See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/273531832

Comparative between FEM Models for FDM Parts and their Approach to a Real
Mechanical Behaviour

Article  in  Procedia Engineering · December 2013

DOI: 10.1016/j.proeng.2013.08.230

CITATIONS READS

34 281

6 authors, including:

J. Martínez Ares Enrique

University of Vigo University of Vigo
10 PUBLICATIONS   61 CITATIONS    66 PUBLICATIONS   429 CITATIONS   

SEE PROFILE SEE PROFILE

A. Pereira Primo Hernández

University of Vigo University of Vigo
30 PUBLICATIONS   93 CITATIONS    10 PUBLICATIONS   75 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

LeanPPD View project

Medical Device Development View project

All content following this page was uploaded by A. Pereira on 20 March 2015.

The user has requested enhancement of the downloaded file.

 Available online at www.sciencedirect.com

ScienceDirect
Procedia Engineering 63 (2013) 878 – 884

The Manufacturing Engineering Society International Conference, MESIC 2013

Comparative between FEM models for FDM parts and their

approach to a real mechanical behaviour.
J. Martíneza,*, J.L. Diégueza, E. Aresb, A. Pereirab, P. Hernándezb, J.A. Pérezb
Escuela de Ingeniería Industrial, Universidad de Vigo, C/ Torrecedeira 86 36208, Vigo (Pontevedra).
Escuela de Ingeniería Industrial, Edificio de fundición, C/Maxwell, Campus Lagoas-Marcosende 36310, Vigo (Pontevedra).

Abstract

Laminated parts made of FDM (Fused Deposition Modeling) are becoming increasingly popular in last years. This increment
can be explained due to their potential in field of manufacturing in order to allow us to make any kind of geometry, but in
another way, these parts have not enough values of stiffness and strength to get an approach to a final constructive part.
However, advantages of nowadays FEM programs allow us to analyze behavior and main manufacturing properties; in order
get previously mentioned structural parts. Then it we must need to accurately predict the response of composite structures to
different load cases. In this way, the adequate modeling tools must be developed.
The motivation of this article is to contribute to this development, making numerical simulations of two different composite
structures with a ply-level approach, and analyze obtained results.

©
© 2013
2013 The
The Authors.
Authors. Published
Published by
by Elsevier
Elsevier Ltd.
Ltd.
Selection
Selection and
and peer-review
peer-review under
under responsibility
responsibility of
of Universidad
Universidad de
de Zaragoza,
Zaragoza, Dpto
Dpto Ing
Ing Diseño
Diseño yy Fabricacion.
Fabricacion

Keywords: Fused deposition modeling, Mechanical characterization, Finite elements modeling, Stiffness matrix.

1. Introduction

Previous studies were focused on building geometrical prototypes; however recent developments in the field of
rapid prototyping enabled some of the prototyping methods to achieve a lithe manufacturing technology to produce
the final product directly (Bellini A., Güçeri S. 2003). This effect allows an important reduction in terms of process

* Corresponding author. Tel.: +34 – 986 - 813648.

E-mail address: jamartinez@uvigo.es

1877-7058 © 2013 The Authors. Published by Elsevier Ltd.

Selection and peer-review under responsibility of Universidad de Zaragoza, Dpto Ing Diseño y Fabricacion
doi:10.1016/j.proeng.2013.08.230
 J. Martínez et al. / Procedia Engineering 63 (2013) 878 – 884 879

time reduction and consequently cost reduction. In the same way, solid freeform fabrication technologies, like
fused deposition modeling, used to get final parts, can be observed as a technology to reach material and energy
saving in compared with another process ¡Error! No se encuentra el origen de la referencia..
Laminated parts made of FDM (Fused Deposition Modeling) are becoming increasingly popular in the last
years. This increment can be explained due to their potential and their capability to make any kind of geometry, but
in another way, these parts have not enough values of stiffness and strength to get an approach to a final
constructive part ( Li L. et al. 2002).
However, advantages of nowadays FEM programs allow us to analyze behavior and main manufacturing
properties; in order get previously mentioned structural parts (Magalhães L.C. et al. 2010). Then it we must need to
accurately predict the response of composite structures to different load cases. In this way, the adequate modeling
tools must be developed ( Martínez J., Diéguez J. Ares J.E., et al. 2011).
The motivation of this article is to contribute to this development, making numerical simulations of two
different composite structures with a ply-level approach, and the analysis of obtained results ¡Error! No se
encuentra el origen de la referencia. et al. 2002).
Before doing this, some basics regarding to laminates should be reviewed. Non homogeneous materials are
formed by two or more dissimilar constituents (also called phases). Prediction of their behavior is one of the main
goals in the field of micromechanics science. It must be determined by each phase behavior (their relevant
properties), and also by the phase arrangement, which is the geometry and topology of each of the constituents. In
the case of a RP (Rapid Prototyping) part, the composition of each layer should be considered in a different way
than other layer composites ( Wang T.M., et al. 2007).
In typical composites, parts are formed by two constituents: matrix and inclusions. The matrix is a topologically
connected constituent, while inclusions are distributed in it. However in FDM technology, we should take into
account that there is a wire raster that should be analyzed as inclusion into a holes matrix (the holes are spaces
between wire rasters and also between layers). So in FDM parts we can consider an orthotropic case, and in our
case, continuous ABS fibers with two different directional properties ( Pilipovi A., Raos P., Šercer M. 2009).

Fig. 1.(a) Layered fiber crossed composite and (b) layered unidirectional composite like FDM parts.

Composite components are in most cases formed by layered laminates that consist in a certain number of layers,
which bond together to form the laminate. In the case that will be studied within this paper, a single ply represents a
unidirectional reinforced thin composite (Rodriguez J. F et al.2001 and Sun Q. et al. 2008). ¡Error! No se
encuentra el origen de la referencia.
This paper presents a comparative between two different FEM models of parts manufactured using fused
deposition modeling and their approach to reality. This should be a first step to observe main parameters in parts
manufacturing and mechanical failure characterization of parts depending on parameters treatment.
880 J. Martínez et al. / Procedia Engineering 63 (2013) 878 – 884

2. Experimental Procedure

In linear elasticity, the relation between stress and strain is described by Hooke’s law. It states that, for small
strains, stress is linearly proportional to strain. Thus, to determine the mechanical properties of a material,
assuming fully linear behavior, the 36 components of the compliance matrix S of equation must be known for the
matrix.

(1)

It`s usual to write Hook`s law as matrix equation:

(2)

In which, i represents normal stress in main directions and ij represents shear stress in main planes; as can be
seen in following figure.

Fig. 2 Main stresses

In the other equation term ij represents tensor shear strain and ij represents shear strain. And finally Sij
represent the terms of the compliance tensor. In following equations are shown that Sij will be defined in order to
Ei (the Young's modulus along axis), ij (Poisson’s ratio that corresponds to a contraction in direction j when an
extension is applied in direction i) and Gij (shear modulus in direction j on the plane whose normal is in direction
i).
For an orthotropic material, which is defined as a material with three mutually perpendicular planes of
symmetry, equation can be simplified in terms of the following nine unknowns:

(3)
 J. Martínez et al. / Procedia Engineering 63 (2013) 878 – 884 881

Where:

; ; ; ; ; ; (4)

In order to de¿ne the mechanical behavior of an orthotropic part, it is necessary to determine these nine
independent constants.

2.1. Tensile tests

In order to obtain the nine independent constant values of the stiffness matrix, five samples built with different
orientations have to be tested. As no special standards exist for RP parts, the samples have been prepared according
to the geometry and dimensions specified in the ISO 527-4 tensile properties of isotropic and orthotropic fiber-
reinforced plastic composites.

Fig. 3 pictures about samples manufacturing positioning in Stratasys.

The material employed was the P400-ABS which contains: 90±99% acrylonitrile butadiene styrene resin, 0±2%
mineral oils and 0±2% wax. The physical properties are: tensile strength 0.034 MPa, flexural strength 0.066 MPa,
tensile modulus 2.48 MPa, flexural modulus 2.62 MPa and 50.00% of elongation, among others. The material
filament is drawn into the FDM head from the spools by drive wheels where it is melted by the liquefier. The
temperature of the liquefier is around 270ºC for the P400 ABS Plastic. The material is extruded out of the FDM
Tips where it solidifies (in around 0.1 s), shrinks, and fuses to previously deposited material as the FDM head is
moved in the horizontal (X±Y) plane. A road width of 0.60 mm was used and a layer height of 0.25mm was
selected.
The dog-bone sample is designed using CATIAV5 and thus it is exported in an STL format. This is the ¿le
format that can be imported in Catalixtex, the program provided by Stratasys to slice the part. After slicing the
parts, oriented as shown.

2.2. Stiffness matrix

Once tests have performed, graphs and values obtained in them allow us to get engineering constants in the
three directions and equation of Hook´s law can be written as:
Equations and formulae should be typed in Mathtype, and numbered consecutively with Arabic numerals in
parentheses on the right hand side of the page (if referred to explicitly in the text). They should also be separated
from the surrounding text by one space.

(5)
882 J. Martínez et al. / Procedia Engineering 63 (2013) 878 – 884

(6)

Where:

(7)

In our case this stiffness matrix can be summarized as:

C= MPa (8)

C is the stiffness tensor for this orthotropic material.

Using this stiffness matrix as input, two different analytic models would be created using ABAQUS, a FEM
software that allow us to compare their approach to a real behaviour.

3. Results

To modeling the geometry of dog bone sample used in experimental tests, a model was designed in ABAQUS.
After geometry definition, mesh and boundary conditions, layer orientations, node and element sets were created,
and then joins them together in an ABAQUS input file. For two compared models, geometry and material
orientation were the same; boundary conditions have been adapted to simulate the same test in both cases.
Originally, this model was developed to accept 0º and 90º layer orientations; however model is easy to expand
to work with another layer orientation by changing some inputs.
Results obtained, shown that failure can be underestimated when the part is designed considering as rigid
orthotropic material. However, when model is designed as a layered part, taking into account the effect of
delamination due to stress and failure transmission between layer interfaces, it can be seen that parts reach higher
Von Mises stress values.
 J. Martínez et al. / Procedia Engineering 63 (2013) 878 – 884 883

Fig. 4 Deformation applies to a laminated sample (von Mises stress criterion).

Fig. 5 Deformation applies to a rigid solid sample (von Mises stress criterion).

As can be observed, under same charge conditions, model built by layers, reach a higher elastic values than
solid case, so it can be said that second case is more restrictive than the first. However differences between both
cases are minimal and in any case, models show a good approximation to reality.

Table 1. Results in layered model.

Stress direction Max. Stress value yield (MPa)
(MPa)
X 8.65 1,6
Y 2.65 0.96
Z 11.6 980
Max. Displacement (mm): 5.11
S under Von Misses criterion: 16.1 MPa

Table 2. Results in solid model.

Stress direction Max. Stress value yield (MPa)
(MPa)
X 2.6 1.43
Y 3.6 900
Z 15 0.40
Max. Displacement (mm): 4.5
S under Von Misses criterion: 15.06 MPa

In the same way, if we change failure criterion to Tresca, taking material as stiffness material, it can be
observed a similar result; also results under Tresca criterion are similar to von Mises in each case.
However, if deformation is observed, we can see several differences between models. This effect is due to
layers interface definition and how cohesive forces (nodes and surfaces) have been programmed.
884 J. Martínez et al. / Procedia Engineering 63 (2013) 878 – 884

4. Conclusions

When a composite laminate is made from layers, its properties are controlled by the number and the different
orientation of the plies. Depending of the final service of the laminate, its strength and stiffness values can be
configured and tailored selecting the orientation of each lamina. However, laminates are usually bonded together
with minimal reinforcement between them in thickness direction, so not only the failure in a single ply can be
considered, but also the failure of the interface between them. In the present study, some different orientation
configurations will be seen and analyzed, and both of these failure modes will be considered.
Parts manufacturing using this technology can be considered as orthotropic.
For FEM models, some deformation and stress failure differences have been found between a model made as a
solid part and the another model made taking account each layer as solid and union forces between layers, to
analyze mechanical characterization.
The type of ply failure is determined by the load case, so if the ply is loaded by tensile stresses in the wire
direction, normally the failure will be fiber dominated (rupture of the fiber).
The other main failure mechanism is failure of the interface between wires, also called delamination. This is
usually originated by out-of-plane stresses due to a transverse loading case, ply drop-off or free edge effects,
among others.
Anyway, it is important to distinguish between the failure of the first layer of the laminate and the overall (or
ultimate) failure of the structure. Once a ply fails, its stresses could be redistributed and carried by the rest of the
plies. The criterion that considers the failure of the first layer of the laminate as the total failure of the laminate is
known as First Ply Failure (FPF).
In any case, part failure is caused by failure of any constituents, and by the interface between them being the
most restrictive model the solid one.

Acknowledgements

Special acknowledgments to Construction and Manufacturing Department from UNED (Madrid), to let us using
their lab to perform tests needed in this study.

References

Ahn S.H., Montero M., Odell D., Roundy S. and Wright P.K. 2002. Anisotropic material properties of fused deposition modeling ABS, , Rapid
Prototyping Vol.8/ No. 4, pp. 248–257.
Bellini A., Güçeri S. 2003. Mechanical characterization of parts fabricated using fused deposition modeling. Rapid Prototyping Journal Vol. 9/
No 4, pp. 252-264.
Li L. et al. 2002. Composite Modeling and Analysis for Fabrication of FDM Prototypes with Locally Controlled Properties. Journal of
Manufacturing Processes Vol. 4/No. 2.
Magalhães L.C. et al. 2010. A Influência dos Parâmetros Construtivos no Comportamento Mecânico de Peças Fabricadas Pela Técnica de
Modelagem por Fusão e Deposição (FDM), Congresso Nacional de Engenharia Mecânica, Campina Grande, Paraíba.
Martínez J., Diéguez J. Ares J.E., et al. 2011. Modelization and structural analysis of FDM parts. Proc. of MESIC 2011: 4th Manufacturing
Society International Conference. Cádiz, Spain.
Pérez C.J.L., Vivancos Calvet J., Sebastian Pérez M.A. 2001. Geometric roughness analysis in solid free-form manufacturing processes. Journal
of Materials Processing Technology, Vol. 119, No. 1-3, pp. 52-57.
Pilipovi A., Raos P., Šercer M. 2009. Experimental analysis of properties of materials for rapid prototyping, International Journal of
Advanced Manufacturing Technology, Vol. 40, No. 11–12, pp. 105–115.
Rodriguez J. F., Thomas J. P., Renaud J. E. 2001. Mechanical behavior of acrylonitrile butadiene styrene (ABS) fused deposition materials,
Rapid Prototyping Journal. Vol. 7, No. 3,pp. 148 – 158.
Sun Q., Rizvi G.M., Bellehumeur G.T., Gu P. 2008. Effect of processing conditions on the bonding quality of FDM polymer filaments, Rapid
Prototyping Journal. Vol. 14 / 2, pp. 72 – 80.
Wang T.M., J Xi.T., and Jin Y. 2007. A model research for prototype warp deformation in the FDM process, International Journal of Advanced
Manufacturing Technology, Vol. 33, No. 11–12, pp. 1087–1096.

View publication stats