The carbon fiber/epoxy composites toughened by fire-resistant glass fiber veils: Flammability and mechanical performance

Abstract

The fire-retardant properties of high-performance fiber-reinforced composites are the crucial benchmark for composite structure stability. However, in the current flame-retardant solution for composites it is difficult to reach the balance between fire resistance and structural performance due to the deteriorating composite interface. In this work, the carbon fiber-reinforced composite was covered with functional glass fiber layers, in which the glass fiber veil had been treated with flame-retardant agents and silicone-modified waterborne polyurethane, in order to be endowed with flame-retardant capability and structure toughness. As such, a significant improvement in the flame retardancy and mechanical structure of the composites could be observed. When compared with the control, the total heat release and total smoke release for composites with 8% silicone-modified waterborne polyurethane treatment could be decreased by 18.5% and 18.1%, while the tensile and flexural strength were significantly increased by 47.3% and 62.2%, respectively. This well-balanced performance is attributable to the structure design with a toughened glass fiber veil to protect the composite surfaces from fire combustion and structure failure. Therefore, this flame-retardant structure design provides a new strategy to achieve high-performance composites with prospective applications for aircraft and aerospace.

Keywords

Fiber-reinforced composites flame-retardant structure design glass fiber veil structure toughness

Carbon fiber (CF)-reinforced polymer composites have been widely used in wind turbine blades, aerospace, rail transportation, and vehicles due to their light weight, high strength, and corrosion resistance.^1,2 However, the polymeric structure of the composites resulted in high flammability when exposed to high temperatures or flames.^3–5

The traditional fire-retardant solution for composites mainly includes the synthesis of a flame-retardant polymer (intrinsic flame retardant) and the addition of a flame-retardant agent (additive flame retardant).^6,7 Compared with intrinsic flame retardants, the approach of additive flame retardants and flame-retardant coatings presents lower cost, simpler forming process, and higher processing creativity. However, additive flame retardants are normally in the form of particles or liquids, which physically dispersed into the resin matrix could lead to a serious effect on the structure performance.^8,9 Inorganic flame retardants, which have the advantages of being nontoxic and smokeless based on a large amount of additional volume, result in serious effects on the interface between fiber and polymer matrix. In addition, the flame-retardant agents of the halogen series show excellent flame-retardant efficiency with toxic gas release during the combustion process.^10–12

Currently, fire-retardant barriers such as insulative fabrics have become an alternative solution to prevent heat transfer into the composites and protect the stability of the composites.^13,14 However, the insulative thermal barrier for composite protection has been reported with unexpected toxic smoke and surface structural failure issues. For instance, Li et al.¹⁵ designed an intumescent flame-retardant mat for CF-reinforced composites and reported a reduction in the heat release rate with side effects on mechanical properties and extra smoke emission. Cong et al.¹⁶ confirmed this issue and proposed that the poor structural stability of the surface protection layer leads to the deterioration of the mechanical properties of the composites. Nonetheless, it remains unclear what strategy to use to balance between fire resistance and structural performance for high-performance fiber-reinforced composites.

In this study, a flame-retardant composite with a sandwich structure was designed, in which the ‘skin’ was for the functional glass fiber layer and the ‘core’ was for the CF/epoxy composite. The glass fiber layer with fiber veil should be functionalized with toughening and flame-retardant capabilities.^17,18 Due to the advantages of flexibility, the silicone-modified waterborne polyurethane (SWPU) was a good option to toughen the glass fiber veil.¹⁹ on the other hand, based on the excellent smoke suppression function for nickel hydroxide (Ni(OH)₂),^20,21 and good compatibility with epoxy for 9,10-dihydro-9-oxa-10-phosphoxafe-10-oxide (DOPO),²² the synergistic flame-retardant agents composed of Ni(OH)₂ and DOPO were prepared to treat the glass fiber veil.

To understand the flame-retardant performance of composites with flame-retardant design, the structure and morphology of functional glass fiber veils were analyzed by Fourier transform infrared (FTIR), X-ray photoelectron spectroscopy (XPS), and scanning electron microscope (SEM), while the flame retardancy and mechanical properties of the composites were evaluated by the limiting oxygen index (LOI), vertical combustion, cone calorimetry, and mechanical testing.

Experimental methods

Materials

Ni(OH)₂ nano flame retardant was synthesized from nickel sulfate hexahydrate (NiSO₄·6H₂O, 98 wt%), and sodium hydroxide (NaOH, 96 wt%) was purchased from Aladdin Co. Ltd. SWPU resin was purchased from Jining Tangyi Chemical Industry. Glass fiber (diameter 2 µm; length 6 mm) was purchased from Nanjing Glass Fiber Design and Research Institute. The unidirectional CF prepregs were provided by Toray Carbon Fiber Co. Ltd. (CF; UD200 (0.111 mm) × 60 cm/100 m³, diameter: 6 µm). The organic flame retardants DOPO and polyethylene oxide (Mv 1,000,000) were supplied by Aladdin Co. Ltd.

Synthesis of Ni(OH)₂

The Ni(OH)₂ was prepared by the hydrothermal method.²³ The 30.9 g nickel sulfate hexahydrate particles were weighed by an electronic balance and poured into a beaker containing 300 ml water. The particles were stirred under the action of a magnetic agitator until the particles were completely dissolved. An aqueous solution of nickel sulfate (300 ml/117.6 mmol) was prepared. At the same time, the electronic balance was used to weigh 2.4 g sodium hydroxide powder, pour it into an empty beaker, weigh 180 ml water and add it, stir until the powder dissolved, prepare sodium hydroxide aqueous solution (180 ml/58.8 mmol), and then pour the sodium hydroxide aqueous solution into the nickel sulfate aqueous solution. After stirring for 2 h with magnetic stirrers, transfer to a 50 ml high-pressure reactor, hydrothermal reaction at 120°C for 12 h, and pour the supernatant to get green Ni(OH)₂ precipitate. It was washed three times in deionized water and then transferred to a freeze dryer to dry, resulting in a dried green Ni(OH)₂ powder.

Preparation of toughened flame-retardant glass fiber veil

The preparation method of the glass fiber veil is similar to the traditional paper-making process, and the produced glass fiber veil was controlled in a thickness of 0.1 mm. The specific steps are as follows: the 2.0 g glass fibers were dispersed in 1000 ml water of fiber dissociator for 2 min, then 1 g of the adhesive polyvinyl oxide was slowly added into the dissociator, and poured into the grinding tool for extraction, and filtration to obtain wet glass fiber veil. After that, the prepared flame-retardant agents and SWPU solution were sprayed on the wet glass fiber veil and then covered with another piece of the same glass fiber veil and placed into a plate dryer for 30 min drying at 60°C. It is worth noting that the toughened flame-retardant fiber veil prepared by adjusting the SWPU with different concentrations (0 wt%, 2 wt%, 4 wt%, 6 wt%, and 8 wt%) was defined as GF-0, GF-2, GF-4, GF-6, and GF-8.

Preparation of composites with flame-retardant structure

The composite material structure design and preparation method are shown in Figure 1. Unidirectional CF mat and flame-retardant glass fiber veils were cut into mold sizes of 200.0 mm × 200.0 mm, and the detailed composition of the composite is recorded in Table 1.

Figure 1.

Flow chart of composite material structure design and preparation method.

Table 1.

Composition of flame-retardant glass fiber veil

Glass fiber veil code	Glass fiber(wt %)	DOPO(wt %)	Ni(OH)₂(wt %)	SWPU(wt %)
GF-0	79.6	10.2	10.2	0
GF-2	78.0	10.0	10.0	2
GF-4	76.8	9.6	9.6	4
GF-6	75.0	9.5	9.5	6
GF-8	73.2	9.4	9.4	8

DOPO: 9,10-dihydro-9-oxa-10-phosphoxafe-10-oxide; Ni(OH)₂: nickel hydroxide; SWPU: silicone-modified waterborne polyurethane.

For the manufacturing process, the CF mats were placed layer by layer to be the CF mat stack in which the flame-retardant glass fiber veils were placed on the upper and lower layers of the CF stack. The whole stack of the fiber veil was placed into the mold and impregnated with epoxy resin by the vacuum infusion method, and then cured in the press at 80°C for 2 h.

In this work, we designed four kinds of SWPUs with different proportions, without adding SWPUs as the control group. The synergistic effect between the modified flame-retardant fiber veil and CF interface was studied, and the fire safety and mechanical performance of five different composites were analyzed. According to the components of the toughened flame-retardant mat (GF-0, GF-2, GF-4, GF-6, and GF-8), the prepared composites were defined as CGF-0, CGF-2, CGF-4, CGF-6, and CGF-8. The parameters of each component are shown in Table 2.

Table 2.

Composition of composites

Composite code	Carbon fiber(wt %)	Epoxy resin(wt %)	Flame-retardant glass fiber veil(wt%)
CGF-0	40.4	50	9.6
CGF-2	40.4	50	9.6
CGF-4	40.1	50	9.9
CGF-6	39.6	50	10.4
CGF-8	39.8	50	10.2

Characterization of flame-retardant glass fiber veil

Chemical structure analysis

In order to have a clear understanding of the chemical structure changes of the glass fiber veil before and after modification, we used a FTIR spectrometer (Thermo Scientific Nicolet iS20, USA) to analyze it. The samples were prepared by the potassium bromide compression method, with wave number range 400–4000 cm⁻¹.

XPS is used for elemental quantitative analysis and identification of chemical structures carried out using AXIS ULTRA DLD (Kratos Analysis Ltd., UK), using an aluminum mono-Kα X-ray source (1486.6 eV) operated at 150 W. Overview scans were obtained with a pass energy of 160 eV and acquisition times of 220 s.

Thermal stability analysis

The thermal stability of the glass fiber veil before and after modification was analyzed by a thermogravimetric analyzer (Netzsch 209F1, Bavaria, Germany). In a nitrogen atmosphere, the nitrogen flow rate was 35 ml/min, the temperature was set at 30°C to 800°C, and the heating rate was constant at 20°C/min.

Surface morphology analysis

The function of a SEM is to observe the microscopic morphology of the glass fiber veil, which can further clarify the relationship between SWPU and the glass fiber veil. The surface of the glass fiber veil was observed with a Hitachi Regulus 8100, and the electron microscope voltage was constant at 10 kv. Before SEM observation, the sample was sputtered with gold-plated thin layers.

The DSA100 contact angle tester was used to measure the static contact angle between the glass fiber veil and epoxy resin. The contact angle before and after the resin contact was recorded, and the final contact angle was determined by the average value of at least five different positions to evaluate the wettability of the glass fiber surface.

Characterization of composites

Flame-retardant performance analysis

The LOI of composite materials was analyzed by the JF-5 automatic oxygen index analyzer (Nanjing Jionglei Instrument Equipment Co. Ltd., China), which was strictly implemented according to the ISO 4589-2 standard. All sample sizes are cropped to 80.0 mm × 10.0 mm × 2.0 mm. Ignite the sample in an oxygen-nitrogen mixture for a maximum of 30 s, and observe whether it continues to burn for 3 min or 50 mm under a certain oxygen concentration. Therefore, the samples were tested at least five times, and the LOI value was obtained after calculating the average value.

Furthermore, according to the UL94-2013 test standard, the vertical combustion test of composite materials was carried out in a vertical combustion chamber (Atlas Materials Testing Technology Co. Ltd., USA), and the sample size was 100.0 mm × 10.0 mm × 2.0 mm. By observing and recording the test results of the samples, and comparing them with the standard data, the fire rating of the composite material can be judged. Therefore, the test samples were tested five times in the vertical position.

The cone calorimeter (Fire Testing Technology Ltd., UK) testing of composites was performed according to ISO 5660-1-2015. Therefore, the sizes of the test samples are all 100 mm × 100 mm × 2 mm, and they are wrapped with aluminum foil and exposed to the external heat flow of 50 KW/m² horizontally, so the samples are tested five times and the average value is calculated. The time to ignition (TTI), peak heat release rate (PHR), total heat release rate (THR), and total smoke release (TSR) of each sample were recorded, respectively.

Raman spectroscopy was tested on a LabRam HR Evolution (France), and the measured wave numbers ranged from 600 cm⁻¹ to 2000 cm⁻¹. By measuring the peak area of the D peak and G peak in the Raman spectrum, the compactness of the carbon layer after the combustion of each sample was judged.

Mechanical properties measurement

Tensile testing of composite materials was carried out on the universal testing machine (LD25.504; Lishi, China), according to the ISO527-2012 standard. All samples were tested at least five times and averaged. The sample size was 200 mm × 25 mm × 2 mm.

For the three-point bending test, a universal testing machine (LD25.504; Lishi, China) was used to test according to the ISO178-2010 standard. The span was 32 mm with a crosshead speed of 2 mm/min, and the dimensions of all specimens were 50 mm × 10 mm × 2 mm.

Furthermore, the interlaminar shear strength (ILSS) was analyzed according to ISO14130-1997, and the sample size was 20 mm × 10 mm × 2 mm. All samples were tested at least five times and averaged.

The plastic pendulum impact testing machine (LZ21.251-B; Lishi, China) was used to test the notch impact strength of the composite material according to the ISO179-2010 standard. There was no notch spline, and the sample size of 50 mm × 10 mm × 2 mm.

Results and discussion

Characterization of flame-retardant glass fiber veil

Contact angle analysis for epoxy resin

With the SWPU coating treatment, the wettability of glass fiber filled with flame-retardant agents could be significantly affected, which is a key factor for resin impregnation during composite manufacture. Therefore, in order to achieve better fire-retardant and mechanical properties, it is necessary to study the interface between the flame-retardant veil and CF. Meanwhile, the morphology of polyurethane coating has a great influence on the bonding between fiber and matrix and the good interaction between fiber and matrix.

As is explicitly demonstrated in Figure 2(a), the contact angle of liquid epoxy resin on the glass fiber veil was presented. Compared with the unmodified glass fiber veil, the contact angle of resin for the modified glass fiber veil was increased with the increase of polyurethane content. For the moment when the resin contacted the fiber veil, the contact angle was increased from 119.5° to 135.6°, when the concentration of SWPU was increased from 0% to 8%, and the increase for contact angle from 29.3° to 44.3° was observed after the resin was soaked for 10 s, and the data are shown in Table 1. The reason for this change is the introduction of a large number of siloxane and oxygen-containing groups, and the enrichment of new carbonyl functional groups on the glass fiber surface.²⁴ Under the condition of external extrusion, the direct interface stability of SWPU coating and resin is better, which can maximize the structural toughness of SWPU. The improvement of the toughness of the fiber veil and the increase of interface binding force can effectively improve the structural stability of composite materials.^25,26

Figure 2.

The structure characterization of: (a) water contact angle; (b) Fourier transform infrared (FTIR) of glass fiber veil; and the X-ray photoelectron spectroscopy (XPS) patterns of (c) fiberglass contact angle with resin; (d) C 1 s and O 1 s of XPS.

Chemical structure analysis for glass fiber veil

In order to study the chemical functional group of glass fiber modified by SWPU and flame-retardant agents, a series of characterizations were carried out. The FTIR results are shown in Figure 2(b). The spectra peek of 3420 cm^–1 is related to the Si-OH bond formed by hydrolysis of Si-O for glass fiber soaked in water, while the absorption peak at 1024 cm⁻¹ corresponds to the stretching vibration peak of the Si-O group.²⁷ In addition, 3320 cm⁻¹ corresponds to the amino bond (–NH) in SWPU, while the weak amino absorption peak was also observed for GF-8. Moreover, the absorption peak at 2945 cm⁻¹ to 2794 cm⁻¹, 1740 cm⁻¹, and 1222 cm⁻¹ corresponded to the functional group of –CH₂ and the carbonyl group of polyurethane.

As such, a significant change in the vibration peak can be found at 800 cm⁻¹ for the glass fiber veil (GF-2 to GF-8) with SWPU treatment with the increase of SWPU content, indicating that more Si–O–Si is formed in the material.²⁸ In addition, this is also attributed to the formation of the new bond of the Si–O–Si group indicating a cross-linked reaction between SWPU and glass fiber, resulting in an improvement of the stability and toughness of the structure.^29,30

Furthermore, in order to verify further the chemical cross-linking between SWPU and glass fiber, the elemental composition and relative contents of each element were analyzed by XPS and are shown in Figure 2(c). The presence of C, O, Si, Ca, N, and other elements was found in the XPS measurement spectrum of the glass fiber veil. When compared with the treated glass fiber veil (GF-2), the presence of the N element was not found for the unmodified glass fiber (GF-0). The atomic percentages observed for the element of C, O, SI, Ca, and N on the surface of the modified veil were 47.86%, 31.29%, 18.06%, 1.83%, and 1.56%, respectively. As shown in Figure 2(d), the C 1 s XPS spectrum of glass fiber could be deconvoluted into three peaks: C–O, COO, and C=O, while the XPS spectrum of O 1 s could be deconvoluted into three peaks: C–O, C=O, and Si–O.^15,31 Based on the O 1 s spectrum, the chemical bond Si–O–Si was confirmed to form between SWPU and glass fiber, with the treatment of SWPU. In addition, the C–N bond also appeared in the C 1 s map, which is the characteristic formation bond of SWPU.

Thermal analysis of glass fiber veil

Thermogravimetric analysis (TGA) was used for the mass loss of different samples in the nitrogen atmosphere. The pyrolysis process is shown in Figure 3, and the specific data are shown in Table 3. As shown in the illustration, the weight of samples started to decrease at 100°C, which could be related to the evaporation of moisture in the glass fiber veil and the decomposition of residual solvent. From the thermogravimetric curves, all the samples presented two main thermal decomposition stages, the first stage (280–310°C) is probably associated with the breakdown of urethane groups (–NH–(C=O)–O) and decomposition products undergo secondary reactions with the main chain. The second stage of decomposition occurs at 390°C to 430°C, which corresponds to the partial decomposition of organosilicon molecules and is mainly the removal of organic side groups from the main chain (–OH), and the decomposition of the main chain occurs at about 410°C.^32,33

Figure 3.

Thermogravimetric analysis (TGA) and derivative thermogravimetric (DTG) curves of flame-retardant glass fiber veil.

Table 3.

Detailed data for TGA

Sample	T_onset(°C)	T_max1 (°C)	T_max2 (°C)	Residue(wt %)
GF-0	/	/	/	98.89
GF-2	232.1	282.9	390.1	90.34
GF-4	244.2	286.4	394.7	84.87
GF-6	256.1	297.4	407.3	79.27
GF-8	265.6	301.9	406.9	73.16
SWPU	269.5	288.6	425.3	2.43

SWPU: silicone-modified waterborne polyurethane; TGA: thermogravimetric analysis.

With the increase of SWPU content (GF-2 to GF-4), the initial decomposition temperature (T_onset) and the temperature corresponding to the maximum thermal decomposition rate (T_max1 and T_max2) of the composites are observed to change as well. The increase of T_onset from 232.1°C (GF-2) to 265.6°C (GF-8) was attributed to the formation of the chemical bond Si–O–Si, which can effectively improve the thermal stability of the material. However, when compared with SWPU, the decomposition temperature for glass fiber was found to reduce with the addition of SWUP, which was related to the flame-retardant agents in the fiber veil improving the thermal stability of SWPU.³⁴

Morphology of glass fiber veil

The morphology of the coating is very important for the toughness of the fiber and its interaction with the matrix. Thus, in order to observe clearly the surface quality with SWPU coating, it is necessary to conduct a SEM observation. Figure 4(a) shows the SEM of the unmodified glass fiber veil, whose surface was smooth and neat, the glass fibers were relatively independent of each other, resulting in loss of structure states. With the treatment of SWPU, most of the coating is uneven, as shown in Figure 4(b). In addition, it can be clearly seen that the integrity of the coating is not good, and the agglomeration of the coating is obvious in Figure 4(c). When the content of SWPU was increased to 6 wt%, the surface area covered by the coating was large with the reduced agglomeration phenomenon, and the integrity of the coating began to improve. As can be seen in Figure 4(e), the fiber is completely covered by SWPU coating, and the integrity of the coating is much better than others. Normally, the better the coating integrity, the better the fiber veil structure stability with higher structure toughness.^24,31

Figure 4.

Scanning electron microscope (SEM) images of glass fiber veil before and after silicone-modified waterborne polyurethane (SWPU) coating treatment.

Fire-retardant performance test

To evaluate the fire safety performance of composite materials in a near-real fire scenario, UL-94 vertical burning tests, the LOI, and the cone calorimeter test was used to evaluate the combustion behavior. The key parameters to assess fire hazards include the TTI, PHR, THR, TSR, CO₂ release rate (CO₂), and fire generation index (FGI = TTI/PHR), and are recorded in Table 4 and Figure 5.

Table 4.

Cone calorimetry detailed data of composites

Sample code	TTI(s)	PHR(KW/m²)	THR(MJ/m²)	TSR(m²/m²)	Average EHC(MJ/kg)	FGI	LOI(%)	UL-94
CGF-0	22	488.5	58.8	1324.9	32.6	0.045	26.6	V1
CGF-2	17	654.1	52.2	1241.7	31.1	0.026	25.5	V1
CGF-4	16	769.8	51.2	1085.1	30.2	0.021	24.3	V1
CGF-6	14	638.8	47.9	1250.1	29.6	0.022	23.6	V1
CGF-8	11	571.1	50.4	1227.3	28.2	0.019	22.7	V1

EHC: effective heat of combustion; FGI: fire generation index; LOI: limiting oxygen index; PHR: peak heat release rate; THR: total heat release rate; TSR: total smoke release; TTI: time to ignition.

Figure 5.

(a) The heat release rate; (b) carbon monoxide release rate of composite during cone calorimetry test; (c) total heat release; (d) total smoke production.

First, the control group sample with an LOI value of 26.6% can burn continuously without dripping during the UL-94 vertical burning test (V-1). With the treatment of SWPU, the LOI value of the composites with different proportions was significantly decreased, and 22.7% was observed for the CGF-8 composite. This is attributed to the surface protection of glass fiber veils with flame-retardant agents. In addition, no significant difference was observed for composites on the UL-94 vertical burning test. Whereas a significant reduction in smoke generation during combustion was observed, due to the free radical trapping, flammable gas phase dilution, and catalytic carbonization effects of composite flame-retardants.^35,36 However, the detailed smoke emission behavior should be discussed in the cone calorimetry analysis.

For the cone calorimetry test, the values of TTI for composites were observed to decrease with the treatment of the flame-retardant agent and SWPU, indicating ignition in advance with a high concentration of SWPU on the surface.

For the control composites CGF-0, a PHR of 488.5 kW/m² with TSP of 11.9 m² could be observed, while the PHR of the SWPU-modified composite CGF-2 was increased to 654.1 kW/m², and the TSP value reached 11.2 m². As such, the addition of SWPU could cause more heat release at the initial stage of combustion, but the amount of smoke release shows a decreasing trend due to Ni(OH)₂ in flame retardants. In addition, the decline in PHR values from 769.8 to 571.1 kW/m² could be observed for composites when the SWPU concentration increased from 4% to 8%. This could be the reason that the high concentration of SWPU is able to form a char layer due to polymer carbonating and catalyzation effects by DOPO. The carbonization of the polymer matrix and SWPU film promotes the stability of the char layer structure, which can effectively isolate heat transfer. As such, the flame-retardant insulation capability of the glass fiber veil with dense residual char could present better thermal insulation performance.³⁷

On the other hand, the THR value is the accumulation of heat released from composites, and the control group CGF-0 composite was 58.8 MJ/m². However, as a result of the formation of a char layer from the carbonation of SWPU, the decrease in THR value from 52.2 MJ/m² (CGF-2) to 51.2 MJ/m² (CGF-4) and 47.9 MJ/m² (CGF-6) was observed. However, it is worth noting that the THR value of 50.4 MJ/m² was recorded for CGF-8. This increased behavior is still unknown yet. Possibly, the increase in the SWPU concentration could provide more pyrolysis items to promote the combustion of composites, resulting in the increase of the THR value.

TSR represents the total smoke produced per unit area, which is the key factor that endangers life safety. For the TSR value of the composite, the CGF-0 composite without SWPU treatment was 1324.9 m²/m², which was reduced to 1085.1 m²/m² for the CGF-4 composite with a 4% concentration of SWPU coating. This could be related to the smoke inhibition effect of flame-retardant Ni(OH)₂ in the composite material, which can be decomposed by heat and promote the formation of carbon, and Ni²⁺ will capture hydroxyl radicals formed during the combustion process, thus reducing the generation of smoke. Insufficient material combustion will also lead to the generation of smoke, but Ni(OH)₂ has an excellent smoke-suppression function, able to capture the smoke particles generated during the combustion process, and the SWPU of the fiber veil could promote the stability of the char layer, to reduce the emission of smoke, with the synergistic effect of the gas phase and solidification phase.^23,38

Effective heat of combustion (EHC) is the ratio of the heat release rate to mass loss rate, which reflects the combustion degree of volatile gas in the gas phase, and has profound significance for evaluating the fire safety performance of materials. As shown in Table 4, the average EHC also decreased significantly, The EHC value of the control composite was 32.66 MJ/kg, and the EHC value of the composite with 8% SWPU treatment (CGF-8) was 28.25 MJ/kg, which decreased by 13.5%. The EHC value of the composite with other additives also showed different degrees of decline, which indicated that the flame retardancy of the gas phase occurred during the surface pyrolysis process.

The FGI is one of the significant indexes to evaluate the fire resistance performance of composite materials, which is the ratio of the ignition time of the composite to the peak heat release rate (FGI = TTI/PHR).³⁹ In theory, the lower the FGI, the safer the material. As depicted in Table 4, the FGI value of the composite was decreased from 0.045 kW/m²s for control CGF-0, to 0.026 kW/m²s, 0.021 kW/m²s, 0.022 kW/m²s, 0.019 kW/m²s for CGF-2, CGF-4, CGF-6, and CGF-8 composites, indicating that the fire resistance of the composites can be improved with SWPU treatment.

In order to clarify further the flame-retardant behavior of the glass fiber veil with SWPU treatment, and confirm the capability of SWPU to form a char layer when exposed to heat, the composite after cone calorimeter test and the SEM image for residual char layer were taken and recorded in Figure 5. It was obvious that the composites were expanded where the surface was covered with the residual char. In addition, with the increased concentration of SWPU from 0% to 8% for the fire-retardant glass fiber veil, the denser residual char could be observed to be more compact, which presented better thermal insulation and prevented pyrolysis gas phase exchange.

For the Raman spectra of the composite char analysis, there were two peaks that could be determined from Figure 6. The D peak at 1332 cm⁻¹ represented the vibration peak of the Sp³ hybrid (crystalline carbon), and the position of peak G at 1569 cm⁻¹ corresponded to the vibration peak of C–C of amorphous graphite or glassy carbon.⁴⁰ Normally, the ratio of these two peak areas (I_D/I_G) can be used to characterize the degree of carbonization. The low I_D/I_G value means a high degree of carbonization. In this case, the calculated I_D/I_G value for CGF-0 after combustion is 1.899 and decreased to 1.713 for a 8% SWPU concentration of CGF-8. As such, with the increasing SWPU concentration, the surface char layer of the glass fiber veil could be further carbonized, as SWPU was able to provide more efficient catalytic carbonization during combustion.⁴¹

Figure 6.

The image of the composite after cone calorimeter test, and the scanning electron microscope (SEM) image of residual char on the burnt composite surface with Raman spectrum analysis.

In theory, the combustion of composite materials involves two processes: the gas phase and the condensed phase, and the generation of smoke is mostly from the pyrolysis of the resin matrix in the gas phase due to partial combustion. According to the behavior of the composites prepared in this paper, a possible flame-retardant mechanism was proposed (Figure 7). In a near-real fire scenario, the organic flame retardants (DOPO) could decompose with free phosphorus oxygen radicals (PO· and PO₂) to capture H· and OH· in the gas phase, resulting in the inhibition of the chain reaction of pyrolysis.²² During the same time, DOPO can quench phosphate groups with hydroxyl groups to promote the formation of carbon layers in the condensed phase, which is able to isolate heat source and oxygen from gas phase infusion, so as to achieve the purpose of protecting the core material.

Figure 7.

The flame-retardant mechanism of composites with the fire- resistant glass fiber veil.

Furthermore, In the process of thermal decomposition of inorganic flame-retardant Ni(OH)₂, the endothermic reaction occurs first to generate water (H₂O), then dilute oxygen (O₂) and other combustible gases, absorb heat, and reduce temperature. At the same time, Ni(OH)₂ can capture the reaction of smoke (carbon particles) generated by thermal cracking or incomplete combustion of composite materials, generate noncombustible gases such as H₂O and carbon dioxide (CO₂), further dilute combustible gases such as O₂, and reduce heat transfer with the resin matrix. At the same time, high melting point particles (NiO and Ni) and resin matrix as well as the free radical adsorption effect will be generated during thermal decomposition to delay or prevent combustion reaction, so as to achieve the flame-retardant effect and protect the substrate.^23,38 More importantly, the glass fiber mat is tightly coated by SWPU, and the glass fiber and SWPU formed a Si–O–Si chemical bond with good heat stability to improve the stability of the overall structure. The synergistic flame retardants and SWPU in the gas phase and solidification phase together improve the flame retardancy of the composites.

Mechanical properties

With the addition of SWPU into the fire-resistant glass fiber veil for composite surface protection from fire hazards, the effect on the mechanical properties was a key factor also related to the application. In this study, the mechanical properties of composites were evaluated by the tensile properties test, bending properties test, notched impact strength, and interlaminar shear strength analysis.

Figure 8(a) shows the tensile test results for composites with different fire-resistant glass fiber veils. It is obvious that the tensile strength and Young’s modulus of the control group were 634.5 MPa and 62.7 GPa, respectively. With an increase in SWPU concentration, the tensile strength and Young’s modulus of the composites were increased to 700.4 MPa and 55.6 GPa for CGF-2 composite, 822.3 MPa, 63.9 GPa for CGF-4 composite, 850.2 MPa and 60.5 GPa for CGF-6, 938.1 MPa and 57.4 GPa for CGF-8, respectively. In addition, as is shown in Table 5, the flexural strength of the composites was improved significantly with the treatment of SWPU for the glass fiber veil. When compared with the control group, the flexural strength was improved from 402.2 MPa to 652.3 MPa, while the flexural modulus varied significantly. With the increase of the SWPU concentration, the modulus shows a decreasing trend. This could be for the fact that the glass fiber layer becomes a part of the composite structure. As such, based on the ‘mixing rule’ of composites, it is known that the Young’s modulus of the composites was decreased with the addition of the glass fiber veil layer due to the lower Young’s modulus for glass fiber.^42,43

Figure 8.

The results of: (a) tensile and (b) flexural test, and the (c) interlaminar shear strength and (d) the impact strength of composites with fire-resistant glass fiber veil.

Table 5.

The mechanical properties of composites

Sample	Flexural strength(MPa)	Flexural modulus(GPa)	Tensile strength(MPa)	Young’s modulus(GPa)	Interlaminar shear strength(MPa)	Impact strength (KJ/m²)
CGF-0	402.2	28.5	635.5	62.7	25.2	852.4
CGF-2	475.5	48.2	700.4	55.6	37.1	960.6
CGF-4	569.2	31.9	822.3	63.9	30.6	1150.8
CGF-6	596.3	28.5	850.2	60.5	29.2	1235.1
CGF-8	652.3	41.0	938.1	57.4	32.1	1359.7

On the other hand, the interlaminar shear strength of composites is presented in Figure 8(c). When compared with the control, the interlamellar shear strength of composites with SWPU treatment was observed to increase, indicating that SWPU was able to improve the interface adhesion. However, with the increase in the SWPU concentration, the side effect on the interlaminar shear strength. As such, the SWPU coating for the glass fiber veil could transfer the main force at the initial stress stage with elastic deformation. However, too much SWPU coated on glass fiber could limit load transfer, resulting in serious delamination at the interface with poor load transfer.⁴⁴ Apart from this, the composites with the surface layer of glass fiber veil exhibited no significant effect on the structural stability of the core region (CF/epoxy) of the composite. This kind of structural design and technology can solve the problem of fiber interface bonding with matrix under the action of flame retardants.

On the other hand, for the notched impact strength shown in Figure 8(d), the impact strength of the composite with 2% SWPU (CGF-2) was 960.6 KJ/m², 12.6% higher than that of the pure composite (CGF-0) with 852.4 KJ/m². With the enhancement of polyurethane content, the impact strength of the corresponding composite also increased. With the 8% SWPU treatment, the impact strength was increased to 1359.7 KJ/m², which was 59.5% higher than that of the control, indicating SWPU could improve the composite toughness for epoxy resin composites.

As shown in Figure 9, the failure morphology of the CGF-0 control composite is relatively smooth, which proves that the epoxy resin is a typical brittle fracture.⁴⁵ However, with the addition of SWPU for the surface fire-retardant glass fiber veil, there are obvious fiber pull-out behaviors. More importantly, after the addition of polyurethane, the surface of CF is coated with a layer of film, and the flexible film is combined with the resin group through adhesion and chemical bonding, which is able to absorb energy under impact and effectively enhances the toughness of the composite.^46–48

Figure 9.

The scanning electron microscopy of composite failure structure.

For fire-retardant treatment coating on composites, there are several classic studies for comparison, Mazzocchetti et al.⁴⁹ prepared a thin poly (arnylon) nanofiber veil for fire-retardant coating of CF-reinforced composites by electrospinning, and confirmed that the flame-retardant effect of the outer surface coating (Cp-PMIA-W) was improved, but the mechanical properties were found to decrease. Khalili et al.⁵⁰ also reported a fiber-retardant mat to protect CF-reinforced composites and mentioned excellent flame retardancy for surface protection, while the decrease of bending strength was difficult to ignore. Therefore, the design of the structure with a flame-retardant layer on the composite surface could be confirmed as an effective composite protection solution, but the structural toughness and interface performance should be balanced. Based on the fire-retardant agent and SWPU treatment for the glass fiber veil, this study achieved both flame-retardant and mechanical properties by self-toughening and improving the interface binding force. The properties of flame retardancy and smoke suppression were excellent (THR decreased by 18.5%, TSP decreased by 18.1%), and the toughening effect was significant (flexural strength increased by 62.2%, impact strength increased by 59.5%).

Conclusions

In this study, an ingenious strategy to improve the fire resistance and mechanical properties of polymer composites was introduced. The SWPU was introduced to treat the glass fiber veil which was designed as a surface protection layer for composite surface protection. It was confirmed that SWPU could form a thin film to cover the glass fiber veil evenly, with a chemical bond (Si–O–Si), resulting in good adhesion of SWPU for interface bonding between the flame-retardant glass fiber layer and the core region, as well as improvement on the structural toughness of the composites. In addition, SWPU has a synergistic effect with flame retardants, which could jointly inhibit the combustion reaction of the gas phase and condensed phase, promoting the stability of the char layer.

In particular, the cone calorimetry results indicated that the THR decreased by 18.5% and the TSR decreased by 18.1%, due to the addition of SWPU forming a dense char layer on the surface preventing exchange between pyrolysis and oxygen. Compared with the control group, the flexural strength, tensile strength, and impact resistance of the composites were significantly increased by 62.2%, 47.3%, and 59.5% when compared to the control, respectively.

In summary, the sandwich structure design for composites with toughened fire-retardant glass fiber veils to protect the core region is an impressive solution to balance the flame retardancy and mechanical performance of high-performance composites, which could potentially be employed for high-value application areas such as aircraft and aerospace.

Footnotes

Declaration of conflicting interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was financially supported by the Zhejiang Provincial Key Research and Development Program (Grant No. 2021C01123); General Scientific Research Projects of the Education Department of Zhejiang (grant no. 21200069-F); ‘Top Soldier’ and ‘Leading Wild Goose’ R&D Project of Zhejiang (Grant No. 2022C01210); the National Natural Science Foundation of China (NSFC; Grant Nos. 51703200 and 51973197); the ‘Ten Thousand Plan’ – Zhejiang Provincial High Level Talents Special Support Plan (Grant No. 2020R52023); and the General Program of Ningbo Natural Science Foundation (Grant No. 2022J022).

ORCID iDs

Jiawei Li

Chenkai Zhu

References

Nawafleh

Celik

Additive manufacturing of short fiber reinforced thermoset composites with unprecedented mechanical performance. Addit Manuf 2020; 33: 101–109.

Shi

Dunn

, et al. Carbon fiber reinforced thermoset composite with near 100% recyclability. Adv Funct Mater 2016; 26: 6098–6106.

Capricho

Fox

Hameed

Multifunctionality in epoxy resins. Polym Rev 2020; 60: 1–41.

Han

M-C

H-W

Kong

W-K

, et al. High-performance electret and antibacterial polypropylene meltblown nonwoven materials doped with boehmite and ZnO nanoparticles for air filtration. Fibers Polym 2022; 23: 1–9.

Uflyand

IЕ

Irzhak

VI.

Formation of fiber composites with an epoxy matrix: state-of-the-art and future development. Mater Manuf Process 2022; 37: 723–747.

Osimitz

Kacew

Hayes

AW.

Assess flame retardants with care. Science 2019; 365: 992–993.

Zhang

Yang

Zhang

, et al. Influence of CeO₂ addition on forming quality and microstructure of TiCx-reinforced CrTi₄-based laser cladding composite coating. Mater Characteriz 2021; 171: 110732.

Gong

Sheng

, et al. A multifunctional flexible composite film with excellent multi‐source driven thermal management, electromagnetic interference shielding, and fire safety performance, inspired by a “brick–mortar” sandwich structure. Adv Func Mater 2022; 32: 2200570.

Yang

Mao

Zhang

, et al. Fast-curing halogen-free flame-retardant epoxy resins and their application in glass fiber-reinforced composites. Text Res J 2019; 89: 3700–3707.

10.

Yang

Wei

Qiu

, et al. A novel bio‐based, flame retardant and latent imidazole compound – its synthesis and uses as curing agent for epoxy resins. J Appl Polym Sci 2022; 139: e53079.

11.

Mohd Sabee

MMS

Itam

Beddu

, et al. Flame retardant coatings: additives, binders, and fillers. Polymers 2022; 14: 2911.

12.

Guo

Zhang

, et al. Synergic realization of electrical insulation and mechanical strength in liquid nitrogen for high-temperature superconducting tapes with ultra-thin acrylic resin coating. Supercond Sci Tech 2022; 35: 075014.

13.

Zhang

Zhao

, et al. A novel and facile strategy for highly flame retardant polymer foam composite materials: transforming silicone resin coating into silica self-extinguishing layer. J Hazard Mater 2017; 336: 222–231.

14.

Zhu

, et al. Fire performance of sandwich composites with intumescent mat protection: evolving thermal insulation, post-fire performance and rail industry testing. Fire Safety J 2020; 116: 103205.

15.

Zhu

Zhao

, et al. Fire properties of carbon fiber reinforced polymer improved by coating nonwoven flame retardant mat for aerospace application. J Appl Polym Sci 2019; 136: 47801.

16.

Cong

Khalili

Zhu

, et al. Investigation of fire protection performance and mechanical properties of thin-ply bio-epoxy composites. Polymers 2021; 13: 731.

17.

Lee

Ghaffarian

Kim

, et al. A study on dislocation mechanisms of toughening in cu-graphene nanolayered composite. Nano Lett 2021; 22: 188–195.

18.

Saka

Moon

W-J

Horiuchi

, et al. Toughening of brittle materials by dislocation cells. Microsc Microanal 2001; 7: 420–421.

19.

Cheng

Fan

, et al. Castor oil based high transparent UV cured silicone modified polyurethane acrylate coatings with outstanding tensile strength and good chemical resistance. Prog Org Coat 2022; 163: 106624.

20.

Chen

, et al. Nickel hydroxide and zinc hydroxystannate dual modified graphite carbon nitride for the flame retardancy and smoke suppression of epoxy resin. Polym Degrad Stab 2020; 182: 109366.

21.

Guo

Wang

Pan

, et al. Polyaniline‐coupled graphene/nickel hydroxide nanohybrids as flame retardant and smoke suppressant for epoxy composites. Polym Adv Technol 2019; 30: 1959–1967.

22.

Wang

and Buhro

WE.

Morphology control of cadmium selenide nanocrystals: insights into the roles of di-n-octylphosphine oxide (DOPO) and di-n-octylphosphinic acid (DOPA).

J Am Chem Soc 2012; 134: 5369–5380.

23.

Feng

Han

Wang

, et al. Multiple synergistic effects of graphene-based hybrid and hexagonal born nitride in enhancing thermal conductivity and flame retardancy of epoxy. Chem Eng J 2020; 379: 122402.

24.

Zhang

Liu

, et al. A novel surface modification of carbon fiber for high-performance thermoplastic polyurethane composites. Appl Surf Sci 2016; 382: 144–154.

25.

Vautard

Ozcan

Meyer

Properties of thermo-chemically surface treated carbon fibers and of their epoxy and vinyl ester composites. Compos Part A: Appl Sci and Manuf 2012; 43: 1120–1133.

26.

Lee

J-U

Cha

H-J

, et al. Partially reduced graphene oxide as a multi-functional sizing agent for carbon fiber composites by electrophoretic deposition. RSC Adv 2013; 3: 25609–25613.

27.

Liang

, et al. Synthesis of waterborne polyurethane by inserting polydimethylsiloxane and constructing dual crosslinking for obtaining the superior performance of waterborne coatings. Compos B Eng 2022; 238: 109889.

28.

Zeng

Shen

Duan

, et al. Effect of silane hydrolysis on the interfacial adhesion of carbon nanotubes/glass fiber fabric-reinforced multiscale composites. Text Res J 2018; 88: 371–391.

29.

Shao

Dong

Zhao

, et al. Tensile and water absorption properties of notched glass fiber/polypropylene and glass fiber/polyoxymethylene injection-molded composites. J Text I 2022: 1–9.

30.

Meng

Yen, Ping

, et al. Effects on the structure and properties of membranes formed by blending polydimethylsiloxane polyurethane into different soft-segment waterborne polyurethanes. J Appl Polym Sci 2006; 102: 210–221.

31.

Fazeli

Liu

Rudd

The effect of waterborne polyurethane coating on the mechanical properties of epoxy-based composite containing recycled carbon fibres. Surf Interfaces 2022; 29: 101684.

32.

Meng

Yen, Ping

33.

Yang

Nie

, et al. Highly stretchable and conductive carbon fiber/polyurethane conductive films featuring interlocking interfaces. ACS Appl Mater 2021; 13: 38656–38665.

34.

Pan

Y-T

Zhang

Yang

The rise of MOFs and their derivatives for flame retardant polymeric materials: a critical review. Compos B Eng 2020; 199: 108265.

35.

Schmidt

Ciesielski

Greiner

, et al. Novel organophosphorus flame retardants and their synergistic application in novolac epoxy resin. Polym Degrad Stab 2018; 158: 190–201.

36.

Nie

Liu

, et al. Decomposition of waste carbon fiber reinforced epoxy resin composites in molten potassium hydroxide. Polym Degrad Stab 2015; 111: 247–256.

37.

Hao

Zhu

Huang

, et al. A phosphorous–aluminium–nitride synergistic flame retardant to enhance durability and flame retardancy of cotton. ChemistrySelect 2019; 4: 13952–13958.

38.

Elmoubarki

Mahjoubi

Elhalil

, et al. Ni/Fe and Mg/Fe layered double hydroxides and their calcined derivatives: preparation, characterization and application on textile dyes removal. J Mater Res Technol 2017; 6: 271–283.

39.

Brzezińska

Bryant

Risk index method – a tool for building fire safety assessments. Appl Sci 2021; 11: 3566.

40.

Hou

Chu

, et al. A review on metal-organic hybrids as flame retardants for enhancing fire safety of polymer composites. Compos B Eng 2021; 221: 109014.

41.

Cui

, et al. Synergistic thermodynamic compatibility of polydimethylsiloxane block in thermoplastic polyurethane for flame retardant materials: super flexible, highly flame retardant and low smoke release. Polymer 2022; 253: 124976.

42.

Zhang

, et al. Enhanced mechanical and thermal properties of hollow wood composites filled with phase-change material. J Clean Prod 2020; 256: 120373.

43.

Saleh

El-Saied

Salaheldin

, et al. Macro-and nanomaterials for improvement of mechanical and physical properties of cement kiln dust-based composite materials. J Clean Prod 2018, 204: 532–541.

44.

Zhang

H-C

Kang

Chen

L-S

, et al. Enhancing toughness of poly (lactic acid)/thermoplastic polyurethane blends via increasing interface compatibility by polyurethane elastomer prepolymer and its toughening mechanism. Polym Test 2020; 87: 106521.

45.

Wang

Zhou

, et al. Measurement and analysis of thermophysical parameters of the epoxy resin composites shape-stabilized phase change material. Constr Build Mater 2019; 223: 368–376.

46.

Zhang

Song

Wang

, et al. Precipitation evolution, strengthening and toughening mechanisms of Fe–3Si–2Cu (in wt.%) steel during aging process. Mater Sci Eng A 2021; 806: 140863.

47.

Gao

Zhang

Liu

, et al. The deformation characteristics, fracture behavior and strengthening-toughening mechanisms of laminated metal composites: a review. Metals 2019; 10: 4.

48.

Askarinejad

Rahbar

Toughening mechanisms in bioinspired multilayered materials. J R Soc Interface 2015; 12: 20140855.

49.

Mazzocchetti L, Benelli T, Maccaferri E, et al. Polym-aramid electrospun nanofibrous mats as highperformance flame retardants for carbon fiber reinforced composites. Compos Part B-Eng 2018; 145: 252–260.

50.

Khalili

Blinzler

Kádár

, et al. Ramie fabric Elium® composites with flame retardant coating: Flammability, smoke, viscoelastic and mechanical properties. Compos Part A Appl Sci Manuf 2020; 137: 105986.

The carbon fiber/epoxy composites toughened by fire-resistant glass fiber veils: Flammability and mechanical performance

Abstract

Keywords

Experimental methods

Materials

Synthesis of Ni(OH)2

Preparation of toughened flame-retardant glass fiber veil

Preparation of composites with flame-retardant structure

Characterization of flame-retardant glass fiber veil

Chemical structure analysis

Thermal stability analysis

Surface morphology analysis

Characterization of composites

Flame-retardant performance analysis

Mechanical properties measurement

Results and discussion

Characterization of flame-retardant glass fiber veil

Contact angle analysis for epoxy resin

Chemical structure analysis for glass fiber veil

Thermal analysis of glass fiber veil

Morphology of glass fiber veil

Fire-retardant performance test

Mechanical properties

Conclusions

Footnotes

Declaration of conflicting interests

Funding

ORCID iDs

References

Synthesis of Ni(OH)₂