Comparison of Fire Behavior of Thermally Thin and Thick Latex Foam Under Bottom Ventilation

Fire behavior of natural rubber latex foam under different thickness conditions(d=1,2and 5cm)were explored though a little of experiments by using the self-built small scale experimental platform. It can be shown that the flame spread law of thermally thin and thermally thickness are different. Natural rubber latex foam with thickness of 2cm show higher fire risk, which value of flame spread rate, maximum flame height, maximum mass loss rate and maximum temperature is 0.00293m/s, 851.875mm, 1.83g/s, 948K,repectively.That may because the thickness of residue formed of thermally thick materials is larger than the thin one, obstructing the contact of the natural rubber latex foam with fresh air .In addition, a special phenomenon is noticed that during the second stage, the bottom unburned zone located in the four edges(thermally thin material) and middle player(thermally thick material).

as polyethylene [7], polymethylmethacrylate(PMMA) [8], XPS foam and polyurethane foam [9,10], epoxy containing ester [11], and so on [12].According to its thermal penetration depth , layered combustibles are often classified as thermally thin and thick materials [13].Similarly, Jinlong Zhao [14] studied whether the four burning period(pre-heating, steady burning, thin-layer burning and extinguishment stage) will appear or not depends on the thickness of the fuel samples, and the results show that the heat flux feedback to sample surface is depend on the thickness of samples.
Further more, some scholars focus on the influence of thickness on the flame spread with solid samples [13,15].For example, the effect of thickness on plywood vertical spread have been studied by Xuanmeng Qing [15], which show that the thickness of samples will affect the regulation of flame spread ,and the flame spread rate will decrease when thickness of samples increases.Lin Jiang [13]give a similar conclusion .The effect of bottom ventilation [16], ignition position [17] on the flame spread of natural rubber latex foam has been studied by previous researches of our groups.It indicated that materials under bottom ventilation(BV) and center ignition position show more higher fire risk by comparing the value of flame temperature, flame height and mass loss rate et al.
However, the flame spread behavior of thermally thin materials and thermally thick materials of natural rubber latex foam are rarely compared.
As a new type of upholstered furniture, natural rubber latex foam has been widely concerned, but there are few studies on its fire risk.Considering that different thickness of latex foam is widely used ,the purpose of studying the fire behavior of natural rubber latex foam under different thickness conditions is to understand the characteristics of fire spread, which can provide theoretical support for fire spread risk assessment of natural rubber latex foam .In this paper, the combustion behavior of latex foam with three different thicknesses (1cm,2cm and 5cm) is compared by using a small scale experiment platform, such as flame propagation speed, flame height, flame temperature, mass loss rate and so on to address the relevant fire risks.The research results provide a theoretical basis for fire risk assessment of typical cushion materials under spontaneous combustion conditions.

2.Experimental methodology
The small-scale platform which made by our group was adopted in this experiment, as shown in fig.2(a),which includes two parts: burning system and measurement system.The introduction of the burning system is detailed in article of our groups [16].The measurement system included thermocouples, HD video cameras, weighbridge, and data acquisition system.During each experiment , the temperatures along the horizontal direction of sample were recorded by some K-type-0.5mmthermocouples with a measurement range of -200℃ to 1300℃.The materials was ignited at the center point, as time go on, flame will spread from this ignition point to the four edges.So those thermocouples of T1-T6 were placed along a half-diagonal of the top surface from the center to the edge, as seen in fig.2(b).The universal several bus(USB) interface of 34908A module data-acquisition is used to connect with the personal computer for easy viewing and recording data.
Moreover, three HD video camera with angle of 45 °,0°,and-45° from the horizontal were used to photograph the experiment procedure.We draw a parallel line every 2.5 cm ,with a black marker, on the surface of the specimen to evenly divide the latex foam into square pieces with sides of 2.5 cm×2.5cm, as seen in Fig2(a)，which is not only helps us to determine the location of the fire front but also facilitates the proportional relationship between the picture space and the actual space.
Based on this, the relevant image processing software can be used to obtain the morphological parameters of the flame, such as flame height, flame front position, etc.Some 25 cm×25 cm natural latex foams with the thickness of 1, 2 and 5cm were prepared as shown in fig.2(b).The samples are divided into 100 small squares with black Mark pen to observe the change of the position of pyrolysis.All samples were selected from the same piece of natural rubber latex foam material with the same air pore size distribution to reduce the effect of uncontrollable parameters on the combustion characteristics of the sample.
Samples were placed in a drying oven for 12 hours before the experiment started under the condition that the values of temperature and humidity are 20℃ and 40%.Subsequently, it was placed on the wire mesh and the surface of the sample is blown with high-speed air to remove surface dash.In the course of the experiment, every time we used the same lighter to ignite the center of sample while ensuring the other conditions are the same.In order to test the accuracy of the results, each experiment repeated twice showing a good repeatability.The combustion process of all three kinds of materials can be divided into this three stages and the biggest difference is the duration of each stage.The duration of each stage under three thickness conditions and the average flame height are highlight in tab.1.As is apparently shown by the pictures above , stage one, the fire began to spread from the center spot to the marginal zone.The three stage combustion duration of natural rubber latex foam with thickness of 5mm is longer(357s) compared with those thin materials(1,2cm).In addition, during stage II, sample with 2 cm thickness burns most fiercely, and average flame height (0.68m) is higher than the materials with thickness of 1cm (0.53m)and 5cm(0.59m),asshown in tab.1 and fig. 3. It shows a similar trend during the last stage that the flame gradually decreased until fire was extinguished when the combustible was gradually consumed.Meanwhile, we can clearly see that some sample residues attached to the stainless steel mesh.m/s respectively.For the test, the flame spread rate of natural rubber latex foam with thickness of 1cm and 2cm is similar and far greater than that of 5cm one , which may indicate that sample with a thickness of 1cm and 2cm showed higher fire risk than others.

Mass loss
Meanwhile, the mass loss rate of samples with d=1,2,5cm is depicted in fig.6. Obviously, the quality of different thickness samples all decrease slowly and then decrease sharply, and the mass loss rate of samples with d=2cm is the most greatest.The peak mass loss rates of samples with d=1,2and 5cm are 1.83,1.91,0.79%wt/s, respectively ,which all appear near the end of the second stage ,and the mass loss rate increased significantly after bottom ignited.thick sample is higher than the others.This may because, like the 1cm one, the bottom of the 2cm sample burns out quickly and has more combustibles than 1cm, it shows a higher flame height.
Meanwhile, an obvious phenomenon deserves our attention, stable stage duration of samples with different thickness is longer with increasing sample thickness and the stable combustion stage of 5cm one is much longer than that of the others.In the last stage (stage III), the value of the flame height dropped sharply with residue, and this process has the same tendency with all thickness(1,2,5cm) natural rubber latex foam.

Temperature profiles
Flame spread is the process that the test specimens are heated and ignited by the energy from the burned region in essential [20].In this paper, understanding the temperature profile is contribute to explore the regulation of flame spread of with different thickness of natural rubber latex foam.
Fig. 8 shows the temperature profiles of the natural rubber latex foam surface with different thickness(d=1,2and5cm) during the flame spread process .As we have seen in fig.8(a)-(c),all the trend of surface temperature increased firstly then declining, which is similar to flame spread regulation of polyurethane foam [19].It can be proved by Fig. 3 and Fig. 8 that flame spreading over natural rubber latex foam surface undergoes three stages.resulting in TC6 separated from the surface and shifted to flame region rapidly.
During stage Ⅱ, according to fig.8(a)and(b),the temperature under the conditions of thickness of 1cm and 2cm reach to peak value, which is around 700-900℃.But for the 5cm one ,it keeps stable, which is about 600℃ except TC6, and it reaches the maximum at the early time of the third stage, as shown in fig.8(c).There is a marked increase change in temperature of all thermocouples under three different conditions, which all increased from 400℃ to the maximum value .Moreover, the jump of temperature of six thermocouples that located on the top surface of samples with different thickness (d=1,2 and 5cm) have been compared in the tab.2, and the average jump of the temperature of six thermocouples is 240.761℃,296.216℃ and 237.765℃ respectively , which can show the intensity of combustion.That is to say the 2cm one burn more fiercely.The reason for these differences is that the 1cm and 2cm materials are burned out quickly, and the top and bottom surfaces of the samples are burned simultaneously, which makes the combustion more intense.On the contrary, because the sample with a thickness of 5 cm is one surface combustion initially, there is a longer stable combustion stage before the bottom combustion.
Based on the experiment video footage recording by the camera c, we have noticed that the bottom ignition time for the three samples is 49s,78s,257s,respectly.Moreover, the maximum temperature of 2cm sample is 948K,which is higher than that of sample with thickness of 1cm(791℃,) and 5cm(893℃) .As time went by, those thermocouples escape from upper surface and shifted to flame zone leading to improvement of temperature.As the combustion came to stage Ⅲ, under all thickness conditions, the material was gradually exhausted and the temperature obtained by TC1 -TC6 dropped sharply.

4.Further discussion
First of all, the value of thermal penetration depth( t  ) in flame spreading is about 3.8cm according to previous studies of our group [16].When the specimen thickness is no less than the heat-penetration thickness, the material is regarded as a thick material.Otherwise, the result turns out contrary.Thus, in this paper, samples with a thickness of 1, 2cm are considered as thin-thermal material and 5 cm thick natural rubber foam are called thermal thick material.The TG test results show that after the combustion of the natural rubber latex foam, the residual residue is about 24.8% of the original, which is not a thermoplastic material [3].Moreover, Due to the different combustion modes, the existence of the residue is found to be different.As shown in fig.9, a half-section view, yellow color represents the natural rubber latex foam and black represents residue after combustion and the some neat holes represent the through-hole of natural rubber latex foam with an aperture of 6mm and a spacing of 30mm.The residue foamed during the combustion of the natural rubber latex foam with d=2cm and the thickness of residue of thermally thick materials is lager than the others.
Differently, seen in Fig. 10,thermally thin materials(d=1,2cm), the residue is thin.Consequently, the relative surface area of air contact with latex foam is larger that the fresh air can help the burning, that is to say thermally thin natural rubber latex foam can burn more fully than the thick one.

Figure 2 .
Figure 2. (a) Schematic of small-scale flame spread setup in this study (b)Tested natural rubber latex foam sample and thermocouples positioned along the horizontal direction (T1 to T6).

Preprints
(www.preprints.org)| NOT PEER-REVIEWED | Posted: 2 November 2018 doi:10.20944/preprints201811.0049.v1Peer-reviewed version available at Polymers 2019, 11, 88; doi:10.3390/polym110100883.Results and discussion3.1 Flame behaviorComparison of the flame shapes of different thickness samples is helpful to explore the difference of the surface flame propagation behavior.The variation of the natural rubber latex foam flame sequences with thickness d=1,2,5cm is shown in fig.3ThreeHD video cameras , placed in the designated area, recorded the whole process of the fire spread, which helped us clearly observed the three different periods of the natural rubber foam combustion: Stage I is initial combustion phases stage(from the moment that the center point of material was ignited to the fire spread across the entire top surface), stage II, so called "stable combustion stage"(from the end of initial combustion phases stage to the moment that flame spread to all natural rubber latex foam surfaces), stage III means fire weaken stage (from the end of stage II to fray-out of flame moment), which is very similar to the combustion process of polyurethane foam[9,18,19].

Figure 3 .
Figure 3. Flame spread process of natural rubber latex foam with different thickness (d=1,2 and 5cm)

Figure 4 .
Figure 4.The flame spreading law at the bottom of the sample with different thickness: (a)1cm and 2cm (b)5cm

Figure 7 Fig. 1 ,
Figure 7 presents the change of flame height with times under different thickness.Based on the Fig.1, the burning process can be detailed and the average flame height for the sample of three different thickness (1cm,2cm,5cm)are 0.53m,0.68m,0.59mrespectively, as shown in Table1.During stage I, the thickness has little influence on the flame height.During stage II, the flame height of the sample reached its maximum for all thickness samples.Clearly, the maximum flame height of 2cm

Figure 7 .
Figure 7. Flame height versus time for the three thickness samples

Figure
Figure indicated that the temperature change trend and the surface average temperature are very similar during stageⅠ for the different thickness samples.The maximum temperature at most of TC point is around 400-500℃.This maybe because that those thermocouples were located on the pyrolysis region.However, a special phenomenon that temperature of TC6,as seen in fig.8(c), rosesharply from the lower one to 680℃ .It probably because that the edge of the sample shrunk

Figure 8 .
Figure 8. Surface temperature field changes of natural latex foam: (a)1cm;(b)2cm;(c)5cm Moreover, because of the different distribution patterns of combustion residue, the flame spreading modes of thermal thick and thin natural rubber latex foam are different.As shown in fig.5andfig.6,the value of flame spread rate and mass loss rate of thermally thin materials is larger than the thick one.At the same times, during this burning process, the value of profile temperature and flame height of samples with d=2cm are the largest, as seen in fig.8and tab.2.In other words, the fire spread speed of thermally thick material may lower than that of thin materials.Preprints (www.preprints.org)| NOT PEER-REVIEWED | Posted: 2 November 2018 doi:10.20944/preprints201811.0049.v1Peer-reviewed version available at Polymers 2019, 11, 88; doi:10.3390/polym11010088

Table 1 .
The duration of the three stages and the average flame height for the three thickness samples Preprints (www.preprints.org)| NOT PEER-REVIEWED | Posted:

Table 2 .
Jump of temperature of thermocouples