Development of fire resistant wool polymer composites: Mechanical perfor-
mance and fire simulation with design perspectives

N.K. Kim, D. Bhattacharyya

PII: S0264-1275(16)30756-0
DOI: doi: 10.1016/j.matdes.2016.06.005
Reference: JMADE 1880

To appear in:

Received date: 1 April 2016

Revised date: 31 May 2016
Accepted date: 1 June 2016

Please cite this article as: N.K. Kim, D. Bhattacharyya, Development of fire resistant
wool polymer composites: Mechanical performance and fire simulation with design per-
spectives, (2016), doi: 10.1016/j.matdes.2016.06.005

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Development of fire resistant wool polymer composites: mechanical performance and fire simulation with
design perspectives

N.K. Kim* and D. Bhattacharyya

Centre for Advanced Composite Materials, Department of Mechanical Engineering, The University of

T
Auckland, Auckland, New Zealand

IP
*
Corresponding author: N.K. Kim, Email: nkim048@aucklanduni.ac.nz, Ph.) +64 9 923 9093

R
Abstract

SC
Wool, as a naturally fire resistant fibre, has been incorporated with other additives, such as ammonium
polyphosphate (APP), talc and maleic anhydride grafted polypropylene (MAPP), to manufacture polypropylene
(PP) based composites. A single-screw extruder, attached with a flat die, has been used to successfully produce

NU
continuous wool-PP composite sheets. Furthermore, the significant roles of nitrogen content of APP and the talc
in forming the effective char to increase the composites’ fire retardancy have been identified through cone
calorimeter tests. In particular, it is highly interesting to note that a reduction of APP content from 20 to 15 wt%
MA
was possible to achieve a direct self-extinguishment (V-0 rating) of the composite. The decrease in APP amount
has also led to the reduction of material cost due to APP’s relatively high price (USD 13-14 / kg) and limit the
possibility of deteriorating mechanical properties due to APP addition. The MAPP has enhanced thermal
D

stability and mechanical properties compared to those without the additive, which could be attributed to the
TE

improved interfacial adhesion between wool and PP. A computational fluid dynamics model has also been
developed by Fire Dynamics Simulator to mimic the cone calorimeter tests, which agreed reasonably well with
the experiments.
P
CE

Keywords: Wool, Ammonium polyphosphate, Flammability, Mechanical properties, Fire Dynamics Simulator

1. Introduction
AC

The intensive burning behaviour of polymer composites has been a critical issue in a broad range of
applications in aircraft, automobiles, civil infrastructure and consumer products as hydrocarbon structures of
polymers have led to poor performance in fire [1]. Extensive studies have been performed over the last few
decades to comprehend and reduce the composites’ flammability. Fire retardation of polymers can be achieved
by breaking or slowing down the chain branching reactions in the combustion cycle, which includes ignition,
pyrolysis, combustion and feedback [2]. However, the resistance to combustion of the composites can be
improved by different methods, such as the incorporation of additive compounds to the polymer matrix [3],
flame retardant coating [4], heat-induced intumescence of the composites surface [5] and fibre treatments [6].
The inclusion of additives, namely fire resistant fillers or flame retardant (FR), has been widely employed as an
effective method due to their direct application during the composites’ fabrication. Flame retardant additives,
such as halogen-based materials, have been intensely utilised, but these additives release highly smoky and toxic
fire effluents by interfering with hydrocarbon oxidation and the conversion of CO into CO2 [7]. Hydrogen
chloride created from chlorine-based FR is of even greater toxicological risk than carbon monoxide [3]. Thus, it

1
 ACCEPTED MANUSCRIPT

is necessary to develop a fire retardant system for polymeric composites which is simultaneously innocuous to
the environment and can be integrated in the present system of manufacturing.
A halogen-free intumescent FR, ammonium polyphosphate (APP), mainly contain phosphorus and
nitrogen, which can improve fire retardancy for polymers under combustion [8]. APP can derive phosphorus-
nitrogen synergism under combustion, producing phosphoric acids and ammonia to form the intumescent char

T
as a barrier, consequently hindering heat and gas transfer between the flame and underlying material [8, 9].

IP
Furthermore, the amount of nitrogen or phosphorus as the main component in APP plays a significant role in
forming the effective char in the condensed phase [10]. Wool is a naturally fire resistant fibre because of its

R
relatively high contents of sulphur (3-4 wt%) and nitrogen (15-16 wt%) [11]. It has been observed that wool

SC
fibres create char in an intumescent manner without melting and dripping when burnt, thus the fibres result in
low heat of combustion (4.9 kcal/g) and high limiting oxygen index (25.2%) [10]. It is also to be noted that
during initial stages of procurement, certain by-products of wool, such as noils and burr, could be combustible if

NU
improperly handled in the processing area [12]. However, wool used in composites manufacturing already went
through the aforementioned processing phase, thus making the existence of by-products unlikely. As the
environmental concerns regarding petroleum-based polymeric composites have increased, the effects of wool as
MA
a natural fibre on the composite’s mechanical and thermal properties have been recently investigated [13-17].
For instance, Kim et al. [15] have found the positive effect of wool with char-forming FR on improving fire
retardant performance of composites and Bertini et al. [16] have also identified the reduction of total heat
D

release and heat release capacity of wool keratin based composites. As nitrogen is a common chemical element
TE

in wool and APP, two types of APP based on different nitrogen contents have been selected in this study to
investigate the effect of nitrogen as a blowing agent on flammability of wool-PP composites.
Several additives can be used in conjunction with APP, which might benefit its char forming ability,
P

consequently improving the fire retardancy. Mineral fillers as inert materials can decrease flame spread by
CE

reducing heat generation and can mitigate the volatile products [18]. Specifically, talc can act as a mass
diffusion barrier for volatile decomposition products during combustion owing to its lamellar structure, which
can simultaneously improve the processibility of polymer manufacturing due to its lubricating nature and
AC

dimensional stability [19]. Levchik et al. [20] have observed that the mixture of talc and APP with nylon creates
magnesium-ammonium polyphosphate (MgNH4(PO3)3) and silicon-ammonium tetrapolyphosphate
(Si(NH4)2P4O13) under linear heating up to 600 °C. These mineral based phosphates can form a ceramic-like
layer at the surface to enhance fire-protective properties. Elsewhere, Duquesne et al. [21] have investigated the
effects of talc and APP on PP flammability using several combinations of their loading amounts. They reported
favourable fire retardant properties under the direct flame application in UL-94 tests but having a proper balance
of the loading amounts was found to be a necessity for the fire performance.
The incorporation of maleic anhydride grafted PP (MAPP) as a compatibiliser enhances the mechanical
properties of natural fibre filled polymer composites by creating potentially reactive groups, such as hydroxyl
and carboxyl groups, upon the wool surface, which are capable of reacting with the PP matrix [22, 23]. Thus,
the adequate amount of the coupling agent has been a governing factor for improving the fibre/polymer
interfacial bonding and fibre dispersion of composites [24]. The interface can also affect the composite’s
flammability- a stronger interfacial adhesion requires more energy to pull apart the constituents in the case of
fire, consequently increasing the decomposition temperature and reducing the decomposition rate [22, 25].

2
 ACCEPTED MANUSCRIPT

However, MAPP is less thermally stable than PP, with almost a 100 °C difference between their respective
onsets of decomposition temperatures [26]. As a result, any free and unreacted MAPP present in the composite
would potentially reduce the overall thermal stability.
The cone calorimeter, as a versatile bench scale instrument, has generated comprehensive fire reaction
properties of materials. A wide range of parameters have been utilised to characterise the combustion behaviour

T
of the tested sample. On the other hand, the experimental processes and measurements to quantify the fire

IP
performance of the materials can be both time consuming and expensive. To overcome this limitation and
achieve the predictive ability of the fire behaviour, computational modelling has been introduced and developed.

R
Computational fluid dynamics (CFD) as the field model is one of the major tools for analysing the full breadth

SC
of fluid flow problems [27, 28]. Fire dynamics simulator (FDS) has been widely used as a CFD code to simulate
the burning process of a single homogeneous material or large scale structure [29-31]. However, only a few
researchers have dedicated their work to the composites fire simulation using the FDS [32, 33].

NU
The overarching aim of the present work is to investigate the effects of additives, such as APP types,
MAPP and talc, on flammability and mechanical properties of the extruded wool-PP composites. The paper
initially includes a comparative study of fire retardant performance of wool-PP-APP composites, based on
MA
different APP types. Furthermore, the burning behaviour of the composite in the presence of MAPP and talc has
also been demonstrated by employing thermogravimetric analysis, cone calorimeter and vertical burn tests. The
fire residue analysis using a scanning electron microscope has revealed the microstructure and chemical
D

elements of char formed during the cone calorimeter test. Moreover, the effects of APP, MAPP and talc on the
TE

mechanical properties of wool-PP composites are presented. Finally, an FDS model has been created to simulate
the combustion of wool-PP composite in the cone calorimeter test. It is to be noted that this is the first
comprehensive study to investigate the effects of APP types in conjunction with talc in wool based polymeric
P

composites.
CE

2. Experimental details
2.1 Materials
AC

Scoured wool fibres having Avg. 45.3 µm thickness were provided by Bloch & Behren Ltd. (New
Zealand) and polypropylene (HP400L, melt flow index: 5.5 g/10 min) as the polymer matrix was supplied by
Lyondell Basell. The PP grade was selected as the suitable thermoplastic polymer to improve the mechanical
properties of wool-PP composites [14]. Furthermore, two types of APP were utilised as the flame retardant to
investigate the effect of nitrogen content on the thermal characterisation of composites. Table 1 presents the
details of the flame retardants. MAPP (Licocene 6452, Clariant Ltd.) and talc (Plustalc N625, Mondo Minerals)
were also chosen as a compatibiliser and mineral filler, respectively.

Table 1 Specification of ammonium polyphosphates

Decomposition Average
Ammonium Phosphorus Nitrogen
Manufacturer temperature particle size
polyphosphate (P) (wt%) (N) (wt%)
(°C) (µm)
Budit® 3167
Budenheim 22 21 >250 14
(APP1)
®
Exolit AP 760 Clariant Ltd. 20 14 >250 10

3
 ACCEPTED MANUSCRIPT

(APP2)

2.2 Preparation of composites

Short wool fibres were dried at 75 °C in an oven to reduce the moisture content and then mixed with
other constituents, such as PP, MAPP, APP and talc. A twin screw extruder (LTE 26-40, Lab Tech Engineering

T
Company Ltd, Thailand) was operated for melt blending of the dry-mixture. The compounding process was

IP
carried out using the screws (length L/ diameter D = 32) co-rotating at a 170 rpm with a temperature profile of

R
172/175/175/175/175/175/175/175/178/179 °C from the hopper to the die across ten heating zones. As a final
moulding process, a single-screw extruder (JWS 45/30 extruder, China) with a slit die (100 x 3.5 mm), was used

SC
to fabricate the continuous composite sheets at a processing temperature of 180 °C and a screw speed of 8 rpm.
A pair of calendering rolls was in operation to obtain a good surface finish and uniform thickness of the

NU
extruded product.

2.3 Characterisation
MA
2.3.1 Thermogravimetric analysis (TGA)
TGA-50 (Shimadzu, Japan) was utilised to record the weight loss of materials, namely fibres, polymer
and the composites, under heating at a constant rate. About 7 mg of sample was heated till 700 °C at a linear
heating rate of 10 °C/min and argon gas with a flow rate of 50 ml/min was maintained in the chamber to reduce
D

any secondary reaction by volatile products from the sample.

TE

2.3.2 Fourier transform infrared (FTIR) spectroscopy

P

In this research, spectroscopy analysis of wool-PP composite’s char formed after combustion tests was
conducted by an IR-Prestige-21 spectrometer (Shimadzu, Japan) with a Pike Miracle Attenuated Total
CE

Reflectance compression clamp. Infrared spectra of the sample were recorded in the range of 4500-450 cm-1
with an optical resolution of 4 cm-1. Data acquisition and analysis were achieved by standard software, Omnic
AC

ESP version 7.1.

2.3.3 Environmental scanning electron microscopy

Micrographs of char residues and fractured cross-sections of composites after cone calorimeter and
tensile tests, respectively, were obtained by an environmental scanning electron microscope (ESEM - FEI
Quanta 200F, Houston, US). Moreover, energy dispersive X-ray spectroscopy, which is a chemical
microanalysis technique commonly used in conjunction with a scanning electron microscope, was also utilised
to detect chemical compositions of samples during the ESEM observation. All samples were coated with
platinum using Quorum Q150RS sputter coater (Quorum Technologies Ltd, East Sussex, UK).

2.3.4 Cone calorimeter study

A cone calorimeter (FTT Ltd, East Grinstead, UK) measured fire-reaction properties of samples
according to ASTM E1354. Samples of 100 x 100 x 3 mm were prepared and tested under an external heat flux
of 50 kW/m2. In this study, TTI (time to ignition), PHRR (peak heat release rate), THR (total heat release), SPR

4
 ACCEPTED MANUSCRIPT

(smoke production rate) and mass loss of samples were selected and average values of three replicated tests
were used to investigate the fire-reaction properties of the wool composites.

2.3.5 Underwriters Laboratory (UL)-94 tests

Flammability of the neat polymer and composites was evaluated using UL-94 vertical burn tests.

T
According to ASTM D3801, samples of 125 x 13 x 3 mm dimensions were prepared and pre-conditioned at 23

IP
°C and 50% humidity for 48 h. Burner flame was applied to the samples in the vertical position and the burning
behaviour, such as self-extinguishment or continuous burning, was recorded. The samples were classified into

R
V-0, V-1, V-2 or NR (no rating) based on the burning characteristics of five repeated tests.

SC
2.3.6 Mechanical tests
Tensile and flexural tests for the composites were performed on the Instron 5567 (10 kN load cell,

NU
SN:C69690) Universal Testing Machine (UTM). ASTM D638 standard was selected to evaluate tensile
properties of composite samples of 3 mm nominal thickness. The composites were cut by a numerical control
machine to prepare standard dumbbell-shaped test samples. The tensile moduli (chord modulus between 0.05
MA
and 0.25% strain) and strengths were measured at a crosshead speed of 5 mm/min and the extension under
tension was measured by a video extensometer. Furthermore, flexural testing was conducted using a 3-point
bending rig on the Instron 5567 UTM based on ASTM D790 standard. For wool-PP composites, the straining
D

rate of the outer surface was 0.1 mm/mm/min (selected due to the large deflections) and the flexural strength
TE

was obtained at 5% strain. The span length of 45 mm and crosshead speed of 12 mm/min were calculated, based
on the sample thickness.
P

3. Fire simulation
CE

3.1 Fire dynamics simulator

Fire dynamics simulator (FDS) is a CFD code developed by the National Institute of Standard and
Technology in the US. FDS numerically solves the Navier-Stokes equations appropriate for low speed,
AC

thermally driven flow with an emphasis on smoke and heat transfer from fire. The software is an efficient and
fast programme designed to calculate the time dependent flows [34, 35]. FDS employs a second-order accurate
finite element method in space and time. The flow variables, such as temperature and heat flux, are updated in
time using an explicit second-order Runge–Kutta scheme. Spatial derivatives in the governing equations are
written as second-order accurate finite differences on the rectilinear mesh [32]. The scalar quantities (e.g.
density) and vector quantities (e.g. velocity) are assigned at the centre of each cell and with the cell surfaces,
respectively [36]. The simulation condition can be designed and calculated within a domain. FDS output files
are visually presented by software called SmokeView. The governing equations and sub-models, such as
turbulence, pyrolysis, combustion and thermal radiation models, for FDS are presented in a supplementary
material.

3.2 Material properties and boundary conditions for FDS

Physical and thermal properties of wool, PP, and wool char were employed as input parameters to
simulate the cone calorimeter test. Table 2 shows the material properties used for the FDS simulation.

5
 ACCEPTED MANUSCRIPT

Table 2 Physical and thermal properties of wool, PP and wool char

Materials
Property Unit Wool Source
Wool PP
char
Supplier (PP) /
Density kg/m3 1390 900 523.4 Measured (wool

T
and char)

IP
Emissivity - 0.85 0.97 0.95 [37, 38]
Differential
Specific heat capacity kJ/kg K 2.68 1.651 2.3 scanning

R
calorimeter
Thermal conductivity W/m K 0.047 0.18 0.1 TC-30

SC
Heat of combustion kJ/kg 2.1E04 4.9E04 - Cone calorimeter
Heat of reaction kJ/kg 1390 1987 - [39, 40]
Reference temperature °C 338.7 452.9 - TGA

NU
Reference rate 1/s 0.00159 0.00406 - TGA
Heating rate °C/min 20 20 - TGA
Residues yield kg/kg 0.24 0 - TGA
Number of reaction - 1 1 - TGA
MA
The FDS domain consisted of a 400 mm wide by 400 mm long by 400 mm high cone calorimeter
combustion region. Fig. 1(a) shows the overall configuration for the FDS cone calorimeter simulation. The
distances between the bottom part of the cone heater and sample and between the top of cone heater and duct
D

fan are 25 and 270 mm, respectively. As boundary conditions, the four side walls of the domain were specified
TE

as closed glasses, whereas the bottom was left open for air supply. A duct fan flow of 24 l/s was also defined for
the gas or flame to flow into a vent at the top of the domain, Fig. 1(a). Furthermore, the wool fibre-PP composite
P

was treated as a single layer composite, containing short wool fibres. Different contents of wool and PP were
defined as the mass fraction on the surface. Additionally, the back side of the sample was specified as insulated,
CE

so any heat loss from the back side is prevented.

The cone heater is the key component of the cone calorimeter for the radiative heat transfer to the
AC

sample. In this thesis, the cone heater was designed by the conversion of a 3D design of the cone heater into the
FDS input code to obtain relatively close shape as shown in Fig. 1(b). The red surfaces indicate the heating
elements, thus the temperature of surfaces can be set as the boundary condition properly. The temperature of the
cone heater for the simulation was selected by a trial and error process [27, 41]. The specific heat flux used for
the cone calorimeter test was 50 kW/m2. In the simulation, a heat flux gauge was located at the centre of the
sample, Fig. 1(c), to measure the heat flux of the sample according to the input temperature in FDS. It was
found that 50 kW/m2 can be obtained from the input temperature of 850 °C.

6
 ACCEPTED MANUSCRIPT

(a) (b)

T
R IP
(c)

SC
NU Heat flux
gauge
MA

Fig. 1. (a) FDS domain, (b) bottom part of the cone heater and (c) heat flux gauge on the sample for the cone
calorimeter simulation
D
TE

A suitable grid size was selected by increasing the grid resolution until the PHRR became independent
on the grid size. To find the appropriate grid setting of the domain, two steps were implemented according to the
axis direction [27]. Firstly, different grid sizes only on vertical axis, Z, were applied. The horizontal axes, X and
P

Y, were then divided into different grid sizes at the determined grid size of the Z axis from the first step. As a
CE

result of the grid sensitivity analysis, the grid sizes of 4 and 3 mm were allocated into horizontal and vertical
grid sizes, respectively, as the desired sizes for the FDS simulation.
AC

4. Results and discussion

4.1 Thermal stability
Thermogravimetric (TG) and derivative thermogravimetric (DTG) curves were analysed to identify the
effects of additives, such as APP types, talc and MAPP, on the thermal decomposition behaviour of wool-PP
composites. Table 3 presents TGA results, such as the onset of decomposition temperature (T5%), temperatures
at the maximum weight loss rate (Tmax1 and Tmax2) and residues at 700 °C, of raw materials and composites.

Table 3 TGA results of raw materials and composites

Residues
T5%
Sample Tmax1 (°C) Tmax2 (°C) at 700 °C
(°C)
(%)
Wool 92.7 - 310.6 24.1
PP 319.2 - 410.5 0
MAPP 351.6 - 456.4 0
APP 300.5 332.3 521.9 29.1

7
 ACCEPTED MANUSCRIPT

Talc - 567.4 682.3 96.4

Wool 30 wt%-PP 267.9 285.8 433 4.93
Wool 30 wt%-PP-MAPP 2 wt% 279.2 322.2 447.8 6.79
Wool 30 wt%-PP-APP1 20 wt% 260.4 328.9 429.7 22.2
Wool 30 wt%-PP-APP2 20 wt% 258.1 324.4 446.8 18.2
Wool 30 wt%-PP-APP1 20 wt%-MAPP 2 wt% 257.3 330.9 440.9 14.6

T
Wool 30 wt%-PP-APP1 20 wt%-MAPP 2 wt%-Talc 270.1 334.2 452.2 27.2

IP
The TG and DTG curves of individual constituents, such as wool, PP, APP, MAPP and talc, are shown

R
in Figs. 2(a) and (b), respectively. Wool shows two major phases of weight loss, Fig. 2(a), - an initial loss

SC
starting from room temperature up to around 100 °C is predominantly due to the water loss. The second weight
loss occurs from around 230 to 370 °C, which is mainly caused by the breakdown of microfibril-matrix structure
and disulphide bonds of wool [42]. Above 370 °C, the loss of various side chains by severe breakage of peptide

NU
takes place as the second decomposition [43]. Furthermore, solid residues of 24.1% remain after the heating
temperature is reached at 700 °C. The carbonaceous substance formed from wool after the test is directly
correlated to the char forming ability and fire resistance of wool. On the other hand, highly flammable
MA
characteristic of the hydrocarbon based PP burns the polymer without any residue. MAPP is also completely
decomposed with the most rapid rate at 456.4 °C due to its low molecular weight [44]. The thermal stability of
APP can also be seen clearly from Fig. 2. Budit 3167 APP is mainly composed of phosphorous, nitrogen and
D

organic compound and can be decomposed into three steps [45]. Firstly, ammonia and water are released by
TE

boiling and swelling reactions to yield highly cross-linked phosphoric acid between 270 and 350 °C [46].
Secondly, the acid reacts with the carbon-rich compound to form phosphate ester between 360 and 420 °C.
During the last step of decomposition at around 500 °C, the phosphate ester catalyses into char formation by
P

dehydration [5]. TG curves, Fig. 2(a), clearly demonstrate the stable nature of talc at a linear ramp of
CE

temperature to 700 °C. Talc loses only around 4 wt% due to the loss of hydroxyl groups and has a very slow
decomposition rate when compared with those of other constituents. Thus, it can be explained that talc is
thermally stable in the temperature range where the other materials, namely wool and APP, are decomposed.
AC

(a) (b) Temperature (°C)

100 0 100 200 300 400 500 600 700

80 -0.003
Relative weight (%)

-0.008
DTG (mg/sec)

60

Wool
40 -0.013
PP
Wool
APP -0.018
20 PP
Talc
APP
MAPP -0.023
0 Talc
0 100 200 300 400 500 600 700 MAPP
Temperature (°C) -0.028

Fig. 2. (a) TG and (b) DTG curves of raw materials

8
 ACCEPTED MANUSCRIPT

Figs. 3(a) and (b) show the TGA results of wool-PP and wool-PP-APP composites, respectively. The
addition of APP into wool-PP composite significantly increases the amount of char at 700 °C and reduces the
mass loss rate, especially Tmax2. TG curves in Fig. 3(a) reveal the insignificant difference in residues between
wool-PP composites having APP1 and APP2. On the other hand, the addition of APP1 appears to be more

T
effective in reducing the decomposition rate of wool-PP composite compared to that of APP2, Fig. 3(b).

IP
(a) (b) Temperature (°C)

R
0 100 200 300 400 500 600 700
100

SC
-0.001
80
Relative weight (%)

-0.003

DTG (mg/sec)
-0.005

NU
60
Wool 30 wt% + PP -0.007
40
-0.009
Wool 30 wt% + PP + APP1 Wool 30 wt% + PP
MA
20 wt%
20 -0.011
Wool 30 wt% + PP + APP2 Wool 30 wt% + PP + APP1
20 wt% 20 wt%
-0.013
0
0 100 200 300 400 500 600 700 Wool 30 wt% + PP + APP2
-0.015 20 wt%
D

Temperature (°C)
TE

(c) Temperature (°C)

(d)
0 100 200 300 400 500 600 700
100
-0.001
P

80
Relative weight (%)

CE

Wool 30 wt% + PP
-0.006
DTG (mg/sec)

60
Wool 30 wt% + PP + MAPP Wool 30 wt% + PP
2 wt%
-0.011
AC

40 Wool 30 wt% + PP + APP1 Wool 30 wt% + PP + MAPP 2

20 wt% wt%

Wool 30 wt% + PP + APP1 Wool 30 wt% + PP + APP1 20

20 20 wt% + MAPP 2 wt% wt%
-0.016
Wool 30 wt% + PP + APP1
Wool 30 wt% + PP + APP1 20
20 wt% + MAPP 2 wt% +
wt% + MAPP 2 wt%
Talc
0
0 100 200 300 400 500 600 700 Wool 30 wt% + PP + APP1 20
-0.021 wt% + MAPP 2 wt% + Talc
Temperature (°C)

Fig. 3. TG, (a) and (c), and DTG, (b) and (d), curves of the various types of wool-PP composites

The effect of MAPP on the thermal decomposition of composites can be recognised by TGA results. In
the case of the wool-PP-MAPP composite, the increase in Tmax1 and Tmax2 could be attributed to higher Tmax of
MAPP than those of wool and PP, Figs. 3(c) and (d). Another possible reason for the result could be related to
the adhesion between wool and PP. The enhanced interfacial bonding due to the presence of MAPP requires
more energy to pull apart the constituents under heat; thus, the decomposition temperature might be incremented
[22]. Furthermore, the thermal stability of talc can also be identified in TG and DTG curves of the composites,
Figs. 3(c) and (d). The incorporation of talc into wool-PP-APP composite significantly increases residues to

9
 ACCEPTED MANUSCRIPT

approximately 27 wt% with the maximum mass loss at a relatively high temperature compared to those required
for other composites. The increase in residues appears to be due to the existence of talc preventing the
constituents from decomposing into volatile fragments. The reaction of the phosphoric acid with talc results in a
mixture of magnesium cyclotetraphosphate and silicon oxymonophosphate under heat, thereby enhancing the
formation of char [20].

T
IP
4.2 Cone calorimeter
Cone calorimeter tests generated quantitative data of fire-reaction properties of wool-PP composites

R
including additives, such as APPs, MAPP and talc, Table 4. Fig. 4 presents heat release rate (HRR) and total

SC
heat release (THR) of PP, wool-PP and wool-PP-APP composites.

Table 4 Cone calorimeter data of the various types of wool-PP composites

NU
CO CO2
TTI PHRR TPHRR THR
Sample (kg/kg) (kg/kg)
(s) (kW/m2) (s) (MJ/m2)
(±0.004) (±0.18)
MA
PP 29 ± 2 1054 ± 120 120 ± 18 117 ± 13 0.029 2.93

Wool 30 wt%-PP 16 807 ± 28 145 110 ± 1 0.026 2.6

Wool 30 wt%-PP-APP1 15 wt% 17.3 ± 0.6 317.1 ± 8 250 ± 5 72.7 ± 2.3 0.047 2.1
D

Wool 30 wt%-PP-APP1 20 wt% 18 ± 0.6 269 ± 14 310 ± 13 73 ± 4 0.043 2.1

TE

Wool 30 wt%-PP-APP2 20 wt% 16.7 ± 0.6 350.9 ± 26 335 ± 28 96.2 ± 6 0.04 2.13

Wool 30 wt%-PP-MAPP 2 wt% 17.3 ± 0.6 766 ± 35 181 ± 7 130 ± 4.6 0.027 2.54
P

Wool 30 wt%-PP-APP1 20 wt%-

17.3 ± 0.6 291.8 ± 4.2 308.3 ± 14 83.7 ± 1.5 0.045 2.2
MAPP 2 wt%
CE

Wool 30 wt%-PP-APP1 20 wt%-

21 ± 2 251.6 ± 7.4 176 ± 71 73.6 ± 4.4 0.031 2.2
MAPP 2 wt%-Talc
AC

The positive effects of wool on reducing HRR, THR and smoke production of PP can be observed from
the cone calorimeter results. Wool filled PP composite shows 23.4% lower peak heat release rate (PHRR) than
one of PP, Table 4. The composite ignites faster than the neat PP; the reason could be attributed to an early
decomposition of wool compared to that of PP, Fig. 2(a).
The incorporation of APP into wool-PP composites significantly decreases PHRR in comparison with
those of neat PP and wool-PP composite, Fig. 4(a). In particular, a composite including APP1 shows 74.5%
reduction of PHRR than that of neat PP. APP1, including the higher amount of nitrogen than APP2, leads to the
lower PHRR (269 kW/m2) of wool-PP-APP composite than that of APP2 (350.9 kW/m2) based composite. On
the other hand, the presence of APP2 within wool-PP composite shows higher and more distinct peaks in the
HRR curve compared to wool-PP-APP1 composite, Fig. 4(a). In particular, the wool-PP composite based on
APP2 has two noticeable peaks of heat release rate at the initial and final stages of the test. As HRR is
dependent on oxygen consumption, high amount of oxygen is consumed during the formation of an intumescent
char structure after ignition of the composite. HRR decreases after the first peak point due to the formation of
char but another obvious peak at around 360 s is observed because of breakages or fissures in the char. As

10
 ACCEPTED MANUSCRIPT

expected, the reduction of APP content from 20 to 15 wt% in wool-PP-APP composite results in increased
PHRR and flame spread rate. In addition, the HRR curve of wool-PP-APP 15 wt% composite in Fig. 4(a)
illustrates two definite curves at the beginning and at the end of the test, indicating relatively weak char
compared to one formed in APP 20 wt% based composite.
Additionally, THR curves, Fig. 4(b), also point out the positive effect of APP1 on heat release of the

T
composite. The addition of APP1 into wool-PP composite produces less heat (73 MJ/m2) than that of wool-PP-

IP
APP2 (96.2 MJ/m2). As the slope of the THR curve can be assumed as representative of flame spread [47], the
lower gradient of the curve of wool-PP-APP1 can indicate that the composite spreads flame more slowly than

R
the wool-PP-APP2 composite. Thus, APP1 has been selected as an effective FR for wool-PP composite to

SC
investigate the effect of APP content and the reaction with other additives, such as MAPP and talc, in cone
calorimeter result analysis. The smoke performance of materials is also an important parameter in the fire safety
field. Yields of major gaseous species, such as CO and CO2, are given in Table 4. CO yield of the composites

NU
increases by adding APP due to the reduction of gas transfer through the carbonaceous char, suppressing the
total oxidation process [11]. Specifically, the wool-PP composites, coupled with APP1, release higher and lower
amounts of CO and CO2, respectively, than those of other composites, Table 4, due to the effective char
MA
formation. The higher CO content at a later stage could be attributed to the lack of oxygen supply through the
well-formed char.
D

(a) PP (b)
TE

1000 120
Wool 30 wt% + PP

Wool 30 wt% + PP + APP1 100

800
Total heat release (MJ/m2)
Heat release rate (kW/m2)

15 wt%
P

Wool 30 wt% + PP + APP1 80

20 wt%
CE

600
Wool 30 wt% + PP + APP2
20 wt% 60 PP

400 Wool 30 wt% + PP

AC

40
Wool 30 wt% + PP + APP1 15 wt%

200 Wool 30 wt% + PP + APP1 20 wt%

20
Wool 30 wt% + PP + APP2 20 wt%
0 0
0 100 200 300 400 500 0 500 1000 1500 2000
Time (s) Time (s)

Fig. 4. Cone calorimeter test results: (a) heat release rate and (b) total heat release

The effects of MAPP and talc on wool-PP composites fire reaction have also been investigated by the
cone calorimeter tests, Fig. 5. The addition of MAPP into the wool-PP and wool-PP-APP composites induces a
change in burning behaviour under the radiant heat flux. MAPP treated wool-PP composite generates a slightly
lower PHRR and takes a longer time to reach PHRR compared to those of the untreated wool-PP composite.
The higher decomposition temperature of wool-PP-MAPP composite compared to wool-PP composite without
MAPP might have led to the prolonged TPHRR to reach the decomposition temperature with fire growth time,
thereby increasing total heat release, Fig. 5(b).

11
 ACCEPTED MANUSCRIPT

On the other hand, the existence of MAPP in wool-PP-APP composites results in slight increases in
PHRR, THR and smoke production rate (SPR), Table 4 and Fig. 5. The unreacted MAPP occurred by adding
APP could change the fire reaction characteristics as MAPP is highly flammable with the rapid decomposition
rate. Furthermore, the MAPP role in enhancing the interfacial bonding between fibre and polymer might be
limited due to the presence of APP. Dvir et al. [44] have reported that maleic anhydride in MAPP may interact

T
with functional or hydroxyl group of the FR rather than fibre reinforcement, thus obtaining lower mechanical

IP
strengths compared to those of composite without FR. Moreover, Ayrilmis et al. [48] have also suggested that
large surface of the area of natural fibre, wood, could be contaminated by crystalline deposits of flame retardants

R
in spite of adding MAPP, leading to a weak bonding. Although some of the untreated MAPP in wool-PP-APP

SC
composite might accelerate the combustion process to an extent, its use mainly enhances the interfacial adhesion
between the constituents.

NU
(a) Wool 30 wt% + PP (b)
900 140
Wool 30 wt% + PP + MAPP
2 wt%
800
MA
120
Wool 30 wt% + PP + APP1
Heat release rate (kW/m2)

Total heat release (MJ/m2)

700 20 wt%
100
Wool 30 wt% + PP + APP1
600 20 wt% + MAPP 2 wt%
80
500 Wool 30 wt% + PP + APP1
D

20 wt% + MAPP 2 wt% +

Talc Wool 30 wt% + PP
400 60
TE

Wool 30 wt% + PP + MAPP 2

300 wt%
40
Wool 30 wt% + PP + APP1 20
200 wt%
Wool 30 wt% + PP + APP1 20
20 wt% + MAPP 2 wt%
100
P

Wool 30 wt% + PP + APP1 20

wt% + MAPP 2 wt% + Talc
0 0
CE

0 100 200 300 400 500 0 200 400 600 800

Time (s) Time (s)

Wool 30 wt% + PP Wool 30 wt% + PP

(c) (d)
AC

Wool 30 wt% + PP + MAPP 2 0.1 Wool 30 wt% + PP + MAPP

100 wt% 2 wt%
0.09
Wool 30 wt% + PP + APP1 20 Wool 30 wt% + PP + APP1
Smoke production rate (m2/s)

wt% 0.08 20 wt%

80
Relative mass (%)

Wool 30 wt% + PP + APP1 20 Wool 30 wt% + PP + APP1

wt% + MAPP 2 wt% 0.07 20 wt% + MAPP 2 wt%

Wool 30 wt% + PP + APP1 20 0.06 Wool 30 wt% + PP + APP1

60 wt% + MAPP 2 wt% + Talc 20 wt% + MAPP 2 wt% +
0.05 Talc

40 0.04
0.03

20 0.02
0.01

0 0
0 100 200 300 400 500 0 100 200 300 400 500
Time (s) Time (s)

Fig. 5. Cone calorimeter results: (a) HRR, (b) THR, (c) relative mass and (d) total smoke production of wool-PP
(MAPP), wool-PP-APP (MAPP), wool-PP-APP-MAPP-talc composites

12
 ACCEPTED MANUSCRIPT

The cone calorimeter measurements identified the positive effect of talc on improving fire retardant
performance. When compared to the composites without talc, a decrease in PHRR/THR/SPR and an increase in
the final mass fraction of composite containing talc are clearly detected in Fig. 5. The results can be attributed to
the creation of a ceramic protective layer at the surface of the composite and the substitution of the flammable
polymer matrix by adding talc. As the cone calorimeter sample reached the maximum temperature of around

T
750 °C, talc particles could retain their physical and chemical structures owing to the late onset of

IP
decomposition at around 950 °C [49]. Therefore, the talc particles improved the char morphology of composite
and acted as microscopic barriers, hindering heat and volatile products diffusion. This resulted in enhanced fire

R
retardancy, especially in respect to PHRR and SPR, as shown in Figs. 5(a) and (d), respectively [19].

SC
4.3 Char analysis
4.3.1 Microstructure of char

NU
The microstructures of char formed from APP1 and APP2 based wool-PP composites after the cone
calorimeter test are demonstrated in Figs. 6(a) and (b), respectively. An addition of APP1 and wool resulted in
compact and closed char structures, as shown in Fig. 6(a), which can restrict the escape of heat and volatile
MA
gases from the char. Unlike the composite containing APP1, the wool-PP-APP2 composite resulted in 100-300
µm sized channels within the char, created by agglomeration of carbonaceous residues. The nitrogen in the
intumescent flame retardant acts an useful component for the improvement of fire retardancy of polymeric
D

materials [10]. Thus, the increase in nitrogen content by adding wool and APP1 leads to the creation of the
TE

effective char having the cohesive structure. However, the smaller amount of nitrogen (14 wt%) in the
phosphorus-nitrogen system (APP2) compared to APP1 (22 wt%) might have led to the formation of the porous
char structure, Fig. 6(b), thereby increasing the composite’s heat release rate.
P

The composite in the presence of talc formed a sticky and ceramic-like carbonaceous layer on the
CE

surface of the burning sample under the heat flux of cone calorimeter. Fig. 6(c) shows the dense char tightly
packed by the residues. Under a higher magnification, Fig. 6(d), the char can be characterised by network-
structured protective layer, without the opening or cracks like those in wool-PP-APP2 char. This structure is
AC

attributed to the mechanical interlocking of the mineral filler and phosphate or remaining talc particles [50].
Therefore, the char layer can be more effective in increasing heat insulation properties and inhibiting the heat
and mass transfer between surface and melting materials.

13
 ACCEPTED MANUSCRIPT

(a) (b)

T
R IP
SC
(c)
NU (d)
MA
D
P TE
CE

Fig. 6. ESEM images of inner surfaces of char of wool-PP composites (a) with APP1 20 wt% (Budit 3167), (b)
AC

with APP2 20 wt% (Exolit 760), (c) and (d) with APP1 20 wt%, MAPP 2 wt%, and talc

Chemical elements and structures of composite and char were identified by energy-dispersive X-ray
spectroscopy (EDX) analysis. The EDX elemental microanalysis in Fig. 7(a) can determine that the major
element of wool-PP-APP is carbon due to the presence of wool and PP. Furthermore, sulphur and phosphorus
peaks are also detected as the chemical elements occurred by wool and APP, respectively. Char formed during
the cone calorimeter test mainly contains phosphorus (P) released from the decomposition of APP, Fig. 7(b).
Fig. 7(c) shows the cross-section of the wool-PP-APP-talc composite before the cone calorimeter test.
Due to PP and fibres, carbon becomes a significant element in the composite. Moreover, magnesium and silica
peaks corresponding to the existence of talc also appear in the spectra. On the other hand, char, as shown in Fig.
7(d), formed under the radiant heat is mainly composed of magnesium, silicate, and phosphorus because of the
presence of talc and phosphorus in the char.

14
 ACCEPTED MANUSCRIPT

(a) (b)

T
R IP
SC
(c) NU (d)
MA
D
P TE
CE
AC

Fig. 7. EDX spectra and ESEM images of (a) cross section and (b) char of wool-PP-APP composite and (c)
cross section and (d) char of wool-PP-APP-talc composite

4.3.2 Fourier Transform-Infrared Spectroscopy

As a continuation of the EDX analysis, FTIR measurement was also conducted to elucidate chemical
compositions of the char. The infrared spectra of wool-PP-APP and wool-PP-APP-talc char, Fig. 8, show peaks
at 3410, 2350, and 1650 cm-1, which correspond to the vibration of N-H or O-H, C=O and C=C groups,
respectively [51, 52]. A peak at around 995 cm-1 indicates the vibration absorption of P-OH group due to the
phosphorus released from APP [51]. The addition of talc revealed new peaks, such as at 1020 and 669 cm-1.
Both peaks depict symmetric Si-O-Si stretching vibrations of talc [53]. The existence of the peaks can also
imply the presence of minerals (e.g. silica) post cone calorimeter test. From a general analysis of the entire
spectra of the composites char, it can be observed that the char from the composite without talc is fairly
aromatised with a dearth of functional groups. This could possibly indicate the existence of carbon molecules

15
 ACCEPTED MANUSCRIPT

having a closed structure. However, the presence of talc in the composite is able to retain its chemical integrity
under the cone heater (~750 °C) due to the high thermal stability.

T
R IP
SC
NU
MA
Fig. 8. FTIR spectra of char
D

4.4 Vertical burn test

Qualitative flammability evaluation was carried out by a UL-94 burn test. The test classifies a sample
TE

by the visual observation instead of quantitative parameters measured by the cone calorimeter. The test of wool-
PP composites without APP showed a slower flame spread and less flaming drips than those of neat PP due to
P

the natural fire resistance of wool after being ignited, Table 5, but could not be rated (NR). On the other hand,
CE

the composite containing 20 wt% APP1 demonstrated an immediate flame-out and no drip after flame
applications, thereby rating V-0, Table 5 and Fig. 9(a). Moreover, the composite after the test, as shown in Fig.
9(a), maintained its shape, which could be attributed to the formation of rigid carbonaceous char during the
AC

flame application, thus protecting the sample from fire. However, the incorporation of APP2 did not reduce the
composites flammability as much as APP1. The wool-PP-APP2 composite burnt with more intense flame and
for a longer time than those with APP1. Fig. 9(b) also shows that more than half of the sample is burnt out. It
can be highlighted that the amount of nitrogen in APP can be crucial to determining the flammability of wool-
PP composites. Furthermore, a significant effect of wool on the reduction of composite’s flammability has been
identified by a decrease in APP content. The composite containing 15 wt% APP1 can achieve a V-0 rating. The
previous research has shown that no natural fibre and polypropylene based composite has obtained the V-0
classification in the UL-94 test when added with 15 wt% APP [46, 54, 55]. The fire resistance of wool,
specifically charring ability, along with 15% APP, largely contributes to the formation of effective char and
inhibiting the flame spread. As the price of APP is relatively high (USD 13-14 / kg) compared to other
constituents, any reduction in the APP content is also beneficial for production cost. Based on weight fraction of
individual components (wool, PP and APP), total production cost of the composite having 20 wt% APP was
USD 10.3 and the composite having 15wt% APP was USD 9.1. Therefore, reduction of APP not only reduced
flammability but also saved about 12% of material cost.

16
 ACCEPTED MANUSCRIPT

Table 5 vertical burn test results of a neat PP and composites

Average burning time after

Sample Dripping Rating
flame applications (s)

PP 88.2 Drips NR

T
Wool 30 wt%-PP 130 Drips NR

IP
Wool 30 wt%-PP-APP1 15 wt% - No drip V-0
Wool 30 wt%-PP-APP1 20 wt% - No drip V-0

R
Wool 30 wt%-PP-APP2 20 wt% 148 Drips NR

SC
Wool 30 wt%-PP-APP1 20 wt%-Talc 120 No drip NR

In comparison, talc does not play a decisive role in reducing the composite’s flammability. Wool-PP-

NU
APP composite containing talc exhibits continuous burning after the second flame application without any drips,
Fig. 9(d). The disparate burning behaviour compared to the cone calorimeter result could be related to the
difference in methods of heat application to the sample. The vertical burn test applies a small calibrated flame
MA
directly to the sample, whereas radiant heat is projected onto the sample in the cone calorimeter. Thermal
decomposition of talc commences around 900 °C, generating magnesium oxide (MgO) and silicon dioxide
(SiO2) along with water vapour, and it melts around 1500 °C [56]. Thus, the burner flame temperature of around
D

1970 °C in the test might melt the talc and facilitate the composite’s burning [21, 56]. Furthermore, the evolved
TE

mineral oxides, such as MgO and SiO2, may competitively react with APP and hinder the good char forming
reaction with wool and polymer. Consequently, the flame is not extinguished due to the continuous flame
propagation into the sample, Fig. 9(d). Moreover, an inadequate weight ratio between APP and talc might also
P

influence the result. Duquesne et al. [21] identified that the different weight ratio between APP and talc would
CE

affect UL-94 and limiting oxygen index values. They reported that an increase in the talc amount had to be
compensated by a corresponding rise in the APP amount to maintain an acceptable fire performance. Thus,
similar to the current research, they also identified a somewhat negative effect of talc under the direct flame
AC

application.

17
 ACCEPTED MANUSCRIPT

T
R IP
SC
(a) (b) (c) (d)

NU
Fig. 9. Residual samples after vertical burn test: (a) wool 30 wt%-PP-20 wt% APP1 (Budit 3167), (b) wool 30
MA
wt%-PP-20 wt% APP2 (Exolit AP 760), (c) wool 30 wt%-PP-15 wt% APP1, and (d) wool 30 wt%-PP-20 wt%
APP1 and 9.7 wt% talc

4.5 Mechanical properties

D

Previous studies have found that although the addition of a flame retardant effectively reduces the
flammability, it can be detrimental to mechanical properties of the composites [54, 57]. Therefore, in this study,
TE

tensile and flexural tests were conducted to investigate the effects of APP on the moduli and strengths of the
wool-PP composites. As shown in Fig. 10, the composites without APP show the significant impact of MAPP
P

on the improvement of tensile and flexural moduli due to the enhanced interfacial bonding between wool and
CE

PP. In wool-PP-APP composites, although mechanical properties decrease by adding APP, an increase in APP
content in the presence of MAPP improves tensile and flexural moduli compared to those of wool-PP-APP
composites without MAPP, Fig. 10. It can be postulated that the APP particles with high stiffness promote the
AC

increase in the modulus [58] and the MAPP still plays a role in bonding the fibre and polymer matrix. On the
other hand, tensile and flexural strengths of wool-PP composites are reduced by the addition of APP. Without
MAPP, the increase in APP content from 15 to 20 wt% leads to reduced tensile and flexural strengths. The
effect of MAPP on the strengths is still positive but the difference in the strengths between wool-PP-15 wt%
APP and 20 wt% APP composites after adding MAPP became insignificant. This could be attributed to the
restricted MAPP role with APP and an inadequate weight ratio between MAPP and APP contents for adhesion
of the constituents. Figs. 11(a), (b) and (c) show the environmental scanning electron micrographs of tensile
fractured cross-sections of wool-PP-MAPP, wool-PP-APP and wool-PP-APP-MAPP composites, respectively.
More holes and fibres pull-out exist in the wool-PP-APP composite compared to the wool-PP composites. Thus,
it is very likely that APP can lead to weak interfacial adhesion between wool fibres and polymer, thereby
reducing the mechanical strengths of the composites [59]. In addition, the APP particles (yellow circles)
observed on wool and PP in Fig. 11(c) could have also interfered with matrix continuity and each particle can be
a site of stress concentration, acting as a microcrack initiator [48].
Apart from enhancing flexural modulus (where deformation takes place in a layer by layer mode
because of strain distribution throughout the cross-section), the incorporation of talc (inorganic mineral filler)

18
 ACCEPTED MANUSCRIPT

generally had an adverse effect on the mechanical properties of the wool-PP-APP composite. The reason for
inferior mechanical properties of talc containing composites can be two-fold in nature: a) incorporation of talc
might hinder the interfacial adhesion of wool and PP; b) talc addition leads to particle agglomeration that acts as
stress concentrators and mechanical failure points, which facilitate sample fracture [60]. Fig. 11(d) illustrates
several fibre pull-outs and large voids after the tensile tests due to weak interfacial bonding. Furthermore, the

T
initial micro-voids caused by debonding between the constituents may also influence the crack propagation,

IP
leading to the reduction of mechanical properties [58].

R
(a)

SC
NU
MA
D
P TE
CE

(b)
AC

Fig. 10. Mechanical properties of wool 30 wt%-PP composites containing APP, MAPP, and talc: (a) tensile and
(b) flexural moduli and strengths

19
 ACCEPTED MANUSCRIPT

(a) (b)

T
R IP
SC
(c)
NU (d)
MA
D
P TE
CE

Fig. 11. ESEM images of tensile fractured cross-sections: (a) wool 30 wt%-PP, (b) and (c) wool 30 wt%- PP-
APP 20 wt% and (d) wool 30 wt%- PP-APP 20 wt%-talc composites
AC

4.6 Simulation result: heat release rate

Heat release rate of wool 30 wt%-PP composite was simulated and compared to the experimental
result, Fig. 12. The appropriate gas species for wool (C3H6O2N) and PP (PROPYLENE, C3H6) for the
combustion model was selected by user and pre-defined FDS species, respectively. The fuel molecule in FDS
can only contain C, H, N and O in its simple chemistry analysis in the combustion model. Therefore, the FDS
model did not consider the flame suppression effect of sulphur. FDS specifies the chemical formula and the
Lennard-Jones potential parameters, such as σ (SIGMALJ) and ε/k (EPSILONKLJ), to compute conductivity,
diffusivity and viscosity of materials [61]. As mentioned in Section 3.2, a single layer was used for the
composite’s simulation and the material properties of wool and PP were applied in the layer for the pyrolysis
and combustion. In particular, thermal conductivity and specific heat capacity of the composite were calculated
using the volume fraction of the constituents.
The simulated HRR curve, Fig. 12, shows higher PHRR (917 kW/m2) than the cone calorimeter result
(807 kW/m2). On the other hand, both FDS and experimental results have more or less similar HRR trends until
fully developed fire stage (PHRR), including the two reaction phases. In the curves, the initial shoulder-like
curve at around 100 s is likely due to the decomposition of wool with char residues during the composite’s

20
 ACCEPTED MANUSCRIPT

combustion. In FDS, lower kinetic parameters of wool, such as activation energy and pre-exponential factor,
than those of PP were calculated based on the reference temperature and rate, Table 2. Thus, the slow reaction
rate of wool contributed to the lower PHRR than the second peak in the curve. Furthermore, the second peak
could be attributed to the effect of PP and the backside boundary condition of sample on further combustion of
the composite, thereby increasing PHRR. In Fig. 12, the experimental HRR curve shows a gradual and slow

T
decrease in HRR after around 180 s. It might be explained that, in reality, the char formed during the test can

IP
influence the sustained flaming at low HRR, thereby delaying the flameout time. On the other hand, the FDS
curve shows a sharp drop after the second PHRR as the simulation could not specify the effect of char on

R
flameout phase of the composite. More properties and accurate boundary condition regarding the char residues

SC
are required to enhance the accuracy of the model.

Wool 30 wt% + PP composite

NU
1000
900
Wool 30 wt% + PP_FDS
800
Heat release rate (kW/m2)

MA
Wool 30 wt% + PP_Experiment
700
600
500
D

400
TE

300
200
P

100
0
CE

0 100 200 300 400

Time (s)
Fig. 12. HRR curves of wool-PP composites under 50 kW/m2
AC

5. Concluding remarks
The following specific conclusions can be drawn from the results of the present work:
- The comparative study of the effects of two APP types on the flammability of wool-PP composites has
demonstrated that APP1 (Budit® 3167) having more nitrogen content than APP2 ( ®Exolit AP 760) is more
effective for fire retardant performance. Wool-PP-APP1 composite has lower PHRR and THR with more
rigid and closed char structure formed under 50 kW/m2 heat flux when compared to that with wool-PP-
APP2 composites. Furthermore, even with a reduced content of APP1 (from 20 to 15 wt%), the V-0 rating
was achieved under the vertical burn test. The combined effect of wool and APP1 can lead to the reduction
of material cost by almost 12%, which would be beneficial for any product design.
- MAPP effects on the thermal properties of composites have been identified by the increase in
decomposition temperatures in TGA. More energy was required to separate the constituents due to MAPP
assisted interfacial adhesion between wool and PP. On the other hand, the cone calorimeter tests have
shown that MAPP treated wool-PP-APP composite has slightly higher PHRR, THR, and SPR in
comparison to those of wool-PP-APP composite without MAPP.

21
 ACCEPTED MANUSCRIPT

- The incorporation of talc into wool-PP-APP composite has formed a closed structure of char network,
leading to lower heat release and smoke production than those of the composite without talc under the
radiant heat transfer.
- The improved interfacial bonding between wool and PP by adding MAPP has also been proven by superior
mechanical properties of composites. The incorporation of MAPP has enhanced tensile/flexural moduli and

T
strengths of wool-PP and wool-PP-APP composites compared to those composites without MAPP.

IP
However, the existence of APP has still deteriorated compatibility between wool and PP. Therefore, the
dissimilar effects of the additives on fire retardant and mechanical properties need to be considered during

R
the design process based on the end uses of the composites.

SC
- The fire dynamics model developed by employing FDS has simulated the cone calorimeter test to specify
heat release rates of materials. The computational domain and boundary conditions have been set up based
on the testing conditions of cone calorimeter. The simulated HRR results of wool-PP composites are shown

NU
to have a reasonable agreement with the experimental results.

Acknowledgements
MA
The authors thankfully acknowledge the financial support of Wool Industry Research Limited of New Zealand
(CP2013_25) and the help from CACM technicians. The authors also thank Dr. Quynh Nguyen and Associate
Professor Tuan Ngo in Infrastructure Engineering Department in Melbourne University for their technical help
D

for FDS.
TE

References
[1] A. P. Mouritz and A. G. Gibson, "Introduction," in Fire properties of polymer composite materials.
P

vol. 143, ed: Springer Netherlands, 2006, pp. 1-18.

[2] D. Price, G. Anthony, and P. Carty, "Introduction: polymer combustion, condensed phase pyrolysis and
CE

smoke formation," in Fire retardant materials, A. R. Horrocks and D. Price, Eds., ed England:
Woodhead Publishing Ltd., 2001.
[3] T. R. Hull, A. Witkowski, L. Hollingbery, Fire retardant action of mineral fillers, Polym. Degrad. Stab.
96 (8) (2011) 1462-1469.
AC

[4] M. Jimenez, S. Duquesne, S. Bourbigot, Multiscale Experimental Approach for Developing High-
Performance Intumescent Coatings, Ind. Eng. Chem. Res. 45 (13) (2006) 4500-4508.
[5] G. Camino, L. Costa, G. Martinasso, Intumescent fire-retardant systems, Polym. Degrad. Stab. 23 (4)
(1989) 359-376.
[6] N. P. G. Suardana, M. S. Ku, J. K. Lim, Effects of diammonium phosphate on the flammability and
mechanical properties of bio-composites, Mater. Design. 32 (4) (2011) 1990-1999.
[7] S. Bocchini, G. Camino, Halogen-containing flame retardants, in: C. A. Wilkie, A. B. Morgan (Ed.),
Fire retardancy of polymeric materials, CRC Press, New York, 2010.
[8] N. M. Stark, R. H. White, S. A. Mueller, T. A. Osswald, Evaluation of various fire retardants for use in
wood flour–polyethylene composites, Polym. Degrad. Stab. 95 (9) (2010) 1903-1910.
[9] B. Chen, W. Gao, J. Shen, S. Guo, The multilayered distribution of intumescent flame retardants and
its influence on the fire and mechanical properties of polypropylene, Compos. Sci. Technol. 93 (2014)
54-60.
[10] L. Benisek, Flame retardance of protein fibers, in: Flame-Retardant Polymeric Materials, Springer,
1975, pp. 137-191.
[11] E. Gallo, B. Schartel, D. Acierno, F. Cimino, P. Russo, Tailoring the flame retardant and mechanical
performances of natural fiber-reinforced biopolymer by multi-component laminate, Compos. Part B.
Eng. 44 (1) (2013) 112-119.
[12] P. Salatino, A. Di Benedetto, R. Chirone, E. Salzano, and R. Sanchirico, Analysis of an explosion in a
wool-processing plant, Ind. Eng. Chem. Res. 51 (2012) 7713-7718.
[13] L. Conzatti, F. Giunco, P. Stagnaro, A. Patrucco, C. Marano, M. Rink, E. Marsano, Composites based
on polypropylene and short wool fibres, Compos. Part A. Appl. S. 47 (2013) 165-171.

22
 ACCEPTED MANUSCRIPT

[14] N. K. Kim, R. J. T. Lin, D. Bhattacharyya, Extruded short wool fibre composites: Mechanical and fire
retardant properties, Compos. Part B. Eng. 67 (2014) 472-480.
[15] N. K. Kim, R. J. T. Lin, D. Bhattacharyya, Effects of wool fibres, ammonium polyphosphate and
polymer viscosity on the flammability and mechanical performance of PP/wool composites, Polym.
Degrad. Stab. 119 (2015) 167-177.
[16] F. Bertini, M. Canetti, A. Patrucco, M. Zoccola, Wool keratin-polypropylene composites: Properties
and thermal degradation, Polym. Degrad. Stab. 98 (5) (2013) 980-987.

T
[17] N. K. Kim, D. Bhattacharyya, R. J. Lin, Multi-functional properties of wool fibre composites, Adv.
Mater. Res. (2013) 8-11.

IP
[18] M. Nikolaeva, T. Kärki, Influence of mineral fillers on the fire retardant properties of wood-
polypropylene composites, Fire. Mater. 37 (8) (2013) 612-620.
[19] S. Bernhard, H. R. Kristin, B. Martin, Synergistic Use of Talc in Halogen-Free Flame Retarded

R
Polycarbonate/Acrylonitrile-Butadiene-Styrene Blends, in: Fire and Polymers VI: New Advances in
Flame Retardant Chemistry and Science, American Chemical Society, 2012, pp. 15-36.

SC
[20] S. V. Levchik, G. F. Levchik, G. Camino, L. Costa, Mechanism of Action of Phosphorus-Based Flame
Retardants in Nylon 6. II. Ammonium Polyphosphate/Talc, J. Fire. Sci. 13 (1) (1995) 43-58.
[21] S. Duquesne, F. Samyn, S. Bourbigot, P. Amigouet, F. Jouffret, K. Shen, Influence of talc on the fire
retardant properties of highly filled intumescent polypropylene composites, Polym. Advan. Technol. 19

NU
(6) (2008) 620-627.
[22] B. K. Kandola, A. R. Horrocks, Composites, in: A. R. Horrocks, D. Price (Ed.) Fire retardant materials,
Woodhead Publishing Ltd, England, 2001.
[23] S. Kalia, B. S. Kaith, I. Kaur, Pretreatments of natural fibers and their application as reinforcing
MA
material in polymer composites—A review, Polym. Eng. Sci. 49 (7) (2009) 1253-1272.
[24] H. Ku, H. Wang, N. Pattarachaiyakoop, M. Trada, A review on the tensile properties of natural fiber
reinforced polymer composites, Compos. Part B. Eng. 42 (4) (2011) 856-873.
[25] M. R. Islam, M. D. H. Beg, A. Gupta, M. F. Mina, Optimal performances of ultrasound treated kenaf
fiber reinforced recycled polypropylene composites as demonstrated by response surface method, J.
D

Appl. Polym. Sci. 128 (2013) 2847-2856.

[26] M. G. Salemane, A. S. Luyt, Thermal and mechanical properties of polypropylene–wood powder
TE

composites, J. Appl. Polym. Sci. 100 (2006) 4173-4180.

[27] Y. Jiang, Decomposition, ignition and flame spread on furnishing materials, Centre for Enviroment
Safety and Risk Engineering, Victoria University, Australia, 2006.
[28] P. J. DiNenno, SFPE handbook of fire protection engineering, National Fire Protection Association,
P

US, 2008.
[29] Y. Xiao, J. Ma, Fire simulation test and analysis of laminated bamboo frame building, Constr. Build.
CE

Mater. 34 (2012) 257-266.

[30] W. Chow, R. Yin, Discussion on two plume formulae with computational fluid dynamics, J. Fire. Sci.
20 (2002) 179-201.
AC

[31] A. Hamins, A. Maranghides, K. McGrattan, E. Johnsson, T. Ohlemiller, M. Donnelly, J. Yang, G.

Mulholland, K. Prasad, S. Kukuck, Report on Experiments to Validate Fire Dynamic and Thermal-
Structural Models for Use in the World Trade Center Investigation, NIST Special Publication, US,
2004.
[32] Q. T. Nguyen, P. Tran, T. D. Ngo, P. A. Tran, P. Mendis, Experimental and computational
investigations on fire resistance of GFRP composite for building façade, Compos. Part B. Eng. 62
(2014) 218-229.
[33] H. Alkhateb, A. Al-Ostaz, M. Nyden, A. Cheng, Experimental Evaluation and Numerical Simulations
of Nanocoatings in Infrastructure Fire Applications, J. Mater. Civil. Eng. 27 (2015) 04015050.
[34] K. McGrattan, R. McDermott, S. Hostikka, J. Floyd, Fire Dynamic Simulator (Version 5) User's Guide,
National Institute of Standards and Technology, Washington DC, US, 2007.
[35] R. N. Meroney, D. W. Hill, R. Derickson, J. Stroup, K. Weber, P. Garrett, CFD Simulation of
ventilation and smoke movement in a large military firing range, J. Wind. Eng. Ind. Aerod. 136 (2015)
12-22.
[36] K. McGrattan, S. Hostikka, J. Floyd, R. McDermott, Fire dynamics simulator (version 6), technical
reference guide, NIST special publication, US, 2014, pp. 1-149
[37] Emissivity table, https://thermometer.co.uk/img/documents/emissivity_table.pdf; 2016 [assessed
16.02.11].
[38] Emissivity of common materials,
http://www.omega.com/literature/transactions/volume1/emissivityb.html#p1 [assessed 15.10.14].
[39] G. T. Linteris, R. E. Lyon, S. I. Stoliarov, Prediction of the gasification rate of thermoplastic polymers
in fire-like environments, Fire. Safety. J. 60 (2013) 14-24.

23
 ACCEPTED MANUSCRIPT

[40] Z. Mei-Fang, H. Yang, Polypropylene Fibers, in: M. Lewin (Ed.) Handbook of Fiber Chemistry, CRC
Press, New York, 2006.
[41] B. H. Chiam, Numerical simulation of a metro train fire: Department of Civil Engineering, University
of Canterbury, New Zealand, 2005.
[42] D. Price, A. R. Horrocks, Polymer Degradation and the Matching of FR Chemistry to Degradation, in:
Fire Retardancy of Polymeric Materials, CRC Press, New York, 2009, pp. 15-42.
[43] D. Rama Rao, V. Gupta, Thermal characteristics of wool fibers, J. Macromol. Sci. B. 31 (1992) 149-

T
162.
[44] H. Dvir, M. Gottlieb, S. Daren, E. Tartakovsky, Optimization of a flame-retarded polypropylene

IP
composite, Compos. Sci. Technol. 63 (2003) 1865-1875.
[45] D. Bhattacharyya, S. K. G. K. De, Fire retardant polypropylene, Google Patents, 2012.
[46] R. Jeencham, N. Suppakarn, K. Jarukumjorn, Flammability and mechanical properties of sisal

R
fiber/polypropylene composites: Effect of combination of flame retardant, Adv. Mat. Res. 123-125
(2010) 85-88.

SC
[47] X. Chen, Y. Jiang, C. Jiao, Smoke suppression properties of ferrite yellow on flame retardant
thermoplastic polyurethane based on ammonium polyphosphate, J. Hazard. Mater. 266 (2014) 114-121.
[48] N. Ayrilmis, T. Akbulut, T. Dundar, R. H. White, F. Mengeloglu, U. Buyuksari, Z. Candan, E. Avci,
Effect of boron and phosphate compounds on physical, mechanical, and fire properties of wood–

NU
polypropylene composites, Constr. Build. Mater. 33 (2012) 63-69.
[49] X. Almeras, M. Le Bras, P. Hornsby, S. Bourbigot, G. Marosi, S. Keszei, F. Poutch, Effect of fillers on
the fire retardancy of intumescent polypropylene compounds, Polym. Degrad. Stab. 82 (2003) 325-331.
[50] G. Levchik, A. Selevich, S. Levchik, A. Lesnikovich, Thermal behaviour of ammonium
MA
polyphosphate—inorganic compound mixtures. Part 1. Talc, Thermochim. Acta. 239 (1994) 41-49.
[51] S. Nie, Y. Hu, L. Song, Q. He, D. Yang, H. Chen, Synergistic effect between a char forming agent
(CFA) and microencapsulated ammonium polyphosphate on the thermal and flame retardant properties
of polypropylene, Polym. Advan. Technol. 19 (2008) 1077-1083.
[52] S. Gao, X. Zhao, and G. Liu, Synthesis of an integrated intumescent flame retardant and its flame
D

retardancy properties for polypropylene, Polym. Degrad. Stab.

doi:10.1016/j.polymdegradstab.2016.05.007.
TE

[53] K. Belgacem, P. Llewellyn, K. NNahdi, M. Trabelsi-Ayadi, Thermal behaviour study of the talc,
Optoelectron. Adv. Mat. Rapid. Comm. 2 (2008) 332-336.
[54] A. Subasinghe, D. Bhattacharyya, Performance of different intumescent ammonium polyphosphate
flame retardants in PP/kenaf fibre composites, Compos. Part A. Appl. S. 65 (2014) 91-99.
P

[55] A. Subasinghe, R. Das, D. Bhattacharyya, Parametric analysis of flammability performance of

polypropylene/kenaf composites, J. Mater. Sci. 51 (4) (2015) 1-11.
CE

[56] M. Wesołowski, Thermal decomposition of talc: A review, Thermochim. Acta. 78 (1984) 395-421.
[57] F. Shukor, A. Hassan, M. Saiful Islam, M. Mokhtar, M. Hasan, Effect of ammonium polyphosphate on
flame retardancy, thermal stability and mechanical properties of alkali treated kenaf fiber filled PLA
AC

biocomposites, Mater. Design. 54 (2014) 425-429.

[58] J. Yin, Y. Zhang, Y. Zhang, Deformation mechanism of polypropylene composites filled with
magnesium hydroxide, J. Appl. Polym. Sci. 97 (2005) 1922-1930.
[59] H. Lin, H. Yan, B. Liu, L. Wei, B. Xu, The influence of KH-550 on properties of ammonium
polyphosphate and polypropylene flame retardant composites, Polym. Degrad. Stab. 96 (2011) 1382-
1388.
[60] Z. X. Zhang, J. Zhang, B.-X. Lu, Z. X. Xin, C. K. Kang, J. K. Kim, Effect of flame retardants on
mechanical properties, flammability and foamability of PP/wood–fiber composites, Compos. Part B.
Eng. 43 (2012) 150-158.
[61] K. McGrattan, S. Hostikka, J. E. Floyd, Fire dynamics simulator (version 6), user’s guide, NIST special
publication, US, 2014, pp. 1-268.

24
 ACCEPTED MANUSCRIPT

Development of fire resistant wool polymer composites: mechanical performance and fire simulation with
design perspectives

N.K. Kim and D. Bhattacharyya

Centre for Advanced Composite Materials, Department of Mechanical Engineering, The University of

T
Auckland, Auckland, New Zealand

R IP
Graphical Abstract

SC
NU
MA
D
P TE
CE
AC

25
 ACCEPTED MANUSCRIPT

Development of fire resistant wool polymer composites: mechanical performance and fire simulation with
design perspectives

N.K. Kim and D. Bhattacharyya

T
Centre for Advanced Composite Materials, Department of Mechanical Engineering, The University of

IP
Auckland, Auckland, New Zealand

R
SC
Highlights

 Short wool fibre filled polypropylene composite sheets of good quality were produced by a continuous

NU
extrusion process.
 Ammonium polyphosphate with high nitrogen content and talc formed rigid and compact char structure,
MA
which effectively reduced the flame propagation.
 Adding wool achieved a V-0 rating in vertical burn test with as little as 15 wt% of ammonium
polyphosphate.
 Heat release rates from fire dynamics simulator and experimental results of cone calorimeter were
D

reasonably close.
P TE
CE
AC

26