Experimental study of moisture content effects on horizontal flame spread over thin cotton fabric

Abstract

In this paper, moisture content effects on horizontal flame spread were experimentally investigated using 0.245 mm thick, 28 cm tall and 28 cm wide untreated cotton fabric sheets with various moisture contents varying from 0 to 34%. The pyrolysis spread rates, flame heights and ignition times were obtained and analyzed. The corresponding results are as follows: as moisture content increases, the flame height and spread rate first increase and then decrease. In contrast, the ignition time shows an opposite trend with moisture content. The extreme values are observed in cases of 2% moisture content samples. Moreover, the flame spread rate in the warp direction is larger than that in the weft direction. For horizontal flame spread, the moisture content has the effect of consuming part of the heat feedback, which can play a role in reducing the flame spread rate; simultaneously, the moisture content can enlarge flame size and increase the convective heat transfer coefficient, thereby resulting in an increase in flame spread rate. The non-monotonous trend in pyrolysis spread rate is the result of competition between these effects.

Keywords

moisture content cotton fabric horizontal flame spread pyrolysis spread rate heat feedback

Textile fabric, a material commonly used in buildings, can be ignited by a relatively faint heat source; the resulting flame may propagate horizontally and vertically, eventually leading to an uncontrollable fire. Over the years, there has been significant progress in our understanding of flame spread over textile fabrics, which is inherently a complex phenomenon involving interactions between flows, heat and mass transfer. Segal and Drake¹ conducted experiments on horizontal flame spread over eight different cotton fabrics and found that there was a significant difference in flame spread rate between the warp and weft directions. Markstein and de Ris² experimentally studied upward flame spread over thin cotton fabric and developed a theoretical model demonstrating that the upward flame spread process asymptotically reached a steady state in a finite length. Kleinhenz et al.³ have performed experiments to investigate one-sided flame spread phenomena over thin composite cotton/fiberglass fabrics, and indicated that one-sided flames and spread rates were smaller than those of two-sided counterparts. Johnston et al.⁴ studied upward flame spread using a thin composite cotton/fiberglass fabric and explored the flame blow off phenomenon, with the remaining flame on the other side spreading all the way with a constant spread rate. However, most previous studies on flame spread over textile fabrics have not taken moisture content (MC) effects into account, but have instead used dry samples.

As we know, the air can be characterized by frequent and significant variation in relative humidity.⁵ Many materials, especially textile fabric, can absorb a certain amount of moisture from the air and experience a different MC accordingly. Therefore, it is necessary to carry out studies on flame spread characteristics over textile fabrics in the presence of MC. Segal and Drake¹ experimentally studied horizontal flame propagation of undyed cotton fabrics with two moisture sets of specimens, and reported that the propagation rate and change in rate were affected by the presence of moisture. Ying⁶ provided a theoretical model to clarify MC effects on flame spread over solid fuels, and stated that the flame propagation process was dominated by the dimensionless propagation speed, evaporation coefficient and combustion heat. It was also found that the flame spread speed could be slowed down by the process of moisture evaporation until the propagation completely stopped. Miller et al.⁷ performed experiments to explore MC effects on the flammability characteristics of textile materials using a flammability analyzer developed by Textile Research Institute (RTI) and demonstrated that there was a negative effect on mass transfer rate with the presence of MC. Gao et al.⁸ found that MC had a non-monotonous effect on upward flame spread over a thin cotton fabric. However, the effects of MC on flame spread over textile fabrics have not yet been adequately explained.

Horizontal flame spread over solid surfaces is relatively slow compared to other flame spread configurations, but is an important area of study due to its usefulness in theoretical development and practical applications.⁹ Recently, more attention has been paid to the study of the characteristics of horizontal flame spread over solid fuels. Fang et al.¹⁰ investigated horizontal flame spread characteristics over thin paper under varied oxygen concentrations and found a transition point for flame illumination, radiative heat flux and flame spread rate. Jiang et al.¹¹ systematically studied the combined effects of sample width and thickness on horizontal flame spread over thin polymethyl methacrylate (PMMA), and demonstrated that the total heat flux first decreased and then increased as the width of samples increased. It can be seen that a minimum flame spread rate exists under a certain range of sample width for steady burning horizontal flame spread. Moreover, the effects of width,^12–15 ambient pressure^13,16,17 and thickness¹⁸ on horizontal flame spread over extruded polystyrene, expanded polystyrene and wood surfaces have been studied quite extensively. However, all of these experimental and theoretical analyses were based on data obtained in the absence of MC.

In previous horizontal flame spreading experiments, the samples were all ignited by simultaneous ignition along one entire edge, and the flame propagated horizontally from one edge of the samples to the other. Moreover, the change range of MC was limited and only certain spreading characteristics were obtained in each MC effect experiment. In the present paper, a series of comparative experiments for untreated cotton fabrics with different MCs ranging from 0 to 34% (0, 2, 4, 6, 10, 14, 18, 22, 26, 30 and 34%) were carried out to study MC effects on horizontal flame spread. The test samples were ignited by a point flame placed beneath and in the middle of the sample. Measurements obtained included the visual observation of flame structure, flame spread rate, flame height and ignition time. The results of this study have implications concerning design for fire safety and may help advance our understanding of MC effects on flame spreading.

Experimental setup

Figure 1 schematically illustrates the experimental support frames. The test fabric was clamped between four pairs of identical aluminum sheet holders with seven screws, which were evenly distributed on the holders in increments of 5.0 cm in order to constrain fabric shrinkage and crinkling.

Figure 1.

Schematic of support frames. (a) Top view and (b) front view.

The data acquisition system was composed of two high-definition (HD) video cameras and a thermal infrared (IR) imager. One HD video camera was set perpendicular to the sample surface to record the flame front position, and the other one was set at the side of the sample to observe the flame structure and track the flame height. The video cameras were capable of imaging 20 frames/s, which were analyzed statistically to determine the flame front and flame height. Zhou,¹³ Tu¹⁶ and Yan¹² similarly employed video cameras to obtain the flame front and flame height in the horizontal flame spread process using statistical analyses in their studies. A high-frequency IR imager (MAGNITY-MAG32HF), set perpendicular to the sample surface, was used to record the sample surface temperature, from which the ignition time can be determined. In order to eliminate the band and continuous emissions, an 8–12 µm bandpass filter was used to obtain the true surface temperature.^19–21 Miller²⁰ and Zhu²¹ previously reported that the absolute surface temperature of PMMA can be successfully tracked by the IR thermography technique with a bandpass filter.

Since cotton fabric can burn without complicated phenomena, such as melting and dripping,² untreated cotton fabric sheets (with lengths of 30 cm, widths of 30 cm and thicknesses of 0.245 mm) were used as test samples. Square test samples were mounted with identical aluminum sheet holders to expose both faces, with 28 cm sample width and 28 cm sample length, to the atmosphere. In the horizontal flame spread experiments, the cotton fabric used was a clean and uncontaminated fabric. The woven fabric was from the same production batch with uniform characteristics, composed of 100% cotton with an area density of 143 g/m². There were 58 and 27 threads/cm in the warp and weft directions, respectively. In the present paper, the fabric samples were not subjected to any treatment, expect for preconditioning in the environmental chamber. To clarify the effects of MC on horizontal flame spread, test samples were preconditioned with different MCs (0, 2, 4, 6, 10, 14, 18, 22, 26, 30, and 34%). The conditioned procedure was as follows. Firstly, the test samples were dried in a forced-draft oven at 90^oC until the sample mass remained unchanged. Subsequently, the samples were removed from the oven and allowed to cool for 2.0 h in a desiccator.^5,7,22 This mass value was defined as the dry fuel mass, $M_{dry}$ . Secondly, each dry sample was placed in the desired humidity within an airtight conditioning chamber to equilibrate for about 24 h until the change in mass was minimal (0.01%). The mass change can be recorded by an electronic balance with an accuracy of 0.1 mg. The equilibrium mass was defined as the wet fuel mass, $M_{wet}$ . The MC can be determined by the following equation:

MC = \frac{M_{wet} - M_{dry}}{M_{wet}} \times 100 %

(1)

where

M_{wet}

represents the weight of wet fabric and

M_{dry}

the weight of dry fabric.

Finally, the conditioned samples were sealed in plastic bags before burning experiments to retain the MC. Each experiment was prepared for about 30 s before point flame application, and the SDs in MC were no more than 0.5%.

The test samples were ignited in the middle of the horizontal surface by a point flame of about 40 mm in length, which was placed about 20 mm beneath the sample surface. In this manner, the test samples with different MCs experienced different ignition time, which was the duration of point flame application. For consistency in comparison with each other, all the samples had an approximately circular pyrolysis zone with a radius of 0.8 cm. In order to verify the repeatability of experimental results, six repeats were performed at 0 and 2% MC conditions, while tests were repeated three times at other conditions. All of the experiments were carried out under controlled temperature and relative humidity conditions at 25±2℃ and 42 ± 3% relative humidity within a combustion chamber.

Results and discussion

Behaviors of flame spread

Figure 2 shows the time sequence images for flame ignition, growth and the horizontal spreading process for 6% MC samples. A similar phenomenon was observed for other MC samples, and this will be discussed as an example. The pyrolysis front, Xp, was defined as the distance from the middle of the fuel to the blackened front, which is shown by the red solid line in Figure 2. Gollner et al. previously defined the pyrolysis zone as the blackened area.²³ It can be seen that the pyrolysis front remains approximately “circular” in shape shortly after ignition, a gradually becomes an “oval” shape. This trend indicates that the horizontal flame spread rates in the warp and weft directions are significantly different, which shows good agreement with the results of Segal and Drake.¹ Based on the above analysis, the flame spreading over the cotton fabric is characterized by anisotropy. In order to compare the MC effects on flame spreading, a simple method is proposed in this work to obtain the characteristic pyrolysis front. As can be seen from the instantaneous image at 35 s, the pyrolysis front (Xp) in the weft direction is defined as the average distance from the middle of the fuel to the green solid points 1 and 2, while Xp in the warp direction is defined as the average distances of points 3 and 4.

Figure 2.

Typical cases of horizontal flame spread for 6% moisture content samples.

Figure 3 shows front view images of the horizontal flame spreading process taken at 20 s for various MC samples. It can be clearly seen that the pyrolysis front propagates more slowly and that the area of the blackened zone is smaller with increasing MC, and that the flame completely fails to propagate at 34% MC. This trend is consistent with Ying's results⁶ that the flame spread can come to a stop when the evaporation coefficient is large enough (i.e. the MC is large enough). However, the flame spread behavior with 2% MC did not follow this trend and a maximum blackened area was produced. The above observations indicate that the presence of MC may have a non-monotonous effect on horizontal flame spread over fabric fuels.

Figure 3.

Front view images taken at 20 s for various moisture content samples.

Flame spread rate

Figure 4 shows the time evolutions of the pyrolysis fronts in the warp and weft directions for 6% MC test samples. For an initial period, the flame remains laminar and the virgin surface can be heated by a small amount of heat flux, resulting in a slow flame spread rate. Due to the initial circular pyrolysis zone and the slow spread rate, X_p in different directions are similar at the early stages. However, from 15 s, X_p in the warp direction grows more rapidly than that in the weft direction, and the difference between the two directions gradually increases as time goes on. At 30 s, X_p in the warp direction roughly exceeds 85.5 mm, while X_p in the weft direction is only 70.5 mm. It is also found that the trends of X_p in the warp and weft directions increase non-linearly until the pyrolysis front reaches the edge of the sample. This trend is different from some other studies using simultaneous ignition along one entire edge,^11,17,24 in which the horizontal flame spreading can attain a steady state after a short initial period. This discrepancy may primarily be due to the difference in ignition source and flame structure. The vertical error bars indicate the SDs of three tests, which are lower than 8.9%.

Figure 4.

Pyrolysis front, X_p, for 6% moisture content samples as a function of time.

While the flame spread does not reach a steady state, the change in spread rate over the relatively small distance is not great.²⁵ Therefore, the average spread rate, V_p, between distances of 10 and 14 cm from the center of the samples is determined by the slopes of linear fit, as shown in Figure 5. It is obvious that the trends of V_p in the weft and warp directions decrease with MC except for MC = 2%. For 0% MC test samples, the values of V_p in the weft and warp directions are 3.83 and 4.10 mm/s, respectively, while the corresponding values for 2% MC samples are 4.12 and 4.29 mm/s (i.e. they speed up by 0.29 and 0.19 mm/s, respectively). As MC increases from 4 to 34%, V_p decreases significantly until the flame cannot propagate continuously at 34% MC. It could be deduced that there is a non-monotonous effect of MC on horizontal flame spread over textile fabrics. The specific contribution of MC in fabric fuels will be addressed in a later section. Moreover, there are notable differences in V_p between the weft and warp directions. The error bars represent the average SDs of repeated tests ranging from 4.9 to 10.1%.

Figure 5.

Pyrolysis spread rate, V_p, as a function of moisture content.

Flame height

The flame height, H_f, is an important parameter characterizing the horizontal flame spread behaviors of textile fabrics. Figure 6 shows the typical flame shapes of 6% MC test samples recorded by the side view video camera. It is found that at the initial time of ignition, the pyrolysis gas is concentrated in the central area due to air entrainment from four sides, forming a peak height flame. At this stage, the flame height (H_f) can be defined as the vertical distance from the surface to the tip of the flame peak, which is shown by the black solid line in Figure 6(a); once the test samples are burned through, the air entrainment can also occurr from the center of the samples leading to multiple-peak height flames, as shown in Figure 6(b). At this stage, H_f is defined by the average value of the multiple-peak flame heights. This phenomenon is caused primarily by a transition in air entrainment.

Figure 6.

Side view images of typical flame shapes for 6% moisture content samples, (a) before and (b) after they are burned through.

Figure 7 shows the development of flame height for 6% moisture content samples as a function of time. It is obvious that the development of flame height over a horizontal fuel surface can be divided into three stages, the growing stage, steady-state stage and decaying stage, which are similar to what has been reported for other MC samples. After the test samples are ignited, the flame height shows an obvious increase followed by a steady-state stage. When most of the fuel is consumed, the flame gradually decreases until the fire fades. The average flame height $\bar{H_{f}}$ in the steady-state stage is obtained by a linear function, which is shown by the black dotted line in Figure 7.

Figure 7.

Flame height, H_f, for 6% moisture content samples as a function of time.

The variation in $\bar{H_{f}}$ , estimated after the time that the steady-state stage is reached, is plotted as a function of MC in Figure 8. It can be seen that the variation in $\bar{H_{f}}$ first increases and then decreases as MC increases. For 0% MC test samples, the value of $\bar{H_{f}}$ is 80.4 mm, while the corresponding value for 2% MC samples is 88.7 mm (i.e. it is lengthened by 8.3 mm). As MC increases from 4 to 34%, $\bar{H_{f}}$ decreases almost linearly until the flame no longer propagates continuously at 34% MC. It is clear that MC has the effect of enlarging the size of the flame to some extent. Prationo et al.²⁶ have suggested that unevaporated moisture could be released with volatile compounds to form a thick cloud layer, thereby enlarging the size of flame significantly.

Figure 8.

Average flame height $\bar{H_{f}}$ as a function of moisture content.

Ignition time

Analysis of the temperature profile is important when studying horizontal flame spread over a solid fuel surface. Figure 9 shows the typical temperature profile for 6% MC test samples recorded by an IR imager. The surface temperature undergoes three stages, including the preheating stage (AC stage), pyrolysis stage (CD stage) and char burning stage (DE stage). It should be noted that the preheating stage can be divided into two zones according to different temperature increase rates. In zone 1 (AB stage), the temperature slowly increases to 100℃ and some of the MC is evaporated. The temperature then rises rapidly until the pyrolysis temperature is reached in zone 2 (BC stage). The duration times of zones 1 and 2 can be defined as $t_{1}$ and $t_{2}$ , respectively, while the overall duration of the preheating stage is defined as the ignition time $t_{ig}$ . Beyond point C, the rise rate of temperature becomes very slow and reaches an almost constant value, corresponding with the pyrolysis stage. This is due to the fact that most of the received heat flux is consumed by the pyrolysis process. Subsequently, the residual char continues to burn in the solid phase, and the temperature first rises and then begins to drop.

Figure 9.

Typical temperature profile for 6% moisture content test samples.

To indicate the effects of MC, the temperature profiles 10 cm from the sample centers in the weft and warp directions for different MC test samples were obtained and analyzed. Figure 10 shows comparisons of $t_{1}$ and $t_{2}$ in the weft and warp directions for different MC test samples. It can be seen that there is a significant increase in $t_{1}$ with increasing MC. The evaporation of moisture can consume part of the heat flux received by the sample surface, eventually resulting in a longer period of time being required for the temperature to increase to 100℃. However, the trend of $t_{2}$ shows a decrease with increasing MC. Such evidence strongly supports the notion that the moisture is not completely evaporated in zone 1, which is in good agreement with the previously reported results of Prationo.²⁶ In zone 2, the unevaporated moisture continues to vaporize, which can increase the convective heat transfer coefficient in the vicinity of the preheating area, thereby strengthening the convective heat transfer. Safronava et al.²⁷ have previously reported that the phase change from bound moisture to steam could significantly reduce the preheating time of polymers compared to dry samples.

Figure 10.

$t_{1}$ and $t_{2}$ as a function of moisture content.

As mentioned above, the ignition time $t_{ig}$ , a sum of $t_{1}$ and $t_{2}$ , is plotted as a function of MC in Figure 11. It can be observed that the trend of $t_{ig}$ shows a non-monotonous increase with MC. For 0% MC test samples, the values of $t_{ig}$ in the warp and weft directions are 24.75 and 25.5 s, respectively. However, for 2% MC samples, the corresponding values of $t_{ig}$ decrease to 23.5 and 24.5 s, and are smaller than those for other MC samples. As MC increases from 4 to 34%, $t_{ig}$ significantly increases until the flame completely stops at 34% MC. Moreover, $t_{ig}$ in the weft direction is larger than that in the warp direction for different MC samples. The vertical error bars indicate the SDs of repeated tests, which are lower than 8.2%.

Figure 11.

Ignition time $t_{ig}$ as a function of moisture content.

Theoretical analysis of moisture effects

To the best of our knowledge, flame propagation over a solid surface is dominated by heat feedback from the flame to the virgin fuel surface.^17,28,29 For a thermally thin material, conductive heat transfer through the solid phase is found to be negligible,^30,31 while heat flux from the gas phase is the dominant heat transfer path.^31,32 The heat transfer in the gas phase from a flame to a virgin surface consists of radiative and convective components.^17,24,31 The energy conservation can be expressed by the following equation:

\overset{\cdot}{q} ″_{f} = \overset{\cdot}{q} ″_{r} + \overset{\cdot}{q} ″_{c} - \overset{\cdot}{q} ″_{loss}

(2)

where

\overset{\cdot}{q} ″_{f}

represents the net heat flux received by the fuel surface,

\overset{\cdot}{q} ″_{r}

and

\overset{\cdot}{q} ″_{c}

are the radiation and convection heat flux from the flame, respectively, and

\overset{\cdot}{q} ″_{loss}

represents the heat loss from the surface by radiation. Since the

\overset{\cdot}{q} ″_{loss}

is not of the same magnitude as the heat feedback from a flame in the gas phase,^33,34 the heat transfer from the flame becomes the key mechanism for horizontal flame spread over a thin solid surface.^12,13,24,35 The radiation and convection heat flux can be expressed as:

\overset{\cdot}{q} ″_{r} = χ ɛ σ {T_{f}}^{4}

(3)

\overset{\cdot}{q} ″_{c} = \bar{h} (T_{f} - T_{s})

(4)

where χ, ɛ and σ represent view factor, emissivity and Stefan–Boltzmann constant, respectively, and

\bar{h}

T_{f}

and

T_{s}

denote convective heat transfer coefficient, flame temperature and fuel surface temperature, respectively.

According to moisture sorption theories for cotton fabrics, the total moisture absorbed by a cotton fabric can be divided into internal and external moisture forms.^36,37 In the preheating stage, the test sample surface is heated by heat feedback from the flame and then evaporation occurs, which consumes part of the heat flux. However, the water vapor in the vicinity of the preheating area can increase the convective heat transfer coefficient, $\bar{h}$ , eventually resulting in an increase in $\overset{\cdot}{q} ″_{c}$ . Li et al.³⁸ have inferred that the existence of water vapor may increase the convection heat transfer coefficient and that convection heat transfer could then be strengthened.

During the preheating stage, the absorbed moisture cannot be completely evaporated.²⁶ In the pyrolysis stage, the remaining moisture is evaporated simultaneously with fuel volatile compounds, which may greatly weaken flame intensity. However, the evaporation of moisture together with fuel volatile compound release can form a relatively thicker flame layer, thereby enlarging the size of the flame significantly. Prationo et al.²⁶ have reported that for wet samples, unevaporated moisture is released together with fuel volatile compounds and that flame height is approximately twice that of the size of the dried samples. Moreover, as the flame height increases, the view factor,χ , between the flame and the preheating zone shows an increase,^12,31 thereby leading to an increase in $\overset{\cdot}{q} ″_{r}$ .

In summary, the effects of MC can be explained as follows: (a) the evaporation of moisture consumes an amount of heat flux; (b) the phase change of moisture can increase the convective heat transfer coefficient, $\bar{h}$ , resulting in an increase in convection heat flux; and (c) the simultaneous release of moisture and fuel volatile compounds leads to an enlargement of flame size, thereby resulting in an increase in the view factor and radiation heat feedback. The horizontal flame spread rate can be reduced by the first moisture effect, but is increased by the last two effects. For 2% MC samples, the last two effects (increasing effects) are dominant and the pyrolysis spread rate is larger than that for other MC samples. However, with an increase in MC, the first moisture effect (decreasing effect) is increasingly dominant and the pyrolysis spread rate decreases until the flame no longer propagates continuously. Therefore, the increasing or decreasing effects of moisture on horizontal flame spread over cotton fabrics depend on the balance of these mechanisms.

Conclusions

A series of experiments were carried out using 0.245 mm thick, 28 cm tall and 28 cm wide untreated cotton fabric sheets, with different MCs varying from 0 to 34%, to investigate non-monotonous moisture effects on horizontal flame spread. Some key characteristic parameters of flame spread were analyzed, such as the flame shape, flame height, ignition time and pyrolysis spread rate. The major conclusions can be drawn as follows:

As MC increases, the flame height and spread rate first increase and then decrease. In contrast, the ignition time shows an opposite trend with MC. Extreme values are observed in the case of 2% MC samples. Moreover, the pyrolysis spread rate in the warp direction is larger than that in the weft direction. A hypothesis of moisture effects is provided: the MC has a decreasing effect by consuming some of the heat flux, and increasing effects by enlarging flame size and increasing the convective heat transfer coefficient. When MC is low, the horizontal flame spread is controlled by increasing effects; as MC increases, the decreasing effects are increasingly dominant. The non-monotonous trend in spread rate is the result of competition between these effects.

In the present paper, the effects of MC on horizontal flame spread over cotton fabric fuels are adequately explained, which is especially important for flame retardant treatment for fabric fuels. Based on the above analyses, when MC exceeds 4%, the decreasing effects of MC are dominant and the flame spread rate decreases. Thus. the MC regain in fabric fuels can act as a natural flame retardant to suppress the flame spreading process. When treated with a hydrophilic flame retardant, fabric fuels could absorb relatively more moisture. Due to the combined effects of moisture and flame retardants, hydrophilic flame retardants may have better flame retardant effects on fabric fuels. Therefore, a further study is in progress to investigate the combined flame retardant effects of MC and flame retardants on horizontal flame spread over cotton fabric fuels.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by National Key R&D Program of China (Project No. 2016YFC0802900), Fundamental Research Funds for the Central Universities (No. 2018BSCXC02), Postgraduate Research & Practice Innovation Program of Jiangsu Province (No. KYCX18_1914), Fire Fighting and Rescue Technology Key Laboratory of MPS Open Project (No. KF201802), Sichuan Science and Technology Project (No. 2018JY0429), National Natural Science Foundation of China (Project No. 51606215), and Natural Science Foundation of Jiangsu Province (Project No. BK20160270).

References

Segal

Drake

Jr . The horizontal flame-propagation rate of undyed cotton fabrics. Text Res J 1976; 46: 238–246.

Markstein

de Ris

. Upward fire spread over textiles. Symp (Int) Combust 1973; 14: 1085–1097.

Kleinhenz

Ferkul

Pettegrew

et al.

One-sided flame spread phenomena of a thermally thin composite cotton/fiberglass fabric. Fire Mater 2005; 29: 27–37.

Johnston

T'ien

Muff

et al.

Self induced buoyant blow off in upward flame spread on thin solid fuels. Fire Saf J 2015; 71: 279–286.

Krasny

Braun

. Textile flammability testing: appropriate levels for moisture content of specimens. Fire Mater 1982; 6: 38–41.

Ying

. Flame propagation of burning solid material with moisture. Fire Technol 1971; 7: 243–250.

Miller

Martin

Goswami

et al.

The effects of moisture on the flammability characteristics of textile materials. Text Res J 1975; 45: 328–337.

Gao

Zhu

et al.

Experimental study of the non-monotonous moisture effect on upward flame spread over cotton fabric. Text Res J. 2018; 88: 2379–2394.

Korobeinichev

Gonchikzhapov

Tereshchenko

et al.

An experimental study of horizontal flame spread over PMMA surface in still air. Combust Flame 2018; 188: 388–398.

10.

Fang

X-z

K-y

et al.

Transition condition and control mechanism of subatmospheric flame spread rate over horizontal thin paper sample. Combust Flame 2018; 188: 90–93.

11.

Jiang

Miller

Gollner

et al.

Sample width and thickness effects on horizontal flame spread over a thin PMMA surface. Proc Combust Inst 2017; 36: 2987–2994.

12.

Yan

Shen

et al.

Experimental study on the width and pressure effect on the horizontal flame spread of insulation material. Int J Therm Sci 2017; 114: 114–122.

13.

Zhou

Xiao

Sun

et al.

Experimental study of horizontal flame spread over rigid polyurethane foam on a plateau: effects of sample width and ambient pressure. Fire Mater 2015; 39: 127–138.

14.

Jiang

Xiao

Zhou

et al.

Theoretical and experimental study of width effects on horizontal flame spread over extruded and expanded polystyrene foam surfaces. J Fire Sci 2013; 32: 193–209.

15.

Zhang

Huang

et al.

Effects of sample width on flame spread over horizontal charring solid surfaces on a plateau. Chinese Sci Bull 2011; 56: 919–924.

16.

Zeng

Fang

et al.

The influence of low air pressure on horizontal flame spread over flexible polyurethane foam and correlative smoke productions. App Therm Eng 2016; 94: 133–140.

17.

Zhang

Huang

Wang

et al.

Experimental study on the characteristics of horizontal flame spread over XPS surface on plateau. J Hazard Mater 2011; 189: 34–39.

18.

Huang

Wang

Zhang

et al.

Thickness effect on flame spread characteristics of expanded polystyrene in different environments. J Thermoplast Compos Mater 2011; 25: 427–438.

19.

Urbas

Parker

Luebbers

. Surface temperature measurements on burning materials using an infrared pyrometer: Accounting for emissivity and reflection of external radiation. Fire Mater 2004; 28: 33–53.

20.

Miller

Gollner

. Upward flame spread over discrete fuels. Fire Saf J 2015; 77: 36–45.

21.

Zhu

Gao

et al.

Experimental studies on the effects of spacing on upward flame spread over thin PMMA. Fire Technol 2017; 53: 673–693.

22.

et al.

The effect of air gaps in moist protective clothing on protection from heat and flame. J Fire Sci, 201231: 99–111.

23.

Gollner

Williams

Rangwala

. Upward flame spread over corrugated cardboard. Combust Flame 2011; 158: 1404–1412.

24.

Zhang

Sun

Huang

et al.

Heat transfer mechanisms in horizontal flame spread over wood and extruded polystyrene surfaces. Int J Heat Mass Transf 2013; 61: 28–34.

25.

Gollner

Huang

Cobian

et al.

Experimental study of upward flame spread of an inclined fuel surface. P Combust Inst 2013; 34: 2531–2538.

26.

Prationo

Zhang

Cui

et al.

Influence of inherent moisture on the ignition and combustion of wet Victorian brown coal in air-firing and oxy-fuel modes: Part 1: The volatile ignition and flame propagation. Fuel Process Technol 2015; 138: 670–679.

27.

Safronava

Lyon

Crowley

et al.

Effect of moisture on ignition time of polymers. Fire Technol 2014; 51: 1093–1112.

28.

Avinash

Kumar

Raghavan

. Experimental analysis of diffusion flame spread along thin parallel solid fuel surfaces in a natural convective environment. Combust Flame 2016; 165: 321–333.

29.

Rakesh Ranga

Korobeinichev

Harish

et al.

Investigation of the structure and spread rate of flames over PMMA slabs. App Therm Eng 2018; 130: 477–491.

30.

Ohlemiller

Shields

. One and two sided burning of thermally thin materials. Fire Mater 1993; 17: 103–110.

31.

Zhang

et al.

Effects of altitude and sample width on the characteristics of horizontal flame spread over wood sheets. Fire Saf J 2012; 51: 120–125.

32.

Ito

Kashiwagi

. Characterization of flame spread over PMMA using holographic interferometry sample orientation effects ⋆. Combust Flame 1988; 71: 189–204.

33.

Huang

Chen

Liu

et al.

Thermal analysis of vertical upward flame spread and dripping behaviors of polystyrene foams at different altitudes. J Macromol Sci B 2017; 56: 517–531.

34.

Quintiere

. Fundamentals of fire phenomena, Chichester: John Wiley & Sons, 2006.

35.

Chen

Liu

Zhou

et al.

Experimental and theoretical analysis on lateral flame spread over inclined PMMA surface. Int J Heat Mass Transf 2015; 91: 68–76.

36.

Hill

CAS

Norton

Newman

. The water vapor sorption behavior of natural fibers. J Appl Polym Sci 2009; 112: 1524–1537.

37.

Ceylan

Landuyt

Meulewaeter

et al.

Moisture sorption in developing cotton fibers. Cellulose 2012; 19: 1517–1526.

38.

X-h

Y-h

et al.

A new approach to evaluate the effect of moisture on heat transfer of thermal protective clothing under flashover. Fibers Polym 2012; 13: 549–554.