Universal evaluation of fire retardant properties of fiber mats by the optimal cone calorimeter

Abstract

By using the optimal cone calorimeter, the universal method to study the fire retardant properties of fiber mats with different categories was explored. The challenge in obtaining accurate and repeatable cone calorimeter data was overcome from two aspects, employing a specific fixing mode and optimizing the measurement parameters. It was confirmed that the optimal grid covering rate of 25% and sample weight of 6 g favored repeatable and reliable data. With the improved cone calorimeter method, the reactions to fire of the selected fibers, involving the common and the inherent fire retardant fibers, are analyzed under different heat fluxes, in which the variations of time to ignition, the heat release rate and total heat release were compared. The fire retardant properties of various fiber mats were efficiently obtained by the improved cone calorimeter, which in turn verified its universality. In addition, the comparison between fire retardant properties and heat resistance of fibers demonstrated that the heat resistance is not the sole factor determining the fire retardant performance; the released fuel is much more important than the decomposition temperature. In conclusion, the universal evaluation method for combustion performance of fiber mats by the optimal cone calorimeter has been successfully established, which can serve as an efficient way in predicting the fire retardant properties of the corresponding fabrics.

Keywords

cone calorimeter universal method fiber mats fire retardant properties heat resistance

Fire, with enormous energy, can greatly benefit humans, but it can also cause disaster. Once in a fire attack, fire retardant (FR) garments are particularly necessary to protect fire fighters, soldiers and the mine workers from danger. For the textiles used in daily lives and the public systems, such as beddings and carpets, should have the ability of reducing the possibility of fire spread from weak to severe. In this condition, a variety of modification methods, such as coatings^1–5 and in situ polymerization,^6–9 have been applied to natural and synthetic textiles to improve FR properties. In addition, some inherent FR fibers, such as poly-m-phenylene isophthal-amide (PMIA),^10,11 polybenzoxazole (PBO),^12,13 poly(1, 3, 4-oxadiazole)s (POD)^14,15 and polybenzimidazole (PBI),¹⁶ also have been extensively investigated.

However, the available evaluation method for flame retardant properties, such as the limiting oxygen index (LOI) and vertical burning testing, are commonly used to test the FR properties of fabrics rather than loose fibers. It is time-consuming and costs a lot of resources in producing fabrics, which is difficult especially for a small amount of chemically modified fibers. In addition, it will limit the efficiency of research in the FR fabrics by blending methods from fibers to fabrics. On the other hand, these traditional methods only provide the burning phenomenon of polymeric materials under a weak flame without reflecting their intrinsic combustion behaviors. Thus, it is of great importance to develop a general method to directly investigate the intrinsic FR performance of fiber mats, which can quickly predict the potential hazards of the corresponding fabrics.

With the development of advanced calorimetric instruments, for example, the cone calorimeter, it has become possible to study the heat and smoke hazards of materials under different fire scenarios. Based on the principle of oxygen consumption,¹⁷ the cone calorimeter was originally designed by Babrauskas¹⁸ to measure the heat release of furniture. Subsequently, three models were proposed by him to predict full-scale furniture combustion behavior by the cone calorimeter.¹⁹ The cone calorimeter can record the released heat rate by calculating the consumed oxygen from the oxygen concentration, with a conversion ratio of 13.1 MJ/kg. From the combustion product stream, the rate of dynamic smoke production rate is evaluated by the analysis of the attenuation of the laser light beam. The cone calorimeter analysis can give comprehensive combustion information of the materials, such as time to ignition (TTI), the heat release rate (HRR), total heat release (THR) and smoke and carbon oxides release, which is a convenient and effective method to evaluate flame retardant properties.^20,21 Based on the heat released during burning, the essential combustion mechanism of material can be studied, which can essentially interpret the results of traditional FR evaluation. Nevertheless, the cone calorimeter was initially employed to assess the flammability properties of building materials, such as thick composites.²² The textiles can be regarded as thermally thin materials,²³ that is, heat can rapidly transform from one side to the other side so that the temperature can almost remain the same in the vertical direction of the material. Therefore, for the textile measurement of the cone calorimeter, the repeatable data is difficult to obtain due to the fact that the textiles are too thin to get a longer burning time. Under this circumstance, in order to investigate the flammability of the textiles with the advanced cone calorimeter, researchers have optimized the test parameters of the fabrics to obtain repeatable data. In the research of commercial polyester fabrics, Tata et al.²⁴ optimized the cone calorimeter by considering various aspects, including the heat flux, ceramic backing pads, retaining grid, sample weight, textile construction and relative humidity. Nazaré et al.²⁵ made reproducibility studies of apparel fabrics by changing the type of the grid, through which optimal results can be obtained by using a 3 × 3 cross wire grid. However, to the best of our knowledge, for fiber mats with different categories, a universal evaluation method to investigate the FR properties has not been established previously, which can remarkably reduce the time of new product development.

With this purpose, in our research, the cone calorimeter method using a specific fixing mode and optimized parameters has been put forward to investigate the FR properties of fiber mats, which can quickly predict the FR properties of the corresponding fabrics. It is important to note that we are only considering the influence of fibers on the FR properties of fabrics here. Different kinds of loose fibers, including common fibers (CFs) and inherent FR fibers (IFRFs), are selected to verify the reliability of the method. The challenge in obtaining accurate and repeatable cone calorimeter data has been solved by optimizing the sample weight and grid type, which is applicable to combustion investigations of various fibers. Finally, the reactions of fire in different fire scenarios are studied by the improved cone calorimeter method and the heat resistance is also analyzed by thermal degradation analysis. In this way, the universal method to evaluate the FR properties of fiber mats has been successfully established, which can serve as the efficient way to study the FR properties of the corresponding fabrics.

Experimental details

Materials

The fibers used are widely employed in general and fire resistant garments. These fibers can be identified as CFs and IFRFs in term of FR property. The CFs include cotton, ramie, wool, FR rayon (product of JiLin Chemical Fiber Group Co., Ltd, China). Since the FR property of FR rayon is much worse than that of inherent FR fibers, the FR rayon is classified as a CF here. The IFRFs include POD (provided by JiangGu POD New Materials Co., Ltd, China), polyamide-imide (Kermel), PMIA (provided by YanTai Tayho Advanced Materials Co., Ltd, China), polysulfonamide (PSA, provided by ShangHai Tanlon Fiber Co., Ltd, China) and polyimide (PI, provided by JiangSu Aoshen Hi-tech Materials Co., Ltd, China).

Sample preparation for the combustion tests

Considering the fact that the stacking of the fibers greatly influences the FR properties, the disordered loose fibers with non-uniform shape will lead to poor repeatable testing results. In this research, to obtain uniform fiber mats, the fibers were firstly carded at least six times by a small-scale carding machine according to the procedure shown in Figure 1(a). In term of the demands in the ISO5660-1 standard, a 100 mm × 100 mm fiber specimen was cut from the uniform fibrous web and put in the middle of aluminum foil with the shiny side towards the specimen; the specimen weight was determined by the experiment required.
Figure 1.
(a) The procedure from loose fiber to uniform fiber mats. (b) Different kinds of grid types. (c) The sample mounting mode.

In order to compare the FR properties of the different kinds of fibers, the specimens must share the same density and grammage, which significantly influence its combustion behavior.^26,27 Since the density of the fiber mats is far below than that of general fabric, the fiber mats should be compressed into fabric status as much as possible. In this way, it is reasonable to predict the FR property of the fabrics based on corresponding loose fiber mats. A cross-steel grid was employed to fix the shape of specimen and prevent the fibers from curling. Four kinds of grid type are shown in Figure 1(b), whose covering rates are 22%, 25%, 28% and 30%, respectively.

The heat-insulated ceramic pad was employed to support the fibrous specimens, for which low thermal conductivity is necessary to avoid influencing the cone calorimeter data of thermally thin specimens. For sample mounting, four holes were drilled in each corner of the ceramic pad, whose size and position are consistent with that in the cross-steel grid. The ceramic pad, the aluminum foil, the sample and cross-steel grid were stacked from bottom to top and then assembled by four bolts. The sample mounting mode is shown in Figure 1(c). The location gasket, between the steel grid and the ceramic pad, was used to control the thickness of the sample. For example, with the specimen weight of 2, 4 and 6 g, the thickness of the location gasket is respectively set as 1, 2 and 3 mm to ensure the same density of the sample.

Cone calorimeter measurement

Calorimetric tests were carried out on a Fire Testing Technology (UK) cone calorimeter, following the ISO5660-1 standard. The sample was located 25 mm below the base of the cone heater. The HRR records were taken every 2 s. For each sample, the tests were repeated six times to ensure the reproducibility of the measurements and the average data is taken. The data reported here include the following information: TTI (s), HRR (kW/m²), 2-min-avg HRR (kW/m²), peak heat release rate (PHRR, kW/m²), time to peak heat release rate (TTP, s) and THR (MJ/m²).

Thermal analysis

Thermal analysis was confirmed to be useful in the investigation of the stability of the material.^28,29 In this research, thermogravimetric analysis (TGA; from 20 to 800℃ with a heating rate of 10℃/min) in a nitrogen atmosphere (200 ml/min) was carried out to investigate the thermal degradation properties of the fibers. The sample weight is about 5 mg for each test.

Results and discussion

Reproducibility investigations on the fiber mats at 75 kW/m² heat flux

Due to the physical feature of loose fibers, it is important to optimize the parameters to test fiber mats by the cone calorimeter so that repeatable and pervasive data can be obtained. According to the previous researches, parameters such as the heat flux, sample weight and grid type can greatly influence the cone calorimeter data, which have been analyzed in detail here. At the beginning of our research, a heat flux of 75 kW/m² was adopted because some IFRFs cannot be ignited in low heat flux. Subsequently, the grid type and sample weight were optimized and verified.

Influence of the steel grid type

The fiber mats were assembled according to the method previously mentioned. In this process, the steel grid type can influence the TTI and PHRR values to some extent. In order to study the influence of the grid type in the cone calorimeter data, four different steel grid types with different covering rates were selected in the test. Since the grids with different covering rates mainly influence the heat radiation area, which can affect the TTI and the corresponding combustion state, the sample weight has less influence on the optimum grid covering rate. The PMIA fiber with a weight of 4 g was employed under the heat flux of 75 kW/m². The resultant cone calorimeter data with respect to the TTI, the PHRR and THR are given in Table 1. Obviously, for the TTI, PHRR and THR values, all of them have the minimum CV% when the grid covering rate is 25%, which demonstrates that the cone calorimeter data has the best repeatability under the grid covering rate of 25%. It can be speculated that, with a covering rate of less than 25%, the fibers cannot be fixed very well, while with a covering rate of more than 25%, the surface of the grid may influence the heat transfer and the ignition of the fibers. Thus, the optimal grid covering rate is set as 25% for all further fiber tests.
Table 1.
The repeatability results of the PMIA fiber under different grid types

Grid covering TTI (s)
PHRR (kW/m²)
THR (MJ/m²)

rate (%) Avg SD CV (%) Avg SD CV (%) Avg SD CV (%)

22 11.0 1.9 17.3 107.9 8.1 7.5 7.2 0.3 4.2

25 13.8 1.2 8.7 101.5 3.6 3.6 6.7 0.2 3.0

28 20.2 1.9 9.4 98.2 3.9 4.0 5.9 0.5 8.5

30 31.0 3.0 9.7 99.8 6.1 6.1 5.7 0.6 10.5

PMIA: poly-m-phenylene isophthal-amide; TTI: time to ignition; PHRR: peak heat release rate; THR: total heat release.

Influence of the sample weight

As mentioned above, the weight and density of the sample remarkably influence the cone calorimeter data. Therefore, the sample weight is a significant parameter to be optimized. For this purpose, fibers with weights of 2, 4 and 6 g (with a fixed thickness of 1, 2 and 3 mm, respectively) were prepared in terms of the method given in the Experimental details section and then measured by the cone calorimeter. It should be emphasized that the optimal steel grid with a covering rate of 25% was used. In order to confirm the optimal sample weight, two kinds of fibers, namely cotton and PI, were selected. The results of the repeatability experiments performed on the fibers are illustrated in Table 2.
Table 2.
Test results for two kinds of fibers under different weights

Weight TTI (s)
PHRR (kW/m²)
THR (MJ/m²)

Sample (g) Avg SD CV (%) Avg SD CV (%) Avg SD CV (%)

Cotton 2 1.4 0.2 14.3 135.5 7.2 5.3 3.1 0.2 6.5

Cotton 4 1.3 0.3 23.1 166.9 6.3 3.6 5.8 0.2 3.4

Cotton 6 1.5 0.2 13.3 171.2 4.6 2.7 8.4 0.3 3.6

PI 2 38.2 3.4 8.9 42.6 5.8 13.6 4.3 0.3 6.9

PI 4 33.0 2.8 8.4 51.4 3.6 7.1 5.7 0.3 5.3

PI 6 33.2 2.7 8.0 53.4 2.8 5.2 6.1 0.2 3.3

TTI: time to ignition; PHRR: peak heat release rate; THR: total heat release; PI: polyimide.

It is interesting to find that, for both fibers, the TTI, PHRR and THR have the minimum CV% value when the sample weight is 6 g. For example, the CV% values of PHRR and THR for the cotton samples are 2.7% and 3.6%, respectively. The test results also show that the PHRR significantly increases when the sample weight ranges from 2 to 4 g, but it almost remains constant with a further increase to 6 g. The probable reason for this phenomenon may be due to the fact that the fiber mats can achieve physical thermal thickness when the sample weight is relatively larger, especially for those better fire resistant fibers, such as PI. Here, the physical thermal thickness can be understood as the temperature gradient in the specimen, that is, the temperature is different between the top and bottom sides of the specimen. The different temperature in the vertical direction in the specimen leads to continuous combustion for long time rather than at the same time, resulting in the relatively stable PHRR. However, the THR, which mainly reflects the total fuel content of the specimen, is raised with the increase of sample weight.

In conclusion, the sample with the weight of 6 g is optimal to have the best repeatability. According to the investigation of the optimal experiments, the steel grid with a covering rate of 25% and the sample with the weight of 6 g are the optimal parameters, which will be used in further tests.

Validation of the optimal cone calorimeter parameters

The cone calorimeter results show that the grid covering rate of 25% and specimen weight of 6 g give the best repeatability. In order to verify the universality and accuracy of the technique in the investigation of the FR properties of other fiber mats, a series of other CFs (ramie, wool, FR rayon) and IFRFs (POD, Kermel, PMIA, PSA) were selected. The corresponding results of the repeatability experiments are shown in Table 3. The results show that, for all these fibers, the CV% with respect to PHRR and THR is less than 6%, which is within the acceptable repeatability range and demonstrates the universality of the improved cone calorimeter technique. On the other hand, the FR properties of these fibers can be easily distinguished based on the cone calorimeter results. That is, the cone calorimeter technique with the optimized parameters we obtained can become a general and reliable evaluation method to study the FR behaviors of fiber mats with different categories.
Table 3.
The verification results of the optimal cone calorimeter parameters with other kinds of fibers

Sample Weight (g) TTI (s)
PHRR (kW/m²)
THR (MJ/m²)

Avg SD CV (%) Avg SD CV (%) Avg SD CV (%)

Ramie 6 1.6 0.6 34.2 234.2 6.9 2.9 9.1 0.3 3.3

Wool 6 1.3 0.5 38.7 250.4 12.9 5.2 12.2 0.4 3.3

FR rayon 6 1.5 0.2 13.3 192.9 3.3 1.7 8.5 0.5 5.9

POD 6 8.0 0.7 8.8 125.9 3.7 2.9 9.1 0.1 1.2

Kermel 6 12.0 1.3 10.5 136.9 3.7 2.7 9.5 0.3 3.2

PMIA 6 16.7 3.1 18.8 123.7 3.9 3.2 8.8 0.3 3.5

PSA 6 22.0 4.7 21.5 107.3 4.8 4.5 8.1 0.2 2.5

TTI: time to ignition; PHRR: peak heat release rate; THR: total heat release; PI: polyimide; FR: fire retardant; POD: poly(1, 3, 4-oxadiazole)s; PMIA: poly-m-phenylene isophthal-amide; PSA: polysulfonamide.

Cone calorimeter studies of the fiber mats at different heat fluxes

Due to the different sizes of fire, it is not possible to cover the whole range of fire types at a certain heat flux. The fire scenarios can be divided into three stages, namely ignition, developing fire and fully developed fire.²¹ In order to study the FR properties of fibers under the different fire scenarios, three external heat fluxes (35, 50 and 75 kW/m²) were employed to evaluate the combustion behaviors of the above-mentioned nine kinds of fibers by the modified cone calorimeter method. The results are given in Table 4; the lost data in the table is due to failed ignition at the corresponding heat flux.
Table 4.
The cone calorimeter data of the fiber samples at different heat fluxes

No. Sample Heat flux (kW/m²) TTI (s) TTP (s) PHRR (kW/m²) 2-min-avg HRR (kW/m²) THR (240 s) (MJ/m²)

1-1 Cotton 35 10.0 63.0 86.1 49.79 7.7

1-2 Cotton 50 2.5 35.5 136.7 59.94 8.7

1-3 Cotton 75 1.5 28.8 171.2 64.08 8.9

2-1 Ramie 35 22.0 61 117.6 65.89 8.7

2-2 Ramie 50 6.5 44.0 174.2 69.86 8.8

2-3 Ramie 75 1.6 31.2 234.2 73.71 9.1

3-1 Wool 35 57 84 127.5 49.95 7.3

3-2 Wool 50 6.5 48.5 165.4 74.57 9.9

3-3 Wool 75 1.3 33.2 245.9 89.22 12.2

4-1 FR rayon 35 6.0 25 106.4 53.89 7.2

4-2 FR rayon 50 2.5 26.0 141.7 57.07 7.4

4-3 FR rayon 75 1.8 16.8 192.9 66.16 8.5

5-1 POD 35 200.0 219.0 89.1 2.4

5-2 POD 50 36.0 62.0 107.9 40.18 6.3

5-3 POD 75 8.0 34.0 125.9 53.05 9.1

6-1 Kermel 35 NI 2.2

6-2 Kermel 50 35.0 73.0 76.5 40.89 7.3

6-3 Kermel 75 12.0 50.3 136.9 58.98 9.5

7-1 PMIA 35 NI 1.3

7-2 PMIA 50 52.5 83.0 88.3 33.96 5.4

7-3 PMIA 75 16.5 46.8 127.8 51.02 8.5

8-1 PSA 35 NI 1.7

8-2 PSA 50 62.0 90.0 73.7 32.15 4.9

8-3 PSA 75 22.0 45.8 107.3 44.28 8.1

9-1 PI 35 NI 0.06

9-2 PI 50 NI 2.13

9-3 PI 75 33.2 62.5 53.4 32.13 6.0

TTI: time to ignition; TTP: time to peak heat release rate; PHRR: peak heat release rate; THR: total heat release; HRR: heat release rate; PI: polyimide; FR: fire retardant; POD: poly(1, 3, 4-oxadiazole)s; PMIA: poly-m-phenylene isophthal-amide; PSA: polysulfonamide.

Time to ignition

Subjected to heat radiation with the surface temperature above the decomposition temperature, the fibers will pyrolyze and then produce combustible gases, leading to ignition. According to previous reports, four explanations for the ignition exist, namely critical external heat flux, critical mass loss, critical surface temperature and time–energy required.³⁰ These four criteria are correlated with each other. For example, the mass loss and surface temperature can be influenced by the heat flux. In the cone calorimeter, the critical heat flux can be used to describe the characteristic and ignitability of a material. Therefore, the TTI can be used to evaluate the FR properties of these samples. The TTI data under different heat fluxes (35, 50, 75 kW/m²) is shown in Table 4 and summarized in Figure 2. It is obvious that, for all fibers, the values of TTI are tremendously shorter at higher external heat fluxes. The reasons are given in the following. On one hand, higher heat flux favors a faster speed to reach the critical temperature for ignition. On the other hand, under the higher heat flux, the pyrolysis of the fiber samples is more liable to occur, leading to release more combustible volatile species.
Figure 2.
The time to ignition (TTI) values of various fibers under different heat fluxes.

At the heat flux of 35 kW/m², the TTI values of cotton, ramie, wool and FR rayon samples are 10.0, 22.0, 57.0 and 6.0 s, respectively, indicating that the wool sample has the best fire resistance in these samples. Based on the TTI, the rank of the FR property is as follows, wool > ramie > cotton > FR rayon. However, for IFRFs, only the POD can be ignited with a TTI of 200 s, which is significantly longer than that of previous CFs. In addition, the Kermel, PMIA, PSA and PI fibers cannot be ignited, which means that the heat flux of 35 kW/m² is too low to reach the critical heat flux. At the heat flux of 50 kW/m², only the sample of PI cannot be ignited, indicating that PI possesses the best fire resistance. For the other fibers, the TTI values under 50 kW/m² are much less than those under 35 kW/m², which is because the fibers can release combustible gases more quickly at a higher heat flux. A similar trend of TTI change can also be observed when the heat flux increases from 50 to 75 kW/m². At the higher heat flux of 75 kW/m², all the samples can be ignited with the lowest TTI values compared with those at other heat fluxes. However, for the four CFs, the TTI values are very close and short (within 2 s). It is difficult to distinguish the FR properties of these four CFs, due to the fact that they can pyrolyze immediately under the higher heat flux. Based on the TTI, the rank of the FR property of IFRFs is as follows, PI (33.2 s) > PSA (22.0 s) > PMIA (16.5 s) > Kermel (12.0 s) > POD (8.0 s).

According to the discussion above, the ignitability of the fiber mats can be evaluated by both the critical heat flux and the TTI values under a certain heat flux. The heat flux influences the TTI values significantly. For the IFRFs, it can be ignited only at the higher heat flux, such as 50 kW/m² or even 75 kW/m². However, for these CFs, the lower heat flux may be more reasonable to compare the FR property on the basis of TTI values at a certain heat flux.

Heat release rate

The HRR of the fibers under three heat fluxes is illustrated in Figure 3, including the HRR of the CFs at three heat fluxes (Figure 3(a), 35 kW/m²; Figure 3(b), 50 kW/m²; Figure 3(c), 75 kW/m²), the heat flux dependence of PHRR for the CFs (Figure 3(d)), the HRR of the IFRFs fibers at 50 kW/m² (Figure 3(e)) and at 75 kW/m² (Figure 3(f)).
Figure 3.
The heat release rate (HRR) curves of common fibers (CFs) under 35 kW/m² (a), 50 kW/m² (b) and 75 kW/m² (c); the relationship between peak heat release rate (PHRR) and heat fluxes of CF (d); the HRR curves of inherent fire retardant fibers (IFRFs) under 50 kW/m² (e) and 75 kW/m² (f). (Color online only).

The value of the HRR depends on the intrinsic properties of the material as well as the fire scenarios (the heat flux). An obvious trend can be seen in Figure 3 and Table 4 that both the peak HRR and 2-min-avg HRR become larger with the increase of external heat flux for all fibers. In addition, the TTP becomes shorter at the higher heat flux, which is ascribed to the fact that the fibers can ignite earlier and reach the peak HRR more quickly under the higher heat flux. For the analysis of the shape of the HRR curve, a sharper peak of the HRR curve appears at the higher heat flux, since a higher percentage of the samples can pyrolyze and burn simultaneously, resulting in the higher heat release.

From Figure 3(a), it can be seen that the time corresponding to the initial increase of the HRR curves (as denoted by red rectangular in Figure 3(a)) varies with different fibers. FR rayon has the lowest time, following by cotton, ramie and wool, indicating that the FR rayon can release combustible gases in the shortest time and is the first one to be ignited. This is consistent with the TTI results. For wool, a larger TTP is favored, but it shares the maximum PHRR (Figure 3(a)), which indicates that the fire spread rate of wool is larger than that of any other fibers after ignition. From Figures 3(b) and (c), it is obvious that the initial time of the HRR increase for the four samples is approximately the same. This is because under the higher heat fluxes (50 and 75 kW/m²), the fibers can pyrolyze more quickly and release combustible gases. Figure 3(d) demonstrates that the average values of PHRR of the four fibers vary with radiant heat flux and the higher heat flux favors larger PHRR. The PHRR curves of the inherent FR fiber samples are shown in Figures 3(e) and (f). Because the PI cannot be ignited under the heat flux of 50 kW/m², the HRR curve of PI is not given here. Comparing Figure 3(e) with Figure 3(f), a similar trend can be observed for all inherent FR fibers, that is, the PHRR values of these samples increase when the heat flux increases from 50 to 75 kW/m². The initial time corresponding to the increase of HRR is different for these samples. The POD has the lowest time, followed by Kermel, PMIA, PSA and PI. The HRR curve demonstrates that the ranking of these inherent FR fibers in turn is PI, PSA, PMIA, Kermel and POD, which also corresponds to the TTI results. Comparing the HRR curves of CFs with IFRFs, the latter indeed has better FR properties than the CFs from all aspects.

Total heat release

The THR is the integration of the HRR as a function of time, which is strongly related to the sample weight and fire scenarios. Among them, the sample weight influences the total heat load and the fire scenarios influence the combustion efficiency in the flame zone. The THR of the fibers under three heat fluxes are illustrated in Figure 4, including the CFs at three heat fluxes (Figure 4(a), 35 kW/m²; Figure 4(b), 50 kW/m²; Figure 4(c), 75 kW/m²), the heat flux dependence of THR for the CFs (Figure 4(d)), the THR of the IFRFs at 50 kW/m² (Figure 4(e)) and at 75 kW/m² (Figure 4(f)). For all fibers, as expected, with the increase of heat flux, the fire hazards increase in terms of THR.
Figure 4.
The total heat release (THR) curves of common fibers (CFs) under 35 kW/m² (a), 50 kW/m² (b) and 75 kW/m² (c); the relationship between THR and heat fluxes of CFs (d); the THR curves of inherent fire retardant fibers (IFRFs) under 50 kW/m² (e) and 75 kW/m² (f). POD: poly(1, 3, 4-oxadiazole)s; PMIA: poly-m-phenylene isophthal-amide; PSA: polysulfonamide; PI: polyimide.

From Figure 4(a), the distinct differences can be seen in CFs. The THR curve of the FR rayon sample is the first one to have an increase, followed by cotton, ramie and wool, which correspond to the discussion in the Heat release rate section (Figure 3(a)). In Figures 4(b) and (c), with the heat flux ranges from 50 to 75 kW/m², and the rising stage (0–50 s) of the THR curves for all common samples are approximately the same. Moreover, the materials can combust completely and rapidly under higher heat flux, and the THR curves can reach the plateau stage in a short time. The relationship between THR and heat flux for CFs samples is illustrated in Figure 4(d). For cotton, ramie and FR rayon, the THR are very close regardless of heat fluxes, whereas the THR of wool increases significantly with the increase of heat flux. This distinct difference arises because wool possesses better FR performance and its THR is more sensible to heat flux. Unfortunately, wool has the largest THR among CFs, which can be regarded as the most hazardous sample. For example, under the heat flux of 75 kW/m², the THR (240 s) of wool is 12.2 MJ/m², much larger than that of ramie (9.1 MJ/m²), cotton (8.9 MJ/m²) and FR rayon (8.5 MJ/m²). The THRs of the IFRFs are shown in Figures 4(e) and (f). It can be seen that, for all IFRFs, the plateau stage cannot be reached at two heat fluxes (50 and 75 kW/m²) and the THR under 75 kW/m² is significantly larger than that under 50 kW/m², which is attributed to the lower combustion efficiency of the IFRFs under a lower heat flux. For all IFRFs, under the heat flux of 75 kW/m², the PI is indicated as the least hazardous sample with a THR (240 s) value of 6 MJ/m², which is smaller than that of PSA (8.1 MJ/m²), PMIA (8.5 MJ/m²), POD (9.1 MJ/m²) and Kermel (9.5 MJ/m²). Compared with CFs, the IFRFs have a low combustion efficiency, indicating better FR properties.

Heat resistance of fibers

In order to investigate the relationship between FR performance and the heat resistance of the fibers, the thermal degradation analysis was carried out under nitrogen and the TGA curves shown in Figure 5. Apparently, IFRFs have a much better heat resistance than CFs, as shown in the Table 5, which can explain the distinct FR performance of the fibers to some extent. The distinct differences in FR properties of the fibers are attributed to the different chemical structures. The IFRFs have highly conjugated structures, such as an aramid ring and a heterocyclic ring, which provides a high level of stability and promotes the resultant high fire resistance. The T_onset_10% values of the IFRFs are higher than those of the CFs (299.3℃ for cotton, 334.3℃ for ramie, 270.8℃ for wool, 253.8℃ for FR rayon, 507.5℃ for POD, 491.1℃ for Kermel, 450.5℃ for PMIA, 464.4℃ for PSA, 588.6℃ for PI). Similarly, the remaining residue at 600℃ of the IFRFs is much higher than that of the CFs (20.03% for cotton, 10.85% for ramie, 28.50% for wool, 27.37% for FR rayon, 51.06% for POD, 61.34% for Kermel, 59.52% for PMIA, 48.80% for PSA, 83.21% for PI). Based on the heat resistance, the FR performance can be predicted to some extent, but some differences and exceptions also exist. For example, the T_onset_10% value of ramie (334.3℃) is higher than that of cotton (299.3℃), but the TTI of ramie is lower than that of cotton under the same heat flux. Moreover, the T_onset_10% of POD (507.7℃) is higher than that of Kermel (491.1℃), PMIA (450.5℃) and PSA (464.4℃), but only the POD can be ignited under the heat flux of 35 kW/m² and the TTI of POD is shorter than that of Kermel, PMIA and PSA exposed to the heat fluxes of 50 and 75 kW/m². The results demonstrate that heat resistance is not the sole factor to determine the FR performance, while the released fuel is much more important than the decomposition temperature.
Figure 5.
Thermogravimetric curves of all the fibers under nitrogen. POD: poly(1, 3, 4-oxadiazole)s; PMIA: poly-m-phenylene isophthal-amide; PSA: polysulfonamide; PI: polyimide.

Table 5.
Thermal degradation analysis of fibers under nitrogen atmosphere

No Sample T_onset10% (℃) T_max (℃) Residue at T_max (%) Residue at 600℃ (%)

1 Cotton 299.3 344.9 50.80 20.03

2 Ramie 334.3 365.8 45.60 10.85

3 Wool 270.8 326.8 66.40 28.50

4 FR rayon 253.8 269.8 57.10 27.37

5 POD 507.7 531.8 68.40 51.06

6 Kermel 491.1 511.5 84.75 61.34

7 PMIA 450.5 458.2 87.23 59.52

8 PSA 464.4 480.2 81.27 48.80

9 PI 588.6 602.7 81.39 83.21

T_onset10%: the corresponding temperature with the weight loss of 10%; T_max: the temperature with maximum degradation rate; FR: fire retardant; POD: poly(1, 3, 4-oxadiazole)s; PMIA: poly-m-phenylene isophthal-amide; PSA: polysulfonamide; PI: polyimide.

Conclusions

In the present research, the universal method to measure the FR properties of fiber mats by the modified cone calorimeter has been successfully established through employing a specific fixing mode and optimizing the test parameters. For different kinds of fibers, including common and inherent FR fibers, the challenge in obtaining accurate and repeatable cone calorimeter data has been solved by optimizing the grid covering rate and sample weight, in which the optimal covering rate and sample weight are confirmed to be 25% and 6 g, respectively. In this way, the FR properties of these selected fiber mats can be successfully evaluated and compared. With the increase of external heat flux, the fibers burn rapidly and completely with a lower TTI as well as higher PHRR and THR. Comparing the CFs with IFRFs, the latter possess better FR properties with a higher TTI and lower PHRR. The results further confirm the universality of the optimal cone calorimeter to evaluate the FR properties of fiber mats. In addition, the IFRFs have a much better heat resistance than CFs, which can explain the superior FR performance of the IFRFs to some extent. However, for some fibers with close FR properties, such as POD and PMIA, the good heat resistance does not represent a good FR property. The results demonstrate that the decomposition temperature is not the sole factor to determine the FR performance; the released fuel is much more important than decomposition temperature. Therefore, the universal method of studying the FR properties of fiber mats by the optimal cone calorimeter has been successfully established, and can serve as an efficient way of investigating the FR properties of corresponding fabrics, providing important information for the design of FR fabrics.

Grid covering	TTI (s)	PHRR (kW/m²)	THR (MJ/m²)
22	11.0	1.9	17.3	107.9	8.1	7.5	7.2	0.3	4.2
25	13.8	1.2	8.7	101.5	3.6	3.6	6.7	0.2	3.0
28	20.2	1.9	9.4	98.2	3.9	4.0	5.9	0.5	8.5
30	31.0	3.0	9.7	99.8	6.1	6.1	5.7	0.6	10.5

	Weight	TTI (s)	PHRR (kW/m²)	THR (MJ/m²)
Cotton	2	1.4	0.2	14.3	135.5	7.2	5.3	3.1	0.2	6.5
Cotton	4	1.3	0.3	23.1	166.9	6.3	3.6	5.8	0.2	3.4
Cotton	6	1.5	0.2	13.3	171.2	4.6	2.7	8.4	0.3	3.6
PI	2	38.2	3.4	8.9	42.6	5.8	13.6	4.3	0.3	6.9
PI	4	33.0	2.8	8.4	51.4	3.6	7.1	5.7	0.3	5.3
PI	6	33.2	2.7	8.0	53.4	2.8	5.2	6.1	0.2	3.3

Sample	Weight (g)	TTI (s)	PHRR (kW/m²)	THR (MJ/m²)
Ramie	6	1.6	0.6	34.2	234.2	6.9	2.9	9.1	0.3	3.3
Wool	6	1.3	0.5	38.7	250.4	12.9	5.2	12.2	0.4	3.3
FR rayon	6	1.5	0.2	13.3	192.9	3.3	1.7	8.5	0.5	5.9
POD	6	8.0	0.7	8.8	125.9	3.7	2.9	9.1	0.1	1.2
Kermel	6	12.0	1.3	10.5	136.9	3.7	2.7	9.5	0.3	3.2
PMIA	6	16.7	3.1	18.8	123.7	3.9	3.2	8.8	0.3	3.5
PSA	6	22.0	4.7	21.5	107.3	4.8	4.5	8.1	0.2	2.5

No.	Sample	Heat flux (kW/m²)	TTI (s)	TTP (s)	PHRR (kW/m²)	2-min-avg HRR (kW/m²)	THR (240 s) (MJ/m²)
1-1	Cotton	35	10.0	63.0	86.1	49.79	7.7
1-2	Cotton	50	2.5	35.5	136.7	59.94	8.7
1-3	Cotton	75	1.5	28.8	171.2	64.08	8.9
2-1	Ramie	35	22.0	61	117.6	65.89	8.7
2-2	Ramie	50	6.5	44.0	174.2	69.86	8.8
2-3	Ramie	75	1.6	31.2	234.2	73.71	9.1
3-1	Wool	35	57	84	127.5	49.95	7.3
3-2	Wool	50	6.5	48.5	165.4	74.57	9.9
3-3	Wool	75	1.3	33.2	245.9	89.22	12.2
4-1	FR rayon	35	6.0	25	106.4	53.89	7.2
4-2	FR rayon	50	2.5	26.0	141.7	57.07	7.4
4-3	FR rayon	75	1.8	16.8	192.9	66.16	8.5
5-1	POD	35	200.0	219.0	89.1		2.4
5-2	POD	50	36.0	62.0	107.9	40.18	6.3
5-3	POD	75	8.0	34.0	125.9	53.05	9.1
6-1	Kermel	35	NI				2.2
6-2	Kermel	50	35.0	73.0	76.5	40.89	7.3
6-3	Kermel	75	12.0	50.3	136.9	58.98	9.5
7-1	PMIA	35	NI				1.3
7-2	PMIA	50	52.5	83.0	88.3	33.96	5.4
7-3	PMIA	75	16.5	46.8	127.8	51.02	8.5
8-1	PSA	35	NI				1.7
8-2	PSA	50	62.0	90.0	73.7	32.15	4.9
8-3	PSA	75	22.0	45.8	107.3	44.28	8.1
9-1	PI	35	NI				0.06
9-2	PI	50	NI				2.13
9-3	PI	75	33.2	62.5	53.4	32.13	6.0

No	Sample	T_onset10% (℃)	T_max (℃)	Residue at T_max (%)	Residue at 600℃ (%)
1	Cotton	299.3	344.9	50.80	20.03
2	Ramie	334.3	365.8	45.60	10.85
3	Wool	270.8	326.8	66.40	28.50
4	FR rayon	253.8	269.8	57.10	27.37
5	POD	507.7	531.8	68.40	51.06
6	Kermel	491.1	511.5	84.75	61.34
7	PMIA	450.5	458.2	87.23	59.52
8	PSA	464.4	480.2	81.27	48.80
9	PI	588.6	602.7	81.39	83.21

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 the Fundamental Research Funds for the Central Universities (CUSF-DH-D-2016012).

References

Alongi

Carosio

Malucelli

. Current emerging techniques to impart flame retardancy to fabrics: an overview. Polym Degrad Stabil 2014; 106: 138–149.

Wang

Romero

Zhang

X-Q

et al.

Intumescent multilayer hybrid coating for flame retardant cotton fabrics based on layer-by-layer assembly and sol-gel process. RSC Adv 2015; 5: 10647–10655.

Apaydin

Laachachi

Fouquet

et al.

Mechanistic investigation of a flame retardant coating made by layer-by-layer assembly. RSC Adv 2014; 4: 43326–43334.

Alongi

Malucelli

. Cotton flame retardancy: state of the art and future perspectives. RSC Adv 2015; 5: 24239–24263.

Carosio

Alongi

Malucelli

. Layer by Layer ammonium polyphosphate-based coatings for flame retardancy of polyester–cotton blends. Carbohydr Polym 2012; 88: 1460–1469.

Bourbigot

Duquesne

. Fire retardant polymers: recent developments and opportunities. J Mater Chem 2007; 17: 2283–2300.

Didane

Giraud

Devaux

et al.

Thermal and fire resistance of fibrous materials made by PET containing flame retardant agents. Polym Degrad Stabil 2012; 97: 2545–2551.

Horrocks

. Flame retardant challenges for textiles and fibres: new chemistry versus innovatory solutions. Polym Degrad Stabil 2011; 96: 377–392.

Sato

Endo

Araki

et al.

The flame-retardant polyester fiber: Improvement of hydrolysis resistance. J Appl Polym Sci 2000; 78: 1134–1138.

10.

Villar-Rodil

Martínez-Alonso

Tascón

JMD

. Studies on pyrolysis of Nomex polyaramid fibers. J Anal Appl Pyrol 2001; 58–59: 105–115.

11.

Villar-Rodil

Paredes

Martínez-Alonso

et al.

Atomic force microscopy and infrared spectroscopy studies of the thermal degradation of Nomex aramid fibers. Chem Mater 2001; 13: 4297–4304.

12.

Y-H

Froelicher

Kaliszewski

et al.

A study of poly(benzo[1,2-d:5,4-d’]bisoxazole-2,6-diyl-1,4-phenylene) reactions at elevated temperatures. Macromolecules 1999; 32: 6565–6569.

13.

Bourbigot

Flambard

Poutch

. Study of the thermal degradation of high performance fibres—application to polybenzazole and p-aramid fibres. Polym Degrad Stabil 2001; 74: 283–290.

14.

Schulz

Bruma

Brehmer

. Aromatic poly(1, 3, 4-oxadiazoe)s as advanced materials. Adv Mater 1997; 9: 601–613.

15.

Yang

Jia

et al.

Thermal degradation behavior and probable mechanism of aromatic poly(1,3,4-oxadiazole)s fibers. Polym Bull 2015; 72: 1067–1080.

16.

Coffin

Serad

Hicks

et al.

Properties and applications of celanese PBI—polybenzimidazole fiber. Text Res J 1982; 52: 466–472.

17.

Huggett

. Estimation of rate of heat release by means of oxygen consumption measurements. Fire Mater 1980; 4: 61–65.

18.

Babrauskas

. Upholstered furniture heat release rates: measurements and estimation. J Fire Sci 1983; 1: 9–32.

19.

Babrauskas

Baroudi

Myllymäki

et al.

The cone calorimeter used for predictions of the full-scale burning behaviour of upholstered furniture. Fire Mater 1997; 21: 95–105.

20.

Schartel

Bartholmai

Knoll

. Some comments on the use of cone calorimeter data. Polym Degrad Stabil 2005; 88: 540–547.

21.

Schartel

Hull

. Development of fire-retarded materials—interpretation of cone calorimeter data. Fire Mater 2007; 31: 327–354.

22.

I. ISO, 5660-1: 2002. Reaction-to-fire tests-heat release, smoke production and mass loss rate–Part 1: heat release rate (cone calorimeter method).

23.

Ohlemiller

Shields

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

24.

Tata

Alongi

Carosio

et al.

Optimization of the procedure to burn textile fabrics by cone calorimeter: Part I. Combustion behavior of polyester. Fire Mater 2011; 35: 397–409.

25.

Nazaré

Kandola

Horrocks

. Use of cone calorimetry to quantify the burning hazard of apparel fabrics. Fire Mater 2002; 26: 191–199.

26.

Alongi

Cuttica

Carosio

et al.

How much the fabric grammage may affect cotton combustion?

Cellulose 2015; 22: 3477–3489.

27.

Jirsak

Sadikoglu

Ozipek

et al.

Thermo-insulating properties of perpendicular-laid versus cross-laid lofty nonwoven fabrics. Text Res J 2000; 70: 121–128.

28.

Wang

Dong

Wang

et al.

High performance long chain polyamide/calcium silicate whisker nanocomposites and the effective reinforcement mechanism. Chin J Polym Sci 2016; 34: 991–1000.

29.

Wang

Dong

Liu

et al.

Microstructure and rheological properties of thermalplastic polyurethane and its composites. Acta Polym Sinica 2013; 3: 367–376.

30.

Wichman

. Material flammability, combustion, toxicity and fire hazard in transportation. Progr Energ Combust Sci 2003; 29: 247–299.

Grid covering	TTI (s)			PHRR (kW/m²)			THR (MJ/m²)
rate (%)	Avg	SD	CV (%)	Avg	SD	CV (%)	Avg	SD	CV (%)
22	11.0	1.9	17.3	107.9	8.1	7.5	7.2	0.3	4.2
25	13.8	1.2	8.7	101.5	3.6	3.6	6.7	0.2	3.0
28	20.2	1.9	9.4	98.2	3.9	4.0	5.9	0.5	8.5
30	31.0	3.0	9.7	99.8	6.1	6.1	5.7	0.6	10.5

Universal evaluation of fire retardant properties of fiber mats by the optimal cone calorimeter

Abstract

Keywords

Experimental details

Materials

Sample preparation for the combustion tests

Cone calorimeter measurement

Thermal analysis

Results and discussion

Reproducibility investigations on the fiber mats at 75 kW/m2 heat flux

Influence of the steel grid type

Influence of the sample weight

Validation of the optimal cone calorimeter parameters

Cone calorimeter studies of the fiber mats at different heat fluxes

Time to ignition

Heat release rate

Total heat release

Heat resistance of fibers

Conclusions

Footnotes

Declaration of conflicting interests

Funding

References

Reproducibility investigations on the fiber mats at 75 kW/m² heat flux