puree: modelling the influence of time, temperature and potato
flakes addition. International Journal of Food Science and Technology, 47, 177-
185.
[10] Ahmed, J., & Ramaswamy, H. S. (2007a). Dynamic rheology and thermal
transitions in meat-based strained baby foods. Journal of Food Engineering, 78,
1274–1284.
[11] Koocheki, A., & Razavi, S. M. A. (2009). Effect of concentration and temperature
on flow properties of Alyssum homolocarpum seed gum solutions: assessment of
time dependency and thixotropy. Food Biophysics, 4, 353-364.
[12] Abu-Jdayil, B. (2003). Modelling the time-dependent rheological behaviour of
semisolid foodstuffs. Journal of Food Engineering, 57, 97–102.
[13] Figoni, P. I., & Shoemaker, C. F. (1983). Characterization of time dependent flow
properties of mayonnaise under steady shear. Journal of Texture Studies, 14, 431-
442.
[14] Alonso, M. L., Larrodé, O., & Zapico, J. (1995). Rheological behaviour of infant
foods. Journal of Texture Studies, 26, 193-202.
[15] Campo-Deaño, L., Tovar, C. A., & Borderías, J. (2010). Effect of several
cryoprotectants on the physicochemical and rheological properties of suwari gels
from frozen squid surimi made by two methods. Journal of Food Engineering, 97,
457–464.
[16] Rao, M. A. (1999). Flow and functional models for rheological properties of fluid
foods. In G. V. Barbosa-Cánovas (Ed.), Rheology of fluid and semisolid foods.
Principles and applications (pp. 25–59). Gaithersburg, MD: Aspen Publishers Inc.
[17] Ahmed, J., & Ramaswamy, H. S. (2007b). Dynamic and steady shear rheology of
fruit puree based baby foods. Journal of Food Science and Technology, 44, 579–
585.
[18] Sttefe, J. F., & Ford, E. W. (1985). Rheological techniques to evaluate the shelf-
stability of starch-thickened, strained apricots. Journal of Texture Studies, 16, 179-
192.
[19] Augusto, P. E. D., Ibarz, A., & Cristianini, M. (2013). Effect of high pressure
homogenization (HPH) on the rheological properties of tomato juice: Viscoelastic
properties and the Cox–Merz rule. Journal of Food Engineering, 114, 57–63.
[20] Alvarez, M. D., Herranz, B., Fuentes, R., Cuesta, F. J., & Canet, W. Replacement
of wheat flour by chickpea flour in muffin batter: effect on rheological properties.
Journal of Food Process Engineering, 2016, 00. doi:10.1111/jfpe.12372.
[21] Dolz, M., González, F., Delegido, J., Hernández, M. J., & Pellicer, J. (2000). A
time-dependent expresión for thixotropic areas. Application to aerosol 100
hydrogels. Journal of Pharmaceutical Sciences, 89, 790-797.
[22] Tárrega, A., Durán, L., & Costell, E. (2004). Flor behaviour of semi-solid dairy
desserts. Effect of temperature. International Dairy Journal, 14, 345–353.
[23] Nguyen, Q. D., Jensen, C. T. B., & Kristensen, P. G. (1998). Experimental and
modelling studies of the flow properties of maize and waxy maize starch pastes.
Chemical Engineering Journal, 70, 165-171.
[24] Campo-Deaño, L., & Tovar, C. A. (2009). The effect of egg albumen on the
viscoelasticity of crab sticks made from Alaska Pollock and Pacific Whiting surimi.
Food Hydrocolloids, 23, 1641–1646.
[25] Dileep, A. O., Shamasundar, B. A., Binsi, P. K., Badii, F., & Howell, N. cK.
(2005). Effect of ice storage on the physicochemical and dynamic viscoelastic
properties of ribbonfish (Trichiurus spp.) meat. Journal of Food Science, 70, 537-
545.
[26] Gabriele, D., De Cindio, B., & D’Antona, P. (2001). A weak gel model for foods.
Rheological Acta, 40, 120-127.
[27] Yoo, B. (2004). Effect of temperature on dynamic rheology of Korean honeys.
Journal of Food Engineering, 65, 459–463.
[28] Oroian, M., Amarieri, S., Escriche, I., & Gutt, G. (2012). A viscoelastic model for
honeys using the time-temperature superposition principle (TTSP). Food and
Bioprocess Technology, 6(9), 2251-2260.
[29] Rosalina, I., & Bhattacharya, M. (2002). Dynamic rheological measurements and
analysis of starch gels. Carbohydrate Polymer, 48, 191-202.
[30] Nickerson, M. T., Paulson, A. T., & Speers, R. A. (2004). A time-temperature
rheological approach for examining food polymer gelation. Trends in Food Science
and Technology, 15, 569-547.
[31] Mendieta-Taboada, O., Sobral, P. J. A., Carvalho, R. A., & Habitante, A. M. B. Q.
(2008). Thermomechanical properties of biodegradable films based on blends of
gelatine and poly (vinyl alcohol). Food Hydrocolloids, 22, 1485-1492.
[32] Meza, B. E., Verdini, R. A., & Rubiolo, A. C. (2012). Temperature dependency of
linear viscoelastic properties of a commercial low-fat soft cheese after frozen
storage. Journal of Food Engineering, 109, 475-481.
[33] Doublier, J. L., Launay, B., & Cuvelier, G. (1992). Viscoelastic properties of food
gels. In: M. A. Rao & J. Steffe (Eds.) Viscoelastic Properties of Foods. London:
Elsevier Applied Science Publishers, pp. 371-434.
[34] Ferry, J. D. (1980). Viscoelastic properties of polymers. Wiley, New York, NY.
[35] Kuhn, K. R., & Cunha, R. L. (2012). Flaxseed oil – whey protein isolate emulsions:
effect of high pressure homogenization. Journal of Food Engineering, 111, 449–
457.
[36] Aguilar, C., Rizvi, S. S. H., Ramirez, J. F., & Inda, A. (1991). Rheological behavior
of processed mustard. I. Effect of milling treatment. Journal of Texture Studies, 22,
59-84.
[37] Canet, W., Alvarez, M. D., Fernández, C., & Luna, P. (2005). Comparisons of
methods for measuring yield stresses in potato puree: effect of temperature and
freezing. Journal of Food Engineering, 68, 143-153.
[38] Witczak, M., Juszczak, L., & Galkowska, D. (2011). Non-Newtonian behaviour of
heather honey. Journal of Food Engineering, 104, 532–537.
[39] Juszczak, L., Witczak, M., Fortuna, T., & Banyś, A. (2004). Rheological properties
of commercial mustards. Journal of Food Engineering, 63, 209–217.
[40] Alvarez, M. D., Fernández, C., Olivares, M. D., & Canet, W. (2010). Rheological
behaviour and functionality of inulin-extra virgin olive oil-based mashed potatoes.
International Journal of Food Science and Technology, 45, 2108-2118.
[41] Sopade, P. A., Halley, P. J., Bhandari, B., D’Arcy, B. R., Doebler, C., & Caffin, N.
(2002). Application of the Williams-Landel-Ferry model to the viscosity-
temperature relationship of Australian honeys. Journal of Food Engineering, 56,
67–75.
[42] Costell, E., Clemente, G., & Durán, L. (1982). Reología físico-química del puré de
albaricoque. II. Caracterización del flujo y relación entre los parámetros reológicos
y las características químicas y físicas del producto. Revista de Agroquímica y
Tecnología de Alimentos, 22, 539-550.
[43] Krokida, M. K., Maroulis, Z. B., & Saravacos, G. D. (2001). Rheological properties
of fluid fruit and vegetables puree products: compilation of literature data.
International Journal of Food Properties, 2, 179-200.
[44] Roopa, B. S., & Bhattacharya, S. (2009). Characterisation and modelling of time-
independent and time-dependent flow behaviour of sodium alginate dispersions.
International Journal of Food Science and Technology, 44, 2583-2589.
[45] Weltman, R. N. (1943). Breakdown of thixotropic structure as a function of time.
Journal of Applied Physics, 14, 343-350.
[46] Ramaswamy, H. S., & Basak, S. (1991). Time dependent stress decay rheology of
stirred yogurt. International Dairy Journal, 1, 17-31.
[47] O’Donell, H. J., & Butler, F. (2002). Time-dependent viscosity of stirred yogurt.
Part I: Coutte flow. Journal of Food Engineering, 51, 249–254.
[48] Butler, F., & McNulty, P. (1995). Time dependent rheological characterisation of
buttermilk at 5°C. Journal of Food Engineering, 25, 569-580.
[49] Abu-Jdayil, B. (2004). Flow properties of sweetened sesame paste (halawa
tehineh). European Food Research & Technology, 219, 265–272.
[50] Clark, A. H. (1992). Gels and gelling. In Physical Chemistry of Foods (H. G.
Schwartzberg & R.W. Hartel, Eds.) pp. 263-305, Marcel Dekker, Inc., New York,
NY.
[51] Katsuta, K., & Kinsella, J. E. (1990). Effects of temperature on viscoelastic
properties and activation energies of whey protein gels. Journal of Food Science,
55, 1296-1302.
[52] Kaur, L., Singh, J., Singh, H., & Mccarthy, O. J. (2008). Starch-cassia gum
interactions: a microstructure-rheology study. Food Chemistry, 111, 1-10.
[53] Fernández-Martín, F. (2007). Bird muscles under hydrostatic high-
pressure/temperature combinations. A DSC evaluation. Journal of Thermal
Analysis and Calorimetry, 87, 285-290.
[54] Rubens, P., & Heremans, K. (2000). Pressure-temperature gelatinization phase
diagram of starch: an in situ Fourier transform infrared study. Biopolymers, 54,
524-530.
[55] Herranz, B., Tovar, C. A., Solo-De-Zaldívar, B., & Borderías, A. J. (2012). Effect
of alkalis on konjac glucomannan gels for use as potential gelling agents in
restructured seafood products. Food Hydrocolloids, 27, 145-153.
[56] Carreau, P. J., Cotton, F., Citerne, G. P., & Moan, M. (2002). Rheological
properties of concentrated suspensions: application to foodstuffs. In Engineering
and Food for the 21st Century (J. Welti-Chanes, G. V. Barbosa-Canovas, & J. M.
Aguilera, Eds.) pp. 327-346, CRC Press, FL, USA.
[57] Chen, H. H., Xu, S. Y., & Wang, Z. (2007). Interaction between flaxseed gum and
meat protein. Journal of Food Engineering, 80, 1051–1059.
[58] Campo-Deaño, L., & Tovar, C. A. (2008). Influence of the starch content in the
viscoelastic properties of surimi gels. Journal of Food Engineering, 84, 140–147.
[59] Alvarez, M. D., Fernández, C., Jiménez, M. J., & Canet, W. (2011). Texture of
extra virgin olive oil-enriched mashed potatoes: Sensory, instrumental and
structural relationships. Journal of Texture Studies, 42, 413–429.
[60] Campo-Deaño, L., Tovar, C. A., Borderías, J., & Fernández-Martín, F. (2011).
Gelation process in two different squid (Dosidicus gigas) surimis throughout frozen
storage as affected by several cryoprotectants: Thermal, mechanical and dynamic
rheological properties. Journal of Food Engineering, 107, 107–116.
[61] Tunick, M. H. (2010). Activation energy measurements in rheological analysis of
cheese. International Dairy Journal, 20, 680-685.
[62] Augusto, P. D. E., Falguera, V., Cristianini, M., & Ibarz, A. (2011). Influence of
fibre addition on the rheological properties of peach juice. International Journal of
Food Science and Technology, 46, 1086–1092.
[63] Augusto, P. D. E., Cristianini, M., & Ibarz, A. (2012). Effect of temperature on
dynamic and steady-state shear rheological properties of siriguela (Spondias
purpurea L.) pulp. Journal of Food Engineering, 108, 283–289.
[64] Nickerson, M. T., Paulson, A. T., & Speers, R. A. (2004). A time-temperature
rheological approach for examining food polymer gelation. Trends in Food Science
& Technology, 15, 569-547.
[65] Hardikar, N. A., Bobba, S., & Jha, R. (2011). Applicability of time temperature
superposition principle to an immiscible blend of polyphenyleneoxide and
polyamide. Journal of Polymer Engineering, 31, 223-236.
[66] Ahmed, J. (2012). Applicability of time-temperature superposition principle:

dynamic rheology of mung bean starch blended with sodium chloride and sucrose –
Part 2. Journal of Food Engineering, 109, 329-335.
[67] Bengtsson, H., & Tornberg, E. (2011). Physicochemical characterization of fruit
and vegetable fiber suspensions. I: Effect of homogenization. Journal of Texture
Studies, 42, 268-280.
[68] Benković, M., Srečec, S., Špoljarić, I., Mršić, G., & Bauman, I. (2013). Flow
properties of commonly used food powders and their mixtures. Food and
Bioprocess Technology, 6, 2525-2537.
[69] Fernández, P., Cofrades, S., Solas, M. T., Carballo, J., & Jiménez Colmenero, F.
(1998). High pressure-cooking of chicken meta batters with starch, egg white, and
iota carrageenan. Journal of Food Science, 63, 267-271.
Chapter 9
RHEOLOGY AS A TOOL FOR FOOD DEVELOPMENT
Suzana Caetano da Silva Lannes*

Biochemical Pharmaceutical Technology Department,
Pharmaceutical Sciences School, University of Sao Paulo, Sao Paulo, Brazil
ABSTRACT
Food designers always come up with ways to best manage the ingredients in their
formulations. A product with good properties are conditioned on good process and
formulation. The interactions between ingredients can lead to several physical situations.
High viscosity impression must be felt when small stresses are applied to the product. At
a certain stress level, the yield stress, the viscosity starts to decrease rapidly. The lower
the yield stress the easier the product can be distributed on the surface, and it can
determine the thickness of the layer of the product on the surface. Rheology, as a tool for
evaluating liquid, semi solid and solid products, can contribute to an optimal result of
structural arrangement of the formulated product. Therefore, this chapter aims to discuss
some rheological concepts to help formulators find a accurate way to get the best profile
of obtained products in selected examples.
Keywords: rheology, food development, viscosity, texture, food structure
INTRODUCTION
Food Development
The development of a new product requires compliance to some criteria that are from
knowledge of the formulators.
*
Corresponding Author: Av. Prof. Lineu Prestes, 580 – bloco 16 -São Paulo-SP/Brazil CEP 05508000, Email:
scslan@usp.br.
290 Suzana Caetano da Silva Lannes
Take into consideration the researcher prior information, some point could be taken
into consideration from consumers: qualities (as lifestyle), emotive levels (as
expectations), experience (as appraisals), emotions (as experience) and behaviors (as
trial). To the product design, the standardization of ingredients, the prototype design, the
regulatory informations about ingredients and the market disponibility and prices, the
production possibilities, the packages, shelf life tests. These points could lead to the
success or failute of the product.
Development is mostly performed in multidisciplinary subjects. The latter
experiences are required to be taken into account to plan a new product. It is suggested
that development occurs with a set of expertises. Some technical skills we may need are:
mathematics and modelling, physics (materials and mechanics), thermodynamics,
kinetics and transport phenomena, chemistry, equipments. Nowadays, there are not
boundaries between food science, nutrition and medicine Nutrigenomics (molecular
nutrition) is beeing studied.
It is necessary to understand the relationship between components of the product, the
process effect upon them, the effect of these on the structural presentation of the product
obtained, and its consequent sensorial and nutritional characteristics.
A food elaboration trajetory since the start till the end could be: formulation plus
process leading to product microstructure, and it to product properties, and the last leads
to mechanical breakdown and digestive processes. It could help to plan a rheologycal
study of the product.
The physics and physicochemistry of each formulation ingredient and its mixture,
such as fats, polysaccharides and proteins, are related to the final food structure and its
properties.
Changes in the formulation of a product can result in significant changes in
mouthfeel and structure. For example, simply taking out sugar and adding a sweetener
can cause noticeable changes. Sugar is also responsible for structuring and viscosity, and
produces a coating on the tongue.
Thus, the points to be considered could be the causal relationship between the
structure of a food product and its performance when assessed visually, organoleptically
and in the delivery of its components to the human digestive tract; and how to predict the
effect of processes on components to form the required product structure [5].
When creating a new food product or redesigning an existing one, food developers
need to pay close attention to the rheological properties of food.
Rheological Concepts
As a common definition, rheology studies deformation in fluid materials, which

present intermediate behavior in transient regions between solid and liquid. Moreover,
Rheology as a Tool for Food Development 291
one would define the relationship between forces acting in materials, resulting in a
deformation [43, 49, 53]. Ideal fluids, as liquids and gases, deform in an irreversible way,
and flow by action of stresses. Solid elastic materials recover their characteristics
completely when the stress is removed. However, there are systems, including foods,
which have simultaneously some solid and liquid characteristics, which are called
viscoelastic materials [24, 46, 49, 52]. Consequently, measuring and specifying their
properties could be a difficult effort.
According to rheology, a material can deform in an elastic, plastic and viscous way.
In rheological food study, it is essential the understanding of composition and structure,
as well as know the process and storage of food [30, 43, 49].
Reiner (1964) has proposed the Deborah number as a manner to distinguish a solid
and a liquid. There is a need to adopt time of process as a factor when a solid or liquid
substance is characterized. It presents a relation between process time (when
characterized a solid or a liquid) and observation time. Higher values are typical of solid
behavior. The food-consumer interaction is relatively short, the Deborah number is large
and, as a consequence, some foods can appear more solid in sensorial evaluation than in
mechanical tests [10].
The flow of a fluid as designed by Newton is known as laminar flow and can be
visualized as a set of infinitely thin parallel planes that slide tangentially over each other
in response to a constant force applied to them when this fluid is confined between a
stationary surface and a parallel surface that moves at a constant speed. Thus, a velocity
gradient is established in the space between the two surfaces, having zero as value at the
point of contact with the stationary area and maximum value in the zone corresponding to
the region that moves. These conditions result in the shearing of the material. The shear
stress () is defined as the shear force per unit area. The velocity gradient, per unit
thickness of the sample, is called the shear rate (). The constant of proportionality ()
between these two quantities was called viscosity coefficient [54] by Isaac Newton.
The physical parameters that affect the viscosity of a system are solvent nature,
solute concentration, molar mass, and suspense particles, despite the shear stress, shear
rate, and temperature. The temperature effect on the rheological behavior of foods are
researched [56, 29], and the Arrhenius equation can quantify it [43], as in the equation 1.
 = η∞exp(Ea / RT) (1)
Where η is the apparent viscosity at a given shear rate, it is a frequency factor; Ea is

the energy of activation of the viscous flow (J mol-1); R is the universal gas constant (J K-
1
mol-1) and T is the absolute temperature of the sample (K). Qualitatively, Ea is a
quantity indicative of the degree of interaction between the various constituent
ingredients of the formulation, since it expresses the size of the energy barrier to be
overcome before the viscous flow begins.
The correlation between stress and shear rate defines the flow behavior of a
substance, and the diagrams showing how these quantities vary between them are called
flow curves. According to the profile exhibited by these curves, the materials may
assume one of the behaviors shown in Figure 1.
Most of food products assume non-Newtonian behavior and the viscosity can rise or
slow with time, even in constant shear rate. In this condition there is a time-dependent
phenomenon. When there is a slow down of the viscosity with shear rate, the system is
called shear thinning or pseudoplastic, and in the opposite way, it is called shear
thickening or dilatant [53]. The dilatancy is not so common as pseudoplasticity, but if it
occurs, it can bring up catastrophic consequences, as impellers or pumps, for example,
could need more energy to their operation, causing damages to the processing system
[26].
The thixotropy phenomenon can be defined as a rise of viscosity when a product is
under rest conditions, and a slow of viscosity when submitted to constant stress. The
entropy of the system rises. Thixotropy provides the information about the behavior of
material microstructure and its effects on its rheology. The microstructure term can be
associated in thixotropic systems to special distribution of bubbles and particles.
Thixotropy in some food products can be described by transition phenomena gel to sol
[49; 7]. The irreversible thixotropy, called the rheopexy, is common in food products and
it can be a determinant factor in the analysis of food behavior. Many food products with
addition of starch are viscous during their storage and show irreversible thixotropy when
agitated before consumption. Eduardo et al. [16], Lannes and Medeiros [32] studied this
occurrence in chocolate milk beverages. Chemical changes as retrogradation can form a
weak structure during the storage, which can be easily disrupted with some movement
[49].
Figure 1. Different flow behaviors verified in materials.

Some materials present rheopexy when submitted to tension or flow and, when in
rest, they tend to disrupt the structure. The behavior in function of time presented by a
flow curve show more viscous product after shearing [7, 43].
Food suspensions, emulsions and foams are time-dependent and show thixotropy or
rheopexy [7, 1].
Viscoelastic Behaviour
Viscoelastic materials can be linear or not. A linear viscoelastic material has

properties dependent only on time and not on the magnitude of applied tension. Which
means that this will always have a linear relationship between tension and deformation,
independent of the magnitude of the applied tension and deformation.
Many food products have a limit of deformation, generally small, bellow of the linear
viscoelastic behavior. After this limit the behavior is no longer linear. In this region, the
rheological properties are independent of the applied tension. To get significative
structural information, the rheological measures are usually performed in linear region to
avoid the breakdown of the structure (linear measurement) [18].
The stress sweep is fundamental for the evaluation of the linear viscoelasticity
interval, the maximum stress at which G′ remains constant is called the critical stress and
defines the limit of the linear viscoelastic region, allowing to measure the maximum
stress value in which there is no breakdown or changes in the sample structure. It is used
to establish the correct parameters for subsequent dynamic testing. The oscillation stress
sweep measurement is used to find a range of stress at which the rheological properties of
the samples are independent of the applied stress. These stress values should not be
exceeded in further measurements.
The viscoelastic region is known to be the region where the structure is preserved,
i.e., the storage or elastic (G’) and loss or viscous (G”) modules are frequency
independent.
When the sample has the elastic shear modulus (G’) much larger than the viscous
shear modulus (G”), both independent of the frequency in the region of viscoelasticity, it
is a material of predominantly solid characteristic. It is therefore a strong gel and the
deformations will be elastic or recoverable.
Frequency sweep is a very versatile rheological test to characterize the viscoelastic
behavior of materials. Studies on the variation of G’ and G” values with frequency have
been made in an attempt to establish the rheological signature of the material in question
The rheological properties which are dependent on time reflect the structural nature
of a system that can be due to the viscoelasticity and/or structural changes. The
viscoelasticity is characterized by a slow answer of the applied force, due to a partially
liquid behavior. By contrast, when the system suffers a structural change, it responds
immediately to applied force. However, the time-dependent behavior exists because the
structural changes are occurring continually, altering the quantity of the tension that the
structure can support and its flow resistance.
Thixotropic Behavior and Yield Stress
The classical method to characterize a structure breakdown is the hysteresis curve,

where the sample is sheared in increasing and decreasing shear rate simultaneously, and
up and down curves are generated. If there is a breakdown of the structure, both curves
are not coincident. The hysteresis is the area between the two curves. The major
disadvantage of this method is that each researcher determines arbitrarily the duration and
the maximum shear rate, making it difficult to compare the results from different
researchers.
As rheological properties of the material are defined by temperature, pressure, shear
rate and shear stress, if these parameters are known the industrial processes, the viscosity
and the elastic properties can be studied in lab with a rheometer. The results can be used
to study the different properties of the materials, such as: stability, consistency, melting
temperature, hardness temperature, stability and molecular weight, or to optimize the
quality of a product and/or predict the impact of a specific formulation in an industrial
process [43, 49].
The availability of viscometers can also permit the study of behavior to materials,
such as dispersions, emulsions and polymers. In the majority of cases, the viscosity
decreases with the rise of shear stress, producing a shear-thinning behavior, or a
temporary loss of viscosity [43, 46].
For many non-Newtonian fluids and for products with plastic characteristics, the
relationship between shear stress and shear rate depends on the duration and flow history.
Thyxotropy (viscosity rise with flow time) and anti-thyxotropy (viscosity decreases with
flow time) [43, 46].
The term tixotropy — a combination of classical Greek “θίξις” (thícsis) - to touch,
and “τροπία” (tropía) – State/mode resulting from turning, yilding “θιξοτροπία”
(thicsotropía) - modification by touch - was used by Peterfi in 1927 in the first
publication that properly described the phenomenon [7].
An immediate consequence of the thyxotropic behavior is the apparent viscosity
affected by handling the material, such as removing it from the original package and
putting it in the test container, and this historical dependence can affect the results of the
rheological test [43, 46].
Many multiphase systems food products, such as sauces, mayonnaises and ice creams
are highly structured fluids or solid solids, and their rheology and microstructure are
complex, showing gel or elastic behavior when submitted to low deformations or stress as
a consequence of their predominant internal tridimensional and heterogeneous structure

[50].
The mechanical response resultant from shearing or deformation of these fluids,
result in various viscoelastics phenomena, either in the linear region where a
microstructure answer in a linear form to the stress and to the deformation, but not alters
itself. In a non-linear region the microstructure alters in response to the stress and
deformation, however reversibly. The thyxotropy complication appears because this
reversibility, as the microstructural change, can take time to occur due to the spacial
rearrangement of the components. This response time make the thyxotropy one of the
biggest challenges to the rheologists, in terms of the correct characterization and
theoretical adequate description [7].
The flow of some materials starts in a limit value of tension, yield value, where the
breakdown of the microstructure occurs. The yield value characterizes the transition of
solid behavior in a low shear and deformations to liquid in a high shear and deformations.
Stability and sensorial perception are attributes of food processing [43, 43, 50].
The materials with yield stress are typically thyxotropics, sensible to thermal and
shear history, and susceptible to aging, because the samples with intact structure has high
values of yield stress, while those with completely destroyed structure have lower yield
stress, further than the shear during the test affect the estimated yield stress values [19].
The magnitude of the static yield stress may be used as an empirical parameter for
controlling the sedimentation on storage and the ease of using or processing a product.
One of the biggest difficulties of determining the yield stress is rarely mentioned,
since it is a time-dependent property. This thyxotropic behavior appears because the
microstructure and it can develop and alter during the applied weigh, as the semi solid
material can flow under small stresses with enough time. Uniform conditions are not
always obtained in materials tests with complex microstructures in short time of shear
applied, normally less than one minute, and the flow curve and the yield stress obtained
can be a transitory measurement due to the observation time. Thus, the determination of
accurate yield value in any system is difficult, since the obtained value depends on the
technique and instrument and/or method used to evaluate the rheological data.
Therefore, there is no standardized proceeding accepted to the assessment, and there
are many different concepts about yield stress [50], since its definition appears
inadequate sometimes and it has many discussions about its existence, the concept is
important to characterize rheologically complex fluids, and understand its industrial
applications.
Many food products have what is known as “slippage” occurring in the flow of
multiphase fluids, viscometers and rheometers, due to the displacement of the dispersed
phase(s) from the solid boundaries, leaving in place a layer of “depleted” liquid “of low
viscosity”. This phenomenon arises from the steric, hydrodynamic, viscoelastic, chemical
and gravitational forces that act in the dispersed phase immediately near solid boundaries.
Creating a low viscosity layer means that any flow at the border is easier because of the
lubricating effect - the fluid actually “slip”. Any geometry in rheometers or viscometers
may show “slippage” under certain circunstances, but those with smooth walls and high
gradients of shear rates are the most vulnerable. The materials that generate large “slip”
effects in the appropriate geometries are concentrated solutions of high molecular weight
polymers, suspensions and emulsions [7].
The existence of “slippage” in the determination of the initial stress with rotating
cylinder viscometers led several researchers to look for other geometries. A direct and
minimally destructive method for measuring the initial stress is the vane sensor, which
consists of a helix with at least four cross-shaped blades that is immersed in the sample
and rotated [57, 19]. Since its introduction, vane has achieved increasing interest in food
research, resulting in some studies of its use in dairy products, and while already well
established as a geometry for measuring structured fluids, it is still used sparingly by
systems and laboratories [50]. Its use when compared to the equivalent concentric
cylindrical sensor produces the same results in the Newtonian region at very low shear
rate but high stress, reducing the amount of “slippage”. Frequently, the initial stress (yield
stress) measurement using vane sensor is conducted at constant and usually low rotational
rates using a controlled rate instrument [44].
Due to the many factors that affect the rheology of food emulsions, rheological
properties cannot be predicted, so experimental studies using rheological models are
necessary. A flow model can be considered as a mathematical equation that can describe
rheological data in a convenient and concise way, with more than one equation
sometimes applicable. In addition to mathematical convenience, it is important to
quantify how the magnitudes of model parameters are affected by food state variables,
such as temperature, structure, composition and solids concentration, in addition to
establishing the applicable conditions that can be called a functional model. For
convenience, the flow models can be divided into time-dependent and flow behavior-
dependent, with models that take into account initial stress known as viscoplastic models
[43].
Characterization of Materials
The characterization of materials occurs as tension and viscosity curves, showing the
characteristics of their behavior when submitted to a determined shear rate [2]. The
rotational test provides the fluid behavior within complex processing systems, as
extruders, mixers and dosers [44].
Viscosity is the internal resistance of the liquid to the flow. The temperature and
composition of the product, as well as the shear rate and shear stresses imposed on the
material. Kinematic viscosity usually refers to Newtonian fluids, which have a linear
relationship between the rate of deformation and shear stress; for non-Newtonian fluids,
viscosity at a given shear rate is called apparent viscosity. All fluids exhibiting non-
Newtonian characteristics under normal conditions are colloidal or macromolecular [49].
Viscosity is one of the characteristics that affect the product [39], being an important
parameter in continuous processes and a way to determine the quality of the process [20].
As an example, the osmotic balance between two phases of foam can determine its
stability, modifying the morphology. The mechanism related to system instability is
directly related to applied emulsifier [34]. The rheological properties of the protein with
the emulsifier used to stabilize the foam have been heavily studied. Measures related to
the dilatant deformation of the interface are of particular interest, because they are
directly related to the formation of the foam and its destabilization [45].
To determine the viscoelastic properties of food products, instead of a constant
tension applied leading in consideration to a stationary state, the submission of samples to
oscillate tension and deformation as a sinusoidal function of time have been used, as seen
in equation 2 [43].
 = 0 sen(t) (2)
Where:  is shear stress (Pa); 0 is yield stress (Pa);  is angular velocity (rad/s).
Tests performed in these conditions are called “dynamical-rheological tests” and
provide viscosity and elasticity data related to the applied frequency: in other words, they
associate the frequency (f) or angular velocity () with the resulting strain or stress and
require measurements over a range of frequencies. For this reason, they usually consume
a lot of time. Performing an oscillatory test means that the rheometer sensor is shifted to
the right and left as a function of time from a very small angle of no more than 1°. This
fact leads to a very important conclusion for the dynamic tests and their applications:
samples of viscoelastic fluids and even solids are not mechanically disturbed nor their
internal structures are broken during the test and the samples are rheologically tested only
with the structures at rest [48].
The obtained results of oscillatory tests are very sensible to chemical composition
and physical arrangement, and they are appropriate in countless applications as gel force
determination, gelatinization of starch in function of temperature, glass transition
characterization, protein denaturation, particles clusters formation in dairy products, shelf
life tests, and correlations between instrumental and sensorial tests [49].
The rheological behavior studied through oscillatory measures allow characterization
of materials with G’ (elastic modulus) and G” (viscous modulus), evaluating as this
modulus varies over time. A frequency variation between 0.1 and 10 Hz, and deformation
between 0.01 and 100% are normally used to measure the viscoelastic properties of the
material [41, 2, 12].
The characteristic shape of the curves of several viscoelastic parameters plotted as a

function of the oscillatory frequency (in logarithmic coordinates) can be qualitatively
associated with different types of molecular responses, relevant in the macroscopic flow
determination [54]. The quantification of these modifications is made from the
parameters listed in Figure 2.
 shear stress; γ deformation; cos δ cosine of phase angle; sen δ sine of phase angle, tan δ tangent of
phase angle; ω angular velocity.
Figure 2. Mathematical relationship used in oscillatory tests.
Interface Rheology
The interface rheology is an indirect manner for obtaining information about the rate
of substances adsorbed. Measurement of interface rheology are yet considered a manner
of understanding the mechanism of bubble stability and formation [23, 42].
In some cases, direct correlations between rise of viscoelasticity in interface and the
rise of stability against the coalescence were found. Also, the addition of polysaccharides
can modify the interface rheology and the interfacial tension, depending on how
integrated it is with present proteins [38, 55, 17].
In general, the reasons why the proteins are good agents of foam formation can be:
they are strongly adsorbed in interface gas/liquid; they promote good sterical and
electrostatical stability; the formed films have cohesive structure with high rheological
interface modulus due to the interactions between adsorbed molecules. The proteins are
adsorbed in gas/liquid interface, because possess hydrophobic regions due to the
hydrophobic amino acids.
In a study system of foam, such as marshmallows, where it is desired the better
hydrocolloid to the formulation, the characteristics of the interface of the bubbles formed
during the aerated process, such as size, shape and composition of the formed film could
influence the stability and the shelf life of the developed product. The identification of
the better interface characteristic as a function of the hydrocolloid used lead to a more
stable product [36].
Texture
The textural evaluation of food products is a complex and dynamic process. The
texture of food is a highly subjective human experience perceived during its
consumption. Its attributes have important role in consumer decision of purchase and
even consumption [24].
The relevant works about texture come from late 19 century and early 20 century.
They involved the construction of simple test instruments and some rudimentary
sensorial evaluation. The primary intention was to eliminate defects. The primary food
products, such as meat, agricultural products and bread were the main targets. The texture
definition was disturbed and each group had its own definition. At the end of 1950s
decade the texture has started to be noted as a concept by a manager research group of
General Foods Corporation from United States. Today there are some principles of
texture visualized as a positive attribute of quality denoting from product freshness to
excellence of food preparation to pleasure to eat [52].
The ISO (International Standardization Organization) has defined texture as “All the
mechanical, geometrical and surface attributes of a product perceptible by means of
mechanical, tactile and, where appropriate, visual and auditory receptors” [8].
In the definition of the texture properties of a food product there are a group of
physical characteristics that come from its structural elements felt by touch, reported by
the deformation, disintegration and flow under a force and measured objectively by the
mass, time and distance functions. Also, the texture was defined as the sensorial and
functional manifestation of the structural, mechanical and superficial properties of food
detected through the senses of sight, hearing, touch and synesthesia, perceived through
instrumental and sensorial methods [46].
The food processors, in a general way, occasionally does not mention the texture with
its real importance. Textural characterizations are extremely important in countless types
of food [28]. Reaching desired textural quality in a food product includes economical
consideration [10], even affecting acceptance or rejection of determined products
depending on this property.
Food products have a variety of molecular interactions that contribute to the
structural network formation. Biopolymer gels, for example, are frequently used to
mechanical investigations due to their relatively low complexity when compared to solid
foods [8]. Thus, understanding the molecular mechanisms responsible for each textural
element can help the food researchers designing specific arrangements, which induce the
microstructural characteristics of a determined product [5, 14].
Experimental Determinations of Texture
Techniques of experimental determinations of texture can be grouped in subjective

and instrumental. The subjective measurements or sensorial evaluation are made by a
panel of trained panelists. The instrumental methods can be grouped as the following
categories: empirics, imitative and fundamentals. Any method of these categories are the
most appropriated to measure the food texture adequately [46].
The cone penetrometer is an instrument to rapid determinations of yield stress
parameter, in an empiric example [31]. Perhaps the imitative methods, where the
mechanic determinations are correlated to sensory analysis without any structural and
molecular valid study, are the main group of instrumental determinations of texture. The
texture profile analysis – TPA – obtained by classification of texture to solid and semi
solid foods, which is broadly used, belong to this group. Excellent correlations between
sensory and instrumental analysis were obtained in initial works with TPA. The
fundamental tests employ techniques of fundamental rheological tests, and the data are
analyzed using rheological, structural and molecular well defined theories [24, 52].
The universal instruments are projected to perform several determinations, only
changing apparatus. Age on foods on different types of force: compression, shear,
extrusion, cut and others. The equipment of universal type, in greater part, present the
same characteristics and concern basically: (a) test cell, which enter in contact with the
sample; (b) system of action, the piston which is connected to the test cell; (c) sensor
element, which detect the resistance of food to applied force; (d) register system, that
produce the force-distance curves or force-time, or the expended energy during the test
when there is a integration system coupled [31].
Bourne [10] pointed the postulates regarding several determinations to characterize
of texture elaborated in 1970s decade: (a) in compression test the force is proportional to
sample mass squared; (b) in shear test the force is proportional to sample; (c) in extrusion
test the force does not depend on the sample mass.
The development of the adequate texture is a continuous effort of the food industry in
a variety of products. Due to the cost limitation, not adequate experience, and the
subjectivity of sensorial panels, efforts are continually made to project instrumental
methods to measure texture [24].
However, few instruments of texture measurement focus on more than one property
at the same time. Therefore, researches in isolated physical aspects result in an
experience, but no one is equivalent to complete sensation [46].
The hardness, one of the texture parameters, has as a proceeding submitted the food
to a compression force, analyzing the force-time resulting curve, and take as parameter
the maximum force applied to samples [34].
There are several probes to lead a texture test. The back extrusion is an example.
Brusewitz and Yu [11] observed that it is particularly useful with materials which have
the pasty consistency, because it can reproduce the high necessary forces to create a flow
that are generally difficult to obtain in a conventional viscometer, and/or have particles in
suspension. It is a method that can be used to quantify the behavior of thixotropic fluids.
Yield stresses can be easily determined additionally.
The back extrusion test, also called annular pumping, include two physical moments:
(a) a cylindrical plunger is forced down into the fluid, and (b) the fluid flows up through
the concentric annular space. Generally, the cylindrical plunger is a solid axis and the
fluid is in a cylindrical cup. The necessary force to extrude the sample in the direction
opposite to that of the rod is measured, or through the formed ring between the cylinder
and the recipient which contain the sample is recorded [25, 40].
These tests have a big potential in the industry, in quality control and/or product
development, since they are simple, fast and of low cost [40].
A typical curve force-distance is shown in Figure 3. In interval A-B, the food product
is deformed and compressed; near the point B, the food product started to be involved by
the water of its composition, and the required force increases; on point C the structure of
the system starts to be disrupted and the flow starts; on the point D the compression
change its direction and the force equals zero in value. The shape and magnitude of the
curve are influenced by viscoelasticity and rupture behavior of the material, as well the
temperature and conditions of the sample [10].
As said, some advantages related to back extrusion test are: facility of execution, fast
and robust testing. The possible disadvantage of back extrusion that can be identified is
the large level of force that could be beyond some texture analyzers.
In mixtures of meat products, the force obtained by back extrusion is an important
parameter, because the mixture needs to be pumped through the tubes and equipments.
Thus, data of force obtained by back extrusion can be useful in projecting the equipment
and selecting pumps in meat industry [25].
Ignacio and Lannes [27] have used back extrusion as a method to analyze twelve
commercial mayonnaises rheologically and texturally. The back extrusion method was
used to determine rheological properties of mayonnaises showing sensible and very
efficient results for providing quantitative and qualitative characteristics. However, a
standardization of samples is necessary to provided repeatability of measurements. The
market offers mayonnaises with a big variety of textural characteristics.
Mayonnaise rheology and texture can depend on the distribution of oil, the
interaction between oil droplets, and the egg yolk emulsifier. Alteration of the
mayonnaise’s microstructure leads to different rheological properties. This means that the
rheology is sensitive to any change in the microstructure of the materials.
In emulsions, droplet size is perhaps the most important factor in determining
properties like consistency, rheology, shelf life stability, color and taste. In general,
emulsions with smaller droplet size result in great stability. Current equipment used for
emulsion preparation includes colloid mills, sonicators or high pressure homogenizers.
The advantage of high pressure homogenizers over other technologies is that more
uniform droplets size distribution are obtained since the product is subjected to strong
shear and cavitation forces that efficiently decrease the diameter of the original droplets
[33].
Figure 3. Characteristic curve of back extrusion [10].
Rheology of mayonnaise can provide valuable information that can be used in quality
control of commercial production, storage stability, sensory assessments of consistency,
knowledge and design of texture, design of unit operations, and knowledge of the effects
of mechanical processing on the structure of the emulsions [13, 18]. Mayonnaise has
been shown to be shear thinning, viscoelastic and thixotropic and it has a yield stress. The
yield stress is the minimum stress required to enable flow or fracture due to the
breakdown or alteration of materials microstructures. One of the many difficulties
encountered during the characterization of such a complex material is to isolate and
quantify individual effects [22].
Researchers have used this method to determine rheological parameters of food

products such as chocolates, where the viscosity index obtained in these tests makes it
possible to predict the flow stress and the apparent viscosity of the chocolates during
manufacture [3, 4].
Another study by Osorio and Steffe [49] demonstrates how to analyze rheological
data of fluids from the Herschel-Bulkley equation (equation 3).
(3)
Where:  is shear stress (Pa); 0 is yield stress (Pa); K is consistency index; n is

viscosity index; is shear rate (1/s)
Gonçalves and Lannes [21] used this method in the analysis of texture of chocolates
in the study of the development of chocolate products. It indicates the advantages of
using this method because it offers the opportunity to test samples without affecting the
structure before the test, and the possibility of investigating samples with large particles
such as spices or mustard seeds. Eduardo and Lannes [15] studied powders of cocoa
products to determine compaction force. Salas, Gonçalves, Lannes [47] determined
viscosity of cake batter with back extrusion plus Herschel-Bulkley equation application.
Su and Lannes [51], Milliatti [35] studied ice cream behavior immediately after
produced, and related with overrun values. Almeida and Lannes [36] analysed chocolate
spreads made with chicken gelatin substituting fat phase.
FINAL CONSIDERATION
The results of the yield stress measurement can be directly related to the sensorial
properties of the products. The texture results can be related to sensorial properties.
Rheological and textural instrumentation and measurements have become essential tools
in the Research Centers for characterizing ingredients and final products, as well as
predicting product performance and consumer acceptance.
REFERENCES
[1] Abul-Jdayil, B. Modelling the time-dependent rheological behavior of semisolid

foodstuffs. Journal of Food Engineering, v. 57, p. 97-102, 2003.
[2] Adams, S; Frith, WJ; Stokes, JR. Influence of particle modulus on the rheological
properties of agar microgels suspensions. Journal of Rheology., v. 48 (6), p. 1195-
1213, 2004.
[3] Afowaka, EO; Paterson, A; Fowler, M. Factor influencing rheological and textural
qualities in chocolate - a review. Trends in Food Science & Technology, v. 18, n. 6,
p. 290-298, 2008.
[4] Afoakwa, EO. Chocolate science and technology. Oxford: Wiley-Blackwell, 2010.
p. 1-2, p. 12-13, 41-51.
[5] Aguilera, JM. Why food microstructure? Journal of Food Engineering., v. 67, p. 3-
11, 2005.
[6] Almeida, PF; Lannes, SCS. Effects of chicken by-product gelatin on the
physicochemical properties and texture of chocolate spread. Journal of Texture
Studies, v. 48, p. 1-11, 2017.
[7] Barnes, HA. Thixotropy: a review. Journal of Non-Newtonian Fluid Mechanics, v.
70, p. 1-33, 1997.
[8] Barrangou, LM; Daubert, CR; Foegeding, EA. Textural properties of agarose gels.
I. Rheological and fracture properties. Food Hydrocolloids, v. 20, p. 184-195, 2006.
[9] Barrangou, LM; Drake, M; Daubert, CR; Foegeding, EA. Textural properties of
agarose gels. II. Relationships between rheological properties and sensory texture.
Food Hydrocolloids, v. 20, p. 196-203, 2006.
[10] Bourne, MC. Food texture and viscosity: concept and measurement. New York:
Academic Press, 2002, 380p.
[11] Brusewitz, GH; Yu, H. Back extrusion for determining properties of mustard slurry.
Journal Food Engeneering, v. 27, p. 259-265, 1996.
[12] Choi, YH; Lim, ST; Yoo, B. Measurement of dynamic rheology during ageing of
gelatine-sugar composites. International Journal of Food Science and Technology.,
v. 39, p. 935-945, 2004.
[13] Dalgleish, DG. Food emulsions – their structure and structure foaming properties.
Food Hydrocolloids., v. 20, p. 415-422, 2006.
[14] Dutta, A; Chengara, A; Nikolov, AD; Wasan, DT; Chen, K; Campbell, B. Texture
and stability of aerated food emulsions––efects of buoyancy and Ostwald rippening.
Journal of Food Engineering., v. 62, p. 169–175, 2004.
[15] Eduardo, MF; Lannes, SCS. Use of Texture Analysis to Determine Compaction
Force of Powders. Journal of Food Engineering, v. 80, p. 568-572, 2007.
[16] Eduardo, MF; Mello, KGPC; Polakiewicz, B; Lannes, SCS. Evaluation of chocolate
milk beverage formulated with modified chitosan. Journal of Agricultural Science
and Technology, v. 16, p. 1301-1312, 2014.
[17] Erni, P; Fischer, P; Herle, V; Haug, M; Windhab, EJ. Complex Interfaces and their
Role in Protein-Stabilized Soft Materials. Chem Phys Chem., v 9, p. 1833 – 1837,
2008.
[18] Gallegos, C; Berjano, M. Linear viscoelastic behavior of commercial and model
mayonnaise. Journal of Rheology, v. 36, n. 3, p. 465-478, 1992.
[19] Genovese, DB; Lozano, JE; Rao, MA. The rheology of colloidal and non colloidal
food dispersions. Journal of Food Science., v. 72 (2), p. 11-20, 2007.
[20] Goetz, J; Balzer, H; Hinrichs, R. Characterisation of the structure and flow
behaviour of model chocolate systems by means of NMR and rheology. Applied
Rheology., v. 15, n.2, p. 98-111, 2005.
[21] Gonçalves, EV; Lannes, SCS. Chocolate rheology: a review. Ciência e Tecnologia
de Alimentos, v. 30, p. 845-85, 2010.
[22] Goshawk, JA; Binding, DM. Rheological phenomena occurring during the shearing
flow of mayonnaise. Journal of Rheology, v. 42, n. 6, p. 1537-1553, 1998.
[23] Grigoriev, DO; Derkatch, S; Kragel, J; Miller, R. Relationship between structure
and rheological properties of mixed BSA/Tween 80 adsorption layers at the
air/water interface. Food Hydrocolloids., v. 21, p. 823-830, 2007.
[24] Gunasekaran, S; Ak, MM. Cheese rheology and texture. Boca Raton: CRC Press,
2003. cap., 5, p. 215-272.
[25] Gujral, HS; Sodhi, NS. Back extrusion properties of wheat porridge (Dalia).
Journal of Food Engeneering, v. 52, p. 53–56, 2002.
[26] Hunter, RJ. Introduction to modern colloid science. New York: Oxford University
Press, 1993, cap., 4, p. 97-130.
[27] Ignacio, RM; Lannes, SCS. Rheological characterization and texture of commercial
mayonnaise using back extrusion. African Journal of Agricultural Research, v. 8, p.
4262-4268, 2013.
[28] Kealy, T. Application of liquid and solid rheological technologies to the textural
characterization of semi-solid foods. Food Research International, v. 39, p. 265-
276, 2006.
[29] Lai, G; Li, Y; Li, G. Effect of concentration and temperature on the rheological
behavior of collagen solution. International Journal of Biological Macromolecules,
v. 42, n. 3, p. 285-291, 2007. Disponível em: http://www.sciencedirect.com/
science/journal/01418130. Acesso em: 12 dez. 2007.
[30] Lannes, SCS. Estudo das propriedades físico-químicas e de textura de chocolates.
São Paulo, 1997. 175p. Tese de Doutorado- Faculdade de Ciências Farmacêuticas-
[Study of the physical-chemical properties and texture of chocolates. São Paulo,
1997. 175p. PhD Thesis- Faculty of Pharmaceutical Sciences] Universidade de São
Paulo.
[31] Lannes, SCS; Medeiros, ML; Gioielli, LA. Physical properties between cupuassu
and cocoa fats. Grasas y Aceites, v. 54, n. 2, 2003.
[32] Lannes, SCS; Medeiros, ML. Rheological Properties of Chocolate Drink from
Cupuassu. International Journal of Food Engineering, v. 4, p. 11, 2008.
[33] Martin-Gonzalez, MFS; Roach, A; Harte, F. Rheological properties of corn oil
emulsions stabilized by commercial micellar casein and high pressure
homogenization. Lebensmittel-Wisseschaft & Technologie, n. 42, p. 307-311, 2009.
[34] Mezzenga, R; Folmer, BM; Hughes, E. Design of double emulsions by osmotic

pressure tailoring. Langmuir., v. 20, p. 3574- 3582, 2004.
[35] Milliatti, MC. Rheological study of ice cream formulations produced with diferente
stabilizers. 2013. Thesis (Master in Biochemical Pharmaceutical Technology) –
Pharmaceutical Sciences School, University of Sao Paulo, Sao Paulo, Brazil.
[36] Miquelim, JN; Lannes, SCS; Mezzenga, R. pH Influence on the stability of foams
with protein polysaccharide complexes at their interfaces. Food Hydrocolloids, v.
24, p. 398-405, 2010.
[37] Munizaga-Tabilo, G; Barbosa-Canovas, GV. Rheology for the food industry.
Journal of Food Engineering. v. 67, p. 147-156, 2005.
[38] Murray, BS. Stabilization of bubbles and foams. Current Opinion in Colloid &
Interface Science, v. 12, p. 232–241, 2007.
[39] Nebesny, E; Zyzelewicz, D; Motyl, I; Li, Z. Properties of sucrose-free chocolates
enriched with viable lactic acid bacteria. European Food Research and
Technology., v. 220, p. 358-362, 2005.
[40] Paoletti, F; Nardo, N; Saleh, A; Quaglia, GB. Back extrusion test on emulsions
stabilized whey protein concentrates. LWT Food Science and Technology, n. 28, p.
616-619, 1995.
[41] Peressini, D; Sensidoni, A; Polini, CM; Gabriele, D; Migliori, M; Cindio, B. Filled
snacks production by co-extrusion-cooking. Part 3. A rheological-based method to
compare filler processing properties. Journal of Food Engineering., v. 54, p. 227-
240, 2002.
[42] Piazza, L; Gigli, J; Bulbarello, A. Interfacial rheology study of espresso coffee
foam structure and properties. Journal of Food Engineering, v. 84, p. 420-429,
2008.
[43] Rao, MA. Rheology of Fluids and Semisolid Foods: Principles and Applications.
Gaithersburg, Aspen Publishers, 1999, 433p.
[44] Roberts, GP; Barnes, HA. Carew, P. Modelling the flow behavior of very shear-
thinning liquids. Chemical Engineering Science., v. 56, p. 5617-5623, 2001.
[45] Romoscanu, AI; Mezzenga, R. Cross linking and rhological characterization of
adsorbed protein layers at the oil-water interface. Langmuir, v. 21, p. 9689-9697,
2005.
[46] Rosenthal, AJ. Food texture measurement and perception. Gaitherburg: Aspen
Publishers, 1999, 311p. (A Chapman & Hall food science book).
[47] Salas, AGV; Gonçalves, EV; Lannes, SCS. Correlation between dough viscosity
and volume increase of gluten and sucrose-free cakes. In: 15th World Congress of
Food Science and Technology-IUFoST, 2010, Cape Town. 15th World Congress of
Food Science and Technology-IUFoST, 2010.
[48] Schramm, G. Reologia e reometria: fundamentos teóricos e práticos. São Paulo:

Artliber [Rheology and Rheometry: theoretical and practical foundations.], 2006,
160p.
[49] Steffe, JF. Rheological methods in food process engineering. East Lansing:
Freeman Press, 1996, 3-47.
[50] Stokes, JR; Telford, JH. Measuring the yield behaviour of structured fluids. Journal
of Non-Newtonian Fluid Mech., n. 124, p. 137-146, 2004.
[51] Su, F; Lannes, SCS. Rheological evaluation of the structure of ice cream mixes
varying fat base. Applied Rheology, v. 22, p. 63871-63877, 2012.
[52] Szczesniak, AS. Texture is a sensory property. Food Quality and Preference, v. 13,
p. 215–225, 2002.
[53] Tabilo-Munizaga, G; Barbosa-Cánovas, GV. Rheology for the food industry.
Journal of Food Engineering., v. 67, p. 147-156, 2005.
[54] Tung, MA; Paulson, AT. Rheological concepts for probing ingredients interactions
in food systems. In: GAONKAR, A.G., ed. Ingredient interactions: effects on food
quality. New York: Marcel Dekker, 1995, cap. 3, p. 45-83.
[55] Turgeon, SL; Schmitt, C; Sanchez, C. Protein–polysaccharide complexes and
coacervates. Current Opinion in Colloid & Interface Science, v. 12, p. 166-178,
2007.
[56] Vidal, JRMB; Pelegrine, DH; Gaspareto, CA. Efeito da temperatura no
comportamento reológico da polpa de manga (mangífera indica L-Leitt). Ciência e
Tecnologia de Alimentos, [Effect of temperature on rheological behavior of mango
pulp (mangífera indica L-Leitt). Food Science and Technology,] v. 24, p. 39-42,
2004.
[57] Weenen, H; Van Gemert, LJ; Van Doorn, JM; Dijksterhuis, GB; De Wijk, RA.
Texture and mouthfeel of semisolids foods: commercial mayonnaises, dressings,
custard dressings and warm sauces. Journal of Texture Studies, v. 34, p. 159-179,
2003.
Chapter 10
USING THE LAPLACE TRANSFORM IN RHEOLOGY

KINETIC MODELLING
I. T. Smykov
Department of Physical Chemistry, All-Russian Scientific Research Institute
for Butter- and Cheese-making, Uglich, Russia
ABSTRACT
This chapter aims to provide a modern state of the art in rheology, balancing
theoretical principles and their practical applications. In the Chapter the utility of direct
and inverse Laplace transform for known rheological model descriptions is demonstrated.
The Laplace transform is considered a very powerful math tool, which is useful in
different fields of science and technology. With increasing complexity of technical
problems, the Laplace transform gives real support in solution of complex problems with
very simple approaches. During the modelling, with the use of the Laplace transform, the
object under study is defined as the object under technological impact: input and
interfering signals evoke output reaction. It is demonstrated that kinetic changes in
rheological parameters with time can be considered as a transfer function (transient
response) for the investigated object changing, to another under influence of input
signals. As examples of the Laplace transform with practical use, some our results are
provided concerning the kinetic changes in rheological parameters during fermentative
phase sol-gel transition in the milk during cheese making. The methodological
significance of the Chapter is provided by examples of deep interconnections between
pure theoretical models and practically important systems.
Keywords: Laplace transforms, modeling, kinetic, milk, gelation, sol-gel transition

Corresponding Author Email: i_smykov@mail.ru.
310 I. T. Smykov
INTRODUCTION
A great number of mathematical models proposed for the description of kinetics of

viscoelastic behaviour of various objects can be found in the literature. Two main groups
of models are distinguished among them. Molecular models, which were created to
explain a mechanism of the occurring processes by the observed curves, are attributed to
the first group. These models allow to understand better the interaction of the individual
components in an object and to get the desired theoretical result. However, these models
represent very cumbersome analytical expressions for the complex objects and they are
not always convenient for computational tools in practical applications. Approximating
models, which describe the rate of curves constructed from experimental data only with
some degree of confidence, are attributed to the second group. These models do not
reflect the physical content of the occurring processes, but they are more convenient to be
used in applied calculations. In all circumstances, the model quality is determined
foremost by its predictive possibilities, i.e., by the model ability to predict an object
behaviour when the influencing factors are changed.
The current state of science and technology is such that these rheological model
groups come close. The technological capabilities for using more and more complex
models, which take into account the impact of many factors, are required for practical
use, and they are available. Further, the improvement of existing and the development of
new mathematical models is required.
One of the most difficult issues for practitioners and interesting subject for
professional rheologists is the sol-gel transition, when sol, which is a liquid, is
transformed into a dense gel, practically a solid, under the influence of external factors.
The change in the viscoelastic behaviour during the gelation is very significant. This can
be determined by the structuring process of the system and by the formation of a three-
dimensional grid due to cross-links between its dispersed components.
The dramatic change in the viscoelastic behaviour during gelation is also interesting
from the practical and physical-chemical point of views and, as a result, many researches
had been carried out studies in this field. Clark and Ross-Murphy gave an overview of the
formation of physical gels [1]. Guenet [2] and Nijenhuis [3] wrote detailed monographs
on thermoreversible gelation. Other authors carried out several research works related to
the structure and properties of polymer gels [4]. Gelation kinetics in food products has
been studied by Scott Blair [5, 6], Douillard [7], Carlson [8-11] and Esteves [12].
The sol-gel transition draws attention of many scientists working in different fields of
science and technology. The used mathematical models for viscoelastic behaviour and
kinetics of their variations with time are usually described by partial differential
equations, having solutions and containing many unknown integration constants that are
determined from the relevant boundary conditions. This quite complicated procedure can
be overcome by applying the Laplace transform, where the unknown constants of
Using the Laplace Transform in Rheology Kinetic Modelling 311
integration are determined when solving with the use of forward algebraic equations. The
Laplace transform is a very powerful math tool used in various fields of practice and
science. Considering the complexity of technical problems, the Laplace transform is very
helpful in their solution through a very simple approach. The Laplace transform is
attributed to the field of the general mathematics, operational calculus, which has an
increased focus on the analysis of linear systems. Currently, there are many computer
programs allowing the effective use of direct and inverse Laplace transform to develop
math models of viscoelastic behaviour kinetics of various objects in theoretical studies
and practical use. Thus, the theoretical paper [13] using the Laplace transform established
solutions for the form of unsteady flows of natural-convection liquid near an isothermal
vertical plate oscillating in its plane. The Prandtl number Pr, the Grashof number Gr, the
viscoelastic parameter α, the phase angle and the time τ were determined. The effective
application of the inverse Laplace transform to study the problem of unsteady flow of
thin liquid film flowing through a porous medium along an inclined plane was also
studied [14]. The Laplace transform was also successfully employed in the work [15] to
solve three problems of Newtonian fluid flow on an infinite plate. It was noted that this
method is very effective and simple, being a valuable tool for engineers and scientists
dealing with real problems.
In the work [16], the Laplace transform is used for rechecking of the experimental
data from rheological experiments on relaxation dynamics and structural transformations
occurring in DNA during gelation. The obtained results have improved the general
understanding of the DNA relaxation dynamics in aqueous dispersions.
The work [17] deals with the measurement of viscoelastic behaviour of highly
hydrated biological materials and it proposes a method for a quick checking of the sample
degradation using the Laplace transform mathematical modelling. The research [18] is
aimed at comprehension of macroscopic viscoelastic behaviour of collagen matrices by
studying the relaxation time distribution spectrum obtained from the stress relaxation
tests by the inverse Laplace transform. The viscoelastic hydrogel networks formed from
low-molecular-weight Fmoc-tyrosine have been investigated in the paper [19] using
microrheology methods. The dynamic pattern of viscoelastic behaviour during the
process of gelation was also investigated and developed using the Laplace transform
mathematical modelling.
The traditional approach has been used in the given examples of mathematical
modelling of kinetic viscoelastic behaviour of various objects as well as in many other
papers: the known model in the form of a differential equation was developed or used;
performed the Laplace transform; determined the parameters; and performed the inverse
Laplace transform. However, there are many practical applications in which it is
impossible or quite complicated to compose such a system of differential equations.
Therefore, the mathematical model follows a "black box" principle, by the change of
output variables under the external influence without consideration of the internal system
312 I. T. Smykov
state [20]. This approach is proposed for viscoelastic behaviour kinetic modelling in the
processes.
1. MATHEMATICAL ANALYSIS
Processing of the experimental results was performed taking advantage of the

following software: “Excel” (Microsoft, Redmond, WA, USA), “Origin” (Origin Lab Co.
Northampton, USA), and “Mathcad” (Parametric Technology Co. Needham, MA, USA).
2. VISCOELASTIC BEHAVIOUR KINETIC MODELLING METHODOLOGY
Many practical cases of sol-gel transition assume the formation of an intermediate or

final product in a process. In the technological viewpoint, a source material - sol and a
product obtained from it - gel might be considered as a single multi-dimensional
controlled object characterized by the existence of a set of points of application of
external process effects (inputs) and a set of variables describing the obtained product
properties (outputs). Moreover, there are a number of factors that are difficult or
impossible to consider, but which can critically affect a product state - these are
disturbance effects.
Prior to and after of process effects, the product state can be considered as static,
while its properties change rapidly during processing, i.e., they have a dynamic
behaviour. Then, on the one hand, the product as a multi-dimensional object can be
represented by a static state function that includes numerical or logical values of the
output variables describing different properties of a product, before and after its
processing: physicochemical, processing, rheological etc. The product state function can
be either generalized, including the description of all or a part of the property groups, or
partial that describes one group or one product property.
On the other hand, the product properties are changed with time, dynamically, and in
accordance with certain regularities upon the process effects. In this instance, the product
as a multi-dimensional object can be described by the differential equation system that
binds the input controlled and output variables with allowance for disturbance effects.
The use of such approach is conditioned by the possibility of simplifying the analysis
of the process effects on the finished product properties. This is due to the transition from
the consideration of the external impacts and dynamic behaviour of the object in the time
domain to their analysis in the frequency domain. Such transition from the differential
equation systems to the ordinary algebraic systems is widely applied in the analysis of
dynamic processes in various fields of science [21]. However, this method has not found
a worthy application for the analysis of viscoelastic behaviour kinetics during gelation.
At the same time, the in-depth analysis of mathematical apparatus of operational calculus
and the availability of modern software allow performing an operative analysis and
exercising control over kinetics of rheological processes occurring in products of
different nature while produced.
It is meant here that the product, while produced, is being a coherent multi-
dimensional controlled object (Figure 1), i.e., all or a part of the input and disturbance
effects are related to all or a part of the output product properties.
Figure 1. Multi-dimensional controlled object diagram (black box).
Consequently, the output properties (y) of that product can be represented in the form
of m-dimension column matrix:
y1
y2 
y  ...  y1 y2 ... ... ym
...
ym
, (1)
k-dimension column matrix of input process signals (u):
u1
u2

u  ...  u1 u2 ... ... uk
...
uk
, (2)
314 I. T. Smykov
and l-dimension column matrix of disturbance effects (z):
z1
z2 
z  ...  z1 z2 ... ... zl . (3)
...
zl
Note here that the transpose of matrices is denoted with the prime.
The output variables should have the same physical dimension, they could be
constructed as vector projections on the axis and a matrix column to be identified with
that vector – an output variable vector. Since the products obtained while processed are
characterized by output variables with different physical dimensions, in general, a
transition from the column matrix to the vector can be performed in order to introduce
weighting factors that equalize the dimensions of the individual elements into the column
matrix. Control vector and disturbance vector could be introduced in a like manner when
equalizing physical dimensions of individual column matrix elements of input process
values and disturbance effects.
Correspondingly, dynamic equations of property change of the product as a multi-
dimensional object while it processed can be represented in the frequency domain using
the Laplace transform in matrix form [22]:
q(s) y(s)  r (s)u(s)  p(s) z(s) (4)
where: q(s) - square matrix of the output variables:
q11( s ) q12 ( s ) ... q1m ( s )

q ( s) q22 ( s ) ... q2 m ( s )
q( s)  q j i ( s)  21 , (5)
mm ... ... ... ...
qm1 ( s ) qm 2 ( s ) ... qmm ( s )
r(s) - rectangular matrix of the input signals:
r11( s ) r12 ( s ) ... r1k ( s )

r ( s) r22 ( s ) ... r2 k ( s )
r ( s )  rj i ( s )  21
m k ... ... ... ...
rm1 ( s ) rm 2 ( s ) ... rmk ( s )
, (6)
p(s) - rectangular matrix of the disturbance effects
p11( s ) p12 ( s ) ... p1l ( s )

p (s) p22 ( s) ... p2l ( s)
p( s)  p j i ( s)  21
m m ... ... ... ... , (7)
pm1 ( s) pm 2 ( s ) ... pml ( s)
If the Laplace transforms goes under zero initial conditions in the expressions (1) -
(7), the matrix equation (4) may be written for Laplace transforms as:
Q(s)Y (s)  R(s)U (s)  P(s)Z (s) (8)
where: Y ( s)  Ly(t ) U ( s)  Lu(t ) and Z (s)  Lz (t ) - corresponds to

the matrices - the Laplace transform columns for output variables, input process signals
and external disturbances.
The equation (8) includes also a square matrix Q( s)  q j i ( s) mm and rectangular
matrices
R( s )  r j i ( s ) and
P( s )  p j i ( s ) .
mk ml
If the matrix Q(s) is non-singular, i.e., the determinant Q( s)  0 , then by

multiplying the left and the right parts (8) by inverse matrix Q-1(s), we will obtain:
Y (s)  G0 (s)U (s)  Gz (s)Z (s) (9)
The transfer function matrices G0(s) of the multi-dimensional object for the input
control actions are introduced here:
~
Q( s)
Go ( s)  G o ji  R( s ) (10)
mk Q( s )
and for the disturbance effects:
~
Q( s)
Gz ( s )  G z
ji  P( s ) (11)
ml Q( s )
316 I. T. Smykov
~
symbol Q( s)  Q j i ( s) mm
is used for denoting an associated matrix Q(s) , and Q j i (s)
is a matrix determinant cofactor, Q(s) .

Equations (9)-(11) allows to produce a connection between output variables and
controlling variables and disturbance effects. For example, for m = 3, k = 2 and l = 0, the
equation system for the Laplace transform of output variables will take the form:
Y1 ( s )  G o11( s )U1 ( s )  G o12 ( s )U 2 ( s ),

Y2 ( p)  G o 21( s )U1 ( s )  G o 22 ( s )U 2 ( s ), (12)
Y3 ( p )  G o
31 ( s )U1 ( s )  G o
32 ( s )U 2 ( s ).
The transfer function of multi-dimensional object may be fully represented in matrix:
G y1u1 ( s ) G y1u 2 ( s ) ... G y1ul ( s )

G y 2 u1 ( s ) G y 2u 2 ( s ) ... G y 2ul ( s )
G yu ( s )  , (13)
... ... ... ...
G ymu1 ( s ) G ymu 2 ( s ) ... G ymul ( s )
The characteristic matrix of multi-dimensional system represents a square matrix of

order m  m:
D( s )  I  G ( s )  d j i ( s ) (14)
mm
where: I – a single matrix of order m  m.

The characteristic equation of multi-dimensional system is obtained by setting a
characteristic matrix determinant equal to zero:
D(s)  I  G(s)  0 . (15)
Thus, considering the product as a multi-dimensional object of the process control

and using the Laplace transform, its partial and generalized transfer functions can be
defined from the known input signals and disturbance effects and output variable
dynamics. On the other hand, knowing the object transfer functions, we can define the
output variables, specifically rheological ones, as its response to input signals and
disturbance effects, and, accordingly, to control the process.
3. METHODOLOGY APPLICATION
When researching the behaviour changes in the viscoelastic systems under the action
of external control and disturbance effects and also when designing the new processes,
the question of precise determination of their rheological behaviour often arises. The
direct analytical or numerical computation of the complex viscoelastic systems presents
severe difficulties of generating and solving the differential equations of mathematical
model of the system. Moreover, computations are complicated by the system
characteristics, which are changed continuously in design. Therefore, a possibility of
mathematical modelling of a specific product based on the experimental data obtained as
a change in the output parameter under the standard external input becomes very
important.
The change of the product output parameter under the applied external inputs is well-
known in a real production environment, or can be determined. The input process effects
ui(t) on a product when manufactured could be of different nature and type. However, the
typical input (control) signals are usually applied when performing dynamic tests of the
complex system behaviour. The unit impulse function, unit step function, ramp function
and sinusoidal function inputs as well as their combinations are attributed to them. When
analysing the kinetics of rheological behaviour of the real systems, the Laplace
transforms of these functions are used, and the most common of which are given below.
3.1. Typical Input (Control) Signals
The unit step function is one of the most important and common types of inputs to
the studied system (Figure 2). This could be, for example, the introduction to a product of
catalyst, enzyme or long-acting solvent during the process.
Figure 2. Unit step function (Heaviside step function).

318 I. T. Smykov
The unit step input follows the Heaviside step function:
0, t  0
H (t )   (16)
1, t  0
The Laplace transform of the Heaviside step function:
LH (t )  Lu1 (t ) 
1
(17)
s
Some lag time is observed in most of the rheological processes. Dead time, τ, of the
process is a moment which elapses between input signals and output changes. This time
shifting can be presented by the delayed transfer function:
L f (t   )  e  s (18)
The graph of Heaviside step function with dead time τ is shown in Figure 3.
Figure 3. Heaviside step function with dead time τ.
The Laplace transform of the Heaviside delayed function:
1  s
h( s )  e (19)
s
The ramp function input is also commonly used in the manufacture (Figure 4). For
example, gradual temperature or pressure variations in the process (u2 (t )  K  t ) .
Figure 4. Unit ramp signal.
The Laplace transform of this input signal:
Lu 2 (t ) 
1
(20)
s2
The measurement of the flow properties of a fluid media is often performed using the
instruments with a prime gauge, being in oscillatory motion. This input signal in the
course of in-process metering can be presented as a sine wave input

(u3 (t )  sin 0t )
shown in Figure 5:
Figure 5. Sinusoidal type input signal (sine wave input).

320 I. T. Smykov
The Laplace transform of this input signal:
0
Lu3 (t )  (21)
s 2  0
2
In practical terms, it is always possible to register changes in the output variables

under the influence of these typical input signals when the experimental rheological
studies are carried out. The obtained data makes possible to pass judgement on general
dynamic behaviour of the tested system regardless of its physical nature and composition.
The transient function, i.e., a tested system response to the unit step function signal, is the
most common and possesses the maximum informative value. The transition function
obtained from experiments allows determining the transfer function of the tested system
by modelling it with the use of the basic elements of dynamic systems. Each element of
the dynamic system has a certain physical meaning and its own transient function, which
allows them to be used for modelling of dynamic characteristics of the tested process.
Further on, some standard systems, which are commonly used in rheology, their
transition functions and the Laplace transform are discussed.
3.2. Basic Elements of Dynamic Systems
A fairly large number of standard systems that describe a dynamic object response to
the input signal is used in operational calculus. The dynamic system order is determined
by the highest-order derivative of its differential equation. Equivalently, this is the
highest degree in the denominator of its transient response. The properties of the main
first- and second-order systems, which are important for rheology, will be discussed
further.
3.2.1. Integrating Systems

As an example, a piston movement in a damper with viscous fluid under the constant
force effect (step input function), the Laplace transform of its transient response can be
attributed in rheology to such a system:
LG (t ) 
1
(22)
s
The Laplace transform of transient response (output signal) of this system is

determined by multiplying the Laplace transform of input step signal by the Laplace
transform of its transfer function:
1 1
h( s )  y ( s )   (23)
s s
The inverse Laplace transform of transient response allows going to the time domain:
y(t )  h(t )  L1h(s)  K  t (24)
The transient response graph of this integrating system is shown in Figure 6.
Figure 6. The transient response of integrating system.
3.2.2. First Order System (Exponential)

The Laplace transform of the transfer function of first-order system (exponential):
LG (t ) 
1
(25)
(T  s  1)
The Laplace transform of transient response of this system:
1 1
h( s )  y ( s )   (26)
s (T  s  1)
The inverse Laplace transform of transient response in time domain:
y(t )  h(t )  L1 h( s)  1  exp(  )

t
(27)
T
322 I. T. Smykov
The transient response graph of this first-order system is shown in Figure 7.
Figure 7. The transient response of first-order system.
3.2.3. Transfer Functions in Series

The Laplace transform of the general transfer function in series (Figure 8) results
from multiplication of the Laplace transforms of their transfer functions.
Figure 8. Block diagram of the integrating system and first-order system in series connection.
Upon the integrating system and first-order exponential system in series connection,
the Laplace transform of their general transfer function will be as follows:
LG (t ) 
1 1
 (28)
s (T  s  1)
The Laplace transform of the transient response of this system:
1 1 1
h( s )  y ( s )    (29)
s s (T  s  1)
Then, a transient response of this connection in time domain will be as follows:

y(t )  h(t )  L1 h( s)  t  T  T  exp(  )

t
(30)
T
The transient curve of general transient response of integrating system connected to
first-order system is shown in Figure 9.
Figure 9. The general transient response of integrating system and first-order exponential system
connected in series.
It is evident that a similar change in the output signal can be obtained, if the ramp
input signal impacts on the first-order system input.
If the differentiated system is connected in series to the first-order exponential
system, the Laplace transform of the transfer function of such connection will be as
follows:
LG (t )  s 
1
(31)
(T  s  1)
The transient response of the systems in series connection in time domain:
y(t )  h(t )  L1 h( s) 

1 t
 exp(  ) (32)
T T
The transient response of the real differentiated system is shown in Figure 10.
324 I. T. Smykov
Figure 10. The transient response of differentiated system and first-order exponential system connected
in series.
3.2.4. Transfer Function in Parallel

The Laplace transform of the general transfer function of two or more different
systems connected in parallel is obtained by adding the Laplace transform of their
transfer functions (Figure 11).
Figure 11. Block diagram of the two first-order systems connection in parallel.
For example, if two first-order exponential systems with different time constants are
connected in parallel, the transform of their general transfer function will be as follows:
LG (t ) 
K1 K2
 (33)
(T1  s  1) (T2  s  1)
The Laplace transform of the general transient response of the system:

1  K1 K2 
h( s )  y ( s )      (34)
s  (T1s  1) (T2  s  1) 
Then, a transient response of such connection in time domain will be as follows:
t t 
y(t )  h(t )  L1 h( s)  K1  K 2  K1  exp    K 2  exp   (35)
 T1   T2 
One of the possible options of the graph of general transient response of two first-
order exponential systems with different time constants connected in parallel is shown in
Figure 12.
Figure 12. The general transient response of two first-order exponential systems connected in parallel.
3.2.5. Second Order Systems (Double Exponential)

The Laplace transform of the transfer function of double exponential system is often
seen in rheological studies:
LG (t ) 
1
(36.1)
(T1  s  1)  (T2  s  1)
The Laplace transform of the transient response of this system:
1 1
h( s )   (36.2)
s (T1  s  1)  (T2  s  1)
326 I. T. Smykov
The inverse Laplace transform of the transient response in time domain:
T1 t  T2 t 
y(t )  h(t )  L1 h( s)  1   exp     exp   (37)
(T2  T1 )  T1  (T2  T1 )  T2 
The second order system transient response is shown in Figure 13.
Figure 13. The second order system transient response.
3.2.6. Feedback Systems

The complex systems (Figure 14), in which an output signal value influences on a
control input value, occur in practice of rheological studies and some processes, which
means there is a feedback.
Figure 14. Block diagram of the feedback system.

The feedback system transfer function is determined by the transfer functions of

direct transform and feedback by the equation:
g d ( s)
g (s)  (38)
1  g d ( s)  g f ( s)
g (s) g (s)
where d - transfer function of direct transform, f - transfer function of
feedback.
The plus sign is used for negative feedback and the minus sign for positive one.
There are only some widespread standard input signals, systems and their
characteristics are presented here as examples. Their complete and detailed description is
provided in the specialized literature [23, 24].
3.2.7. Response to Sinusoidal Input

A response of the tested object to external vibration input is very important in the
rheological studies. Let’s consider as an example a response of first-order exponential
system.
The Laplace transform of a first-order exponential system response (25) to sinusoidal
input (21) is determined by multiplication of its Laplace transform by the system transfer
function:
 1
h( s )   (39)
s 
2 2
(T  s  1)
The inverse Laplace transform of the transient response allows to go to the time
domain:
  t  1 
y(t )  h(t )   exp     sin(  t )  T  cos(  t ) (40)
 T 1   T  
2 2

The sinusoidal input and transient response of first-order exponential system are
shown in Figure 15.
These graphs are quite typical for rheological studies when viscometers and
rheometers with oscillating indenters are used. The result obtained allows determining
both a signal down value and a phase shift in the real object model and comparing it with
the results of experimental tests of objects with viscoelastic behaviour.
328 I. T. Smykov
Figure 15. First-order exponential system response to sinusoidal input.
Rheological measurements with oscillating indenters are conducted on certain

frequency ranges. The dynamic measurement results can be presented, for example,
through a complex shear modulus G:
G  ( )  G( )  iG() (41)
where: G() – the real part of complex shear modulus (storage modulus), and G() –
the imaginary part of complex shear modulus (loss modulus), i   1 , or in the form
of:
G  ( )  G  ( ) cos   i G  ( ) sin  (42)
where  - the phase angle between the sinusoidal input and the output signal.
Considering the process that takes place when measuring, it is easy to see that the
amplitude and phase frequency characteristics of the tested object are determined in fact.
The frequency characteristic can be obtained from the partial process transfer function
(10) or (13) by means of simple substitution with a complex value i instead of a Laplace
variable s, i.e.:
W ( s) si  W (i ) (43)

Thus, the frequency transfer function W(i) represents a complex number, which
absolute value is equal to the output variable amplitude to input variable amplitude ratio,
and the argument – to the output variable phase shift in relation to input variable:
mod W (i )  W (i )  ymax umax ,

arg W (i )   (44)
The frequency transfer function is a Fourier transform of impulse response, i.e., the
integral transformation takes place:

W (i )   w(t )e it dt (45)
0
and the frequency transfer function can be represented as follows:
W (i )  A( )e i  U ( )  iV ( ) , (46)
where: A()- a frequency transfer function module, () – a phase, U() and V()- real
and imaginary elements of the frequency transfer function.
From which the amplitude frequency and phase frequency characteristics of the
tested process model are easily defined:
A( )  U 2 ( )  V 2 ( ) ,
V ( )
 ( )  arctg (47)
U ( )
Comparing (46), (47) and (42), the complex shear modulus can be compared with the
frequency process transfer function, the storage modulus - with the real part of transfer
function, and the loss modulus – with the imaginary element.
More generally, the frequency process transfer function can be represented as the
ratio of the Fourier transform (in frequency domain) of the output variables and input
signals under the input effect of any form (unit pulse or unit step etc.) and physical
nature:
W (i)  y(i) u(i) (48)

330 I. T. Smykov
which is being a direct consequence of the Laplace transform transition to the Fourier
transform.
The isotropic object transfer function obtained from the results of research with the
use of one-type device can be used for modelling of measurements on a device operated
using a different principle of measurement, and different external inputs. Thus,
measurements of objects with viscoelastic behaviour are taken quite often using the
devices in which the input signal (compression or tension) represents a ramp. Irrespective
of the input signal type, the transfer function is a property of the object itself and, if
determined on one device, it will be possible to compute assumed measuring results on a
different-type device, e.g., an oscillating viscometer.
4. USING THE LAPLACE TRANSFORM IN RHEOLOGY
The viewed standard systems can be used for the development of a

phenomenological mathematical model of specific product based on the results of
experimental studies. The real standard control inputs on a product are used for that
purpose when carrying out the tests as well as the rheological behaviour kinetic is
recorded. An adequate model is created by means of combining the standard systems
based on the obtained dynamic test results and a priori known occurring processes.
As an example, let's consider such a well-known and very important process step as
production of rennet curd from milk.
4.1. Kinetic Gelation Special Model Synthesis
The feed milk is a dispersed system characterized by the initial state function Yo,
which includes parameters such as acidity, density, temperature, fat and protein content,
microorganism content characteristic, logical functions of compliance with any samples,
etc. In the present instance, milk is a process control object (Figure 1). The introduction
of rennet into the milk for coagulation is being a control input u(t) and corresponds to an
input step signal (Figure 2). Let’s assume that the modulus of elasticity of resulting milk
curd G(t) is considered as an output signal. The process response to the input step signal
is being a transient response h(t), characterizing the process transition dynamics from the
initial state function Yo to the final state Yk,.
The transient response can be defined, if the process transfer function is known, or
vice versa, it is possible to define a transfer function, if the transient response is known. It
is obvious that we do not know the transfer function. However, the modulus of elasticity
of resulting curd and its behaviour in time, i.e., a transient response, can be easily
measured in practice and such data is widely found in the literature. Figure 16 shows the
storage modulus curves at renneting under the different conditions.
Figure 16. Change in the storage modulus at renneting: tg - gelation points, G - storage modulus.
The empirical approximating model of the change of storage modulus of resulting

rennet curd - G(t) proposed by Scott-Blair and Barnett [5] in 1958, and refined later [6],
was one of the first attempts for the mathematical description of that process:
G(t )  G [1  exp[ k (t  t g )]] (49)
However, it is based on the use of a priori unknown limiting storage modulus values
(G’∞). This model is incapable to predict the gel behaviour under the temperature
variations, рН, enzyme concentration and the effects of other input factors, and it does
not describe the process occurring prior to gelation point.
Let's consider this model in detail. Its Laplace transform is:
 1 exp k  t g 
LG(t )  G  
s s  k   (50)
Then the Laplace transform of the process transfer function will be defined following
(12):
332 I. T. Smykov
Lg ( p)  LG(t )/ Lu(t ) (51)
or:
 1 s exp k  t g 
Lg 0 ( p)  G  
 C0 s  k   (52)
We will assess a process dynamic component without getting into specifics of

influence of the constants G, Со and tg:
 s 
F (t )    (53)
 s  k  
It is evident that this is a transfer function of the standard differentiated system with
delayed action. However, it is well-known that the processes occurring at coagulation of
milk are of integrative nature – first, casein micelles form the aggregates, and then the
aggregates are combined into a single spatial structure of curd. At the same time, these
are the delayed processes, i.e., there is a lag phase. Hence, it is not difficult to conclude
that the result obtained here does not correspond to the real processes occurring at
coagulation of the milk. The analysis demonstrates that this is due to the incorrect choice
of function (49) describing the kinetics of rennet curd formation. Other models of Scott-
Blair do not also provide a correct description of the nature of coagulation, though they
approximate the experiments results accurately.
Further modelling of the enzymatic gelation of milk was described in the Carlson's
works [8-11], and the model proposed by him was presented as a sum of two
exponentials:
kl kf
G  G [1  ( )  exp( k f  t  )  ( )  exp( kl  t  )] (54)
kl  k f kl  k f
Basically, this model provides the appearance of "active" sites on a casein micelle,
which subsequently react with each other in the formed bonds of gel structure. Both
reactions, activation and deactivation, are presented here as the first-order processes.
Activation emerges in the enzymatic hydrolysis of the κ-casein, and therefore it is
proportional to the enzyme concentration and activity. The statement that the deactivation
reaction of sites on a micelle should also be of first order is the empirical observation,
which is associated with the necessity to include small aggregates and free micelles
which remained inactive by that time in the gel structure. At a high enzyme
concentration, κ-casein hydrolysis occurs rapidly compared to micelles deactivation, and
the Carlson's model is transformed into single exponential model. According to some
authors [12], the Carlson's model, in general, describes experimental data better than the
Scott-Blair's model. However, it requires to be further improved so that the model could
consider the induction (flocculation) gelation phase and a greater number of influencing
factors.
We have earlier reviewed a standard double exponential system (9) and (11). If to
compare formulas (37) and (54), it will be easy to notice that the Carlson's kinetic
gelation model differs only with the terms and symbols from the transient response of
double exponential second-order system, which is suggestive of the relevance to use the
Laplace transform and appropriate standard systems for the considered rheological
process kinetic modelling.
The main drawback of the Carlson's model is that it does not consider such important
part of the gelation process as a flocculation. However, this drawback can be avoided by
using the Laplace transform.
The flocculation phase duration is characterized by the coagulation time (τ), and it
can be viewed, when modelling, as a delayed system that corresponds to the Heaviside
step function with the time delay, which we have discussed earlier in (18) and (19).
Then, the modified Carlson's kinetic gelation model, including the flocculation phase,
will be as follows:
 ( t  )  ( t  )
 
Lm
1
h(s)  H (t   )  K 1 
 T1
e T1

T2
 e T2 

(55)
 T 2T1 T2  T1 
However, it was established based on the results of our tests that the three-
dimensional stochastic micro flow field is generated in the system during gelation and it
contributes to increased quantity of interacting components in the gel structure per time
unit. This feedback covers the coagulation phase only (38). The feedback is characterized
by depth (β), which, in this instance, demonstrates which part of energy used for the gel
formation is spent for the micro flow field excitation in the process volume.
Consequently, the elements covered with a non-identity feedback can be described by the
following transfer function:
g d ( s)
g ( s)  (56)
1    g d ( s)
334 I. T. Smykov
In addition, a rating instantaneous element (K) with transfer function, which provides
coordination in variable and dimension of the input signal and the output parameter is
required for the model.
Therefore, the functional disperse system diagram (of milk) as a process control
object in the formalized form used in operational calculus will be as such shown in
Figure 17.
Figure 17. Functional bloc diagram of milk model as a process control object.
The general transfer function of this milk model as a control object is determined by
multiplication of the elements connected in series and the part of which is covered with a
feedback, and appears as follows:
g m s   e  s 
T1  s  1  T2  s  1
(57)
 K 
1    
 T1  s  1  T2  s  1
The Laplace transform of the function of transient process h(s) from the disperse
system state to the gel state, i.e., a response to introduction of some amount of enzymatic
agent, or otherwise rheology kinetic at gelation in milk, will be as follows:
1 
hs     g m ( s)  (58)
s 
or:
Gs   e  s  
1 T1  s  1  T2  s  1 (59)
s  K 
1    
 T1  s  1  T2  s  1
The inverse Laplace transform from the expression (59) is applied for the transition
in time domain using a relevant application (for example, Mathcad). The resulted
expression is very cumbersome to write and so it is not given here. However, using the
simplest algebraic transformations and introducing intermediate factors A and B, and
without any functional reductions, it is written as follows:
Gt  
  
K  H t     2  A  T1  T2  1  coshA  t    e Bt    T1  T2   sinhA  t    e Bt    (60),
2  T1  T2  1  K   
T1  T2 2  4  K  T1  T2    T1  T2 
where: A  and B  intermediate factors,
2  T1  T2 2  T1  T2
H(t - τ) - Heaviside delayed step function, cosh[*] – hyperbolic cosine, sinh[*] –
hyperbolic sine.
The obtained model of elasticity modulus kinetic during gelation remains
cumbersome in the arrangement, but the use of common application software for
computation does not make difficult its application, because it mechanically derived from
the expression (59). It is worth to underscore here that the resultant model is not a
speculative construction, but a result of mathematical modelling using the Laplace
transform.
4.2. Kinetic Gelation Model Verification
The studied results in terms of the effect of acidity and enzyme concentration on
rheological behaviour of milk gel described in the paper [25, Figure 1] have been taken as
a modelling object. The gelation curve graphing provided in this paper was digitized with
the aim of verification. Then, using the iteration method, the values of rating coefficient
(K), time constants (Т1, Т2) and feedback factor (β) were determined. The verification
result of the developed model (60) of enzymatic gelation kinetics is shown in Figure 18.
The correlation coefficient value between experimental and design data were R2 =
0.999 at P = 0.95. Figure 19 shows the enlarged view of the initial gelation phase,
characterizing the coagulation time τ = 8.83 min, which practically coincides with a value
given in this paper, and with a smooth transition from flocculation phase to coagulation
phase. Therefore, the obtained gelation model not only describes the coagulation process
adequately, as it runs in other well-known models, but also includes the flocculation
process phase.
336 I. T. Smykov
Figure 18. Gelation model verification. Circles – experimental data [25], continuous line – model (11).
The similar verification of the developed kinetic gelation model has demonstrated,
which is consistent with many literature and own experimental data, that this model is
capable with a high confidence to describe the results of various experiments dedicated to
the assessment of external effects on rheological gel parameters.
It means that it is being an adequate approximating model and suitable for descriptive
purposes. However, this model (60) does not allow to determine the impact of individual
input signals and/or disturbance effects on gelation kinetics and a final product
rheological behaviour. This particular task is the main goal for the development of the
complete mathematical gelation model, and the approximating model (60) may serve as a
base for its solution.
Figure 19. Initial coagulation phase of gelation. Circles – experimental data [25], continuous line –
model (11).
4.3. Kinetic Gelation Special Model Synthesis
To be used effectively, the mathematical enzymatic kinetic gelation model must be

minimum written as follows:
Y (t )  F C r , C с , T, pH  (61)
where: Y(t)- output gel parameter (shear modulus, viscosity etc.), Cr – enzymatic agent
concentration, Cc – calcium chloride concentration, T – temperature, pH – acidity.
In fact, the formed gel parameters depend considerably on many factors and they are
quite difficult to consider at once. Moreover, not more than three – four factors are often
varied in practice. Therefore, the most significant and frequently used factors have been
initially chosen for the model. Eventually, the synthesized kinetic gelation model should
take into account the influence of other factors, i.e., to be scalable by factors.
There are many works dedicated to the studies of the impact of different factors and
external effects on the relation of kinetic and rheological characteristics. The results of
these studies contribute to the development of relevant mathematical models as well as
the discovery of effective control methods for the production process.
The work [26] is dedicated to the study of the milk acidity effect on the gelation time,
and it demonstrates that the effect is of a nonlinear nature and the gelation time is reduced
dramatically as it gets close to isoelectric point. The work [25] is dedicated to the study
of the joint effect of the acidity and enzyme concentration on the rheological curd
behaviour. The two-factor experiments demonstrated that the joint effect of these factors
on gelation is not being additive. The joint acidity and enzyme concentration effect has a
nonlinear nature.
Milk acidity influences drastically on the rennet coagulation capability and the
obtained curd properties. The work [27] deals with the study of that influence on
coagulation time and storage modulus. Acidification results in the coagulation time
decrease, and such decrease is quicker as closer the isoelectric point. The gel elastic
modulus is reduced likewise with acidification. It may be concluded from this study that,
for the considered experiment, as coagulation time decreases so modulus of elasticity
reduces upon unchanged enzyme concentration. The directly opposite picture is observed,
if the enzyme concentration is modified at unchanged acidity. In this instance, modulus
of elasticity increases with the decrease of coagulation time. Consequently, the gel elastic
modulus and coagulation time relationship is determined by the initial milk acidity.
The impact of the calcium and phosphorus ions concentration on the milk
coagulation time and the storage modulus of the gel was previously studied [28]. It was
shown that under the joint effect of these salts, coagulation time will have a relevant
338 I. T. Smykov
minimum at mean values of their concentration, while modulus of elasticity slightly will
change.
The milk temperature during the coagulation is one of the most important and
accessible control factors, which is extensively used in a production environment. This is
a good reason why a great number of works is dedicated to the studies of temperature
effect on coagulation. The results of studies of the temperature effect on coagulation time
and gel elastic modulus are given in [29]. The nature of the temperature effect on these
parameters is also nonlinear. Time is decreased exponentially with the increase of
coagulation temperature, while gel storage modulus increases at the beginning and then
drops, i.e., the dependency is close to quadratic function. Apparently, the optimal gelation
temperature exists under such conditions.
The influence of several external effects on the milk coagulation kinetics cannot be
presented as a simple arithmetical sum of the effects of the individual factors. There are
pairwise, triple and multiple interactions between factors and gelation kinetics. Many
papers are dedicated to their study, in particular, the joint effect of acidity; calcium
chloride concentration; temperature and enzyme concentration were investigated in [30].
The coagulation time dependencies from individual factors were graphically presented. A
multiple regression equation of the coagulation time and all the external effects were also
plotted. The multi-factor study demonstrated that acidity has the greatest impact on the
coagulation time, while the temperature contributes to the prediction of the curd tension
parameters. Except for the curd tension, the coagulation parameters are determined to a
larger extent by acidity, temperature and calcium concentration interactions. The
coagulation time depends on two- and three-factor interactions. Acidity - calcium
concentration interaction was the only combination of factors related to the curd tension.
The obtained linear regression model in that paper is of some interest, but despite the
experimental results evidently influence nonlinearly on coagulation time, the paper does
not take this into account. So, its wide application is in doubt.
The work [31] discusses the complex study of the influence of various factors on
parameters that on one way or another characterizes coagulation. As in the paper [30], the
results of the experimental studies are presented and a number of mathematical models
for determination of the coagulation time and an elastic modulus limit value for the curd
are provided. However, the reliability of these equations cast doubts, because the milk at
рН < 5 (below the isoelectric point) should be curdled quickly (immediately), but
according to the provided equations, the coagulation time is quite long. The proposed
formulas for the limit curd elastic modulus are open to question.
The coagulation time and modulus of elasticity presented in one or another form are
being the main and reliably measurable parameters of the gelation process. The work [32]
has also discussed the influence of various factors on these parameters and some
mathematical models of that effect. Based on the results of our approbation, the
dependencies provided in this paper allow calculating the milk coagulation time
accurately under the specified conditions. At the same time, these dependences lack the
effect of calcium chloride concentration on the coagulation time, which is critically
relevant. Moreover, the nature of the effect of the factors considered should be also
clarified based on the data published by other authors.
We corrected and added the empirical dependency [32] after the performance of
thorough comparative theoretical and experimental studies. As a result, the influence
model of the foregoing external effect on the coagulation time (τ) took the following
synthetic form:
log   0.0176  log C r   0.446  log C r  0.084   pH   0.84  pH

2 2
0.0272  Θ (62)
 0.0053  Θ2  0.414  Θ  0.25  C c  0.3  C c  11.6
1.5 0.5
log C r
The formula (62) has an intercept term (= 11.6), which serves for normalization and
can be changed to either side, excluding the influence of any factors or introduction of
additional factors, for example, the concentration of casein, fat, salt etc.
Milk coagulation time, which runs over the log value from the specified values of the
external effects (62), is one of the main factors determining the values of kinetic gelation
model parameters (60).
The development of kinetic gelation model contains four variable parameters: K, T1,
T2 and .
 K parameter (static transition coefficient) is being a rating coefficient in this
model, it is a constant number, and it does not depend on the milk properties and
external effects, K  5000. It can be changed to either side, but only
dependencies of other parameters from external effects will change, while the
modelling result is preserved;
 T1 parameter – time constant characterizing the flow intensity of inter- and cross-
micellar biochemical reactions;
 T2 parameter – time constant characterizing the decay of coagulation process,
which is a result of depletion of the disperse system reactive elements;
 β parameter – feedback factor determining effectiveness of interaction between
micelle clusters, free micelles, between clusters and micelles (intensity of
stochastic micro flows).
Time constant T1 in the model (60) also depends as coagulation time on internal
factors and external effects, but it is bound with it by a simple relationship:
T1  k1   (63)
340 I. T. Smykov
where: τ - measured in minutes, k1 - first coefficient depending on the internal milk

properties, and based on the study results its value is within the range: 0.3 < k1 < 0.4:
Time constant T2 is interconnected with the influencing factors like T1 via
coagulation time:
T2  k 2   
1/ 2
(64)
where k2 - second coefficient, which also depends on the internal milk properties, and it
is usually within the range: 300 < k2 < 1100.
Feedback factor  is strongly influenced by the acidity and static transfer coefficient
K:
k r  1/ 3
 (65)
K   pH  5.44
where: kr – a coefficient depending on a set of internal properties of the specific milk

sample, it is within the range: 10 < kr < 16.
Consequently, by setting the milk temperature, its acidity, concentration of
introduced rennet and calcium chloride, the coagulation time is determined using the
formula (62). Further, time constants and feedback are calculated using the formulas (63)
- (65). Substituting the values received in (60), we get a numerical model for elastic
modulus change during gelation for a specific type of milk.
In general, the obtained model dependencies are close to the results of experimental
studies of various authors and were considered above. This proximity is conditioned by
the fact that the development of the gelation model was based on the analysis and
synthesis of the results of these works, but the dependencies may differ from the results
obtained in other studies, where another milk or another milk pre-treatment was used.
4.4. Kinetic Gelation Special Model Verification
Numerical verification of the special kinetic model of sol-gel transition to milk has
been conducted using the results of experimental studies published in the work [28]. This
paper studies the effect of acidity value on the coagulation time and gel elastic modulus
with the unchanged intrinsic milk properties and the same external conditions. The
objective of verification carried out by us was to establish adequacy of the kinetic
gelation model with the fixed intrinsic milk properties, i.e., with constant values of
coefficients k1, k2 and kr and a broadly varied acidity value.
The graphical presentation of the modelling results using the kinetic gelation model
(60) and the empirical model of the external effects on the gelation time (62) are
illustrated in Figure 20.
Figure 20. Modelling of acidity effect on coagulation time and gel storage modulus (continuous lines)
based on experimental data of the work [28]: pH 6.7 (■); 6.4 (□); 6.0 (▲); 5.8 (△); 5.6 (●), and рН 5.4
(○).
Correlation coefficients between modelling results and experimental data are close to
one, which makes possible to speak about the descriptive adequacy of kinetic model for
the processes occurring at rennet gelation under the described experiment conditions. We
see here that gelation process, upon approximation of the acidity value to the isoelectric
point, worsens considerably and this confirms the existence of its acidic instability in the
option of enzymatic milk curdling.
It is worth to mention that the obtained modelling result is only substantial, effective
and unbiased for the considered case. The use of the model for prognostic evaluations is
only for the strict compliance with the defined experiment conditions.
The maintenance of the constancy of initial intrinsic milk properties, including its
storage and pre-treatment, bacterial content etc., which more or less have an effect on the
form of dependencies (63) – (65) and, eventually, on predicted results, are attributed here.
At the same time, the basic kinetic gelation model (60) remains unchanged in any case.
The algorithm of kinetic gelation model usage aimed at receiving prognostic
evaluation of the results is the following:
342 I. T. Smykov
 Depending on the set task, the type of initial milk (model, reconstituted from dry
milk, from specific animal, a group of animals, farm etc.) is chosen and it is
treated further on as a test one and a reference sample;
 Specific gelation conditions in the test milk sample (enzyme and calcium
chloride concentration, temperature and acidity) are chosen, and their values will
be used subsequently in (62);
 Using the available equipment, modulus of elasticity is measured in the course of
the milk gel formation and gelation time is determined;
 Substituting specific gelation conditions in the test for milk sample in (62), a
theoretical value of the coagulation time is determined and, if it is necessary, it is
corrected to the experimental value by changing the value of intercept term;
 Using the iteration method, coefficients k1, k2 and kr are revised in formulas
(63) – (65) within the limits ± 10% in order to have experimental points (curves)
and theoretical values to be maximum close to each other. These coefficients
remain unchanged further on for the given type of milk;
 Once the reference milk sample properties were tested, it is possible to proceed to
the modelling of external effects on gelation;
 If it is necessary to assess the influence of other external effects on the gelation
by experimental or literature data, the additional terms and their dependencies for
account of the reduction of the intercept term are introduced to the formula (62).
The example of the formalization procedure for the sol-gel transition process in the
dispersed system is provided here. The process formalization is a basis for the
development of the relevant application software, which can provide the effective process
control and gain an additional profit to the production. We have developed one of the
variants of such application software and use it in practice.
The gelation processes are widely spread in many modern biochemical and chemical
technologies of manufacture of various materials and finished products. First, this relates
to polymerization processes, which theory and relevant mathematical models have been
for the first time discussed in the Flory's works [33, 34]. His proposed approaches are still
valid and used by many researches. For example, the work [35] considers the temperature
effect on the process of polyacrylamide gel formation from aqueous acrylamide solution.
The analysis has been performed using the vibration rheometer, which registers modulus
of elasticity and loss modulus in the test sample. Our digitalized results on conducted
researches from this work such as polyacrylamide gel elastic modulus and loss modulus
behaviour in time depending on temperature are illustrated in Figure 21.
If these graphs are compared with the results from the enzymatic gelation in milk,
their evident similarity can be noticed, which means that the acrylamide polymerization
process can be describe using the elaborated gelation model. Figure 21 demonstrates the
modelling result received using the model (60) in which the time constants of the gelation
process activation and decay, coagulation time and feedback factor were specifically
selected. Time values in minutes were used for the modelling.
Figure 21. Temperature effect on polyacrylamide gel formation. Symbols – data from the work [35],
continuous lines – the result of modelling using the model (60).
CONCLUSION
The Laplace transform is a powerful mathematical tool for solving many complex
scientific and practical issues associated with the viscoelastic behaviour kinetics of
various objects. Moreover, the object can be considered as a single multi-dimensional
controlled object – a “black box,” characterized by the existence of a set of points of
application of external process effects (inputs), a transfer function and a set of output
variables describing the properties of the product obtained. The kinetics of rheological
(viscoelastic) behaviour of the object is characterized by a transient function or a
response characteristic of the object to the external process effect.
The typical input effects presented in this Chapter as well as the transfer and transient
functions in the form of the Laplace transforms obviously demonstrate the kinetics of
rheological (viscoelastic) behaviour of objects and allow modelling it without regard to
the essence of occurring internal processes.
The development of two mathematical models of sol-gel transition kinetic using the
Laplace transform was taken as an example. The first one is general and approximating it
can be widely applied in science and technology. The second one is special and possesses
prognostic properties, and it is intended for controlling rheological behaviour kinetic in
the process of enzymatic phase of the sol-gel transition in the milk during production of
cheese based on the introduced enzyme and calcium chloride concentration values and
344 I. T. Smykov
the temperature and acidity variables. The models have been verified using the published
experimental data of different authors. An algorithm for using a special kinetic model of
sol-gel transition in milk for prognostic evaluation of the milk gel behaviour for non-
standard milk was proposed.
The developed models of gelation kinetics can also be used for mathematical
modelling of polymerization processes of various chemical compounds.
The use of such methodology approach allows intensifying the work on optimization
of processing modes of various products, development of new and modernization of
existing processes, elaboration of control methods, and predicting the products quality.
REFERENCES
[1] Clark, A. H., Ross-Murphy, S. B. 1987, Structural and mechanical properties of

biopolymer gels. Adv. Polym. Sci., 83, 55-192.
[2] Guenet, J. M. 1992, Thermoreversible gelation of polymers and biopolymers.
Academic Press, London, UK, p. 278.
[3] Nijenhuis, K. 1997, Thermoreversible Networks. Viscoelastic Properties and
Structure of Gels. In Advances in Polymer Science. Springer Berlin Heidelberg, p.
267.
[4] Stepto, R. F. 1998, Polymer Networks: Principles of their Formation, Structure and
Properties, Chapman & Hall, Glasgow, p. 290–318.
[5] Scott Blair, G. W., Burnett, J. 1958, Physical changes in milk caused by the action
of rennet. J. Dairy Res. 25, 297–303.
[6] Scott Blair, G. W., Oosthuizen, J. C. 1961, A viscometric study of the breakdown
of casein in milk by rennin and rennet. J. Dairy Res. 28, 165–173.
[7] Douillard, R. 1973, Rheological analysis of curd formation. J. Text. Stud., 4, 158–
165.
[8] Carlson, A, Hill, Jr. C. G, Olson, N. F. 1987, Kinetics of milk coagulation: I. The
kinetics of kappa casein hydrolysis in the presence of enzyme de-activation,
Biotechnol. Bioeng., 29, 582–589.
[9] Carlson A, Hill Jr. C. G, Olson, N. F. 1987, Kinetics of milk coagulation. II.
Kinetics of the secondary phase: micelle flocculation, Biotechnol. Bioeng., 29, 590–
600.
[10] Carlson, A., Hill Jr., C. G., Olson, N. F. 1987, Kinetics of milk coagulation. III.
Mathematical modelling of the kinetics of curd formation following enzymatic
hydrolysis of k- casein – parameter estimation. Biotechnol. Bioeng., 29, 601–611.
[11] Carlson, A., Hill Jr., C. G. and Olson, N. F. 1987, Kinetics of milk coagulation. IV.
The kinetics of the gel-firming process. Biotechnol. Bioeng., 29, 612–624.
[12] Esteves, C. L. C., Lucey, J. A. and Pires, E. M. V. 2001, Mathematical modelling of

the formation of rennet-induced gels by plant coagulants and chymosin. J. Dairy
Res., 68, 499–510.
[13] Ali, F., Khan, I., Shafie, S. 2014, Closed Form Solutions for Unsteady Free
Convection Flow of a Second Grade Fluid over an Oscillating Vertical Plate. PLoS
ONE, 9(2), e85099.
[14] Ali, M., Awais, M. 2014, Laplace Transform Method for Unsteady Thin Film Flow
of a Second Grade Fluid through a Porous Medium, J. Mod. Phys., 5, 103–106.
[15] Mathur, R., Poonia, S. 2013, Application of Laplace Transform to Newtonian Fluid
Problems. Int. J. Sci. and Res., 2(7). 358–361.
[16] Arfin, N., Aswal, V. K., Kohlbrecher, J., Bohidar, H.B. 2015, Relaxation dynamics
and structural changes in DNA soft gels. Polymer, 65, 175–182.
[17] Tirella, A., Mattei, G., Ahluwalia, A. 2014, Strain rate viscoelastic analysis of soft
and highly hydrated biomaterials. J. Bio. Mat. Res., 102(10), 3352–3360.
[18] Xu, B., Li, H., Zhang, Y. 2013, Understanding the viscoelastic behavior of collagen
matrices through relaxation time distribution spectrum. J. Biomatter, 3(3), e24651.
[19] Aufderhorst-Roberts, A., Frith, W. J., Kirkland, M., and Donald, A. M. 2014,
Microrheology and Microstructure of Fmoc-Derivative Hydrogels. Langmuir, 30,
4483−4492
[20] Grimard, J., Dewasme, L., and Wouwer, A. V. 2016, A Review of Dynamic Models
of Hot-Melt Extrusion. Processes 4(2), 19−47.
[21] Dyke, P. P. G. 2000, An Introduction to Laplace Transforms and Fourier Series.
Springer, p. 250.
[22] Riley, K. F., Hobson, M. P., Bence, S. J. 2010, Mathematical methods for physics
and engineering, 3rd ed., Cambridge University Press, UK, p. 455.
[23] Doetsch, G. 1974, Introduction to the Theory and Application of the Laplace
Transformation, Springer-Verlag Berlin Heidelberg, p. 327.
[24] Dorf, R. C. and Bishop, R. H. 2010, Modern Control Systems, 12th ed., Pearson,
Prentice Hall, p. 1104.
[25] Lopez, M. B., Lomholt, S. B. and Qvist, K. B. 1998, Rheological Properties and
Cutting Time of Rennet Gels. Effect of pH and Enzyme Concentration. Int. Dairy
J., 8, 289–293.
[26] Ezeh, V. N. Lewis, M. J. 2011, Milk reversibility following reduction and
restoration of pH. Int. J. of Dairy Tech., 64(2), 179–187.
[27] Choi, J., Horne, D. S., Lucey, J. A. 2007, Effect of Insoluble Calcium
Concentration on Rennet Coagulation Properties of Milk. J. Dairy Sci., 90, 2612–
2623.
[28] Udabage, P., McKinnon, I. R. Augustin, M. A. 2001, Effects of Mineral Salts and
Calcium Chelating Agents on the Gelation of Renneted Skim Milk. J. Dairy Sci.,
84, 1569–1575.
346 I. T. Smykov
[29] Esteves, C. L. C., Lucey, J. A., Hyslop, D. B., Pires, E. M. V. 2003, Effect of
gelation temperature on the properties of skim milk gels made from plant
coagulants and chymosin. Int. Dairy J., 13, 877–885.
[30] Najera, A. I., de Renobales, M., Barron, L.J.R. 2003, Effects of pH, temperature,
CaCl2 and enzyme concentrations on the rennet-clotting properties of milk: a
multifactorial study. Food Chem., 80, 345–352.
[31] Daviau, C., Famelart, M. H., Pierre A., Goudedranche H., Maubois J.-L. 2000,
Rennet coagulation of skim milk and curd drainage: Effect of pH, casein
concentration, ionic strength and heat treatment. Lait, 80, 397–415.
[32] Hori, T. 1985, Objective Measurements of the Process of Curd Formation during
Rennet Treatment of Milks by the Hot Wire Method. J. of Food Sci., 50, 911–917.
[33] Flory, P. J. 1941, Molecular size distribution in three dimensional polymers. (I, II,
III). J. Am. Chem. Soc. 63, 3083–3090, 3091–3096, 3096–3100.
[34] Flory, P. J. 1941, Principles of Polymer Chemistry, Cornell University Press:
Ithaca, NY, 1953.
[35] Calvet, D., Wong, J. Y., Giasson, S. 2004, Rheological Monitoring of
Polyacrylamide Gelation: Importance of Cross-Link Density and Temperature.
Macromolecules, 37, 7762–7771.
Chapter 11
THE DESIGN OF VIBRATION ABSORBERS USING

THE RHEOLOGICAL PROPERTIES OF
VISCOELASTIC MATERIALS
Juan Carlos Jauregui-Correa*

Universidad Autonoma de Queretaro, Queretaro, Qro. Mexico
ABSTRACT
The design of vibration absorbers, in general, combine materials with high rigidities
and very elastic materials. The main purpose of these designs is to reduce the
transmission of vibration waves along structures and diminish the damage in machinery.
There are many methods for reducing the transmission of vibration waves; but the most
common are passive vibration isolators. These isolators are constructed with a metal base
and an elastomer that acts as a damper. Vibration isolation is the most widely used
technique for vibration and noise protection, because it breaks the transmitted wave by
modifying the structural stiffness of the system, and by adding a higher damper. The
design of these devices represents a big challenge, even though these devices have been
used for decades, the theoretical models and numerical solutions are valid only within
limited geometries and specific designs. There are two major limitations: one is the
development of an appropriate elastomer model that relates the rheological properties
with the actual dynamic behavior, and the numerical formulation of the dynamic model
that includes the elastomeric properties and the rigid behavior of the metal components.
There have been different approaches for solving these limitations, but the analytical
results are far from the experimental result, except in well-known cases. In this chapter, a
detailed analysis of the different formulations for design vibration absorbers is presented.
In this analysis, the definition of the dynamic behavior of viscoelastic materials based on
their rheological properties is included and two numerical formulations are presented,
*
Corresponding Author: Cerro de las Campanas S/N, Col Las Campanas, Queretero, Qro, 76010, Mexico, Email:
jc.jauregui@uaq.mx.
348 Juan Carlos Jauregui-Correa
namely: Finite Differences and Finite Elements. The results are illustrated with practical
examples and the basis of more sophisticated tools that allows the design of complex
vibration isolators are included.
Keywords: viscoelastic materials, vibration absorbers, finite element analysis
INTRODUCTION
Vibration absorbers are devices attached to a mechanical system, or structure, with

the purpose of reducing the transmission of vibrations and sound. Vibration absorbers
have been used as passive vibration control systems for decades and they consist of a
metal support and an elastomer. Their performance depend on the elastomer properties
and geometry, nevertheless, designing a vibration absorber could be a challenging task.
The main difficulty comes from the rheological properties of the elastomer, since they
respond elastically and they also flow as a reaction of the applied forces. This behavior is
unique and there are several models for representing the relationship between the applied
forces and the deformation, but these models differ from the actual performance and their
applicability is limited to certain particular cases. Finite element techniques are good for
determining the elastic behavior but they are unable to have a proper damping model,
therefore, if the damping ratio is inaccurate, it is difficult to obtain a realistic design.
There is a vast amount of publications regarding the design of vibration absorbers.
Ketema [1] presented a study of the application of viscoelastic absorbers for suppressing
vibrations. He developed the model for a single degree of freedom and he included the
dependency of the material properties on temperature. He determined the dynamic
coefficients for the absorber from the transmissibility function. Zhang and Richards [2]
identified the dynamic parameters for an elastomeric isolator using a Maxwell Voigt
material model. They only presented numerical examples and they did not define the
rheological properties of a typical elastomeric material. Hasheminejad and Kazemirad [3]
developed a method for determining the wave propagation in a cylinder with a
viscoelastic core. For the viscoelastic model, they applied the relaxation modulus that are
frequency dependent. Espíndola et al. [4] optimize a viscoelastic absorber using a
Frobenius norm for the objective function. They used fractional derivatives and the
relaxation model and they applied their method to different shapes. Martínez and
Elejabarrieta [5] presented a work on the characterization of the damping effect on
sandwich structures. Chazot et al. [6] included a definition of the elastic modulus as real
and complex functions and they defined the damping as the difference between them.
Saksa et al. [7] presented the analysis of a viscoelastic panel moving in one direction.
Their model is similar to a traveling wave on a viscoelastic belt presented by Lozano et
al. [8]. Ribeiro et al. [9] presented a model for optimizing vibration absorbers for rotor
The Design of Vibration Absorbers using the Rheological Properties … 349
dynamic machines. They proposed a design based on a viscoelastic support. They

represent the viscoelastic model as fractional order derivatives with four material
properties. The material was represented by a complex function that depends on the
excitation frequency. Karttunen and Hertzen [10] developed a model for studying the
steady state behavior of a viscoelastic cylinder confined within two rigid walls. The
constitutive equation was based on a Maxwell material, this consideration contrasts with
other publications that, in general, utilize the Kevin-Voigt model. Faridani and Capsoni
[11] presented a similar formulation considering viscous and elastic moduli. They use the
complete tensor for viscosity instead of just the shear stresses. They applied their model
to a sandwich and they used the FE modelling for the solution. Xu et al. [12] applied a
sandwich plate FEM model and they optimized the design with a multi-objective
function. They only presented numerical results.
There are other applications reported, some of which are referring to special
applications, whereas others model different materials such as Kelvin-Voigt or Maxwell
types. Wang et al. [13] developed a simulation of a viscoelastic vehicle mount using
Finite Elements. In their simulation, they defined material properties using a commercial
program. In this work, a multimodal analysis is conducted, and they presented a method
for simulating vibration absorbers that isolate several frequencies. The basic concept is to
find the absorber modes according to the excitation frequencies that need to be isolated.
Muscolino and Palmeri [14] and Chen [15] developed a method for modelling
Timoshenko beams and determined the damping coefficient by using the Kelvin-Voigt
material concept. The energy dissipated by the viscous term is similar to other
formulations and to the one presented in this chapter. Regarding the application to soil
mechanics, Boaga et al. [16] considered that soils can be modeled as a viscoelastic
material. Their model is based on a Kelvin-Voigt approach, and the viscous coefficient is
determined from the damping ratio. Braga et al. [17] described the effect of friction
dampers as vibration absorbers. They generate different models and they included a
viscoelastic model using a Kelvin-Voigt material. They validate their model with
experimental results. Gong et al. [18] and Bi and Hao [19], from the Curtin University in
Australia, study the application of viscoelastic absorbers for pipe lines. Keramat and
Shirazi [20] developed a method for calculating the dynamic behavior of solids. In their
method, they defined the constitutive equation for a solid using Maxwell and Kelvin-
Voigt concepts. The stress function is defined as a convolution between the material
properties and a differential strain tensor. Their analysis is applied to creep. Cepo, Manin
and Boltezar [21] developed a method for analyzing the damping effect in belt-drive
systems. They estimated the dynamic coefficients using the same free damping test as we
did. They modeled the belt as a Kelvin-Voigt.
Designing a passive vibration absorber, as any design, consists of selecting an
appropriate material and the definition of the geometry. Both aspects are interrelated
since they determine the main dynamic parameters, namely the stiffness and the damping
coefficients. In every case, these parameters must be selected based on the vibration
frequency that will be isolated. It could be one or more frequencies, or vibration modes,
thus, the complexity of the geometry will determine the number of principal modes that
the new design will have. Since the absorber design requires 3D models, there are no
close form solutions and only complex numerical techniques can be applied. From them,
Finite Elements is the most utilized one. Nevertheless, due to the material characteristics,
this technique provides solutions that contrast from experimental results.
In this chapter, the basis for formulating mathematical models for designing vibration
absorbers are presented, the models are derived approximating the rheological properties
as a viscoelastic material that behaves as a Kelvin-Voigt type material. Although it is an
approximation, it simplifies the numerical complexity and the results are close enough for
many engineering applications. Based on this concept, different solution techniques are
described; finite differences and finite elements. To link the design with specific
applications, the concept of transmissibility is included and in order to show the
advantages of these techniques, some examples are presented.
VISCOELASTIC MATERIALS
The materials used in vibration absorbers, besides metals, are different types of
natural rubbers, neoprene, silicon rubbers and other artificial polymers that have very low
elastic coefficients. Most of them are engineered to behave at very specific conditions
and their formulation is design to absorb a specific range of vibration frequencies.
From an engineering point of view, the relationship between the external forces and
the material behavior is defined as the constitutive equation; this equation is not simple
because the stress-strain relationship depends on the strain and the strain rate [22], [23],
[24]. In order to reduce the complexity of this relationship, materials are considered as
continuous, meaning that every point within the body formed by the material has the
same properties, or in other words, every point within the boundaries of the material has
similar properties. In this way, there is a body constituted by certain material that
occupies a certain region. Thus, each particle occupies a single point and its position is
denoted with a position vector 𝑥, ̃ where 𝑥̃ is referred to a reference coordinate system.
When the body is submitted to a set of forces, every particle will move and this
movement is represented as a displacement or a change in position. The new region is
defined as the deformation of the body and it has a new configuration. This deformation
occurs in space and time, thus the body will have a deformation or strain and a velocity of
deformation or strain rate. Every point moves along a trajectory that is assumed as a
smooth function. This function has a sufficient number of derivatives in time and space.
Using Timoshenko’s notation, the displacement of a point along an orthogonal
direction is expressed as:
𝜕𝑢
𝑃𝑥 = 𝑢 + 𝜕𝑥 𝑑𝑥 (1)
The strain is defined as the change of length with respect to the original length.
Therefore, the strain along an orthogonal direction can be expressed as:
1 𝜕𝑢
𝜀𝑥 = 𝑑𝑥 (𝑃𝑥 − 𝑢) = 𝜕𝑥 (2)
Similarly, the strains in the other two directions are expressed as:
1 𝜕𝑣
𝜀𝑦 = (𝑃 − 𝑣) = (3)
𝑑𝑦 𝑦 𝜕𝑦
1 𝜕𝑤
𝜀𝑧 = 𝑑𝑧 (𝑃𝑧 − 𝑤) = 𝜕𝑧
(4)
where 𝑢, 𝑣, and 𝑤 are the displacements along the Cartesian coordinates. The body also
deforms angularly and this deformation is expressed as:
𝜕𝑢 𝜕𝑣
𝛾𝑥𝑦 = 𝜕𝑦 + 𝜕𝑥 (5)
𝜕𝑢 𝜕𝑤
𝛾𝑥𝑧 = + (6)
𝜕𝑧 𝜕𝑥
𝜕𝑤 𝜕𝑣
𝛾𝑦𝑧 = + (7)
𝜕𝑦 𝜕𝑧
Using the tensor definition:
𝜀𝑥 𝛾𝑥𝑦 𝛾𝑥𝑧
𝛾
𝜺 = [ 𝑦𝑥 𝜀𝑦 𝛾𝑦𝑧 ] (8)
𝛾𝑧𝑥 𝛾𝑧𝑦 𝜀𝑧
which is a symmetric tensor.

The strain rate is defined as the rate of deformation. Considering the smoothness of
the deformation function, the strain rate can be defined as:
𝑑𝜺
𝜺̇ = 𝑑𝑡 (9)
For example, the strain rate in the x direction is:

1 𝜕𝑢 𝜕𝑢̇
𝜀̇𝑥 = 𝑑𝑡 (𝜕𝑥 ) = 𝜕𝑥 (10)
Therefore, the strain rate tensor becomes:
𝜀𝑥̇ 𝛾𝑥𝑦̇ 𝛾𝑥𝑧̇

𝜺̇ = [𝛾𝑦𝑥̇ 𝜀𝑦̇ 𝛾𝑦𝑧̇ ] (11)
𝛾𝑧𝑥̇ 𝛾𝑧𝑦̇ 𝜀𝑧̇
It is a linear transformation of the velocity of each point of the body.

From the continuum mechanics theory, there are three fundamental equations:
conservation of mass, conservation of energy and the balance of momentum.
Nevertheless, these equations are incomplete and it is necessary to define another set of
equations that describe the relationship between the external works with the material
properties. This set of equations is known as the constitutive equation.
Since elastomeric materials behave differently than other solids, their constitutive
equations could be very complicated, because they depend on the strain and the strain
rate. To overcome this difficulty, there are several approaches, but the most accepted
formulation is to treat them as viscoelastic materials.
Viscoelastic materials are those that can be represented as a combination of elastic
materials and viscous fluids. The elastic behavior obeys Hook’s law, whereas Newton´s
viscous law defines the viscous behavior. As a simplistic representation, the elastic term
is modeled as a linear spring and the viscous component as a damper. The combination of
springs and dampers represent different materials. Elastomers and rubbers for vibration
absorbers have a better representation with a Kelvin-Voigt model [24]. (Figure 1). The
advantage of using the Kelvin-Voigt model is that it separates the stress-strain function
into an elastic term and a viscous term, and these terms depend on the displacement and
the velocity independently. In this way, if the rate of change is slow the material behaves
almost elastically, whereas if there is no deformation, the damper has no effect on the
strain. Other combinations (as the Maxwell model) cannot separate the viscous and
elastic effects.
The constitutive is determined from the simplified model, the stress-strain function is
found from the spring function, using Hook’s law:
𝐹 = 𝐾𝑥 (12)
From a unidimensional stress function, it can be expressed as:
𝜎 = 𝐸𝜀 (13)
F(t)
C K
Figure 1. Representation of a Kelvin-Voigt material.
The stress-strain rate function can be determined considering that the material
behaves as a Newtonian fluid, meaning that the shear stress is proportional to the velocity
gradient. It is also considered that the material is incompressible, thus the volumetric
deformation is only related to the elastic deformation, therefore:
𝜏 = 𝜇∇𝑢̇ (14)
Adding the spring and damping effects, a general tensor can be expressed as:
𝝈 = 𝑬𝜺 + 𝝁𝜸̇ (15)
where 𝑬 is the linear elastic tensor and 𝝁 is the viscous tensor. For a unidimensional
system, the constitutive equation becomes:
𝑑𝑢 𝑑𝑢̇
𝜎 = 𝐸 𝑑𝑥 + 𝜇 𝑑𝑥 (16)
Assuming that the displacement and velocity are linear:
𝑢 = 𝑢0 + 𝑢1 𝑥 (17)
and
𝑢̇ = 𝑢̇ 0 + 𝑢̇ 1 𝑥 (18)
With this assumption, the equation of motion can be determined using Lagrange’s
equation:
𝑑 𝜕𝐿 𝜕𝐿
( ) − 𝜕𝑢
𝑑𝑡 𝜕𝑢̇
=0 (19)
where 𝐿 = 𝑇 − 𝑉
𝑇 is the kinetic energy:
1
𝑇 = 2 (∫Ω 𝜌 𝑑Ω) 𝑢̇ 2 (20)
And 𝑉is the potential energy due to the deformation:
1
𝑉 = 2 ∫Ω 𝜎𝜀 𝑑Ω (21)
Substituting the potential energy,
𝜕𝐿 1 𝑑𝜀
= (∫Ω 𝐸𝜀 𝑑Ω + ∫Ω 𝜇𝜀̇ 𝑑Ω) (22)
𝜕𝑢 2 𝑑𝑢
And the kinetic energy:
𝑑 𝜕𝐿
( ) = (∫Ω 𝜌 𝑑Ω) 𝑢̈ (23)
𝑑𝑡 𝜕𝑢̇
Figure 2. Unidimensional model.

If the material is homogeneous and the element has a constant area along the
direction of interest (Figure 2), the equation of motion becomes:
𝜇𝐴 𝐸𝐴
𝑚𝑢̈ + 2𝐿 𝑢̇ + 𝐿
𝑢 = 𝑓(𝑡) (24)
This simple equation can be used for determining the material properties by applying
the standard ASTM D-5992-96. The viscoelastic properties are determined by applying a
free vibration test to a straight specimen of length 𝐿 and cross sectional area 𝐴. It is
necessary to attach a mass to the free end and a fixed one to the other extreme. Figure 3
shows the displacement as a function of time. The measured data can be represented as a
second order differential equation:
𝐾
𝑢̈ + 2𝜁(√𝐾𝑚)𝑢̇ + √𝑚 𝑢 = 0 (25)
From the analysis of the data,
4𝜋2
𝐾=𝑚 (26)
𝑇2
and
1
𝜁= 2
(27)
√1+4𝜋2
𝛿
where
𝑎𝑛
𝛿 = ln (𝑎 ) (28)
𝑛+1
Making a comparison between the dynamic model and the analysis of the
measurements, the viscoelastic parameters can be calculated as:
𝐿
𝜇 = 4𝜁 𝐴 √𝐾𝑚 (29)
And
𝐾𝐿
𝐸= (30)
𝐴
It is also possible to find the natural frequency by applying the Fast Fourier
Transform (FFT) to the measurements.
Typical values for rubber materials are:
Table 1. Test Results
Property Value
𝐸 5.0  0.032 MPa
𝜇 65.0  0.13 KPa-s
x
1
0.8
0.6 a1
Normalized displacement
0.4 a2
0.2
-0.2
-0.4
-0.6
-0.8
T Time [s]
Figure 3. Displacement measurements. Free vibration test.
Table 2. Elastomer properties
Property Value
Elastic modulus 5.1 MPa
Poisson’s ratio 0.49
Tensile Strength 15.2 MPa
Elongation at Break 280%
Hardness, "A" 68
Density 795 kg/m3
In general, elastomers have a wide range of elastic modulus and viscous coefficients.
A recommended material for isolating vibrations is the ethylene acrylic elastomer. It is a
terpolymer of ethylene, methylacrylate, and a cure site monomer cured using an amine-
based vulcanization system. One of the main advantages of this material is its high-
temperature durability and oil resistance (30°C to 160°C). Its typical properties are:
These values are taken as a reference for the examples included in this chapter.
Before describing the most common technique for solving complex designs, the concept
of transmissibility is described. The design of an absorber starts from the definition of the
frequencies that need to be isolated. There could be a single frequency or multiple
frequencies, in which case a complex 3D modeled is required, and the designer must
consider that each of the absorber modes will isolate a particular frequency.
TRANSMISSIBILITY
A vibration absorber is a mechanical element that is mounted between a vibration

source and a device that will be isolated. The vibration source has a characteristic
frequency 𝜔, or excitation frequency, and the device and the vibration absorber will have
a natural frequency 𝜔𝑛 . In this case, the device mass will be supported by the absorbers
which will be characterized by a spring and a damper (Figure 4).
Device
Absorbers
Basament
Figure 4. Schematic representation of a device mounted on vibration absorbers.
Therefore, the absorber’s design should shift the natural frequency of the new system
far away from the excitation frequency. The elastomer’s rheological properties will
determine the stiffness 𝑘 and damping 𝑐 values, such that:
𝑐
𝜁 = 2√𝑘𝑚 𝑎𝑛𝑑 𝜔𝑛 = √𝑘⁄𝑚 (31)
where 𝑚 is the mass of the isolated device.

Figure 5 represents the device mounted on the absorber and the excitation force. The
excitation force has an excitation frequency that must be higher than the natural
frequency.
The design of a vibration absorber is based on the transmissibility function that is

defined as:
𝜔 2
1+4𝜁 2 ( )
𝜔𝑛
𝑇𝑓 = √ 2 (32)
𝜔 2 𝜔 2
(1−( ) ) +4𝜁 2 ( )
𝜔𝑛 𝜔𝑛
Figure 6 shows the transmissibility as a function of 𝜔⁄𝜔𝑛 . It is impossible to have

zero transmissibility, but an acceptable value is below one. From Figure 6 it is clear to
find that 𝜔⁄𝜔𝑛 > √2.
F= cos(ωt)
Device
Basament
Figure 5. Schematic representation of the dynamic model.
Trasmissibility
3
Damping coefficient
2.5 0.2
0.4
0.6
2
0.8
Amplitude
1
1.5
1.2
0.5
0
1 1.5 2 2.5 3 3.5 4 4.5 5
/
Figure 6. Transmissibility as a function of the excitation frequency.

If the excitation has more than one frequency, a multi-modal absorber is needed. The
design of this type of absorber requires a 3D model design. Such a design must be done
with a Finite Element model or another numerical algorithm.
Once the design conditions are determined and the material properties selected, the
geometry of the absorber must be defined. For that reason, the solution of the strain
distribution can be obtained only for simple geometries and the stiffness and damping
values must be found numerically. The reason is that there is no exact solution for
differential equation 19. The simplest technique for numerically solving the differential
equation is the Finite Difference approach.
FINITE DIFFERENCE APPROACH
Finite difference is a very well-known technique for solving differential equations.

The basis for this technique is to approximate a differential with a finite difference,
avoiding applying the limits, thus:
𝑑𝑦 Δ𝑦
≈ (33)
𝑑𝑥 Δ𝑥
Having this concept in mind, there are a wide variety of possible approximations to a
differential equation. To illustrate the application of this technique to a viscoelastic
problem, the dynamic response of a travelling belt is described next.
Beam Model
To solve this problem, it is assumed that the travelling belt can be represented as an
elastic beam with axial tension (Figure 7). The beam is assumed to have a constant cross
section and the dominant displacement is in the y direction. The general equation of
motion is found by applying the beam theory, assuming an Euler-Bernoulli beam type. If
the displacement in the y direction is 𝑣, then [8]:
𝜕2 𝑣 𝜕2 𝑣 𝜕2 𝑣 𝐸𝐼 𝜕4 𝑣 𝜇𝐼 𝜕5 𝑣 𝜇𝐼 𝜕 𝜕4 𝑣 𝑃
𝜕𝑡 2
+ 2𝑢 𝜕𝑥𝜕𝑡 + (𝑢2 − 𝑐 2 ) 𝜕𝑥 2 + 𝐿𝑚 𝜕𝑥 4 + 𝑢 𝐿𝑚 𝜕𝑥 5 + 𝐿𝑚 𝜕𝑡 (𝜕𝑥 4 ) = 𝑚 cos(Ω𝑡) (34)
In this case 𝐿 is the belt´s span, 𝑚 is the mass per length, 𝑢 is the travelling speed, 𝑃
is the tension on the belt, 𝑐 = √𝑃⁄𝑚 and 𝐼 is the moment of inertia of the cross section.
Figure 7. Travelling belt.
This equation can only be solved using finite difference. Due to the complexity of the
finite difference equation, the solution is found using the Crank-Nicholson scheme. Thus,
the partial derivatives are approximated sequentially for the time derivative:
𝜕𝑓(𝑥,𝑡) 𝑓𝑥𝑡+1 −𝑓𝑥𝑡

𝜕𝑡
= Δ𝑡
(35)
And, for the space derivative:
𝑡+1 −𝑓 𝑡+1 𝑡 𝑡
𝜕𝑓(𝑥,𝑡) (𝑓𝑥+1 𝑥−1 )+(𝑓𝑥+1 −𝑓𝑥−1 )
𝜕𝑥
= 2Δ𝑥
(36)
𝑡+1 −2𝑓 𝑡+1 +𝑓𝑡+1 𝑡 𝑡 𝑡

𝜕2 𝑓(𝑥,𝑡) (𝑓𝑥+1 𝑥 𝑥−1 )+(𝑓𝑥+1 −2𝑓𝑥 +𝑓𝑥−1 )
𝜕𝑥 2
= 2(Δ𝑥)2
(37)
With this method, a set of linear equations in terms of individual points along the
belt, are obtained and can be applied for determining the dynamic response at different
travelling speeds. Figure 8 shows the deformation of the belt at different time intervals
when 𝑢 < 𝑐. From these results, it is clear that the viscoelastic properties increase the
damping effect, and the travelling speeds shift the first vibration mode. Figure 9 shows
the center point’s displacement as a function of time, whereas Figures 10 and 11 show
equivalent results, but for 𝑢 > 𝑐. In this case, the belt tends to a stable position and the
oscillations are reduced.
Modelling the dynamic equation is relatively simple, but the numerical solution
depends on the ration Δ𝑡/Δ𝑥:
-8
x 10 Deformation as function of travelling speed
12
11
10
9
Magnitude
5
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
x/L
Figure 8. Vertical displacement v as a function of x. u < c.
-8
x 10
10.15
10.1
10.05
Displacement center point
10
9.95
9.9
9.85
9.8
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
time
Figure 9. Vertical displacement of the center point (x = L/2) as a function of time, u < c.
-8
x 10 Deformation as function of travelling speed
12
11
10
9
Magnitude
5
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
x/L
Figure 10. Vertical displacement v as a function of x. u > c.
-7
x 10
1.3
1.25
1.2
Displacement center point
1.15
1.1
1.05
0.95
0.9
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
time
Figure 11. Vertical displacement of the center point (x = L/2) as a function of time, u > c.
Cylindrical Coordinates
There are many applications where the vibration absorber has a cylindrical shape. For
these applications, the dynamic equations can be obtained using cylindrical coordinates.
The stress function in cylindrical coordinates are:
𝜕𝜎𝑟 𝜎𝑟 −𝜎𝜃 1 𝜕𝜏𝑟𝜃

𝜕𝑟
+ 𝑟
+𝑟 𝜕𝜃
+𝑅 = 0 (38)
𝜕𝜏𝑟𝜃 2𝜏𝑟𝜃 1 𝜕𝜎𝑟

𝜕𝜃
+ 𝑟
+𝑟 𝜕𝜃
+𝑆 =0 (39)
The external forces 𝑅 and 𝑆 are defined as forces per volume; in this case, they are
only calculated as inertial forces:
𝜕2 𝑢
𝑅 = 𝜌 𝜕𝑡 2 (40)
and
𝜕2 𝑣
𝑆=𝜌 (41)
𝜕𝑡 2
The strain function is calculated as:
𝜕𝑢
𝜀𝑟 = 𝜕𝑟
(42)
1 𝜕𝑣 𝑢
𝜀𝜃 = + (43)
𝑟 𝜕𝜃 𝑟
1 𝜕𝑢 𝜕𝑣 𝑣
𝛾𝑟𝜃 = + − (44)
𝑟 𝜕𝜃 𝜕𝑟 𝑟
Combining the strain and stress function with the constitutive equation, a set of
differential equations is found. The constitutive equation in cylindrical coordinates is
determined from the theory of elasticity and, considering that the viscous part behaves as
a Newtonian fluid, the viscoelastic term affects only the shear stress, therefore, the
constitutive relation is expressed as:
1
𝜀𝑟 = 𝐸 (𝜎𝑟 − 𝜈𝜎𝜃 ) (45)
1
𝜀𝜃 = 𝐸 (𝜎𝜃 − 𝜈𝜎𝑟 ) (46)
𝜕𝛾𝑟𝜃
𝜏𝑟𝜃 = 𝐺𝛾𝑟𝜃 + 𝜇 𝜕𝑡
(47)
Since the differential equation cannot be solved analytically, the solution is found
using finite difference. In the following section, an example of a crankshaft damper is
presented.
Example
Crankshaft vibration dampers are designed to reduce torsional vibration in internal

combustion engines. The crankshaft undergoes angular oscillations due to the
thermodynamic cycle (ignitions inside engine cylinders), changes on the external load,
and oscillations due to the piston-rod mechanism, among others.
Inner ring
Elastomer ring
Inertia ring
Figure 12. Construction of a torsional damper.
In elastomer dampers (Figure 12), the outer inertia ring is fixed relative to the
mounting hub through a ring (or rings) of elastomeric material (usually rubber). The
inertia ring and elastomer system is designed to isolate the fundamental torsional natural
frequency of the crankshaft (as described before). The properties of the elastomer are
selected considering the transmissibility function in order to convert the mechanical
energy into heat. The capacity of a torsional absorber depends on the amount of energy
that it dissipates, and its life will depend on the amount of overheating that it can support,
especially in engines with high load variations [25].
The design process consists of selecting the thickness of the elastomer band and its
radial location. The external ring is considered as a concentrated mass and the natural
frequency and damping value are calculated numerically. For this example, the material
properties are the same as in previous examples. The elastomer ring has an inner radius of
140 mm, a thickness of 12.5 mm and an axial depth of 41.5 mm. The external ring is
made of steel and its outside diameter is 210 mm.
The results are included in Figures 13 and 14. Figure 13 shows the tangential
displacement (𝑣) as a function of the radius and time. Figure 14 shows the torsional
displacement of the steel ring. From this figure, the natural frequency and damping
coefficient of the torsional absorber are found, and they are:
Table 3. Dynamic parameters for the torsional damper
Property Value
𝐾 1821 N/m
𝐶 12 Ns/m
0.8
0.6
Rotation
0.4
0.2
0
10
8 100
80
6
60
4 40
2 20
0
Radial Position [%] Time [s]
Figure 13. Torsional deformation as a function of time and radial position.
0.7
0.6
0.5
0.4
Torsional
Displacement
0.3
0.2
0.1
0
0 20 40 60 80 100 120
Time [s]
Figure 14. Torsional displacement of the inertial ring.
Designs that are more sophisticated can be obtained using Finite Elements.
FINITE ELEMENTS APPROACH
In previous sections, the design of vibration absorbers was based on solving the
differential equations numerically. The differential equations were derived from the
theory of continuous mechanics. In the Finite Element Approach (FEA), the continuum is
divided into a finite number of regions connected though points or nodes. The regions, or
elements, have simple geometry that can easily be integrated (Figure 15). Any point
within the region is assumed to move following a polynomial function. The number of
degrees of freedom determines the polynomial degree. The limitation of applying FEA
for the solution of vibration absorbers is that it produces a set of ill-conditioned matricies,
which can introduce unacceptable numerical errors. To avoid this, a special FEA
formulation is required and different techniques should be incorporated in the analysis, as
well as adequate matrix construction.
The field displacement is approximated by a linear polynomial:
𝑢(𝑥) = ∑𝑛𝑖=1 𝑎𝑖 𝒙𝑖−1 (48)
where 𝑛 is the number of nodes of each individual element. 𝑎𝑖 are the coefficients of the
polynomial function and 𝒙 is a vector formed by the space variables. The coefficients 𝑎𝑖
are determined evaluating the displacement function at the node coordinates, and it can
be represented as a matrix function:
𝑢({𝑥}) = [𝑁]{𝑢} (49)
where, {𝑥} represents the space coordinates vector, [𝑁] is called the form matrix and {𝑢}
represents the nodal displacements. The norm of {𝑢} is equal to the number of degrees of
freedom of each element.
v Volumetric
v
Linear Higher order
Nodes
Planar
Figure 15. Examples of finite elements.

The strain and the strain rate functions can be expressed in terms of the polynomial
coefficients as:
𝜕
{𝜀} = 〈[𝑁]{𝑢}〉 = [𝐵]{𝑢} (50)
𝜕{𝑥}
𝜕
{𝜀̇} = 〈[𝑁]{𝑢̇ }〉 = [𝐴]{𝑢̇ } (51)
𝜕{𝑥}
In this way, the deformation energy is expressed as:
1 1 1
𝑉 = 2 ∫Ω 𝜎𝜀 𝑑Ω = 2 ∫Ω 〈{𝑢}𝑇 [𝐵]𝑇 [𝐸][𝐵]{𝑢}〉 𝑑Ω + 2 ∫Ω 〈{𝑢}𝑇 [𝐵]𝑇 [𝜇][𝐴]{𝑢̇ }〉 𝑑Ω
(52)
1 1
𝑇 = 2 (∫Ω 𝜌 𝑑Ω) 𝑢̇ 2 = 2 ∫Ω 𝜌〈{𝑢̇ }𝑇 [𝑁]𝑇 [𝑁]{𝑢̇ }〉 𝑑Ω (53)
where Ω is the volume of each individual element.

Expressing Lagrange’s equation in terms of the element matrices:
𝜕 1
𝜕{𝑢}
𝑉 = ∫Ω 〈[𝐵]𝑇 [𝐸][𝐵]{𝑢}〉 𝑑Ω + 2 ∫Ω 〈[𝐵]𝑇 [𝜇][𝐴]{𝑢̇ }〉 𝑑Ω (54)
𝑑 𝜕𝐿
( )
𝑑𝑡 𝜕𝑢̇
= ∫Ω 𝜌〈[𝑁]𝑇 [𝑁]{𝑢̈ }〉 𝑑Ω (55)
Therefore, the equation of motion of a single element is:
[𝑀]{𝑢̈ } + [𝐶]{𝑢̇ } + [𝐾]{𝑢} = {𝑓} (56)
where:
[𝑀] = ∫Ω 𝜌〈[𝑁]𝑇 [𝑁]〉 𝑑Ω (57)
1
[𝐶] = ∫Ω 〈[𝐵]𝑇 [𝜇][𝐴]〉 𝑑Ω (58)
2
[𝐾] = ∫Ω 〈[𝐵]𝑇 [𝐸][𝐵]{𝑢}〉 𝑑Ω (59)
In this way, the dynamic model depends on the material properties, 𝜌, 𝐸, 𝜇.

Previous matrices are computed for each element, and then are integrated into a
whole matrix. In this way, a complex system of partial differential equations is converted
into a simple linear algebra problem.
Plane Elements
Many problems can be solved using planar elements. Planar elements can be applied
to plane stress, plane strain or axis-symmetric problems where only two space
coordinates are needed. Since the degree of the displacement polynomial depends on the
node numbers, a linear solution is obtained with a triangular element (Figure 16). If the
field displacement is defined as:
𝑢𝑥
{𝑢} = {𝑢 } (60)
𝑦
uy
k ux
Figure 16. Linear planar element.
Then, the interpolating polynomial will have form of:
𝑢𝑥 (𝑥, 𝑦) = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑦 (61)
𝑢𝑦 (𝑥, 𝑦) = 𝑏0 + 𝑏1 𝑥 + 𝑏2 𝑦 (62)
Evaluating these polynomials at the nodes (using node coordinates), the coefficients
are:
𝑖
𝑎0 𝑁1 𝑁2 𝑁3 𝑢𝑥
1
{𝑎1 } = ∆ [𝑁4 𝑁5 𝑁6 ] {𝑢𝑥𝑗 } (63)
𝑎2 𝑁7 𝑁8 𝑁9 𝑢𝑘
𝑥
and
𝑖
𝑏0 𝑁1 𝑁2 𝑁3 𝑢𝑦
1
{𝑏1 } = ∆ [𝑁4 𝑁5 𝑁6 ] {𝑢𝑦𝑗 } (64)
𝑏2 𝑁7 𝑁8 𝑁9 𝑢𝑘
𝑦
where
∆= (𝑥𝑗 𝑦𝑘 − 𝑥𝑘 𝑦𝑗 ) − (𝑥𝑖 𝑦𝑘 − 𝑥𝑘 𝑦𝑖 ) + (𝑥𝑖 𝑦𝑗 − 𝑥𝑗 𝑦𝑖 ) (65)
and
𝑁1 = 𝑥𝑗 𝑦𝑘 − 𝑥𝑘 𝑦𝑗
𝑁2 = 𝑥𝑘 𝑦𝑖 − 𝑥𝑖 𝑦𝑘
𝑁3 = 𝑥𝑖 𝑦𝑗 − 𝑥𝑗 𝑦𝑖
𝑁4 = 𝑦𝑗 − 𝑦𝑘
𝑁5 = 𝑦𝑘 − 𝑦𝑖 (66)
𝑁6 = 𝑦𝑖 − 𝑦𝑗
𝑁7 = 𝑥𝑘 − 𝑥𝑗
𝑁8 = 𝑥𝑖 − 𝑥𝑘
𝑁9 = 𝑥𝑗 − 𝑥𝑖
Summarizing, the displacement field, in terms of the nodal displacements is:
𝑢𝑥𝑖
𝑁1 0 𝑁2 0 𝑁3 0
𝑢𝑦𝑖
0 𝑁1 0 𝑁2 0 𝑁3
𝑗
1 𝑁4 0 𝑁5 0 𝑁6 0 𝑢𝑥
{𝑢} = {1 𝑥 𝑦} (67)
∆ 0 𝑁4 0 𝑁5 0 𝑁6 𝑢𝑦
𝑗
𝑁7 0 𝑁8 0 𝑁9 0
𝑢𝑥𝑘
[ 0 𝑁7 0 𝑁8 0 𝑁9 ] 𝑘
{𝑢𝑦 }
{𝑢} = [𝑁]{𝑢𝑖 } (68)
The strain function can be express as a vector:
𝜀𝑥
𝜕
{𝜀} = { 𝜀𝑦 } = 〈[𝑁]{𝑢}〉 = [𝐵]{𝑢𝑖 } (69)
𝜕{𝑥}
𝛾𝑥𝑦
where
𝑁4 0 𝑁5 0 𝑁6 0
[𝐵] = [ 0 𝑁7 0 𝑁8 0 𝑁9 ] (70)
𝑁7 𝑁4 𝑁8 𝑁5 𝑁9 𝑁6
And the strain rate:
𝜕
{𝜀̇} = 〈[𝑁]{𝑢̇ }〉 = [𝐴]{𝑢̇ } (71)
𝜕{𝑥}
Similarly, [𝐵] and [𝐴] can be obtained. After developing the element matrices, the
integration and matrix reduction depends on each particular problem, and in many
commercial programs, this process is done automatically.
The solution of the equation of motion (Eq.56) is obtained by applying a finite
difference scheme, but in the design of vibration absorbers, the mass and stiffness
matrices are ill conditioned because the steel properties are several orders of magnitude
larger than the elastomer properties. To find a solution for a real case, it is necessary to
modify the matrices. For this purposes, the Guyan matrix reduction is applied. If the
nodes corresponding to the steel elements are grouped in {u2}, the system of equations
can be reorganized as:
𝑀11 𝑀12 𝑢̈ 1 𝐶 𝐶12 𝑢̇ 1 𝐾 𝐾12 𝑢1 𝑓

[ ] { } + [ 11 ] { } + [ 11 ]{ } = { } (72)
𝑀21 𝑀22 𝑢̈ 1 𝐶21 𝐶22 𝑢̇ 2 𝐾21 𝐾22 𝑢2 0
Assuming that the sub-matrices 𝑀21 = 𝑀12 = 0 and that the motion is harmonic, the
partition method is defined as:
{𝑢̇ 2 } = −[𝐶22 ]−1 [𝐶21 ]{𝑢̇ 1 } (73)
and
{𝑢2 } = −[𝐾22 ]−1 [𝐾21 ]{𝑢1 } (74)
Taking into account these considerations, the equation of motion becomes:
[𝑀11 ]{𝑢̈ 1 } + 〈[𝐶11 ] − [𝐶12 ][𝐶22 ]−1 [𝐶21 ]〉{𝑢̇ 1 } + 〈[𝐾11 ] − [𝐾12 ][𝐾22 ]−1 [𝐾21 ]〉{𝑢1 } = {𝑓}
(75)
This approximation eliminates the ill-condition and the numerical solution will be
free of calculation errors.
As an example, a Finite Element Model of a center bearing support for an automobile
drive shaft is included (Figure 17). The boundary conditions were set in a way that the
rubber was confined within a metal support. Since the model is linear, the overall
damping was determined by applying a unit value force. In this model, the material
properties were modified and the average-damping coefficient was calculated as a
function of the logarithmic decrement. Table 4 shows the results of the absolute values. It
is clear that the geometry influences these results and the damping coefficient in the
vertical direction is different from the horizontal value. From the transmissibility curve,
the desired damping factor can be selected, and afterwards, the corresponding logarithmic
decrement. As discussed in previous sections, the viscoelastic properties can be
determined and the rubber properties can be specified.
Table 4. Damping values
Logarithmic decrement C11 C22 C12 C21

5 0.0578 0.0922 0.0252 0.0252
10 0.1127 0.1802 0.0492 0.0492
15 0.1651 0.2641 0.0722 0.0722
20 0.2153 0.3444 0.0941 0.0941
25 0.2635 0.4215 0.1152 0.1152
30 0.3098 0.4954 0.1354 0.1354
Figure 17. Rubber support for an automobile drive shaft.
Even though Finite Elements are fully developed and they are the most accepted
analysis technique, they have some limitations derived from the complexity and size of
the element mesh that needs to be created in some applications. To overcome this
difficulty, a new method that reduces the size and complexity of the mesh and provides
similar results with less computer effort is been developed.
Isogeometric Analysis
The evolution of the analysis methods is been driven by the speed and capacity of
modern computers. This characteristic has allowed the design of sophisticated parts and
components. Most of these parts are modeled in CAD systems using spline functions.
The reason behind using spline functions, or B-splines, is that they define complex
geometries with a minimum amount of data. Meanwhile, engineering analysis has
evolved based on representing the parts with a large number of discrete elements. Thus,
there is a strong contradiction between representing complex geometries with a minimum
number of points (coordinates of the points that define each B spline) and then creating
an analysis model with a very large amount of discretization points. In order to eliminate
this contradiction a new analysis method has been created, namely Isogeometric Analysis
[26].
Isogeometry is based on creating the interpolation functions with B-splines (like the
polynomial functions in the Finite Element Method). The displacement space has the
same base dimension as the geometry space, then the displacements are approximated
using NURBS (Non-Uniform Rational B-Splines).
Making an analogy to Finite Elements, the B-spline parametric space forms
subdomains called “patches” rather than elements. Within the patches, the discretization
function and the material properties are assumed to be uniform.
In the Finite Elements Method, continuity is determined by connecting adjacent
elements through the same nodes. In this way it is assured that the solution will be
continuous throughout the entire model. In the case of Isogeometry, the continuity is
defined with the aid of another vector that defines continuity conditions between adjacent
B-splines. This vector is defined as the knot vector. The knot vector is a set of
coordinates in the parametric space and, for one dimension NURBS, it is written as:
𝑁 = {𝑛1 , 𝑛2 , . . . , 𝑛𝑛+𝑝+1 } (76)
where, 𝑛 is the number of basis functions and 𝑝 is the polynomial order.

The basis B-spline function is defined as:
𝐶(𝜉) = ∑𝑛𝑖 𝑁𝑖,𝑝 (𝜉)𝐵𝑖 (77)
where
1 𝑖𝑓 𝜉 𝑖 ≤ 𝜉 < 𝜉𝑖+1
𝑓𝑜𝑟 𝑝 = 0: 𝑁𝑖,0 (𝜉) = { (78)
0
𝜉−𝜉𝑖 𝜉𝑖+𝑝+1 −𝜉
𝑓𝑜𝑟 𝑝 ≥ 1 ∶ 𝑁𝑖,𝑝 (𝜉) = 𝜉 𝑁𝑖,𝑝−1 (𝜉) + 𝜉 𝑁𝑖+1,𝑝−1 (𝜉) (79)
𝑖+𝑝 −𝜉𝑖 𝑖+𝑝+1 −𝜉𝑖+1
In this way, the B-spline curves are constructed by taking a linear combination of B-
spline basis functions. The coefficients of the basis functions are defined as control
points, and as explained before, the continuity along the entire curve is determined with
the knot vector.
A single B-spline with three control points and a second order polynomial will have
the form shown in Figure 18.
Figure 18. A simple B-spline.
For this function the knot vector will be:
Ξ = {0,0,0, 1,1,1} (80)
Moreover, the interpolation functions are represented as the combination of the

curves represented in Figure 19.
Figure 19. Basis functions for a simple B-spline.

Figure 20 shows a similar B-spline, but in this case, four control points define its
form and the knot vector is:
Ξ = {0,0,0,0, 0.5,1,1,1} (81)
Figure 20. B-spline with four control points.
In addition, the basis function is the combination of four curves (Figure 21).
Figure 21. Basis function for a four control points B-spline.
For example, in Figure 22 the blue curve is composed of three B-splines. The first B-
spline is defined by control points 1, 2 and 3. Control points 3, 4, 5 and 6, determine the
second curve and control points 6, 7 and 8 form the third curve. In this example, the
resulting curve is smooth except at point 6, where there is only continuity in the position
but not in the slope. This continuity condition is defined with the knot vector.
There are two approaches for solving the dynamic model of a viscoelastic absorber;
the first approach is based on the Boundary Element Method (BEM). Geometrically, it
only needs to define the boundary of the body with a close B-spline and applying the
BEM theory. The difficulty is in the integration process. The second approach is using
the same variational principle as in Finite Elements, but instead of solving for the nodal
displacements, the variational principle is based on the NURBS functions.
4 5
7 8
3 6
Figure 22. Example of a NURB.
For plane problems, the construction on the NURB is determined considering that the
interpolation function is also a two degree function. It is constructed using the notation
shown in Figure 23. Following this approach, the displacement function
is defined as [27]:
𝑢 = ∑𝑛𝑖 ∑𝑚
𝑗 𝑁𝑖,𝑝 (𝜉)𝑁𝑗,𝑝 (𝜂)𝒖𝑖𝑗 (82)
and the velocity:
𝑢̇ = ∑𝑛𝑖 ∑𝑚
𝑗 𝑁𝑖,𝑝 (𝜉)𝑁𝑗,𝑝 (𝜂)𝒖̇ 𝑖𝑗 (83)
where 𝜉 and 𝜂 are the local coordinates and 𝒖𝑖𝑗 is the displacement vector of each control
point and 𝒖̇ 𝑖𝑗 their velocities.
In this case, there are two knots vectors (for each local coordinate):
Ξ = {𝜉1 , 𝜉2 , … , 𝜉𝑛+𝑝+1 } (84)
and
Υ = {𝜂1 , 𝜂2 , … , 𝜂𝑛+𝑝+1 } (85)

x
Control points
Knot span
Figure 23. Configuration of a planar patch and a two dimensional B-spline.
The development of the equations of motion are similar to the equations developed in
previous sections, thus, the strain function is:
𝜕
{𝜀} = 〈{𝑢}〉 (86)
𝜕{𝑥}
𝜕
{𝜀̇} = 〈{𝑢̇ }〉 (87)
𝜕{𝑥}
Substituting the displacement function:
𝜕𝑢
𝜕 −1 𝜕𝜉
〈{𝑢}〉 =𝐽 {𝜕𝑢} (88)
𝜕{𝑥}
𝜕𝜂
and
𝜕𝑢̇
𝜕 𝜕𝜉
𝜕{𝑥}
〈{𝑢̇ }〉 = 𝐽−1 {𝜕𝑢̇ } (89)
𝜕𝜂
where 𝐽 is the Jacobian:

𝜕𝑥 𝜕𝑦
𝜕𝜉 𝜕𝜉
𝐽= [𝜕𝑥 𝜕𝑦
] (90)
𝜕𝜂 𝜕𝜂
Expressing the strain and strain-rate functions in a matrix form:
𝜕𝑢
𝜕𝑥
0
𝜕𝑢
{𝜀} = 0 𝜕𝑦
= [𝐵]𝑢 (91)
𝜕𝑢 𝜕𝑢
[𝜕𝑦 𝜕𝑥 ]
similarly
{𝛾̇ } = [𝐵]𝑢̇ (92)
Substituting this expressions into the Lagrange’s equation:
𝜕𝐿 1 𝑑𝜀
= (∫Ω 𝐸𝜀 𝑑Ω + ∫Ω 𝜇𝜀̇ 𝑑Ω) (93)
𝜕𝑢 2 𝑑𝑢
The system becomes:
𝜕𝐿 1
𝜕𝑢
= (∬[𝐵]𝑇 [𝐸][𝐵]det[𝐽]𝑑𝜉𝑑𝜂)𝑢 + (2 ∬[𝐵]𝑇 [𝜇][𝐵]det[𝐽]𝑑𝜉𝑑𝜂) 𝑢̇ (94)
With these expressions, a similar system of equations is derived. The big difference
with respect to the Finite Element Method is that the expressions must be numerically
integrated. The big advantage is that few control points are needed for defining a
complex geometry.
This note is included at the end of the chapter as a reference, future work is needed in
order to have software that will solve dynamic problems using NURBS.
FINAL REMARKS
The design of vibration absorbers is based on knowing the rheological properties of

elastomers and the actual dynamic behavior of complex geometries. The easiest
formulation is obtained by assuming that the elastomer follows a Kelvin-Voigt material.
Assuming this condition, the dynamic equations of any configuration is linearized and
can be solved with relatively simple models. The complexity of the solution depends on
the geometry of the elastomer and combining metal and elastomer properties within the
same formulation.
With the aid of modern Finite Element programs, it is possible (but not always
straightforward) to design absorbers capable of isolating several frequencies. This is
achieved by finding geometries with different vibration modes that coincide with the
excitation frequency of the system.
In this chapter, the basic concepts for modeling vibration absorbers were presented.
The formulations are easily implemented for simple geometries and they can be
elaborated for complex applications. These concepts are valid only for the assumption of
Kelvin-Voigt materials, nevertheless, the errors in practical applications are reasonably
low.
Further developments with new algorithms that will simplify numerical solutions and
test procedures for determining the viscous parameters will be available in the near
future.
REFERENCES
[1] Y. Katema, “A Viscoelastic Dynamic Vibration Absorber with Adaptable

Suppression Band : A Feasibility Study,” J. Sound Vib., vol. 216, pp. 133–145,
1998.
[2] J. Zhang and C. M. Richards, “Dynamic Analysis and Parameter Identification of a
Single Mass Elastomeric Isolation System Using a Maxwell-Voigt Model,” J. Vib.
Acoust., vol. 128, no. December, pp. 713–721, 2006.
[3] S. M. H. Ã and S. Kazemirad, “Dynamic viscoelastic effects on sound wave
scattering by an eccentric compound circular cylinder,” J. Sound Vib., vol. 318, pp.
506–526, 2008.
[4] C. A. Bavastri and E. M. Lopes, “Design of Optimum System of Viscoelastic
Vibration Absorbers with a Frobenius Norm Objective Function,” J. Brazilian Soc.
Mech. Eng., vol. XXXI, no. 3, pp. 210–219, 2009.
[5] J. Elejabarrieta and M. Martinez-agirre, “International Journal of Mechanical
Sciences Characterisation and modelling of viscoelastically damped sandwich
structures,” Int. J. Mech. Sci., vol. 52, pp. 1225–1233, 2010.
[6] J. Chazot, B. Nennig, and A. Chettah, “Harmonic response computation of
viscoelastic multilayered structures using a ZPST shell element,” Comput. Struct.,
vol. 89, no. 23–24, pp. 2522–2530, 2011.
[7] T. Saksa, N. Banichuk, J. Jeronen, M. Kurki, and T. Tuovinen, “International
Journal of Solids and Structures Dynamic analysis for axially moving viscoelastic
panels,” Int. J. Solids Struct., vol. 49, no. 23–24, pp. 3355–3366, 2012.
[8] Lozano-Guzmán, C. Lopez-Cajun, and J. C. Jáuregui-Correa, “Viscoelasti Beam

Vibration Under Axial Tension Variations,” in Proceedings of the 9th World
Congress IFTOMM, 1993.
[9] E. Afonso, J. Tomás, and C. Alberto, “Passive vibration control in rotor dynamics :
Optimization of composed support using viscoelastic materials,” J. Sound Vib., vol.
351, pp. 43–56, 2015.
[10] T. Karttunen and R. Von Hertzen, “European Journal of Mechanics A/Solids
Steady-state vibration of a viscoelastic cylinder cover subjected to moving loads,”
Eur. J. Mech./A Solids, vol. 58, pp. 202–210, 2016.
[11] H. Moghadasi and A. Capsoni, “Investigation of the effects of viscous damping
mechanisms on structural characteristics in coupled shear walls,” Eng. Struct., vol.
116, pp. 121–139, 2016.
[12] C. Xu, M. Wu, and M. Hamdaoui, “Mixed integer multi-objective optimization of
composite structures with frequency-dependent interleaved viscoelastic damping
layers,” Comput. Struct., vol. 172, pp. 81–92, 2016.
[13] L. Wang, Z. Lu, and I. Hagiwara, “Finite element simulation of the static
characteristics of a vehicle rubber mount,” Proc. Inst. Mech. Eng. Part D J.
Automob. Eng., vol. 216, pp. 965–973, 2002.
[14] G. Muscolino and A. Palmeri, “Response of beams resting on viscoelastically
damped foundation to moving oscillators,” Int. J. Solids Struct., vol. 44, pp. 1317–
1336, 2007.
[15] W. Chen, “Bending vibration of axially loaded Timoshenko beams with locally
distributed Kelvin – Voigt damping,” J. Sound Vib., vol. 330, no. 13, pp. 3040–
3056, 2011.
[16] J. Boaga, S. Renzi, R. Deiana, and G. Cassiani, “Soil damping influence on seismic
ground response : A parametric analysis for weak to moderate ground motion,” Soil
Dyn. Earthq. Eng., vol. 79, pp. 71–79, 2015.
[17] M. Braga, H. Tronconi, F. Paulo, and L. Neto, “Assessment of semi-active friction
dampers,” Mech. Syst. Signal Process., vol. 94, pp. 33–56, 2017.
[18] J. Gong, A. Zecchin, M. Lambert, and A. Simpson, “Study on the frequency
response function of viscoelastic pipelines using a multi-element Kevin-Voigt
model,” Procedia Eng., vol. 119, no. 0, pp. 226–234, 2015.
[19] K. Bi and H. Hao, “Numerical simulation on the effectiveness of using viscoelastic
materials to mitigate seismic induced vibrations of above-ground pipelines,” Eng.
Struct., vol. 123, pp. 1–14, 2016.
[20] Keramat and K. Heidari, “Finite element based dynamic analysis of viscoelastic
solids using the approximation of Volterra integrals,” Finite Elem. Anal. Des., vol.
86, pp. 89–100, 2014.
[21] L. Manin, M. Bolte, and C. Gregor, “Introduction of damping into the flexible
multibody belt-drive model : A numerical and experimental investigation,” J. Sound
Vib., vol. 324, pp. 283–296, 2009.
[22] J. M. Franco and P. Partal, “The Newtonian Fluid,” Encyclopedia of Life Support
Systems. [Online]. Available: http://www.eolss.net/sample-chapters/c06/e6-197-04-
00.pdf.
[23] C. W. Macosko, RHEOLOGY Principles, Measurements and Applications. New
York: Wiley-VCH, 1994.
[24] J. Vincent, Basic Elasticity and Viscoelasticity. Princeton University Press, 2017.
[25] J. C. Jauregui-Correa, A. Lozano-Guzman, and J. Becerril, “Viscoelastic Torsional
Damper Model,” in ASME Power Transmission and Gearing Conference, 1996, pp.
777–780.
[26] T. J. R. Hughes and J. A. Evans, “Isogeometric Analysis,” ICES REPORT 10-18,
The Institute for Computational Engineering and Sciences, The University of Texas
at Austin, 2010.
[27] W. A. Wall, M. A. Frenzel, and C. Cyron, “Isogeometric structural shape
optimization,” Comput. Methods Appl. Mech. Eng., vol. 197, no. 33–40, pp. 2976–
2988, 2008.
Chapter 12
A VISCOELASTIC FLUID DUE TO A NONLINEAR

ACCELERATING ELASTIC SHEET
U. S. Mahabaleshwar
Department of Mathematics, Government First Grade College for Women,
Hassan –India
ABSTRACT
The present chapter deals with the steady, laminar boundary layer flow of non-
Newtonian fluid flow due to a nonlinear accelerating sheet. The accelerating sheet
presumed to be at least quadratic along the axial axis. The study encompasses two classes
of viscoelastic fluids namely, Walters’ liquid B and second order liquid. Similarity
solution is obtained for velocity profiles which are governed by nonlinear differential
equation. Analytical expressions are obtained for stream function and also the
expressions for velocity profiles in terms the accelerating and viscoelastic parameters.
The flow filed is influenced by the physical parameters like Chandrasekhar number,
nonlinear accelerating parameter and mass transpiration parameter. The results so
obtained are in line with the classical works dealt with the flow of fluid past a nonlinear
accelerating sheet moving away from the slit. When the liquid is at rest, the effect of
nonlinear stretching parameter is negligibly small. The expression for the stream function
 coincides with that of Rajagopal (1984) and Siddheshwar and Mahabaleshwar (2005).
The presence of viscoelastic and nonlinear accelerating sheet parameter has significant
impact on physical properties of the laminar boundary layer flows. This physical
phenomenon has got wide applications in the industrial processes such as cooling of the
liquid in the extrusion processes and fabrication of glass, manufacturing of polymer
fibers by means of the metal spinning, coiling of wires and other allied industrial
processes.
Keywords: Walters’ liquid B, second order liquid, similarity solution, nonlinear

accelerating sheet, mass transpiration parameter
382 U. S. Mahabaleshwar
NOMENCLATURE
a, b, C constants
B0 magnetic field (T-Tesla)
f similarity function for velocity
Q Chandrasekhar’s number
u velocity component along x-axis (m s-1)
v velocity component along y-axis (m s-1)
vw mass transpiration parameter
Vc dimensionless mass transpiration parameter
x axial axis (m)
y transverse axis (m)
Greek Letters
 constants rate of accelerating (s-1)

 nonlinear accelerating constant
 electrical conductivity
μ viscosity (kg m-1s-1)
 kinematic viscosity,      (m2 s-1)
 density (kgm-3)
 unknown
 physical stream function (m2 s-1)
θ inclined angle in degrees
Subscripts/Superscripts
0 pole
w wall condition
 far from the sheet
f

first order derivative with respect to  '  d
d 
A Viscoelastic Fluid Due to a Nonlinear Accelerating Elastic Sheet 383
INTRODUCTION
The viscoelastic flow of a steady, laminar boundary layer accelerating sheet has
significant applications in the manufacturing processes and polymer industries (see Fisher
1970 and Altan et al. 1979). As a result of this fact, many of the researchers around the
world are continuously evolving with various mathematical models for cooling of the
liquid as it plays an important role in obtaining the desired output. Pioneering works of
Sakiadis and Blasius are the fundamental works on the fluid flows in theoretical fluid
dynamics based on the dynamics of velocity of the fluid. Sakiadis (1961a, b) is the first to
discover the effects of continuously stretching sheet on the flow field. The flow of a
liquid due to a moving surface is called Sakiadis flow. The problem of fluid flow with
steady speed was examined by Blasius (1908) and this flow is known as Blasius flow of
the fluid. Since, the solutions obtained by Sakiadis are not exact in nature; Crane (1970)
provided the exact analytical solution of the Navier-Stokes equation due to accelerating
sheet. This classical work of Crane has attracted many researchers to undertake problems
in boundary layer flows. Further, works of Crane is extended by Pavlov (1973 and 1974)
with the inclusion of important physical parameter, magnetohydrodynamics (MHD),
since the flow geometry is greatly influenced by MHD. The important applications in
engineering processes such as MHD generators, pumps and other equipment whose
boundary layer control are affected by MHD (see Andersson 1992, Chaim 1997, Chen
1998, Gorder and Vajravelu 2011). Gupta and Gupta (1974) extended the works of Crane
and Pavlov with the inclusion of the mass transpiration parameter. In recent times, the
viscoelastic fluids (Walter’s liquid B and second order liquid) were widely considered in
modeling of the problem due to their enormous applications in polymer processing
industries. Rajagopal (1984 and 1987), investigated the flow of a viscoelastic fluid over a
stretching sheet and also obtained the analytical solutions of a non-Newtonian fluid flow
due to stretching sheeting with uniform free stream function. Subsequently, this work was
extended by Siddeshwar and Mahabaleshwar (2005), with the inclusion of with inclined
magnetic field. The present problem deals with the nonlinear accelerating sheet in the
presence of inclined magnetic field. The influence of physical parameters such as
Chandrasekhar number, mass transpiration, inclined MHD and nonlinear accelerating
parameters on the flow geometry analyzed graphically. The limiting cases of the present
results yields that of Crane (1970), Pavlov (1974), Gupta and Gupta (1977), Kumaran and
Ramanaiah (1996), Rajagopal (1984) and Siddheshwar and Mahabaleshwar (2005).
MATHEMATICAL FORMULATION OF THE PROBLEM
The two-dimensional steady, laminar boundary layer flow of an electrically

conducting viscoelastic fluid flow due to nonlinear accelerating sheet is considered (see
Figure 1). The inclined magnetic field of strength B0 is applied normal to the direction of
x-axis with acute angle θ. Two forces of same magnitude and of opposite directions are
applied along the axial axis, so that the acceleration of the sheet takes place with the fixed
origin. With the Lorentz force acting on the sheet, the porous laminar boundary layer
sheet, is accelerated with a large force. The resulting velocity acting on the sheet far from
the origin is a quadratic polynomial, u   x   x .
2
Figure 1. Schematic diagram of the flow geometry.
The steady, laminar boundary layer flows of a viscoelastic fluid past a nonlinear
accelerating sheet with the governing boundary layer conditions for the conservation of
mass and the momentum boundary layer equations is given by (see Beard and
Walters,1964):
u v
  0, (1)
x y
u u  2u   3u  3u u  2u u  2u   B02
u v   2  k0 u  v    u sin 2  ,
x y y  x y
2
y x y y x y  
3 2
(2)
where, u and v are longitudinal and transverse velocity components and equation (1) is
equipped with the following boundary conditions.
u  x, 0    x   x 2 , 


v  x, 0   vw   x,  , (3a, b, c)
lim u  x, y   0 
y  

here, α and β are the constants of the specified rate of accelerating with dimension
inverse of time and for rest of the parameters, one can refer nomenclature.
To facilitate the analysis, the dimensionless variables,  and f   are introduced
into the model and the system of nonlinear differential equations so obtained are analyzed
via conventional similarity transformations (see Kumaran and Ramanaiah, 1996). The
physical stream function is defined as,

 ( x, y)   x f    x 2 f '   , (4)
2

where,   y

The axial and transverse velocity components in terms of stream functions are given
by,
   2 
u   xf '    x f ''   and v      f     x f '   .
y 2  x
(5a, b)
Using equations (4) and (5a, b) in (2), the governing boundary layer equation for
physical stream function (4) is obtained as,
      3     2 
   
      
y 
, , ,
 3    y 3    y y
2
    Q sin 2   0 , (6)
  K1  
y 3   x, y     x, y    x, y   y

 

here, second and third term denotes the Jacobians and the imposed boundary conditions
fulfilled by the stream function (4) is given by,
 
y
 
 x   * x2 
y 0 

  (7a, b, c)
 2  0 x ,
x y 0 

 
0
y y 

Using equations (4), (5a, b) and (6) subject to the boundary conditions (7a, b, c), the
similarity solution of the dimensionless stream function is taken as,
  x, y   x f ( )  0 x2 f '( ) , (8)
where, superscript indicates the derivative with respect to ɳ. Using equations (6) - (8) and
equating the coefficients of x, x2 and x3, the following nonlinear ordinary differential
equations are obtained:
 f 
2
 
 f f   f   K1 2 f  f   f f    f    Qf  sin 2  ,
2
(9)
f  f   f f   f   K1  f  f   f f   Qf  sin 2  (10)
and
 f 
2
 
 f  f   K1 f  f   2 f  f    f   ,
2
(11)
k 2 
where, K  is the viscoelastic parameter and Q  B0
0
is the Chandrasekhar
1
 
number.
The corresponding imposed boundary conditions are,
f (0)  1, f (0)   , 

f (0)  Vc ,  (12 a, b, c)
f ()  0, f ()  0,  
2  * v
here, f (0)      and Vc  w .
  0 
The exact analytical solution of two-dimensional Navier-Stokes equation in the
absence of nonlinear accelerating sheet due to Crane (1970) is,
f    1  exp    . (13)
Motivated by this, Kumaran and Ramanaiah (1996) obtained the exact analytical
solution for Navier-Stokes equation in the presence of nonlinear stretching sheet as,
f    a  b exp    , (14)
where, a, b and  are arbitrary constants.

It can be easily seen that, f     exp    is the solution of the ordinary
differential equation (10) and satisfies the derivative boundary conditions (11).
The exact solution of the equation (9) is given by,
 1  exp( ) 
f ( )  Vc   , (15)
  
which satisfy the boundary condition (14).

Upon using equation (15) in equation (9), the following algebraic equation is
obtained:
 1  K1  2  1   1  Q sin 2  
 
3
      0 . (16)
 Vc K1   K1   Vc K1 
To analyze the characteristics by means of stream function (4), the following

dimensionless variables are introduced:
* 
 , x  x  and 0   . (17a, b, c)
  2
Using equation (17a, b, c) in equation (4), the expression for the streamline pattern is
obtained as,
 *  x f    0x 2 f    , (18)
The streamline function,   C (constant) can be written as,

*
  x 
    0x 2 
  log  
1
 . (19)
   xVc  x  C 
   
Substituting Eq. (17a, b, c) into Eq. (5a, b), the expression for dimensionless velocity
components are derived as,
u  x f   0x 2 f  and v    f  20x f  . (20a, b)
The physical model of the steady, laminar boundary layer MHD flows of a non-
Newtonian (Walters’ liquid B) past a nonlinear accelerating sheet is investigated. Here,
the generalization of classical works of Rajagopal (1984) and Siddeshwar and
Mahabaleshwar (2005) is considered. In this model, the fluid flow considered is due to
the nonlinear accelerating sheet with the applied external uniform inclined magnetic field.
The presence of physical parameters greatly influences the flow field. The flow field is
characterized with the help of physical stream function and also the longitudinally as well
as transverse velocity profiles are characterized. The exact analytical solutions of the
transformed equations are obtained and the velocity distribution patterns are analyzed
graphically.
Figures 2-7, depicts that the velocities both longitudinal as well as transverse is a
decreasing function of ɳ, as it is an exponential function with negative argument. It is
clear from these plots that, the flow region is influenced by the physical parameters like
Chandrasekhar number (Q), inverse Darcy number (Da-1), inclined angle (θ) and mass
transpiration parameter (Vc). The increase in these physical parameters, results in the
decrease of laminar boundary layer thickness.
Figures 8-10, shows the effect of various physical parameters on streamline patterns.
It can be noted that the increasing values of physical parameters forces streamline
function to shift towards the stretching surface.
Figure 2. Effect of Chandrasekhar number Q on transverse velocity for mass transpiration parameter Vc
= -0.5, viscoelastic parameter K1 = 0.2 and angle parameter θ = 900.
Figure 3. Effect of Q on transverse velocity for Vc = -0.5, K1 = 0.2 and θ = 900.
Figure 4. Effect of Vc on transverse velocity for Q = 2, K1 = 0.2 and θ = 900.
Figure 5. Effect of Vc on axial velocity for Q = 2, K1 = 0.2 and θ = 900.

Figure 6. Effect of θon transverse velocity for Q = 2, K1 = 0.2 and Vc =-0.5.
Figure 7. Effect of θon axial velocity for Q = 2, K1 = 0.2 and Vc =-0.5.
Figure 8. Streamline pattern for different values of nonlinear accelerating parameter β0 with C= 4, Q =
1, K1 = 0.2, θ = 900 and Vc = 0.6.
Figure 9. Streamline pattern for different values of θ with β0 =0.2, C= 5,Q = 2, K1 = 0.2 and Vc = 0.5.
Figure 10. Streamline pattern for different values of Q with θ = 900, β0 =0.2, C = 2, K1 = 0.2 and Vc =
0.5.
CONCLUSION
The inclined magnetic field on the laminar boundary layer flows due to an
accelerating sheet in the presence of mass transpiration is investigated. Both, longitudinal
and transverse velocity profiles are influenced by the presence of various physical
parameters. It can also be seen in the reduction in the laminar boundary layer thickness
with the increasing values of these physical parameters. It is clearly evident, that the
inclination angle of the magnetic field plays a major role in regulating the effects of
MHD on the fluid flow geometry. Also, these physical parameters influence the
streamlines to fall steeply near the accelerating sheet. The limiting cases are provided as
the present results are in good agreement with the classical works.
REFERENCES
Altan, T. Process modeling: fundamentals and applications to metals, Proceedings of

American Society for metals process modeling sessions, materials and processing
congresses, USA, 1978 & 1979; 448 pages.
Andersson, H. I. 1992. MHD flow of a viscoelastic fluid past a stretching surface, Acta
Mech., Vol. 95, pp. 227-230.
Beard, D. W., & Walters, K. 1964. Elastico-viscous boundary-layer flows. I. Two-
dimensional flow near a stagnation point, Proc. Camb. Phil. Soc., Vol. 60, pp. 667-
674.
Blasius, H. 1908. Grenzschichten in Flüssigkeiten mit kleiner Reibung, Z. Angew. Math.
Phys.,Vol.56, pp.1–37.
Chaim, T. C. 1997. Magnetohydrodynamic heat transfer over a non-isothermal stretching
sheet, Acta Mechanica, Vol.122, pp.169-179.
Chen, C.H., 1998, Laminar mixed convection adjacent to vertical, continuously stretching
sheet, Int. Journal of Heat and Mass Transfer, Vol. 33, pp. 471-476.
Crane, L. J. 1970. Flow past a stretching plate, Z. Angew. Math. Phys. Vol. 21, pp. 645-
647.
Fisher, E.G. Extrusion of plastics, Newnes-Butterworld, London, UK, 1976; Third
edition, 355 pages.
Gorder, R.A.V., & Vajravelu, K., 2011, Multiple solutions for hydromagnetic flow of a
second grade fluid over a stretching or shrinking sheet, Quarterly of Applied
Mathematics, Vol. LXIX (3), pp. 405–424.
Gupta, P.S. & Gupta, A.S. 1974. Radiation effect on hydromagnetic convection in a
vertical channel, International Journal of Heat and Mass Transfer, Vol. 17, pp. 1437-
1442.
Gupta, P.S. & Gupta, A.S., 1977, Heat and mass transfer on a stretching sheet with
suction or blowing, The Canadian Journal of chemical engineering, Vol. 55(6), pp.
744-746.
Kumaran, V. & Ramanaiah, G. 1996, A note on the flow over a stretching sheet, Acta
Mechanica, Vol. 116(1), pp. 229-233.
Pavlov, K.B. 1973. Oscillatory magnetohydrodynamic flows of viscoplastic media in
plane ducts, Magnetohydrodynamics, Vol. 9(2), pp. 197–199.
Pavlov, K.B. 1974. Magnetohydrodynamic flow of an incompressible viscous fluid
caused by deformation of a plane surface, Magnitnaya Gidrodinamika, Vol.4,
pp.146–147.
Rajagopal, K. R., Na, T. Y., & Gupta, A. S. 1987. A non-similar boundary layer on a
stretching sheet in a non-Newtonian fluid with uniform free stream, J. Math. Phys.
Sci., Vol. 21, pp. 189–200.
Rajagopal, K.R., Na, T.Y., & Gupta, A.S.1984. Flow of a viscoelastic fluid over a
stretching sheet, Rheol.Acta., Vol.23, pp.213-215.
Sakiadis, B.C. 1961a. Boundary-layer behavior on continuous solid surfaces: I.
Boundary-layer equations for two-dimensional and axisymmetric flow, AIChE
Journal, Vol. 7(1), pp. 26–28.
Sakiadis, B.C. 1961b. Boundary-layer behavior on continuous solid surface: II.
Boundary-layer equations for two-dimensional and axisymmetric flow, AIChE
Journal, Vol. 7, pp. 221–225.
Siddheshwar, P. G. and Mahabaleshwar, U. S. 2005. Effects of radiation and heat source
on MHD flow of a viscoelastic liquid and heat transfer over a stretching sheet,
International Journal of. Non-linear mechanics, Vol. 40, pp.807-820.
Chapter 13
AN MHD BOUNDARY LAYER VISCOELASTIC FLUID

FLOW PAST A STRETCHING/SHRINKING SHEET IN
A POROUS MEDIUM
P. N. Vinay Kumar
Department of Mathematics, Sri H.D. Devegowda Government First Grade College,
Paduvalahippe, India
ABSTRACT
The present chapter deals with the steady, laminar boundary layer flow of an
incompressible viscoelastic fluid (Walters’ liquid B) due to a porous stretching/shrinking
sheet in the presence of magnetohydrodynamics (MHD). The flow of viscoelastic fluid
due to a stretching /shrinking sheet is widely considered. The applications can be found
in the industrial processes such as wire drawing, manufacture of polymer fibres and
plastic fibres, in the glass blowing as well as to increase the performance of the
lubricants. This fluid flow problem is mathematically modeled into a nonlinear system of
partial differential equations which are then transformed via suitable similarity
transformation reducing it to nonlinear ordinary differential equation. The exact
analytical solution is obtained. The presence of MHD greatly influences the flow
geometry as the Lorentz force acts against the flow field. In addition to this, the presence
porosity in the medium, results in the Darcy drag force which further supports the
opposition to the flow of the fluid. This combined effect of Lorentz as well as Darcy drag
force is much advantageous as these two forces acts as one. Also, the outcome depends
on the physical parameter which accounted into the practical problems of chemical and
metallurgical industries. The effects of physical parameters like Chandrasekhar number
(Q), inverse Darcy number (Da-1) and suction/injection (Vc) parameters on the flow

Corresponding Author Email: vinaykumarpn1981@gmail.com.
396 P. N. Vinay Kumar
geometry are demonstrated graphically. The results so obtained are in line with that of the
classical works and the limiting cases are provided to give good comparison.
Keywords: MHD flow, viscoelastic fluid (Walters’ liquid B), inverse Darcy number,
stretching/shrinking sheet
NOMENCLATURE
B0 magnetic field
f similarity function for velocity
k0 viscoelastic constant
K1 dimensionless viscoelastic parameter
Da-1 inverse Darcy number
k permeability of the porous medium
Q Chandrasekhar number
u velocity component along x-axis (m s-1)
v velocity component along y-axis (m s-1)
Vc suction/injection parameter
x axial axis (m)
y transverse axis (m)
Greek Letters
 & specified constants

ɳ dimensionless similarity variable
μ viscosity (kg m-1s-1)
 kinematic viscosity (m2 s-1)
 density (kgm-3)
 stretching/shrinking coefficient
 physical stream function (m2 s-1)
Subscripts/Superscripts
0 Pole
w wall condition
 for from the sheet
An MHD Boundary Layer Viscoelastic Fluid Flow … 397
f first order derivative with respect to 
f second order derivative with respect to 
f third order derivative with respect to 
f fourth order derivative with respect to 
INTRODUCTION
The laminar, steady boundary layer flows of viscoelastic fluid past a

stretching/shrinking sheet is modelled into two-dimensional Navier-stokes equations. The
Navier-stokes equations are fundamental in theoretical fluid dynamics and they are
nonlinear partial differential equations with exact solutions. The flow geometry of non-
Newtonian fluid flows due to a continuously stretching/shrinking sheet plays vital role in
many practical problems, as the outcome of the product depends on the rate of cooling of
the liquids. This type of flow problems arises in the industrial processes like production
of glass fibre, paper production and the extrusion of molten polymers.
Motivated by the physical applications of aforementioned flow problems many
authors have presented different mathematical models by accounting the effects of
different physical parameters (see Walters’ 1960, Berker 1963, McCormak and Crane
1973, Cheng 1977 and Davidson 2001).
In recent years, the study of viscoelastic fluid flow in the presence of MHD in flow
model has attracted many researchers due to realistic applications. The problems arising
in the industries dealing with the production of petroleum products like varnish, paints
and lubricants, metallurgical process and polymer fibre production are influenced by the
MHD parameter. The controlling of momentum in the flow field can be realized by the
presence of Lorentz force. The classical works of Crane (1970) provides the exact
analytical solution of the boundary layer flow due to a continuously stretching surface.
Pavlov (1973 and 1974) was the first to observe the effect of MHD in the flow geometry.
Andersson (1992) has obtained the exact analytical solution of the laminar boundary
layer flows due to stretching in the absence of porous media. For related works, see,
Rajagopal et al.(1984), Chaim (1997), Andersson (2002).
Another important parameter to be accounted in considering the flow model is
porosity in the medium. The heat dissipation area in the porous medium is comparatively
larger than that of the conventional medium. Due to this fact, porous medium is termed to
be more advantageous in modeling the real world problem. For convection of heat in the
porous medium related problems one can refer, Ingham and Pop (2005), Vafai (2005),
Neild and Bejan (2013). Initially, it is Cheng (1977) who investigated the blowing of
convective heat transfer past a transverse permeable surface embedded in a porous media.
Later, Rajagopal et al., (1986) have investigated the non-Newtonian fluid flow past an
infinite porous plate. Vajravelu (1994) provided the analytical solutions for the fluid flow
with heat transfer in a saturated porous medium due to a stretching sheet. The combined
effects of MHD and porosity, plays a significant role, as this suits better for the realistic
problems. These are of great interest from both theoretical and practical applications such
as geothermal systems, thermal insulations and ground water flow (see Abel and Veena
1998, Siddeshwar and Mahabaleshwar 2005, Cortel 2007a, b, Sedeek 2007).
The elasticity of the sheet is another physical parameter which plays major role in the
flow geometry (see Gorder and Vajravelu 2011). The viscous fluid flow due to shrinking
sheet was first investigated by Miklavcic and Wang (2006). This work was further
extended by Fang (2008) with the inclusion of power law of fluids. Later, many authors
have contributed in developing the flow models involving shrinking sheet. In view of all
the physical parameters above mentioned, the present problem is mathematically
modeled to throw light on the combined effects of MHD, porous medium, viscoelastic
parameters, suction/injection parameter and stretching/ shrinking parameters on the
laminar boundary layer flows problem. The exact analytical solutions are investigated via
similarity transformations.
MATHEMATICAL MODELLING OF THE PROBLEM
The two-dimensional steady laminar flows of an incompressible fluid in a porous

medium of permeability k due to a stretching/shrinking sheet in an electrically conducting
elastic Walter’s liquid B considered is as shown in the Figure 1.
Two forces of opposite directions with equal in magnitude are applied along the x-
axis which results in the linear stretching/shrinking of the sheet. A uniform magnetic field
of strength B0 is imposed normal to the sheet. The fluid flow on two-dimensional plane is
chosen by taking x-axis (longitudinal) along the sheet and y-axis (transverse) is
considered normal to the sheet. The conservation of mass and the momentum equations
governing the steady, laminar boundary layer fluid flows within the porous matrix (see
Beard and Walters’ 1964) are given by,
qi
 0, (1)
xi
 qi qi  p 
  qj    ij' , j  qi   Bo2qi , (2)
 t x j  xi k
here, the pressure gradient is negligible to stretching/shrinking of the sheet and for the
remaining physical parameters refer the listed nomenclature.
Figure 1. Schematic diagram of the physical problem of boundary layer flows in porous medium.
The constitutive relationship of Walters’ liquid B in equation (2) for  ij is given by,
'
 d    q q j  
 ij'  1  ko     i   , (3)
 dt    x j xi  
 
d  
where,   qj and k0 is the viscoelastic coefficient.
dt t x j
The resulting governing boundary layer equations for conservation of mass and
conservation momentum transforms into:
u v
  0. (4)
x y
u u  2u   3u  3u u  2u u  2u    H 02  
u  v   2  k0  u v 3      u, (5)
x y y  x y
2
y x y 2 y x y    k
with the subsequent boundary conditions on the velocity,
u  x, 0   uw ( x)   x, v  x, 0   vc ,
lim u ( x, y )  0, lim u yy  x, y   0
. (6)
y  y 
Here, u and v represents the velocity components along the longitudinal and
transverse axis along the sheet.
In support of the analysis, the following dimensionless variables  and f   are
introduced in the similarity transformation:
   x f ( ) , (7)

and   y , (8)

where,  represents the dimensionless distance from the origin and ψ is the induced
physical stream function. The transformed velocity components related with stream
function are given by,
u   y   xf , v   x    f . (9)
here, subscript denotes the derivative with respect to  .

Using eq. (9) and (5), the nonlinear differential equation with constant coefficient is
derived as:
f  f2  ff   Q Da 1  f  K1  2 f f  f2  ff  , (10)

  B02
where, Da 1  is the inverse Darcy number, Q  is Chandrasekhar number
k 
 k0
and K1  is the viscoelastic parameter.

Using equations (5) and (9), the governing boundary conditions for dimensionless
variable f is obtained as,
vc
f (0)  Vc   , f (0)   , (11)

f ()  0 and f ()  0 . (12)
where, Vc is the suction/injection parameter. When Vc  0 , represents suction, Vc  0

represents injection and Vc  0 indicates the impermeable case. Based on the choice of
 , the suction/injection parameter, we come across three different cases namely,
stretching sheet, shrinking sheet and for fixed sheet depending on   0 ,   0 and
  0 respectively. The exact analytical solution of the nonlinear differential equation
(10) is as follows:
 1  exp(  ) 
f ( )  Vc    , (13)
  
which satisfies the governing boundary layer conditions.

Upon using the equation (13) in (10), one can arrive at,
 1  K1  2  1     Q  Da 1 
3           0 . (14)
 Vc K1   K1   Vc K1 
The cubic equation so obtained characterizes the laminar boundary layer flows past a
stretching/shrinking sheet. Also, it can be analyzed by means of physical stream function
via elementary transformations.
The mathematical model of the laminar boundary layer MHD flows of a viscoelastic
fluid (Walters’ liquid B) in a porous plate due to a stretching/shrinking sheet with
suction/injection is considered. Initially, this problem is a system of nonlinear partial
differential equations (4)-(5) which are then transformed into nonlinear ordinary
differential equation (10) with governing boundary conditions (11)-(12) via similarity
transformations. The exact analytical solution of equation (10) is obtained in (13). The
analytical solution of the laminar boundary layer flows can be characterized by means of
physical stream function in the cubic equation (14). The plots in the chapter depicts the
effects of physical parameters like Chandrasekhar number (Q), inverse Darcy number
(Da-1) and the suction/injection parameter (Vc) on the longitudinal and transverse velocity
components.
The effect of Darcy drag force on the transverse and longitudinal velocity
components due to a stretching sheet case is shown in Figure (2a) and (2b), by keeping
all other physical parameters constant. It clearly demonstrates that the momentum
boundary layer thickness is inversely proportional to the Darcy drag force. Figure (3a)
and (3b) shows the resistance offered by the Darcy drag force to the momentum boundary
layer flows in transverse and longitudinal velocity components in shrinking sheet case
with all other physical parameters kept constant.
Figures (4a)-(4b) and (5a)-(5b) depicts the effect of Lorentz force acting against the
momentum boundary layer flows in transverse and longitudinal velocity components for
stretching and shrinking sheet cases respectively. Figures (6a)-(6b) and (7a)-(7b) outlines
the effect of suction/injection parameter on the velocity components in stretching and
shrinking cases respectively. In all, these plots it is visible that the increase in the
suction/injection parameter results in the decrease of momentum boundary layer
thickness
Figure 2a. Velocity profiles f (η´) for various values of inverse Darcy number (Da-1) when Walters’
liquid B K1= 0.2, Chandrasekhar Q = 1 and suction/injection Vc = -0.5.
Figure 2b. Velocity profiles f η for various values of Da-1 when K1=0. 2, Q = 1 and Vc = - 0.5.
Figure 3a. Velocity profiles f (η) for various values of Da-1 when Q = 1, K1 = 0.2 and Vc = -0.5.
Figure 3b. Velocity profiles f η for various values of Da-1 when Q = 1, K1 = 0.2 and Vc = -0.5.
Figure 4a. Velocity profiles f (η´) for various values of Q when Da-1= 1, K1 = 0.2 and Vc = -0.5.
Figure 4b. Velocity profiles f η for various values of Q when Da-1= 1, K1 = 0.2 and Vc = -0.5.
Figure 5a. Velocity profiles f (η´) for various values of Q when Da-1= 1, K1 = 0.2 and Vc = -0.5.
Figure 5b. Velocity profiles f η for various values of Q when Da-1= 1, K1 = 0.2 and Vc = -0.5.
Figure 6a. Velocity profiles f (η´) for various values of Vc when Da-1= 1=Q and K1 = 0.2.
Figure 6b. Velocity profiles f (η´) for various values of Vc when Da-1= 1=Q and K1 = 0.2.
Figure 7a. Velocity profiles f (η´) for various values of Vc when Da-1= 1=Q and K1 = 0.2.
Figure 7b. Velocity profiles f η for various values of Vc when Da-1= 1=Q and K1 = 0.2.
CONCLUSION
This chapter systematically deals with the steady, laminar boundary layer flows of an
electrically conducting viscoelastic fluid (Walters’ liquid B) due to a stretching/shrinking
sheet in porous plate. The flow geometry is influenced by the physical parameters like
Chandrasekhar number (Q), inverse Darcy number (Da-1) and suction/injection
parameter. The proper choice of velocity distribution yields the desired outcome. In
summary, the result clearly demonstrates that the proper choice of the fluid and the
velocity distribution, cooling of the liquid can be controlled based on the expected
outcome.
REFERENCES
Abel M. S., & Veena, P. 1998. Visco-elastic fluid flow and heat transfer in a porous
medium over a stretching sheet, International Journal of. Non-linear mechanics, Vol.
33, pp. 531-538.
Andersson, H. I. 1992. MHD flow of a viscoelastic fluid past a stretching surface, Acta
Mech., Vol. 95, pp. 227-230.
Andersson, H. I. 2002. Slip flow past a stretching surface, Acta Mechanica, Vol. 158,
pp.121-125.
Beard, D. W., & Walters, K. 1964. Elastico-viscous boundary-layer flows. I. Two-
dimensional flow near a stagnation point, Proc. Camb. Phil. Soc., Vol. 60, pp. 667-
674.
Berker, R. 1963. Integration des equations du mouvement d’un fluide visqueux
incompressible. In Handbuch der Physik, ed. S. FIiigge, VIII/2: 1-384. Berlin:
Springer Verlag.
Chaim, T. C. 1997. Magnetohydrodynamic heat transfer over a non-isothermal stretching
sheet, Acta Mechanica, Vol. 122, pp.169-179.
Cheng, P. 1977. Combined free and forced convection flow about inclined surfaces in
porous media, Int. J. Heat Mass Transfer, Vol. 20, pp. 807–814.
Cortell, R. 2007. MHD flow and mass transfer of an electrically conducting fluid of
second grade in a porous medium over a stretching sheet with chemically reactive
species, Chem. Eng. Process., Vol. 26, pp. 721–728.
Cortell, R. 2007. Toward an understanding of the motion and mass transfer with
chemically reactive species for two classes of viscoelastic fluid over a porous
stretching sheet, Chem. Eng. Process: Process Intensification, Vol. 46, pp. 982–989.
Crane, L. J. 1970. Flow past a stretching plate, Z. Angew. Math. Phys. Vol. 21, pp. 645-
647.
Davidson, P. A. Introduction to magnetohydrodynamics, Cambridge university press,
2001; First edition, pp.572.
Fang, T. 2008. Boundary layer flow over a shrinking sheet with power- law velocity, Int.
J. Heat Mass Transfer, Vol. 51(25/26), pp. 5838-5843.
Gorder, R. A.V., & Vajravelu, K., 2011, Multiple solutions for hydromagnetic flow of a
second grade fluid over a stretching or shrinking sheet, Quarterly of Applied
Mathematics, Vol. LXIX (3), pp. 405–424.
Ingham, D. B., & Pop, I. Transport Phenomena in Porous Medium III., Elsevier, Oxford,
UK, 2005; First edition, pp. 450.
McCormack, P. D., & Crane, L. J. Physical Fluid Dynamics, Academic Press, New York,
USA, 1973; First edition, pp.487.
Milavcic, M., & Wang, C.Y. 2006. Viscous flow due to a shrinking sheet, Quarterly of
Applied Mathematics, Vol. LXIV, pp. 283-290.
Nield, D. A., & Bejan, A. Convection in porous media, Springer-Verlag, USA, 2013;
Fourth edition, 778 pages.
Pavlov, K. B. 1973. Oscillatory magnetohydrodynamic flows of viscoplastic media in
plane ducts, Magnetohydrodynamics, Vol. 9(2), pp. 197–199.
Pavlov, K. B. 1974. Magnetohydrodynamic flow of an incompressible viscous fluid
caused by deformation of a plane surface, Magnitnaya Gidrodinamika, Vol.4,
pp.146–147.
Rajagopal, K. R., Na, T. Y., & Gupta, A. S. 1984. Flow of a viscoelastic fluid over a
stretching sheet, Rheol. Acta. Vol. 23, pp. 213-215.
Rajagopal, K. R., Szeri, A. Z., & Troy, W. 1986. An existence theorem for the flow of a
non-Newtonian fluid past an infinite porous plate, International Journal of Non-
Linear Mechanics, Vol. 21, pp. 279-289.
Sedeek, A. 2007. Heat and mass transfer on a stretching sheet with a magnetic field in a
viscoelastic fluid flow through a porous medium with heat source and sink, Comput.
Math. Sci., Vol.38, pp.781–787.
Siddheshwar, P. G. and Mahabaleshwar, U. S. 2005. Effects of radiation and heat source
on MHD flow of a viscoelastic liquid and heat transfer over a stretching sheet,
International Journal of. Non-linear mechanics, Vol. 40, pp. 807-820.
Vafai K. Handbook of Porous media, CRC Press, Boca Raton, 2005; Second edition,
pp.784.
Vajravelu, K. 1994. Flow and Heat Transfer in a Saturated Porous Medium over a
Stretching Surface, ZAMM, Vol. 74(12), pp. 605-614.
Walters, K. 1960, The motion of an elastico-viscous liquid contained between coaxial
cylinders (II), Q. J. Mech. Appl. Math. Vol.13, pp. 444-461.
Chapter 14
THERMAL RADIATION EFFECT ON FULLY

DEVELOPED LAMINAR MIXED CONVECTION FLOW
IN A VERTICAL POROUS STRATUM BY USING
DIFFERENTIAL TRANSFORM METHOD
Patil Mallikarjun1, and U. S. Mahabaleshwar2,†

1
Department of Studies and Research in Mathematics,
Tumkur University, Tumakuru, India
2
Department of Mathematics,
Government First Grade College for Women,
Hassan, India
ABSTRACT
An analysis is carried out to study the flow of Newtonian liquid. The heat transfer in
the medium is analyzed for observing the effects of thermal radiation on both free as well
as forced convection flow in a parallel-plate vertical channel in the porous medium with
symmetric and asymmetric wall heating conditions. The wall heating boundary
conditions are classified into three types and are analyzed. These thermal boundary
conditions are isothermal-isothermal, isoflux-isothermal and isothermal-isoflux for the
left-right walls of the channel. Further, comparison amongst the boundary conditions are
presented in the form of plots and also by taking into account of various governing
parameters such as porous parameter, the ratio of Grashof number to Reynolds number,
radiation parameter, Brinkman number and perturbation parameter for equal and different
wall temperatures. Analytical solutions of the coupled nonlinear governing equations is

Corresponding Author Email: mbp1007@yahoo.com.
†
Corresponding Author Email: ulavathi@gmail.com.
410 Patil Mallikarjun and U. S. Mahabaleshwar
derived with the help of regular perturbation method. The solutions obtained in as
analytical and by differential transform method are in good agreement with the classical
works for small values of perturbation parameter. In the present study, the DTM and
perturbation method were used to find analytical solutions of mixed convection problem
in the presence of porous medium, radiation effect and viscous dissipation.
Keywords: mixed convection, porous medium, radiation, perturbation method,

differential transform method
INTRODUCTION
Buoyancy is of considerable importance in technological applications, such as in

heated rooms or reactor configurations, where variation in temperature coefficient results
in abstract flow patterns. Indeed, it is a well known fact that the heat exchangers
technology often involves convective flows in vertical channels, where these flows in
most cases imply thermal conditions of uniform heating of channel walls either by
isothermal or iso-heat flux boundary conditions. Due to this wide applications in heat
exchangers technology, it has attracted many researchers to model such problems. A
theoretical study of fully developed, mixed convection in vertical channel was conducted
by Aung and Worku [1] and Cheng et al. [2] including flow reversal. Further, Zanchini
[3] and Barletta [4] investigated the effects of viscous dissipation on mixed convection
flow in vertical channels by taking thermal boundary conditions as prescribed uniform
heat fluxes on both walls and the case of prescribed uniform temperatures on both walls,
or isothermal-isoflux boundary conditions. Boulama and Galanis [5] obtained the
analytical solution for mixed convection in vertical channel with heat and mass transfer.
Barletta et al. [6] discussed the flow of dual mixed convection in vertical parallel-plate
channel. Prathap Kumar et al. [7] have studied the mixed convection of composite porous
medium in a vertical channel with asymmetric wall heating conditions. Later, Umavathi
et al. [8, 9] analyzed the effect of mixed convection of two immiscible fluids in a vertical
wavy channel.
The radiative effects have important applications in physics and engineering, space
technology and industries dealing with high temperature processes. The high temperature
plasmas, cooling of nuclear reactors, liquid metal fluids, and power generation systems
are some important applications of radiative heat transfer from a vertical wall to
conductive gray fluids. For more related works on thermal radiation involving various
physical parameters refer, Siegel and Howell [10], Chamkha [11], Raptis [12, 13], Bakier
[14], Raptis and Perdikis [15], Bég et al. [16], Ghosh and Bég [17]. Umavathi et al. [18]
Mahesha et al. [19] The nonlinear partial differential equations are transformed into
ordinary differential equation by differential transformation method (DTM). The DTM is
a semi-numerical–analytical method for solving problems of this type which uses Taylor
Thermal Radiation Effect on Fully Developed Laminar … 411
series to approximate the solution. Chu and Lo [2] formulated the solution of nonlinear
transient heat conduction problems using hybrid differential transformation finite
difference method. Joneidi et al. [28] obtained the analytical solution of fin efficiency of
convective straight fins with temperature-dependent thermal conductivity by DTM. Jang
et al. [29] investigated and characterized the two-dimensional thermal conductive
boundary value problem with discontinuous boundary and initial conditions. Rashidi
et al. [30] applied DTM to obtain the analytic solution of mixed convection about an
inclined flat plate embedded in a porous medium. Later, Yaghoobi and Torabi [31]
analyzed the application of differential transformation method to nonlinear equations
arising in heat transfer (For more refer, Zhou [20, 21–24], Bert [25], Chu and Chen [26]).
In all the above works, the combined convection and radiation problem in the fully
developing region are analyzed. However, The main objective of the the present study is
to obtain a semi-numerical-analytical solution for the combined mixed convection and
radiation in a fully developed vertical channel, by taking into account of viscous
dissipation and different thermal boundary conditions.
GOVERNING EQUATIONS
The two dimensional plane with axial axis is parallel to the gravitational acceleration

vector g , but with the opposite direction. The transverse axis orthogonal to the channel
walls, and the origin of the axes is such that the positions of the channel walls are at
Y = -L/2 and Y = L/2, respectively. The schematic diagram of the present problem is
depicted in Figure 1. The thermal conductivity, the dynamic viscosity and the thermal
expansion coefficient are assumed to be constant. The Oberbeck-Boussinesq
approximation is valid for the present case.
  0  1    T  T0   (1)
The stream-wise and the transverse momentum balance equation using Brinkman
model yields (Barletta, 1998),
1 dP d2U 
g  T  T0    2  U  0. (2)
0 dX dY 
X
g
O Y
Y  L 2 YL 2
Figure 1. Physical configuration.
The Y-momentum balance equation can be expressed as,
P
0 (3)
Y
where, P  p  g X is the difference between the pressure and the hydrostatic pressure.
0
The walls of the channel are isothermal in nature. In particular, the temperature of the
boundary at Y = -L/2 is T1, and the temperature at Y = L/2 is T2 with T2 T1. These
boundary conditions are compatible with equation (2) if and only if dP is independent of
dX
X. This results in the constant A, with
dP
A (4)
dX
On differentiating (2) and also, by using equation (3), we get,
dT
0 (5)
dX
By taking into account, the effect of viscous dissipation and radiation effect the
energy balance equation transforms into,
d 2 T   dU  
2
1 dq R
      U2  0 (6)
dY Cp  dY  0 Cp dY 0 Cp 
2
Equations (2) and (6) give a differential equation for U as,
d 4 U g  dU  g dq R 1 d 2 U g 2
2
      U 0 (7)
dY 4 Cp  dY  0 Cp dY  dY 2 
The relevant boundary conditions on U with no slip are,
 L L
U   U   0 (8a)
 2 2
The thermal boundary conditions are obtained using equations (2) and (4) as follows,
d2 U A  g  T1  T0 
  (8b)
dY 2 y L/2  
d2 U A  g  T2  T0 
  (8c)
dY 2 yL/2  
The transformed equation with the help of dimensionless parameters (6)-(8c) is

given by,
U T  T0 Y g   T D3 Gr T  T1
u ;  ; y ; Gr  ;  ; RT  2 ;
U0 T D  2
Re T
U0 D  U02 CD2 D2
Re  ; Br  , E 
2
,  
2
, (9)
 KT K 
where, D = 2L is the hydraulic diameter. The reference velocity U 0 and the reference
temperature T0 are given by,
AD2 T1  T2
U0   ; T0  . (10)
48 2
The reference temperature difference T is given by,
T  T2  T1 if T1  T2 (11)
2
T  if T1  T2 . (12)
Cp D 2
The dimensionless parameter R T is zero for symmetric heating (T1  T2 ) and is one
for asymmetric heating (T1  T2 ) .
Using equations, (9) and (10), we get
 du 
2
d2 
 Br    E 2   2 Br u 2  0 (13)
 dy 
2
dy
 du 
2
d4u d2u
   2
 E 2
   Br     Br2 u 2  2 E 2 u  48E 2  0 (14)
 dy 
4 2
dy dy
and the imposed boundary conditions (6) – (8c) yields,
 1 1 d2u R T d2u RT 

u   u   0   48  ,  48  (15)
 4 4 dy 2 y
1 2 dy 2 y
1 2
4 4
On substituting the equations (9) and (10) in equation (2), we obtain the temperature
field as,
1 d2u 
    48  2 u  2  (16)
 dy 
SOLUTIONS
Special Cases
If the viscous dissipation is negligible, then Br  0 , the dimensionless temperature θ

and dimensionless velocity u are uncoupled. In this case, the solution of equation (14) is
obtained using the boundary conditions (15) and the velocity field transforms into,
48  cosh(y)  R T   sinh(Ey) sinh(y) 

u 2 
1     
  cosh( / 4)  2    E   sinh(E / 4) sinh( / 4) 
2 2
(17)
R T sinh(Ey)
 (17a)
2 sinh(E / 4)
If the radiation parameter is not considered i.e.,   0 , the velocity field becomes,
3 2R T   sinh(Ey) 
u  24y 2  y  (18)
2 E 
2
4sinh(E / 4) 
The temperature profile remains same as equation (17a) which corresponds to the
velocity and temperature profiles determined by Prasad, Mallikarjun and Vaidya [32].
If E  0 , the velocity field becomes,
R y  1 
u  T  24    y 2  (19)
 3   16 
Temperature equation is given by,
  2R T y (19a)
which corresponds to the velocity and temperature profiles determined by Aung and
Worku [1]. In the case of asymmetric heating, when buoyancy forces are dominated i.e.,
when     , equations (17), (18) and (19) gives,
u 1  sinh(Ey) sinh(y) 
    (20)
 2    E   sinh(E / 4) sinh( / 4) 
2 2
u 1  sinh(Ey) 
 2 y  (20a)
 E  2sinh(E / 4) 
u y1 2
  y  (21)
 3  16 
which is Batchelor’s [33] velocity profile for free convection. When buoyancy forces are
negligible and viscous dissipation is relevant, i.e.,   0 , so that a purely forced
convection occurs, solutions of equations (14) and (13) becomes
3
u  24y 2 (22)
2
 b1 cosh(Ey)  b2 sinh(Ey)  l1 cosh(2y)  l2 cosh(y)  l3 (23)
2304Br 4608 2304Br

l1  l2  2 2 l 
  4  E  cosh ( / 4) ;
2 2 2 2
    E 2  cosh( / 4) ; 3 2 E 2
1 R
b1   l1 cosh( / 4)  l2 cosh( / 4)  l3  b2  2sinh(Et / 4)
cosh(E / 4) ;
Solutions of equations (14) and (13) for clear viscous fluid in the absence of
buoyancy force and radiation effect leads to the Hagen-Poiseuille velocity profile.
1 
u  24   y 2  (24)
 16 
and
3Br
  192 Br y4  2 R T y  (25)
4 ,
which agree with the results obtained by Cheng and Wu [34] in the case of forced
convection with asymmetric heating.
DIFFERENTIAL TRANSFORM METHOD
The differential transformation method, provides a simple approach to derive high

order Taylor’s coefficient by iterative procedures. The kth differential transformation of
an analytical function u  y  is defined as,
1  d u  y 
k
U k     , (26)
k !  dy k  y  y
0
and the inverse differential transformation is given by,


u  y   U  k  y  y0  ,
k
(27)
k 0
Combining Eqs. (26) and (27), we obtain
 y  y0  d ku  y 
k

u  y   , (28)
k 0 k! dy k y  y0
The DTM function, u  y  is expressed by a finite series as,
N
u  y   U  k  y  y0  ,
k
(29)
k 0
where, the value of N depends on the convergence requirement in real applications.
FORMULATION OF THE DTM
The transformed differential equations of Eq. (14) are
U  k  4 
1
 k  1 k  2  k  3 k  4 
 ( 2
 E 2 )  k  1 k  2 U  k  2 
(30)
  Br   k  r  k  r  1U  k  r U  k  r  1  48E 2   k  
The corresponding initial conditions are given by,
U  0   a1 , U 1  a2 , U  2   a3 , U  3  a4
where a1 , a2 , a3 and a4 are constants and are evaluated using the condition (15).
PERTURBATION METHOD
Equation (14), along with the boundary conditions (15) is nonlinear in nature and
hence it is difficult to find the closed form solutions. We employ a perturbation series
method by using the dimensionless parameter,
g D
  Br   Re Pr (31)
Cp
as the perturbation parameter. The solution of equation (14) can be expressed by the
perturbation series method and is given by,
u  y   u 0  y    u1  y    2 u 2  y      nun  y (32)
n 0
The second and higher order terms of  gives a correction to u 0 , u1 accounting for the
viscous dissipation effects.
Isothermal-Isothermal Walls (T1  T2 )
Substituting equation (32) in equations (14), (15) and equating the coefficients of like
powers of  on both side, we obtains the boundary value problems for n  0 and n  1
as,
d4u0 d2u0
   2
 E 2
  E 2 2 u 0  48E 2 (33)
dy 4 dy 2
d 2 u o  du 0 
2
d 4 u1
   2
 E 2
     u 0   E u1
2 2 2 2
(34)
dy 4 dy 2  dy 
 1 1 d2u0  RT
u0     u0    0 ,  48  ,
 4 4 2
dy y   1 2
4
d2u0  RT
 48  (35)
dy 2 y
1 2
4
 1 1 d 2 u1 d 2 u1
u1     u1    0 , , 0 (36)
 4  4 dy 2 y 
1 dy 2 y
1
4 4
Equation (33) is ordinary linear differential equation and its exact solution can be
found. This solution obviously coincides with the solution of equation (14) in case of
Br  0 . The solution of equation (33) subjected to the boundary conditions (35) is given
by,
48
u 0  c1 cosh y  c2 sinh y  c3 cosh Ey  c4 sinh Ey  (37)
2
Equation (34) can also be solved exactly by using above equation (37) and the
relevant boundary conditions are given by equations (35) and the solution is,
u1  c5 cosh y  c6 sinh y  c7 cosh Ey  c8 sinh Ey  l 4 cosh 2y  l5 sinh 2y  l6 cosh 2Ey
 l7 sinh 2Ey  l8 cosh    E  y  l9 cosh    E  y  l10 sinh    E  y  l11 sinh    E  y (38)
 l12 sinh y  l13 cosh y  l14 sinh Ey  l15 cosh Ey  l13
Evaluation of exact solution for n  2 needs higher order computations and hence
neglecting the terms for n  2 the solution is obtained up to O  as,
1
 
u  u 0   u1 . (39)
The dimensionless temperature field is obtained from equation (16) considering

velocity field defined as in equations (37) and (38) as,
 E 2  2  c3   c7  cosh Ey   E 2   2  c 4   c8  sin Ey  
 
1  3  l1 cosh 2y  l 2 sinh 2y   (4E   )(l3 cosh 2Ey  l 4 sinh 2Ey) 
2 2 2
(40)
   
  (E 2  2E)(l5 cosh(  E)y  l7 sinh(  E)y  (E 2  2E)(l 6 cosh(  E)y 
 l8 sinh(  E)y)  (E 2   2 )(l11 sinh Ey  l12 cosh Ey)   2l13 
  
Isoflux-Isothermal (q1  T2 ) Walls
For this case, the thermal boundary conditions for the channel walls can be written in
the dimensional form as follows,
dT
q1   K (41)
dY Y  L
2
L
T    T2 (42)
2
The dimensionless form of above equations can be obtained by using the equation (9)
with T  q1D / K to gives rise to,
d
1
 1 (43)
dy y 
4
1
    R qt (44)
4
where, R qt  (T2  T0 ) / T is the thermal ratio parameter for differentiating Eq. (2) with
respect to Y with dP / dX  A gives,
d 3 U  g dT
 0 (45)
dY3  dY
Eq. (46) is non-dimensional and is analyzed by using Eq. (9) to obtain,
d3u d
3
 0 (46)
dy dy
Evaluating Eq. (47) at the left wall (y  1/ 4) yields,
d3u  1 
    (47)
dy3  4 
The other boundary condition at the right wall can be shown to be the same as that
given for the isothermal-isothermal wall with R T replaced by R qt such that,
d2u  1   R qt
2    48  (48)
dy  4  2
Isothermal-Isoflux (T1  q 2 ) Walls
For this case, the thermal boundary conditions for the channel walls can be written in
the dimensional form as,
dT
q 2  K (49)
dY L
2
 L
T     T1 (50)
 2
The dimensionless form of above equations can be obtained by using the equation (9)
with T  q 2 D / K to gives us,
d
 1 (51)
dy 1
4
 1
     R tq (52)
 4
where, R tq  (T1  T0 ) / T is the thermal ratio parameter. Similar to the procedure done
in the previous section on isoflux-isothermal walls, the dimensionless form of the
boundary conditions obtained from eqn. (2) and by using Eqn. (51a) as follows,
d3u  1 
  (53)
dy3  4 
The other boundary condition at the right wall can be shown to be the same as that
given for the isothermal-isothermal wall with R T replaced by R tq such that
d2u  1   R tq
2 
   48  (54)
dy  4  2
In this chapter, we investigate analytically, the effect of radiation on mixed

convective flow and heat transfer characteristics associated with the steady, laminar
incompressible viscous flow between two vertical parallel plates. The viscous dissipation
term is included in the energy equation. The nonlinear ordinary differential Eq. (14) with
boundary conditions (15) is solved analytically using the Differential Transform Method
and regular perturbation method. The analytical solutions are found using the regular
perturbation method with the product of mixed convection parameter    Gr Re  and
Brinkman number Br as the perturbation parameter. The perturbation solutions are valid
only for small values of the perturbation parameter    1 . As there are no exact
solutions for nonlinear equations, a comparison of analytical solutions is conducted and
also compared the results with the available data.
The objective is to develop a better understanding of the effects of radiation on
velocity and temperature field between the two vertical plates for a range of the mixed
convective parameter  , the radiation parameter E , the perturbation parameter  , and
the Brinkman number Br . The parameters  , Br , E and  are fixed as 500 , 0.1,
5.0, and 0.1 respectively except the varying one. The flow field in the case of
asymmetric heating  RT  1 and symmetric heating  RT  0  for cases with
isothermal-isothermal, isoflux-isothermal, and isothermal-isoflux are obtained and
depicted in Figures 2 to 11.
Plots of u versus y for asymmetric heating ( RT  1 ) conditions for   500
(both assisting and opposing flows) and   0,  8,  12 are shown in Figure 2. It can
be observed that the velocity field increases for increasing the values of  for assisting
flow and reversal effect is observed on opposing flow. It can also be seen that for large
values of  , flow reversal occurs both at the cold and hot walls. The effect of  and 
on temperature is shown in Figure 3. As  increases the temperature increases for both
assisting and opposing flow also, as we expected in the mixed convection when the value
of   0 the profiles are linear due to the presence of radiation in the fluid the nature of
the temperature profiles are not linear.
The effects of radiation parameter E on the velocity and temperature fields are
present in Figures 4 and 5 respectively. As the radiation parameter increases the velocity
increases in the left half of the region and decreases in the right half of the region for
assisting flow whereas the reversal effect is observed on the opposing flow as seen in
Figure 4. That is, for opposing flow velocity decreases in the left half of the region and
increases in the right half of the region as the radiation parameter increases. The variation
of radiation parameter E on the temperature field is observed in the Figure 5. The
radiation parameter increases the temperature in the left half of the region and decreases
in right half of the region.
Figures 6 and 7 depict the variation of radiation parameter on the velocity and
temperature fields respectively for the isothermal-isothermal symmetric wall boundary
conditions. It is can be observed from the Figure 6 that, the effect of radiation parameter
on the velocity field is invariant forboth assisting and opposing flows. The effect of
radiation parameter is to decreases the temperature field and it is observed that the effect
is similar for both assisting and opposing flows.
Figures 8 to 11 display the effect of radiation parameter for assisting and opposing
flow for isoflux-isothermal and isothermal-isoflux wall conditions for asymmetric
heating. The velocity increases for opposing flow and decreases for assisting flow as the
radiation parameter increases for the cases of isoflux-isothermal wall condition as seen in
Figure 8. As the radiation parameter increases the temperature decreases for both
assisting and opposing flows as seen in Figure 9. It is also demonstrates the effect is
highly significant near the cold wall for isoflux-isothermal wall condition. The effect of
radiation parameter on velocity and temperature for isothermal-isoflux wall condition are
depicted in Figures 10 and 11 respectively. The effect of radiation parameter velocity is
opposite to that of isoflux-isothermal wall conditions. The temperature increases for both
assisting and opposing flows but the magnitude of promotion is significant for the
opposing flow as seen in Figure 11. It can also be observed that the effect is highly
significant near the hot wall.
 RT = 1
2.5
E=5

2.0 
RT = 1
E=5
1.5 
u
1.0
0.5
0.0
-0.5
-0.2 -0.1 0.0 0.1 0.2
y
Fig. 2. Velocity profiles for different values of and 
Figure 2. Velocity profiles for different values of  and .
0.4 RT = 1
E=5
 = -500

0.2  = 500
 0.0 
-0.2
-0.4
-0.2 -0.1 0.0 0.1 0.2

y
Fig. 3. Temperature profiles for different values of and 
Figure 3. Temperature profiles for different values of  and .
2.5
 = 0.1, 
RT = 1
 = -0.1, 
2.0
E = 1,5,10
1.5
 1.0 E = 1,5,10
0.5
0.0
-0.5
-0.2 -0.1 0.0 0.1 0.2
y
Fig. 4. Velocity profiles for different values of radiation parameter E.
Figure 4. Velocity profiles for different values of radiation parameter E.
 = 0.1, 
0.4
 = -0.1, 
RT = 1
E = 1,5,10
0.2
 0.0
-0.2
E = 1,5,10
-0.4
-0.2 -0.1 0.0 0.1 0.2

y
Fig. 5. Temperature profiles for different values of radiation parameter E.
Figure 5. Temperature profiles for different values of radiation parameter E.
1.6
E = 1,5,10
1.4
1.2
1.0
 0.8
RT = 0
0.6 = -0.1, 
= 0.1, 
0.4
0.2
0.0
-0.2 -0.1 0.0 0.1 0.2
y
Figure 6. Velocity profiles for different values of radiation parameter E for RT = 0.

0.16
0.14
RT = 0 E=1
0.12 = -0.1, 
= 0.1, 
0.10
E=5
-3
x 10
0.08
0.06
0.04 E = 10
0.02
0.00
-0.2 -0.1 0.0 0.1 0.2

y
Figure 7. Temperature profiles for different values of radiation parameter E for RT = 0.
14 Rqt= 1 E=1
12 =0.1, 
=-0.1,  E=2
10
E=3
8
6
4
u 2
0
-2
-4 E=3
-6
E=2
-8
E=1
-10
-0.2 -0.1 0.0 0.1 0.2

y
Figure 8. Velocity profiles for different values of radiation parameter E of isoflux-isothermal.

1.0
Rqt = 1
=0.1, 
0.9 =-0.1, 
0.8
 0.7
E=1,2,3
0.6
0.5
0.4
-0.2 -0.1 0.0 0.1 0.2

y
Figure 9. Temperature profiles for different values of radiation parameter E of isoflux-isothermal.
14 Rqt= 1
12 =0.1,  E=1
=-0.1,  E=2
10
8 E=3
6
4
u 2
0
-2
-4 E=3
-6 E=2
-8
-10 E=1
-0.2 -0.1 0.0 0.1 0.2

y
Figure 10. Velocity profiles for different values of radiation parameter E of isothermal- isoflux.
-0.4
-0.5
-0.6 E=1,2,3

-0.7
Rqt = 1
=0.1, 
=-0.1, 
-0.8
-0.9
-0.2 -0.1 0.0 0.1 0.2
y
Figure 11. Velocity profiles for different values of radiation parameter E of isothermal- isoflux.
CONCLUSION
In the present study, the DTM is used to find analytical solutions of mixed
convection problem in the presence of radiation effect and viscous dissipation. It was
found that DTM is a powerful method for solving problems consisting of systems of
nonlinear differential equations. In this paper, a reliable algorithm is presented based on
the DTM to solve highly nonlinear equations. A comparison was made between
perturbation method solutions, available published results and the present approximate
solutions. The numerical results indicate that the DTM gives more accurate in
comparison to shown methods. The method has been applied directly without requiring
linearization, discretization, or perturbation. The obtained results certify the reliability of
the algorithm and give it a wider applicability to nonlinear differential equations. Also
some major observations are: the velocity and temperature increases for mixed
convection parameter and perturbation parameter whereas it decreases for increasing the
radiation parameter.
REFERENCES
[1] Aung, W. and Worku, G. “Theory of Fully Developed, Combined Convection

Including Flow Reversal,” ASME J of Heat Transfer, Vol. 108, pp. 485-488 (1986).
[2] Cheng, C. H., Kou, H. S. and Huang, W. H. “Analysis on the Flow Reversal and
Heat Transfer of Fully Developed Mixed Convection in Vertical Channels,” AIAA
Journal of Thermophysics and Heat Transfer, Vol. 4, pp. 375-383 (1990).
[3] Zanchini, E. “Effect of Viscous Dissipation on Mixed Convection in a Vertical
Channel with Boundary Conditions of the Third Kind,” Int J Heat Mass Transfer,
Vol. 41, pp. 3949-3959 (1998).
[4] Barletta, A. “Laminar Mixed Convection with Viscous Dissipation in a Vertical
Channel,” Int J Heat Mass Transfer, Vol. 41, pp. 35013513 (1998).
[5] Boulama, K., and Galanis, N. “Analytical Solution for Fully Developed Mixed
Convection between Parallel Vertical Plates with Heat and Mass Transfer,” J Heat
Transfer, Vol. 126, pp. 381388 (2004).
[6] Barletta, A., Magyari, E. and Keller, B. “Dual Mixed Convection Flows in a
Vertical Channel,” Int J Heat Mass Transfer, Vol. 48, pp. 48354845 (2005).
[7] J. Prathap Kumar, J. C. Umavathi and Basavaraj M. Biradar, “Mixed convection of
composite porous medium in a vertical channel with asymmetric wall heating
conditions,” J. Porous Media, vol. 13, No. 3, pp. 271-285, 2010
[8] J. C. Umavathi and M. Shekar. “Mixed convective flow of two immiscible viscous
fluids in a vertical wavy channel with traveling thermal waves”, Heat Transfer
Asian Research, vol. 40(8), pp. 721-743, 2011.
[9] J. C. Umavathi, I. C. Liu, and M. Shekar. “Unsteady mixed convective heat transfer
of two immiscible fluids confined between a long vertical wavy wall and a parallel
flat wall,” Appl. Math. Mech.-Engl. Ed., vol. 33(7), pp. 931–950, 2012.
[10] Siegel, R. and Howell, J. R. Thermal Radiation Heat Transfer, International
Student Edition, New York: McGraw-Hill (1972).
[11] Chamkha, A. J. “Solar Radiation Assisted Natural Convection in a Uniform Porous
Medium Supported by a Vertical Flat Plate,” ASME J Heat Transfer, Vol. 119, pp.
8996 (1997).
[12] Raptis, A. “Radiation and Free Convection Flow through a Porous Medium,” Int
Commun. Heat Mass Transfer, Vol. 25, pp. 289295 (1998).
[13] Raptis, A. “Radiation and Flow through a Porous Medium,” J Porous Media, Vol.
4, pp. 271273 (2001).
[14] Bakier, A. Y. “Thermal Radiation Effects on Mixed Convection fromVertical
Surfaces in Saturated Porous Media,” Int Comm Heat and Mass Transfer, Vol. 28,
pp. 243248 (2001).
[15] Raptis, A., and Perdikis, C. “Unsteady Flow through a Highly Porous Medium in
the Presence of Radiation,” Transport Porous Media, Vol. 57, pp. 171179 (2004).
[16] Bég, O. A., Zeuco, J., Takhar, H. S., and Bég, T. A. “Network Numerical
Simulation of Impulsively-Started Transient Radiation-Convection Heat and Mass
Transfer in a Saturated Darcy-Forchheimer Porous Medium,” Non Linear Analysis:
Modelling and Control, Vol. 13, pp. 281303 (2008).
[17] Ghosh, S. K., and Bég, O. A. “Theoretical Analysis of Radiative Effects on

Transient Free Convection Heat Transfer Past a Hot Vertical Surface in Porous
Media,” Non Linear Analysis: Modelling and Control, Vol. 13, pp. 419432 (2008).
[18] J. C. Umavathi, A. J. Chamkha, M. H. Manjula and A. Al-Mudhaf. “Radiative heat
transfer of a two-fluid flow in a vertical porous stratum.” Int. J. Fluid Mechanics
Research, vol. 35, pp. 510-543, 2008.
[19] Mahesha Narayana, Ahmed A. Khidir, Precious Sibanda, P. V. S. N. Murthy.
Viscous dissipation and thermal radiation effects on mixed convection from a
vertical plate in a non-Darcy porous medium, Transp Porous Med (2013) 96:419–
428.
[20] J. K. Zhou. Differential Transform and its Applications for Electrical Circuits,
Wuhan, Huarjung University Press, 1986.
[21] F. Ayaz. Solutions of the system of differential equations by differential transform
method, Applied Mathematics and Computation 147 (2004) 547–567.
[22] I. H. Abdel-Halim Hassan. Application to differential transformation method for
solving systems of differential equations, Applied Mathematical Modelling 32
(2008) 2552–2559.
[23] S. S. Chen, C. K. Chen. Application of the differential transformation method to the
free vibrations of strongly non-linear oscillators, Nonlinear Analysis: Real World
Applications 10 (2009) 881–888.
[24] Arikoglu, I. Ozkol. Vibration analysis of composite sandwich beams with
viscoelastic core by using differential transform method, Composite Structures 92
(2010) 3031–3039.
[25] C. W. Bert. Application of differential transform method to heat conduction in
tapered fins, ASME Journal of Heat Transfer 124 (2002) 208–209.
[26] H. P. Chu, C. L. Chen. Hybrid differential transform and finite difference method to
solve the nonlinear heat conduction problem, Communications in Nonlinear
Science and Numerical Simulation 13 (2008) 1605–1614.
[27] H. P. Chu, C. Y. Lo. Application of the hybrid differential transform-finite
difference method to nonlinear transient heat conduction problems, Numerical Heat
Transfer Part A 53 (2008) 295–307.
[28] A. A. Joneidi, D. D. Ganji, M. Babaelahi. Differential transformation method to
determine fin efficiency of convective straight fins with temperature dependent
thermal conductivity, International Communications in Heat and Mass Transfer 36
(2009) 757–762.
[29] M. J. Jang, Y. L. Yeh, C. L. Chen, W. C. Yeh. Differential transformation approach
to thermal conductive problems with discontinuous boundary condition, Applied
Mathematics and Computation 216 (2010) 2339–2350.
[30] M. M. Rashidi, N. Laraqi, S. M. Sadri. A novel analytical solution of mixed

convection about an inclined flat plate embedded in a porous medium using the
DTM-Padé, International Journal of Thermal Sciences 49 (2010) 2405–2412.
[31] Hessameddin Yaghoobi, Mohsen Torabi, The application of differential
transformation method to nonlinear equations arising in heat transfer, International
Communications in Heat and Mass Transfer 38 (2011) 815–820.
[32] K. V. Prasad, P. Mallikarjun and H. Vaidya. Mixed convective fully developed
flow in a vertical channel in the presence of thermal radiation and viscous
dissipation, Int. J. Applied Mechanics and Engineering, 22 (2017) 123-144.
[33] Batchlor G. K. Heat transfer by free convection across a closed cavity between
vertical boundaries at different temperatures, Quarterly of Applied Mathematics, 12
(1954) 209-233.
[34] K. C. Cheng and R. S. Wu. Viscous dissipation effect on convective instability and
heat transfer in plane Poiseulle flow heated from below, Applied Scientific
Research 32 (1976) 327-346.
ABOUT THE EDITOR
María Dolores Torres Pérez

Senior Researcher
Department of Chemical Engineering and Biotechnology,
University of Cambridge, UK
Dr. Torres was awarded her PhD in the Department of Chemical Engineering at the
Universidade de Santiago de Compostela (Spain) working in the field of the thermo-
rheological characterization of gluten-free products. She also has developed her research
activity in other prestigious centers such as University of Cambridge, UK; INRA, France;
ISA, IPC and FEUP, Portugal; CSIC-IATA, Spain. As a result of her intensive and
sustained scientific activity, Dr Torres possesses a demonstrated track record of scientific
publications in reputable international journals and numerous communications to
scientific international Conferences on Chemical Engineering, Food Technology and
Rheology. She has been internationally recognized with several awards for her
contribution in rheology of gluten-free products.
INDEX
branched polymers, vi, vii, xi, 193, 194, 196, 205,

#
207, 213, 215, 216
branching, xi, 193, 194, 200, 201, 202, 203, 204, 216
3D printer, 1, 3, 7
bread, 8, 64, 100, 105, 106, 107, 108, 299
brewers spent grain, 105, 113
A
acrylonitrile (ACN), v, x, 119, 121, 123, 138, 171, C

200, 201, 202, 203, 214, 215, 216
cheese(s), xiv, 3, 64, 66, 76, 89, 90, 91, 92, 109, 110,
Altan et al. 1979, 383
111, 117, 263, 264, 285, 287, 305, 309, 343
angular oscillations, 364
coefficients, 30, 32, 153, 154, 171, 185, 235, 239,
ARES, 197, 198
249, 260, 340, 341, 342, 348, 349, 350, 356, 366,
arrhenius model, xii, 219, 220, 235, 236, 250, 251,
368, 373, 386, 418
252
conservation of mass, 352, 384, 398, 399
artificial polymers, 350
consolidation, viii, 13, 16, 17, 18, 21, 22, 26, 37, 38,
39, 40
B constitutive equation, 46, 349, 350, 352, 353, 363
continuity, 372, 373, 374
B spline, 372 continuous, xi, 11, 68, 72, 73, 78, 81, 89, 106, 121,
baby foods, vii, xii, 219, 220, 221, 222, 223, 226, 146, 175, 177, 180, 191, 243, 271, 278, 297, 301,
230, 231, 232, 235, 237, 239, 243, 245, 247, 248, 336, 341, 343, 350, 366, 372, 393
249, 250, 251, 253, 255, 257, 258, 259, 260, 261, continuum mechanics theory, 352
262, 263, 264, 265, 266, 267, 268, 269, 270, 271, control points, 373, 374, 377
272, 273, 274, 275, 276, 277, 279, 280, 281, 282, cookies, 64, 92, 93, 94, 95, 96, 97, 98, 109, 110, 112,
283, 284 114, 116
balance of momentum, 352 Cox-Merz rule, 209
biconical rotor, 198, 199, 206 Crank-Nicholson, 360
botanical origin, 176, 177, 182, 189, 190 Crankshaft vibration dampers, 364
boundary conditions, xv, 310, 370, 384, 385, 386, creep/recovery, 197
387, 400, 401, 402, 409, 410, 411, 412, 413, 414, cylindrical coordinates, 363
417, 419, 421, 422, 423
boundary element method, 374
436 Index
D F
damping, 8, 203, 204, 206, 348, 349, 353, 357, 359, fat mimetic, 73, 115
360, 364, 365, 371, 379, 380 finite difference, xiv, 348, 350, 359, 360, 364, 370,
damping coefficient, 349, 350, 365, 371 411, 430
deformation, xi, xii, 15, 45, 95, 104, 175, 180, 186, finite element, xiv, 348, 349, 350, 359, 365, 366,
187, 189, 194, 202, 207, 208, 212, 213, 216, 220, 370, 371, 372, 375, 377, 378, 379
221, 227, 253, 254, 283, 290, 293, 295, 297, 298, finite element analysis, 348
299, 348, 350, 351, 352, 353, 354, 360, 365, 367, flow behavior, xii, 5, 39, 76, 181, 219, 220, 222, 225,
392, 407 226, 229, 231, 232, 235, 239, 251, 292, 296, 306
deformation energy, 187, 221, 367 flow curves, xi, 4, 30, 32, 71, 175, 176, 177, 180,
denaturation, 69, 70, 71, 79, 80, 114, 249, 257, 297 181, 183, 198, 224, 225, 231, 233, 234, 237, 238,
design process, 364 292
desserts, 80, 81, 86, 88, 108, 114, 117, 284 fluid flow due to, xiv, 381, 383, 398
dietary fibre, 86, 94, 111, 114 food development, vii, 67, 289
differential transform method, vii, xv, 410, 430 food structure, ix, 63, 64, 65, 80, 94, 289, 290
discretization points, 372 food-rheology, 64
displacement, 72, 295, 350, 352, 353, 355, 356, 359, form matrix, 366
360, 361, 362, 365, 366, 368, 369, 372, 375, 376 formulation, x, xiv, 6, 36, 64, 65, 66, 79, 88, 90, 93,
displacement function, 366, 375, 376 95, 98, 105, 116, 120, 220, 222, 232, 263, 289,
Doi-Edwards theory, 194 290, 291, 294, 299, 347, 349, 350, 352, 366, 377,
dough extensibility, 107 383, 417
dough resistance, 107 free vibration, 355, 356, 430
droplet size distributions, 70 frequency sweeps, 88, 90, 96, 197, 226
droplets, 68, 70, 72, 83, 275, 279, 302 friction, viii, 28, 39, 43, 44, 53, 54, 55, 56, 57, 58,
dynamic model, xiv, 347, 355, 358, 367, 374 59, 61, 122, 123, 194, 220, 349, 379
dynamic response, 359, 360
G
E
gelation, 5, 77, 80, 82, 110, 113, 114, 116, 285, 287,
effective viscosity, viii, 43, 44, 47, 48, 49, 50, 55 309, 310, 311, 313, 330, 331, 332, 333, 334, 335,
elastic beam, 359 336, 337, 338, 339, 340, 341, 342, 344, 345, 346
elastic materials, xiv, 291, 347, 352 gels, ix, xii, 10, 11, 63, 64, 75, 76, 77, 78, 79, 80, 81,
elastomer dampers, 364 82, 83, 84, 85, 86, 87, 88, 90, 102, 104, 109, 110,
elastomers, xi, 193, 194, 196, 214, 215, 352, 356, 111, 112, 113, 114, 115, 116, 219, 221, 253, 257,
377 262, 264, 282, 284, 285, 286, 287, 300, 304, 310,
element matrices, 367, 370 344, 345, 346
elements, 16, 17, 22, 279, 299, 314, 320, 329, 333, glass transition, x, 93, 119, 124, 126, 132, 134, 135,
334, 339, 350, 366, 368, 370, 372 136, 138, 140, 143, 148, 154, 155, 157, 171, 205,
emulsion polymerization, 200, 204 297
emulsion-polymerized elastomers, xii, 194 grouts, v, vii, viii, 13, 14, 17, 18, 19, 21, 22, 23, 25,
emulsions, ix, 37, 63, 64, 67, 68, 69, 70, 71, 72, 73, 26, 27, 28, 30, 32, 36, 37, 38, 40
74, 75, 76, 111, 112, 113, 115, 116, 285, 293, Guyan, 370
294, 296, 302, 304, 305, 306
entangled polymer, 194, 212
H
ethylene acrylic, 356
Euler-Bernoulli, 359
high pressure pipeline rheological test, 44
excitation frequency, 349, 357, 358, 378
Index 437
honey, v, vii, xi, 175, 176, 177, 178, 180, 181, 182, Maxwell material, 349
185, 188, 189, 190, 191, 273, 286 mayonnaise, 68, 284, 302, 304, 305
Hook’s law, 352 MCR501, 197
mechanical features, 1, 3
mechanical spectra, 69, 70, 74, 78, 91, 98, 226, 254,
I
255, 256, 261, 262, 276
MHD flow, 388, 392, 393, 396, 402, 407, 408
ill conditioned, 370
microalgae, ix, 63, 72, 73, 74, 82, 83, 84, 85, 90, 94,
ill-condition, 366, 370
97, 98, 99, 100, 101, 102, 108, 109, 112
inclined angle in degrees, 382
microstructure, 5, 9, 18, 19, 20, 21, 47, 80, 93, 110,
injection, v, vii, viii, xv, 13, 14, 16, 17, 21, 25, 26,
114, 220, 229, 277, 279, 287, 290, 292, 294, 295,
36, 37, 39, 40, 44, 395, 396, 398, 401, 402, 406
302, 304, 345
interaction parameter, 120, 122, 123, 124, 130, 131,
milk, xiv, 8, 83, 89, 90, 110, 114, 223, 292, 304, 309,
132
330, 332, 334, 335, 337, 338, 339, 340, 341, 342,
interpolation functions, 372, 373
343, 344, 345, 346
inverse Darcy number, xv, 388, 395, 396, 401, 402,
miscibility, vii, x, 119, 120, 121, 123, 124, 125, 126,
406
128, 129, 130, 131, 132, 134, 136, 138, 139, 140,
iso-free volume, x, 120, 161, 162, 165, 166, 171
146, 149, 151, 169, 171, 172
isogeometric, 372, 380
mixed convection, vii, xv, 392, 410, 411, 422, 428,
429, 430, 431
K modeling, vii, xiii, xv, 10, 37, 40, 189, 268, 309,
378, 383, 392, 397
Kelvin-Voigt, 349, 350, 352, 353, 377, 378 Mooney, xi, 27, 28, 36, 39, 194, 195, 196, 198, 199,
kinetic, vi, vii, xii, xiii, 20, 81, 83, 219, 225, 242, 200, 201, 202, 205, 206, 207, 209, 216, 217
244, 245, 246, 248, 250, 251, 252, 255, 309, 311, Mooney analysis, 27, 28, 36
312, 330, 333, 334, 335, 336, 337, 339, 340, 341, Mooney Torque, xi, 194, 200, 201, 202
343, 354 Mooney viscometer, 195, 198, 199, 200
kinetic energy, 20, 255, 354 multi-modal absorber, 359
knot vector, 372, 373, 374
N
L
natural hydraulic lime, viii, 13, 14, 18, 22, 23, 36
Lagrange, 353, 367, 377 natural rubbers, 350
Laplace transforms, 309, 315, 317, 322, 343 NBR, xi, 193, 196, 197, 199, 200
LDPE, 196, 198, 211 neoprene, 350
linear and nonlinear viscoelasiticity, 194 Newton´s viscous law, 352
linear equations, 360 Newtonian, xi, xiv, 30, 37, 44, 52, 57, 58, 71, 114,
linear viscoelasticity, 74, 109, 115, 123, 196, 197, 148, 175, 191, 231, 232, 236, 286, 292, 294, 296,
293 304, 307, 311, 345, 353, 363, 380, 381, 383, 388,
liquid CO2, v, vii, viii, 43, 44, 46, 47, 48, 50, 51, 52, 392, 397, 398, 407, 409
55, 59, 60, 61 Newtonian fluid, xiv, 44, 57, 58, 236, 294, 296, 311,
353, 363, 381, 383, 392, 397, 398, 407
M nodal displacements, 366, 369, 375
nodes, 366, 368, 370, 372
masonry, viii, 13, 14, 15, 16, 17, 21, 22, 26, 37, 38, nonlinear accelerating sheet, xiv, 381, 383, 384, 386,
39, 40 388
mass transpiration parameter, xiv, 381, 382, 383, 388 nonlinear viscoelasticity, 196, 198
Maxwell, 9, 213, 348, 349, 352, 378
438 Index
non-uniform rational B-splines (NURBS), 372, 375, rheology, v, vi, vii, viii, ix, xiii, 1, 2, 4, 6, 7, 8, 9, 11,
377 13, 18, 19, 21, 28, 37, 38, 39, 40, 41, 43, 61, 62,
numerical solution, xiv, 347, 360, 370, 378 63, 64, 65, 68, 75, 79, 80, 82, 83, 85, 86, 88, 89,
90, 93, 94, 95, 97, 98, 99, 105, 106, 108, 109,
110, 111, 112, 113, 151, 173, 175, 176, 177, 189,
O
190, 191, 216, 221, 276, 283, 284, 285, 286, 287,
288, 289, 290, 291, 292, 294, 296, 298, 302, 303,
oil resistance, 357
304, 305, 306, 307, 309, 320, 330, 334, 433
outstanding technologies, 1
rotational rheometry, vii, viii, 14, 38
rubber, xi, 193, 195, 198, 200, 216, 217, 356, 364,
P 371, 379
particle size, xiii, 7, 20, 21, 220, 222, 228, 230, 243,
274, 275, 276, 277, 282, 283 S
passive vibration absorber, 349
salts, 88, 337, 345
pasta, 3, 64, 99, 100, 102, 103, 104, 105, 109, 110,
SBR, xi, 193, 195, 196, 197, 198, 199, 204, 205,
111, 112, 113, 115, 116, 220
206, 207, 209, 210, 211, 214, 216
patches, 372
second order liquid, xiv, 381, 383
PBR, 193, 196, 199, 200, 201, 202
shear startup, vi, xii, 193, 194, 196, 199, 200, 201,
perturbation method, xv, 410, 422, 428
202, 204, 205, 206, 207, 208, 209, 210, 211, 212,
phase behavior, x, 119, 120, 121, 123, 132, 136, 149
214, 215, 216
planar elements, 368
shear stress, viii, xii, 8, 14, 28, 30, 31, 32, 45, 46, 69,
planar patch, 376
185, 194, 195, 196, 198, 199, 212, 213, 215, 219,
plane problems, 375
221, 224, 225, 226, 229, 231, 233, 237, 239, 240,
plateau modulus, x, 70, 74, 120, 151, 152, 166, 168,
241, 242, 246, 249, 250, 251, 252, 282, 291, 294,
169, 170, 171, 205
296, 297, 298, 303, 349, 353, 363
poly(-caprolactone), 119, 121, 155
shear stress growth, 194, 196, 198
poly(-methyl styrene)-co-acrylonitrile, 119
shift factor, x, xiii, 119, 143, 144, 152, 153, 154,
poly(methyl methacrylate), x, 119, 121
155, 157, 165, 171, 220, 228, 273, 274, 282
poly(styrene-co-acrylonitrile), vii, x, 119, 120, 121
Siddheshwar and Mahabaleshwar, xiv, 381, 383
poly[styrene-co-(maleic anhydride)], x, 119, 121
silicon rubbers, 350
poly[styrene-co-(N-phenylmaleimide)], x, 119, 121
similarity solution, xiv, 381, 386
polynomial coefficients, 367
slow phase separation, x, 119, 132, 171
polynomial function, 268, 366, 372
small amplitude oscillatory shear, xi, 175, 197
pompom model, 196, 211
sol-gel transition, xiv, 77, 309, 310, 312, 340, 342,
porous medium, vii, xv, 311, 396, 397, 398, 399,
343
408, 409, 410, 411, 429, 430, 431
stiffness, xiv, 9, 227, 347, 349, 357, 359, 370
potential energy, 354
strain, xi, xii, 8, 175, 179, 185, 194, 195, 197, 200,
psyllium, ix, 64, 94, 95, 96, 111, 113, 115
201, 202, 204, 208, 214, 227, 297, 345, 349, 350,
351, 352, 353, 359, 363, 367, 368, 369, 370, 376,
R 377
strain function, 352, 363, 369, 376
radiation, vi, vii, xv, 392, 393, 408, 409, 410, 411, strain rate, 345, 350, 351, 352, 353, 367, 370, 377
412, 415, 416, 422, 423, 424, 425, 426, 427, 428, stress overshoots, vi, 193, 200, 212
429, 430, 431 stretching/shrinking sheet, xv, 395, 396, 397, 398,
recovery, xi, 4, 6, 113, 175, 179, 185, 186, 187, 189, 401, 402, 406
194, 196, 197, 208, 209, 211, 212, 214, 216 surface is calleds, 383
Index 439
viscoelastic materials, vii, xiv, 187, 291, 293, 347,

T
348, 352, 379
viscoelastic model, 285, 348, 349
texture, ix, 3, 4, 63, 64, 65, 66, 67, 68, 73, 74, 75, 76,
viscoelastic properties, vii, x, xi, xii, 64, 83, 90, 91,
80, 82, 86, 88, 89, 90, 91, 93, 94, 95, 97, 98, 99,
97, 111, 119, 122, 123, 146, 148, 151, 152, 159,
100, 101, 102, 103, 104, 105, 106, 107, 110, 112,
160, 171, 175, 219, 221, 222, 264, 268, 273, 282,
115, 116, 283, 284, 285, 287, 288, 289, 299, 300,
283, 284, 285, 286, 287, 297, 355, 360, 371
301, 302, 303, 304, 305, 306, 307
viscoelasticity, vi, vii, 1, 72, 123, 184, 219, 220, 268,
thermoreversible gelation, 79, 310, 344
285, 293, 298, 301, 380
time-temperature superposition, x, xiii, 119, 140,
viscosity, vi, vii, viii, x, xi, xii, xiii, 3, 4, 5, 6, 7, 21,
146, 157, 165, 171, 208, 220, 222, 227, 285, 288
23, 28, 30, 32, 34, 35, 36, 40, 43, 44, 47, 48, 50,
Timoshenko, 349, 350, 379
52, 55, 57, 60, 64, 67, 68, 71, 72, 73, 75, 76, 77,
Torque, Mooney, xi, 194, 200, 201, 202
95, 120, 122, 143, 148, 157, 161, 166, 175, 176,
torsional damper, 364, 365
179, 180, 181, 184, 185, 186, 187, 188, 189,
torsional displacement, 365
216,219, 220, 221, 222, 225, 226, 231, 232, 234,
torsional natural frequency, 364
235, 236, 243, 244, 245, 247, 249, 251, 255, 264,
transmissibility, 348, 350, 357, 358, 364, 371
276, 282, 283, 286, 289, 290, 291, 292, 294, 295,
transmissibility function, 348, 358, 364
296, 297, 303, 304, 306, 337, 349, 382, 396, 411
travelling belt, 359, 360
viscous, x, xv, 4, 6, 19, 20, 21, 45, 55, 77, 89, 119,
travelling speeds, 360
146, 148, 180, 186, 187, 205, 220, 221, 227, 235,
triangular element, 368
238, 253, 264, 291, 292, 293, 297, 320, 349, 352,
353, 356, 363, 378, 379, 392, 398, 407, 408, 410,
U 411, 412, 414, 416, 418, 422, 428, 429, 430, 431
viscous fluids, 352, 429
unfavorable interaction, x, 119, 123, 171
W
V
wall condition, 382, 396, 423
variational principle, 375 wall slip, viii, 13, 27, 28, 30, 31, 35, 36, 38, 41, 198,
vibration absorber(s), vii, xiv, 347, 348, 349, 350, 199
352, 357, 358, 363, 366, 370, 377, 378 Walters’ liquid B, xiv, 381, 388, 395, 396, 399, 402,
vibration frequency, 350 406
vibration source, 357 water content, xii, 23, 25, 115, 176, 178, 180, 181,
viscoelastic, v, vi, vii, x, xi, xii, xiv, xv, 4, 62, 64, 65, 219, 232, 235, 236, 260, 268, 273, 283
67, 68, 69, 75, 76, 77, 78, 79, 80, 83, 85, 89, 90, waterless fracturing fluid, 44, 61
91, 93, 95, 97, 98, 111, 119, 122, 123, 125, 140, WLF equation, x, 119, 120, 143, 144, 153, 156, 157,
146, 148, 151, 152, 159, 160, 171, 175, 177, 187, 166, 171
193, 197, 199, 204, 211, 213, 219, 221, 222, 226,
227,256, 260, 264, 268, 273, 276, 282, 283, 284,
285, 286, 287, 291, 293, 295, 297, 298, 302, 304, X
310, 311, 312, 313, 317, 327, 343, 344, 345, 347,
 parameters, x, 119, 130, 138, 139, 171
348, 349, 350, 352, 355, 359, 360, 363, 371, 374,
378, 379, 380, 381, 383, 384, 386, 388, 392, 393,
395, 396, 397, 398, 399, 401, 402, 406, 407, 408, Z
430
viscoelastic absorbers, 348, 349 zero-shear viscosity, x, 120, 122, 152, 157, 158, 159,
viscoelastic fluid (Walters’ liquid B), 395, 396, 402, 160, 161, 163, 165, 169
406

Advances in Rheology

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Advances in Rheology

Hochgeladen von

Copyright:

Verfügbare Formate

PHYSICS RESEARCH AND TECHNOLOGY

ADVANCES IN RHEOLOGY RESEARCH

Additional books in this series can be found on Nova’s website

Additional e-books in this series can be found on Nova’s website

ADVANCES IN RHEOLOGY RESEARCH

MARÍA DOLORES TORRES PÉREZ

NOTICE TO THE READER

Library of Congress Cataloging-in-Publication Data

ISBN:  H%RRN

Published by Nova Science Publishers, Inc. † New York

Chapter 7 Unusual Nonlinear Rheological Behavior

contribute to decision-making on product development stages. It will be emphasized how

Chapter 6 - Honey is a complex fluid with viscoelastic properties predominantly

emulsion-polymerized elastomers, characterized by broad Molar Mass Distribution and

RHEOLOGY AND 3D PRINTING

María Dolores Torres Pérez* and Javier Seijo

Keywords: 3D printer, mechanical features, outstanding technologies, viscoelasticity

3D printing is more than just a novelty – it is the future of manufacturing industries

TYPES OF PRINTING MATERIALS: NATIVELY PRINTABLE,

CURRENT TRENDS IN THERMO-MECHANICAL CHARACTERIZATIONS

Rheological measurements, in terms of steady-shear flow curves (apparent viscosity),

examples of different experimental set-ups available for rheological measurements,

THERMORHEOLOGICAL FEATURES OF 3D PRINTABLE MATERIALS:

RHEOLOGICAL CHARACTERISATION OF INJECTION

Luis G. Baltazar1,*, Fernando M. A. Henriques1

The overall goal of this chapter is to contribute to the understanding and

Keywords: rotational rheometry, injection grouts, masonry, natural hydraulic lime

(a) (b) (c) (d)

Figure 3. Deformation and failure of a multi-leaf masonry wall.

A multi-leaf masonry wall is usually characterized by its constructive weaknesses,

Grout Injection as Consolidation Technique

Grout injection or grouting is a process of introducing a fluid mixture into the

be compatible with the in situ materials in order to ensure a successful consolidation

RHEOLOGY OF CEMENTITIOUS SUSPENSIONS

Grouts can be seen as a colloidal suspension consisting of suspended binder particles

Interactions Influencing the Rheology of Grouts

In cementitious grouts, colloidal interparticle forces and coagulation processes

Table 1. Chemical composition of natural hydraulic lime

Table 2. Physical properties of natural hydraulic lime

Label Compression resistance Setting time Specific gravity Fineness (Blaine)

Figure 5. Grain size distribution of natural hydraulic lime.

The latest generation of HRWR is based on carboxylic polymers (PC) or carboxylic

Figure 6. Representation of a suspension dispersed through coupled mechanisms exercised by

Another important effect associated to the use of HRWR is called depletion

The rheological characterization of cementitious suspensions can be challenging

Yield Stress and Plastic Viscosity

In the frame of rheology of cementitious suspension knowledge, there are different

where c is a constant with no physical meaning.

Stress controlled measurements at various gaps of 1 mm, 1.5 mm, 2 mm and 3 mm

RESULTS AND DISCUSSION

Yield Stress and Plastic Viscosity

Ways to Minimize the Wall Slip Effects

The authors decided to eliminate the slippage effects by performing rheological

Table 3. Rheological properties measured with rough and smooth plate

Measurements with rough plate Measurements with smooth plate

In this chapter, the influence of wall slip on rheological measurements of natural

Billberg, P. (2006). Form Pressure Generated by Self-Compacting Concrete — Influence

Ran, Q. et al. (2010). Adsorption Mechanism of Comb Polymer Dispersants at the

Wallevik, J. (2009). Introductions to Rheology of Fresh Concrete. Course compendium,

LIQUID CO2-BASED GEL FLUID

Weixiang Cui*, Mingyue Cui and Chunpeng Wang

fracturing fluid. Temperature on the rheological parameters influence is bigger, whether

Compared with traditional fracturing, pure liquid CO2 fracturing, a waterless

Rheological properties of non-Newtonian fluid can be measured by a variety of

ISBN: H%RRN