Original article

Textile Research Journal

2018, Vol. 88(17) 1954–1964

Simulation of the effects of structural ! The Author(s) 2017

Article reuse guidelines:
sagepub.com/journals-permissions
parameters of glass fiber fabric on the DOI: 10.1177/0040517517715081
journals.sagepub.com/home/trj
thermal insulation property

Zhenrong Zheng1,2, Hongmei Wang1, Xiaoming Zhao1

and Nannan Zhang1

Abstract
The impact of different structural parameters on the thermal insulation of glass fiber fabrics, by using numerical simu-
lation, was investigated. The structural parameters investigated were the weave structure, warp density and yarn fine-
ness. A series of structure models were calculated under the same boundary and initial conditions, which combine
thermal radiation, conduction and convection. The simulation results have been validated by experiments. The experi-
mental results and the predictions from numerical simulations were in good agreement. The results show that thermal
insulation for the samples is in the order plain < 2/1 twill < 2/2 twill < 2/3 twill < 3/3 twill fabric, when using constant
structure density and yarn fineness. Increasing the warp density, from 110 to 160 ends/10 cm, leads to a significant
decrease in heat transfer performance. When the warp density increases to more than 160 ends/10 cm, the thermal
insulation property shows a decreasing trend with an increase in warp density. Furthermore, thermal insulation per-
formance of the samples dramatically increases as yarn fineness goes from 129 to 280 tex, and then decreases for yarn
fineness greater than 280 tex. This study will provide the theoretical basis for the thermal design and application of
fibrous materials.

Keywords
glass fiber fabric, structural parameters, numerical simulation, thermal insulation

In recent years, the safety of industrial workers and Cubric et al.5 have drawn conclusions about the correl-
firefighters operating under thermally hazardous condi- ation between the thermal resistance of knitted fabric
tions has been a major concern.1 Thermal insulation and thickness, mass per unit area, cover factor and por-
textiles are a crucial and important route to decrease osity, then the optimal structure parameters of the
thermal hazards.2 The thermal protective performance fabric are selected to meet the requirements set by the
of fabrics is affected by many parameters. The effects of purpose of the product.6 Li et al.7 evaluated the effect
structural parameters of fabrics, including fiber diam- of different material combinations and design features
eter, bulk density, thickness, cover factor, porosity, on thermal insulation and the moisture permeability
weave structure, weave density and yarn fineness, on index of firefighter turnout clothing by using a sweating
their thermal resistance properties were investi- manikin.
gated.3–10 Raeisian et al.3 have reported that the ther-
mal resistance properties of needle-punched nonwoven
fabrics can be affected by different processing and fiber 1
College of Textiles, Tianjin Polytechnic University, China
2
diameters, the bulk density of the layer and the thick- Key Laboratory of Advanced Textile Composites of Ministry of
ness and surface pressure on layers of hollow fibers. Education, Tianjin Polytechnic University, China
Zhang et al.4 analyzed the effects of different materials
Corresponding author:
and different structural parameters on the dynamic Xiaoming Zhao, Tianjin Polytechnic University, Binshuixidao Road 399,
thermal conductivity performance of fabrics by using TianJin 300387, China.
a self-developed dynamic heat transfer testing device. Email: [email protected]
Zheng et al. 1955

Structural parameters, including fiber diameter, bulk Materials and methods

density, thickness, cover factor, porosity, weave struc-
ture, weave density and yarn fineness of fabric, need to
Materials
be taken into account to develop a new protective In order to study the influence of fabric structural par-
fabric. However, most of the previous studies have ameters on the thermal insulation property of glass
focused on the effects of structural parameters on the fiber fabric, 14 pieces of fabric were woven using
heat transfer in fabric, and are based solely on a large ASL3000 rapier looms. The weft density of the glass
number of experimental tests. However, this approach fiber fabrics is 100 picks/10 cm. The glass fiber yarns
requires a high-temperature working environment, and (high-silica glass fiber) were produced by JiangSu
results in added costs and energy consumption, and Special Glass Fiber Co. Ltd, China. The yarns of
operational risk.11–13 Therefore, it is necessary to estab- fabric were twisted from four glass fiber filaments,
lish a heat transfer model to directly predict the effects and the twist of yarn is 165 turns/m. The structural
of structural parameters on the heat transfer in fabric, parameters of the glass fiber fabrics are shown in
thus providing a quick way to find the optimal values of Table 1.
structural parameters of the fabric. It also provides an Heat transfer through fabric is influenced by many
important theoretical basis for the design, optimization factors, such as the fiber material, thickness, porosity,
and application of the thermal insulation fabric in the the pore size, tightness, weave, yarn linear density, warp
field of thermal protection. (weft) density and so on. Pore size, density (porosity)
The most widely used inorganic fibrous insulating and thickness are the primary determinants of heat flow
materials are glass fiber textiles.14 Glass fiber textiles through fabric. However, if these three parameters are
can be used as high-temperature thermal insulation determined or designed, it is difficult to weave a fabric
materials due to their excellent properties, such as with the expected parameters accurately. A single-layer
high temperature resistance, low cost, flame retardancy, woven fabric is made from a set of warp and weft yarns.
corrosion resistance, high strength, good chemical sta- By arranging interlacement between the warp and weft
bility, etc. In this paper, we propose a series of structure yarns, different types of woven patterns can be
models to predict the thermal insulation of glass fiber obtained. Difference in weave can lead to different
fabric under various structure densities, yarn fineness pore size and porosity of fabric. Yarns are the basic
and weave structures. Furthermore, we have discussed meaningful structural elements of interlaced fabrics.
the effects of different structural parameters on the heat For the glass fiber yarn, warp (weft) density and yarn
transfer performance of the fabric. Finally, the numer- linear density are two major factors that affect the
ical simulation results are compared with the experi- fabric porosity and fabric thickness. Therefore, the
mental values in order to verify the validity of the weave, warp density and yarn linear density were con-
proposed approach. sidered in our work. If weave, warp (weft) density and

Table 1. Structural parameters of the glass fiber fabrics

Sample Count of warp/ Linear density Weave Average length Fabric Fabric areal
No. ends/10 cm of yarns/tex Structure of float thickness/mm density/g/m2 Porosity%

1 120 280 Plain 1 0.930 632.3 63.2

2 120 280 2/1 twill 1.5 0.947 628.2 64.1
3 120 280 2/2 twill 2.0 1.030 617.7 67.6
4 120 280 2/3 twill 2.5 1.230 604.2 73.4
5 120 280 3/3 twill 3 1.457 593.9 77.9
6 110 280 3/3 twill 3 1.446 576.6 78.4
7 140 280 3/3 twill 3 1.523 645.9 77.1
8 160 280 3/3 twill 3 1.527 700.0 75.2
9 180 280 3/3 twill 3 1.544 725.0 74.6
10 160 129 3/3 twill 3 0.660 302.2 69.5
11 160 160 3/3 twill 3 0.801 400.0 68.0
12 160 198 3/3 twill 3 1.030 455.7 72.4
13 160 220 3/3 twill 3 1.170 555.1 73.2
14 160 320 3/3 twill 3 1.531 814.0 72.2
1956 Textile Research Journal 88(17)

yarn linear density are determined, the fabric can be

weaved successfully. Then, the fabric thickness, poros-
ity and the fabric density were tested, and these major
parameters were used to analyze the variation of
thermal properties of fabrics with different weave
structures.
To investigate the effect of weave on the thermal
insulation of the fabric, five different glass fiber fabrics
(Nos. 1–5) with the same fabric density and yarn fine-
ness were investigated. Fabric samples 6–9 differed in Figure 1. Schematic diagram of yarn measurement parameters.
warp density (110, 120, 140, 160, 180 ends/10 cm), and
they have the same weft density of 100 picks/10 cm,
yarn fineness of 280 tex and a 3/3 twill weave pattern.
With a constant weave structure and fabric density, six Table 2. Physical and thermal properties of materials for
kinds of glass fiber fabrics (Nos. 8, 10–14) with different simulation
yarn fineness – 129, 160, 198, 220, 280 and 320 tex –
Thermal Specific heat
were used to compare the thermal insulation properties conductivity capacity Density
of glass fiber fabric. Yarn (tex) (W/mK) (J/kgK) (kg/m3)

129 (Sample No. 10) 0.1671 810.395 1500.00

Methods 160 (Sample No. 11) 0.1782 784.535 1560.61
The fabric density was determined according to EN 198 (Sample No. 12) 0.1853 765.468 1602.74
1049-2:1993. 220 (Sample No. 13) 0.1909 742.90 1770.49
The thickness values of fabrics were tested using a 280 (Sample No. 8) 0.1950 719.641 1848.48
YG141L digital thickness gauge at a pressure of 320 (Sample No. 14) 0.1983 694.85 1913.04
20 CN/cm2, a pressure area of 25 cm2 and a pressure Air 0.0242 1006.00 1.225
time of 10 s, in accordance with GB/T 7689.1-7689.5,
2001. The final measurement result was the average of
the 10 test points, accurate to 0.01 mm.
The fabric area density was measured according to that indicates its ability to conduct heat. It is important
GB/T 4669-2008. for the heat transfer through the fabric. In Table 2, the
Fabric porosity was determined using the following thermal conductivity increases with the increase of the
equation15 linear density, because the higher the linear density is,
the thicker the yarn is. The number of fibers per unit
m volume increases, and the thermal conductivity of yarn
P¼1  100% ð1Þ
h is higher than air. Hence, the thermal conductivity of
yarn increases as the yarn linear density increases. The
where P is the fabric porosity (%), m is the fabric area relationship of the thermal conductivity (Y) and the
density (g/m2),  is the yarn density (kg/m3) and h is the yarn linear density (X) was investigated by regression
fabric thickness (mm). analysis. The thermal conductivity of yarn at different
The yarn cross-sections were cut using a laser beam yarn linear densities can be calculated according to the
razor blade and images of the fabrics were initially following equation: Y ¼ 0.11183 + 5.47101  104X
taken using a Scanning Electron Microscope (SEM).  8.73134  107X2. The regression coefficient is 0.99,
To obtain the structural parameters of the fabric, the and it represents that the relationship of the thermal
images were imported to ImageJ software for measure- conductivity and yarn linear density is fitted very well.
ments to be taken. Schematic diagrams of yarn meas- The emissivity of the fabric was measured using a
urement parameters are illustrated in Figure 1. TIR100-2 Thermal Emissivity Instrument. The emissiv-
The thermal properties were measured using a Hot ities of the fabrics were taken and the average was
Disk 2500S Thermal Constant Analyzer. The thermal found to be 0.82. This was adopted as the emissivity
conductivity and specific heat capacity of the glass fiber of the fabric.
yarns were measured in the radial direction of the yarns Experiments have been carried out to validate the
at 20 C. The density of yarns was measured using a numerical simulation results. An illustration of the
pycnometer.16 The physical and thermal properties of experimental process is shown in Figure 2. An asbestos
materials for simulation are shown in Table 2. The ther- gauze was placed above the flame, and a circle of radius
mal conductivity of a substance is an intensive property 3 cm was cut out in its center. Then the fabric
Zheng et al. 1957

Figure 3. Schematic diagram of the fabric.

Figure 2. Schematic diagram of the verified experiment.

(H  W ¼ 15 cm  15 cm) was placed on the asbestos

gauze. The front face of the fabric was ablated using
an alcohol blast burner at 750 C, Test conditions:
ambient temperature 20  1 C, 65% relative humidity
and wind velocity of 1.5 m/s. An infrared thermocouple
(Raytek, USA) was used to measure the variation of the
temperature of the back face of the fabric during
the heating process. The details of fabrics used in this
validation experiment are presented in Table 1.

Modeling
Texgen software was used to develop the fabric model
for finite element analysis; the specifications of the
fabric, such as yarn cross-sections, yarn widths, yarn Figure 4. Contact regions between the yarns and air domain.
heights and yarn spaces, were required to establish
the geometry model of fabrics. The paths of the yarns
are described by specifying discrete points (control differential equations. FEM analysis is a very effective
nodes) and a Bezier interpolation function was used numerical analysis method for extremely complex
to describe the yarn path. A two-dimensional (2D) engineering structures. After the fabric geometry
binary matrix was used to represent the interlacing of models were obtained, they were imported into
the warp and weft yarns.17 According to the experimen- ANSYS Workbench. As shown in Figure 3, the fabric
tal data, a realistic fabric model was then generated by domain consisted of yarns and dry air. The physical
defining the widths, heights, cross-section shape, yarn properties of yarn and air, such as thermal conductiv-
spacing and thickness of the warp yarn and weft yarn, ity, specific heat capacity and density, were assigned
respectively. When a domain had been created, yarns respectively. Because the yarn surface was contacted
were trimmed to fit inside the domain that represents with the air, the contact area was defined as in
the limits of the unit cell, as shown in Figure 3. Figure 4. Following the definition of the contact
regions, meshing was carried out. A tetrahedron elem-
ent was chosen as the element type and the mesh size
Finite element analysis was 1.5  104 m. The total number of elements and
The finite element method (FEM) is a computerized nodes for a series of three-dimensional (3D) fabric
method for predicting how an object reacts to real- models analysis are shown in Table 3. A meshed unit
world forces, vibration, heat, fluid flow and other phys- cell of 3/3 twill fabric with warp density of 110 ends/
ical effects. It subdivides a large problem into a finite 10 cm is shown in Figure 5.
number of non-overlapping units for finding approxi- Heat conduction, convection and radiation occur at
mate solutions to boundary value problems for partial the surface of the fabrics and act as boundary
1958 Textile Research Journal 88(17)

Table 3. The total number of elements and nodes for a series of three-dimensional fabric models
Sample No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Number of 56,813 30,919 67,365 76,381 155,831 125,538 122,866 134,499 112,881 69,638 76,207 126,787 117,834 117,295
elements
Number of 95,336 52,755 111,741 128,116 251,336 206,026 199,055 216,489 183,030 123,763 133,499 209,446 194,683 193,653
nodes

The temperature boundary conditions can be

described as follows:
the initial temperature Tðx, 0Þ ¼ T0 ;
temperature boundary conditions Tðx, tÞ ¼ T.
The ambient temperature (T0 ¼ 20 C) was taken as
the initial condition. The temperature along the bottom
face of the fabric model was a constant temperature
T ¼ 750 C.
Convection is caused by the movement of air mol-
ecules and the convective boundary acts on the outer
surface of the fabric. The following equation is used to
express the convective heat transfer between the fabric
surface and the environment

Q ¼ hAðTW  T1 Þ ð2Þ

where h is the convective heat transfer

Figure 5. Meshed unit cell of 3/3 twill fabric with warp density coefficients (W/(m2  C), A is the surface area of heat
of 110 ends/10 cm. convection (m2), Tw is the surface temperatures of the
fabric ( C) and T1 is the ambient temperature of
air ( C).
Considering that the fabric thickness is much smaller
than the fabric width and length, it is reasonable to
assume that the fabric is simplified to an infinite flat
plate.
The air flow condition of the boundary layer is deter-
mined by

u1 L u1 L
Rayleigh numbers Re ¼ ¼ ð3Þ
 v

The velocity of airflow is 1.5 m/s, which was mea-

sured using a Testo 425 Thermal Anemometer. The
dimensions of the fabric sample are 15 cm  15 cm.
Based on these conditions, the Rayleigh numbers (Re)
Figure 6. Schematic diagram of heat transfer through the were calculated at different surface temperatures
fabric. (20–780 C). The results showed that the Rayleigh num-
bers were all lower than 5  105. Namely, Re < 5  105.
Thus, the flow pattern of air on the entire fabric surface
conditions. Heat transfer through fabrics involves con-
boundary layer was laminar flow. The convective heat
duction through the solid yarn and air space between
transfer coefficient was proposed by Karleker and
the yarns, as shown in Figure 6. In our studies, constant
Desmond18 and is expressed as the following equations
temperature was selected, and the fabric was subjected
to a fixed temperature on the front face. Both the con- Z L  1=2
vection and radiation were defined on the other five 1 k u1 x
h¼ 0:331 P1=3
r dx
ð4Þ
sides of the fabric, except for the bottom face. L 0 x 
Zheng et al. 1959

Figure 7. Boundary conditions for a unit cell model of 3/3 twill fabric with warp density of 110 ends/10 cm: (a) temperature;
(b) convection; (c) radiation.

From the above equation

600
 
k u1 L 1=2 1=3
h ¼ 0:662 Pr ð5Þ 500
L v

)
400

Temperature(
where L is the qualitative dimensions (m), k is the ther-
300 plain-sim
mal conductivity of airflow (W/(m2 C)),  is the dens- plain-exp 2/3-sim
ity of air (kg/m3), u1 is the speed of airflow (m/s),  is 200 2/1-sim 2/3-exp
2/1-exp 3/3-sim
the dynamic viscous coefficient of airflow, Pr is the 100 2/2-sim 3/3-exp
Prandtl number and v is the kinematic viscosity of air- 2/2-exp
flow (m2/s). 0
0 5 10 15 20 25
The thermal physical parameter of air, such as the
Time(s)
thermal conductivity k, the kinematic viscosity of air-
flows v and the Prandtl number Pr , were all obtained at
the different temperatures.18 By applying the physical Figure 8. Simulation and experimental temperature curves of
parameter to Equation (4), the calculated convective fabrics with different weave structures.
heat transfer coefficient on the surface of the fabric is
12 W/(m2  C) in the range 20–780 C.
Radiation heat transfer from the fabric to the Results and discussion
environment was also considered. For the one-
Model validation
dimensional case, the radiation exchange was between
the faces of the fabric and the environment. The In order to validate the numerical simulation results,
heat loss by radiation was estimated according to experimental measurements have been carried out.
Equation (5) 19 The experimental test for each sample was repeated
five times, and the average was taken as the final
  result. For the numerical simulation, the mean tem-
qr ¼ " T4W  T41 ð6Þ
perature was calculated from five nodes selected ran-
domly on the top surface of the fabric for each model.
where qr is the heat flux by radiation, " is the emissivity Figures 8–10 indicate the temperature evolution during
of the fabric, s is the Stefan–Boltzmann constant, Tw is the first 25 s from the simulation and experimental
the surface temperatures of the fabric and T1 is the results of glass fiber fabric under different structure
ambient temperature of air. weave, warp density and yarn fineness, respectively.
The boundary conditions, including temperature, The relative error of each fabric between the simu-
convection and radiation, were applied to the fabric lated and experimental temperature values was calcu-
model (see Figure 7). The simulations were run with a lated every second. So the number of the relative errors
constant time interval of 0.01 s. It took approximately for every fabric is 25. Then, the average of them is the
30 s to complete the computation. average relative error for this fabric. For one structure
1960 Textile Research Journal 88(17)

same variation tendencies. During the simulations and

500 experiments, the rate of temperature rise on the surface
of the glass fiber fabric increased rapidly within the first
)
400
2 s, because of exposure to a high-temperature load. It
Temperature(

300 then increased gradually with time, until a stable tem-

110-sim 110-exp
200 120-sim 120-exp perature was reached. As shown in Figures 8–10, in the
140-sim 140-exp
160-sim 160-exp
heat transfer after 25 s for both the simulations and
100 180-sim 180-exp experiments, there are significantly different perform-
0
ances for different weave structure, warp density and
0 5 10 15 20 25 yarn fineness of fabric. That is, the different structural
Time(s) parameters of a glass fiber fabric play an important role
in its heat transfer properties. In Figure 8, where the
Figure 9. Simulation and experimental temperature curves of fabric density and fineness are constant, fabric with
fabrics with different warp densities. weave patterns of 3/3 twill showed the highest ability
of heat insulation. In Figure 9, where weave structure,
weft density and yarn fineness are constant, the results
show that the thermal insulation of fabric with warp
density of 160 ends/10 cm is the best, since its tempera-
600
ture rise is the least. Figure 10 shows the variations in
500 temperature of fabrics with the same weave pattern and
)

400
fabric density, where the change of yarn fineness is a
Temperature(

significant factor affecting heat transfer performance,

129tex-sim 129tex-exp
300 and fabric with yarn fineness of 280 tex shows a good
160tex-sim 160tex-exp
200 198tex-sim 198tex-exp ability to prevent heat transfer. The increase or
220tex-sim 220tex-exp decrease of thermal insulation with reference to the fac-
100 280tex-sim 280tex-exp tors (weave pattern, warp density, yarn fineness) will be
320tex-sim 320tex-exp
0 analyzed and explained in detail in the Effects of weave
0 5 10 15 20 25 structure, Effects of structure densities and Effects of
Time(s) yarn fineness sections, respectively.
Conduction heat comes from the flame of the fire,
Figure 10. Simulation and experimental temperature curves of and it can increase the temperature of fabric.
fabrics with different yarn fineness. Convection is generated due to the movement of air
molecules. When the temperature of the fabric is
higher than the ambient temperature, due to the con-
Table 4. Average relative error for the simulation and duction, the heat of fabric will be taken away by the
experimental results movement of air molecules. At the same time, when the
temperature of the fabric is higher than ambient, radi-
Average relative
Structure parameter error range (%) ation heat will transfer from the fabric to the
environment.
Weave pattern 2.31–4.45 In order to investigate the contributions due to con-
Warp density 1.69–4.60 duction, convection and radiation, three types of ther-
Yarn fineness 2.00–3.95 mal loads were defined step by step. The numerical
simulation results are shown in Figure 11. This can be
implemented by the simulation method, but it is diffi-
parameter (weave, warp density or yarn linear density), cult to implement in most cases, as the conduction,
a series of fabrics were weaved and measured. Then, the convection and radiation may exist simultaneously,
average relative error range of each structure parameter and they are difficult to separate into three parts.
would be obtained. The maximum and minimum aver- As shown in Figure 11, when only conduction was
age relative error for each structural parameter of fab- loaded on the fabric, the temperature of the fabric
rics were taken as the average error range, as shown in increased rapidly, and it was 748 C at 30 s. When
Table 4, which indicates a good agreement between the both the conduction and convection were loaded, the
numerical calculations and the experimental results. temperatures of fabrics were obviously lower, and it
The results show that the model predictions and was 640 C at 30 s. When conduction, convection and
experimental data for the temperature curves have the radiation were all loaded on the fabric, the temperature
Zheng et al. 1961

800 500
700
)
Temperature( 600 490
500
400 480
300 only conduction
conduction and convection 470
200 conduction, convection and radiation
100
0 5 10 15 20 25 30 460
Time (s) 1.0 1.5 2.0 2.5 3.0
The average length of yarn float

Figure 11. Relative contribution due to the three types of

thermal transfer. Figure 13. Statistical model of the temperature and the aver-
age length of the yarn float.

length of the yarn float is plain < 2/1 twill < 2/2
550
The average length of yarn float

the average length of yarn float

sim
3.0 twill < 2/3 twill < 3/3 twill fabric and the temperature
exp 2.5 values of the samples after 25 s are plain > 2/1
)

500 twill > 2/2 twill > 2/3 twill > 3/3 twill. Figure 12 shows
Temperature(

2.0
that the longer the average float length, the lower the
1.5
temperature value, indicating increased thermal resist-
450 1.0
ance. For woven fabrics, the thermal resistance of twill
0.5 weave fabric is higher than that of plain weave fabric.20
400
l l l l
0.0 Figure 12 shows that when the average float length
n
plai 2/1twil 2/2twil 2/3twil 3/3twil of fabric increases from 1 to 3, the temperature of the
Fabric weave back face of fabric samples decreases in both the
numerical simulations and the experiments. Since the
Figure 12. Comparison of temperature variation after 25 s for larger the average float result in the fewer interlacing
samples with different weave structures. times of yarns, more soft and better fluffy fabrics21 and
more still air is contained in the fabric. As air has lower
thermal conductivity than the fibers, it can diminish the
of fabric was about 522 C at 30 s. In the period up to heat transfer rate through the fabric. According to
5 s, the differences of the fabric temperatures were very Figure 12, plain and 3/3 twill fabrics have markedly
small when heat transferred through fabrics in these different temperatures due to their different weave
three situations (only conduction, conduction and con- structure. This suggests that the average float length
vection, conduction and convection and radiation). of fabric affects the heat transfer in the fabric. As
However, these differences become obvious after 5 s, shown in Figure 13, the relationship between the tem-
because the fabric temperature was higher than 400 C perature and the average length of the yarn float can be
after it obtained the conductive heat for more than 5 s. described by a linear regression model
This indicates that convection and radiation heat is
important when the fabric temperature is high. T1 ¼ 511:15  16:59X1

where T1 is the temperature of the fabric sample and X1

Effects of weave structure
is the average length of the yarn float. The correlation
The weave structure of woven fabrics plays a crucial coefficient is R2 ¼ 0.94.
role in their thermal insulation performance. To inves- In addition, when the fabric density and yarn fine-
tigate the effect of weave structure on the thermal insu- ness are constant, the average length of the yarn float is
lation of the fabric, five different glass fiber fabrics as follows: plain < 2/1 twill < 2/2 twill < 2/3 twill < 3/3
(Nos. 1–5, as shown in Table 1) with the same fabric twill fabric. As shown in Table 1, with the average
density and yarn fineness were investigated by changing length of the yarn float increasing, the porosity grad-
the weave structure of the samples. The temperature ually increases from 63.2% (plain) to 77.9% (3/3 twill),
variations of the samples with different weave struc- and the fabric thickness gradually increases from
tures after 25 s are shown in Figure 12. When the fabric 0.930 mm (plain) to 1.457 mm (3/3 twill). The thickness
density and yarn fineness are constant, the average and porosity are two major parameters that affect the
1962 Textile Research Journal 88(17)

decrease in heat transfer performance.22 When the warp

sim density continues to increase to a certain level (160
500 exp ends/10 cm), this decreases the size of pores within the
Temperature( )

yarns, which may cause smaller air channels in the

sample and, as a result, the convective heat transfer
450 decreases.3 The temperature of the fabric increases
slowly, and the heat insulation performance is
improved. When the warp density is higher than 160
400 ends/10 cm, with the increase of density, the tempera-
ture on the surface of fabric rises faster, and the thermal
110 120 140 insulation of the glass fiber fabric is decreased. This is
160 180
because, as the density increases to a certain threshold,
Warp density ends/10cm the space between yarns is already very small.
Furthermore, the structure is being more compressed
Figure 14. Temperature comparison for samples with different by increasing fabric density, which can result in a
yarn densities. greater contact surface area between the yarns and
lower porosity, resulting in a decrease in the amount
heat transfer in the fabric. With higher porosity, more of still air in the fabric and better conduction by yarns
still air will be contained in the fabric. With higher in the heat transfer process and, so, heat transfer is
fabric thickness, the heat insulation properties will be improved.4
better. Hence, the temperature values of the samples The relationship between the temperature and the
after 25 s are plain > 2/1 twill > 2/2 twill > 2/3 warp density can be described by the regression
twill > 3/3 twill. model as follows

T2 ¼ 23580:56  679:94X2 þ 7:46X22  3:61  102

Effects of structure densities
X32 þ 6:50  105 X42
Structure density is a parameter that must meet the
usage of heat insulation fabric, and it describes the
number of yarns in a unit length. Fabric samples 6–9 where T2 is the temperature of the fabric sample and X2
differed in warp density, as shown in Table 1. The warp is the warp density of the fabric. The correlation coef-
densities of the fabrics are 110, 120, 140, 160, 180 ends/ ficient is R2 ¼ 1.
10 cm, they have the same weft density of 100 picks/
10 cm, yarn fineness of 280 tex and 3/3 twill weave pat-
tern. Figure 14 shows that the tendency of temperature
Effects of yarn fineness
variation is consistent between the numerical simula- With a constant weave structure and fabric density, six
tions and the experimental results. It is also evident kinds of glass fiber fabrics (sample Nos. 8, 10–14 in
that with an increase of warp density from 110 to 160 Table 1) with different yarn fineness – 129, 160, 198,
ends/10 cm, the temperature on the surface of the fabric 220, 280 and 320 tex – were used to compare the ther-
decreases gradually, indicating that the heat insulation mal insulation properties of glass fiber fabric. As shown
is more effective. When the warp density increases to in Figure 15, as the fabric samples were heated for 25 s,
180 ends/10 cm, the temperature of the fabric rises rap- the thermal insulation properties of the samples dra-
idly and the heat insulation effect is decreased. The matically increases as yarn fineness increases from 129
fabric with warp density of 160 ends/10 cm achieves to 280 tex, and then decreases with yarn fineness greater
the lowest temperature and exhibits the best thermal than 280 tex. The reason for this phenomenon could be
insulation. From Table 1 it can also be seen that, the fact that the fabrics have different yarn fineness.
when weft density is constant and with an increase of Thicker yarns with an increase in width of yarns lead
warp density from 110 to 140 ends/10 cm, the thickness to a decrease in the size of pores within fabrics of the
of the fabric increases and the porosity decreases. With same fabric density, which may cause lower air perme-
a further increase from 140 to 180 ends/10 cm, the ability of the samples and, as a result, the ability of heat
thickness shows a slight change, and the porosity getting through the fabric was weakened and the tem-
decreases. This is because the fabric has a sparse struc- perature on the surface of the fabric was lower at the
ture at low warp density, and the air permeability of the same time. The thicker yarns create smaller void spaces
fabric increases, which leads to a larger convection heat in the woven structure. Decreasing the size of voids
transfer within the fabric. This causes the temperature within the fabric may cause reduction of the air flow
of the fabric to increase quickly, resulting in a significant through samples, which leads to more trapped air.
Zheng et al. 1963

During an unsteady-state condition, the structural

sim
550 parameters of glass fiber fabric play an important role
exp
Temperature(ć) in their thermal insulation performance. The results
500
indicate that the influence of the weave structure is
related to the thermal insulation of fabric. The figures
shows that thermal insulation properties increase as the
450
average float length increases. When the fabric density
and yarn fineness are constant, the temperature values
400 of the samples after 25 s are in the order plain > 2/1
129 160 198 220 280 320 twill > 2/2 twill > 2/3 twill > 3/3 twill fabric. The influ-
Yarn fineness tex ence of warp density on thermal insulation properties is
significant. Increasing the warp density (>110 ends/
10 cm) leads to a decrease in air permeability and a
Figure 15. Temperature comparison for samples with different
significant decrease in heat transfer performance.
yarn fineness.
When the warp density continues to increase to a cer-
tain level (160 ends/10 cm), the thermal insulation
property shows a decreasing trend. There is a specific
As air has lower thermal conductivity in comparison fabric density (160 ends/10 cm), for glass fiber fabric
with fibers, the thermal insulation of the glass fiber with constant weft density of 100 picks/10 cm, yarn
fabric increases. When yarn fineness increases to more fineness of 280tex and 3/3 twill weave pattern, to
than 280 tex, thicker yarns are denser within the limited obtain optimum thermal insulation. The results of
fabric structure and have greater contact surface area, numerical simulations and experiments prove the influ-
which increases heat conduction. In addition, the ence of fineness of yarn on changes in thermal insula-
denser structure has less air in the void spaces, resulting tion properties. The results clearly show that the
in decreased thermal insulation. The relationship thermal insulation performance of samples dramatic-
between the temperature and the yarn fineness can be ally increases with yarn fineness from 129 to 280 tex
described by the regression model, as follows and then decreases for yarn fineness greater than
280 tex. Finally, it is anticipated that the optimal par-
T3 ¼ 2929:64  51:47X3 þ 0:41X23  1:39  103 ameters for good thermal insulating properties of glass
fiber fabric are weave structure of 3/3 twill, warp dens-
X33 þ 1:72  106 X43
ity of 160 ends/10 cm and yarn fineness of 280 tex, when
using a constant weft density of 100 picks/10 cm and
where T3 is the temperature of the fabric sample and X3 boundary conditions.
is the yarn fineness of the fabric. The correlation coef- The numerical simulation of the heat transfer
ficient is R2 ¼ 0.94. through the fabric can provide the theoretical basis
for the design and application of thermal protective
fabric. It can be widely used in the design of thermal
Conclusion
fabrics for firefighters, welders, hot pipelines, oil refin-
For high-temperature materials, high-silica glass fiber ing equipment and so on. Nevertheless, further studies
fabric has the advantages of excellent heat resistance, are necessary, since the thermal properties of fibrous
ablation resistance, good stability in alkali and acid materials are changing during the process of thermal
(hydrofluoric acid, phosphoric acid and hydrochloric loading. This work can be further extended to improve
acid excluded), no harm to the human body and the accuracy of calculations.
lower price relative to other high-temperature resistant
fabrics. This paper proposes a new numerical modeling Declaration of conflicting interests
technique that combines radiation with conduction and The authors declared no potential conflicts of interest with
convection in a series of glass fiber fabric models for the respect to the research, authorship and/or publication of this
prediction of the impact of different structure param- article.
eters on the thermal insulation of the fabrics, and it
can be extended to the thermal design of any other Funding
fibrous fabric. The numerical model was validated The authors disclosed receipt of the following financial sup-
by comparison with experimental values. The experi- port for the research, authorship, and/or publication of this
mental results and the predictions from numerical article: This work was supported by the National Natural
simulation were in good agreement, and a number of Science Foundation of China (Grant 51206122), the Natural
conclusions were made. Science Foundation of Tianjin (Grant 13JCQNJC03000) and
1964 Textile Research Journal 88(17)

the Technical Guidance Project of the China National Textile 11. Zheng ZR, Zhao XM, Wang CH, et al. Investigation of
and Apparel Council (Grant 2012002). automated geometry modeling process of woven fabrics
based on the yarn structures. J Text Inst 2015; 106:
References 925–933.
1. Wan XF and Fan JT. Heat transfer through fibrous 12. Li XH, Guan MH and Li J. Evaluation on thermal pro-
assemblies incorporating reflective interlayers. Int J tective performance of fabric for firefighter protective
Heat Mass Transf 2012; 55: 8032–8037. clothing. J Text Res 2015; 36: 110–115.
2. Li N and Zhu FL. Testing of flame resistance and ther- 13. Song GW, Cao W and Gholamreza F. Analyzing stored
mal protective performance of high-temperature resistant thermal energy and thermal protective performance of
clothing. Dyeing Finishing 2016; 9: 47–50. clothing. Text Res J 2011; 81: 1124–1138.
3. Raeisian L, Mansoori Z, Hosseini-Abardeh R, et al. An 14. Cheng L. Study on the thin flexible multilayer membrane
investigation in structural parameters of needle-punched and its thermal insulation. Shanghai: Donghua University,
nonwoven fabrics on their thermal insulation property. 2007.
Fibers Polym 2013; 14: 1748–1753. 15. Jhanji Y, Gupta D and Kothari VK. Thermo-physiolo-
4. Zhang RQ, Li JQ, Xu JJ, et al. Dynamic heat conduct- gical properties of polyester–cotton plated fabrics in rela-
ivity property of fabrics. Adv Mater Res 2011; 332: tion to fibre linear density and yarn type. Fashion Text
845–849. 2015; 2: 1–14.
5. Cubric IS, Skenderi Z, Mihelic-Bogdanic A, et al. 16. Gao WQ, Chen LH and Du BB. The density of basalt
Experimental study of thermal resistance of knitted fab- fibers was determined by the pycnometer method. China
rics. Exp Therm Fluid Sci 2012; 38: 223–228. Fiber Inspect 2010; 7: 52–53.
6. Lizak P and Mojumdar SC. Influence of the material 17. Lin H, Zeng XS, Sherburn M, et al. Automated geomet-
structure on the thermal conductivity of the clothing tex- ric modelling of textile structures. Text Res J 2012; 82:
tiles. J Therm Anal Calorimetr 2015; 119: 865–869. 1689–1702.
7. Li J, Barker RL and Deaton AS. Evaluating the effects of 18. Karleker BV and Desmond RM. Engineering heat trans-
material component and design feature on heat transfer fer. New York: West Publishing Co, 1980.
in fire fighter turnout clothing by a sweating manikin. 19. Kitamura K, Mototani K and Kimura F. Heat transfer
Text Res J 2007; 77: 59–66. of combined forced and natural convection from horizon-
8. Fan JT and Cheng XY. Heat and moisture transfer with tal cylinder to air. Heat Transf Asian Res 2007; 36:
sorption and phase change through clothing assemblies. 474–488.
Text Res J 2005; 75: 187–196. 20. Udayraj, Talukdar P, Das A, et al. Heat and mass trans-
9. Li YB and Fan JT. Transient analysis of heat and mois- fer through thermal protective clothing-a review. Int J
ture transfer with sorption/desorption and phase change Therm Sci 2016; 106: 32–56.
in fibrous clothing insulation. Numer Heat Transf A Appl 21. Wu HJ, Qian K and Cao HJ. Effect of fabric structure
2007; 51: 635–655. parameters on its heat transfer properties. J Text Res
10. Mao AH and Li Y. Numerical heat transfer coupled with 2007; 28: 21–31.
multidimensional liquid moisture diffusion in porous 22. Zhu GC, Kremenakova D, Wang Y, et al. Study on air
textiles with a measurable-parameterized model. Numer permeability and thermal resistance of textiles under heat
Heat Transf A Appl 2009; 56: 246–268. convection. Text Res J 2015; 85: 1681–1690.