Prepregs for Temperature Resistant Composites

In this paper, carbon fabric reinforced inorganic matrix composites were prepared. The inorganic matrix based on alkali activated aluminosilicate was used because of its resistance to fire and the temperatures up to 1000 °C. Influence of heat treatment of fabric, high temperature treatment of composite and preparation method on the mechanical properties and morphology of the composites were studied. The preparation of composites with the subsequent steps of impregnation, layering and curing of the composites was compared with the prepreg preparation method, which separates the impregnation of the reinforcement from the production of the composite. The SEM photographs show no differences in morphology between composites prepared from heat treated fabric and composites prepared from original fabrics. All four series of samples were comparatively saturated with matrix. Despite this, tensile properties of heat-treated fabric composites were negatively affected. While composites with heat-treated fabric reached the tensile strength up to 274 MPa, composites prepared without heat-treated fabric exhibited strengths higher than 336 MPa. Samples exposed to temperatures reaching 600 °C retained up to 40% of their original strength. The effect of composite preparation method on the tensile properties of the composites has not been proved.


Introduction
Fiber-reinforced polymer composites are widely used as lightweight structural materials in aerospace, naval, navigation, electronics, and automotive industries. In most cases, composites are made from organic matrix and carbon, basalt, or glass fibers. These materials exhibit excellent properties such as high tensile and flexural strength, low density, and corrosion resistance [1][2][3]. The use of nanofiber structures as composite fillers is a recent trend in the field of composites [4,5]. The disadvantage of composites prepared from organic matrix is that they cannot be used at temperatures above 200 • C. This is one of the reasons why an inorganic matrix has been studied in recent years [3].
One of the perspective substitutes for an organic matrix for the preparation of composites is inorganic material based on the alkali-activated aluminosilicates (A-matrix), material that can be called geopolymer under certain conditions. A-matrix is formed by mixing powdered aluminosilicates with a liquid alkaline activator. A liquid alkali silicate and alkali metal hydroxide solution are usually used to dissolve material containing Si and Al such as metakaolin [6][7][8][9]. A-matrix is often reinforced by solid materials to improve its properties [9]. The properties of prepared A-matrix depend on the type and amount of aluminosilicate, activator, additive, water, Si/Al molar ratio, Na/Al molar ratio, Na + , or K + content and the curing conditions [9][10][11][12]. A-matrix can be cured at the low temperatures (even at a room temperature). This material has good mechanical properties and it is resistant to chemicals and temperatures up to 1000 • C.    Alkaline activator was prepared with composition of molar ratio SiO2/K2O = 1.15 and K2O/B2O3 = 5.15 by mixing the commercial potassium water glass, potassium hydroxide solution (KOH: H2O = 1:1 in weight ratio), solid boric acid, and distilled water. Activator was mixed in a blender (Kenwood KVL8400S Chef XL Titanium, Havant, United Kingdom) for 24 hours and stored in the fridge at 5 °C for two days. Then, the metakaolinite-rich material and silica fume were added to the alkaline activator. This mixture of A-matrix with composition of SiO2/Al2O3 = 33.9, K2O/Al2O3 = 3.98, H2O/K2O = 12.1 (molar ratio) was blended c. 30 min, stored in a freezer at -18 °C for 24 hours and then used to prepare four six-layer composite plates.
Four six-layer composite plates were prepared from 24 pieces of 50 cm × 30 cm carbon fabric. Composite series were different in the combination of fabric pre-treatment and preparation method. For clarity, the scheme of composites preparation with the sample marking system is shown in Figure  3. Twelve source pieces of carbon fabric for preparation of the first and second composite plates were kept in the air (CA) until the composite preparation. The remaining twelve pieces of fabric were placed into an oven at 300 °C for one hour to remove the epoxy layer from the surface of the carbon fibers (CO).    Alkaline activator was prepared with composition of molar ratio SiO2/K2O = 1.15 and K2O/B2O3 = 5.15 by mixing the commercial potassium water glass, potassium hydroxide solution (KOH: H2O = 1:1 in weight ratio), solid boric acid, and distilled water. Activator was mixed in a blender (Kenwood KVL8400S Chef XL Titanium, Havant, United Kingdom) for 24 hours and stored in the fridge at 5 °C for two days. Then, the metakaolinite-rich material and silica fume were added to the alkaline activator. This mixture of A-matrix with composition of SiO2/Al2O3 = 33.9, K2O/Al2O3 = 3.98, H2O/K2O = 12.1 (molar ratio) was blended c. 30 min, stored in a freezer at -18 °C for 24 hours and then used to prepare four six-layer composite plates.
Four six-layer composite plates were prepared from 24 pieces of 50 cm × 30 cm carbon fabric. Composite series were different in the combination of fabric pre-treatment and preparation method. For clarity, the scheme of composites preparation with the sample marking system is shown in Figure  3. Twelve source pieces of carbon fabric for preparation of the first and second composite plates were kept in the air (CA) until the composite preparation. The remaining twelve pieces of fabric were placed into an oven at 300 °C for one hour to remove the epoxy layer from the surface of the carbon fibers (CO). Four six-layer composite plates were prepared from 24 pieces of 50 cm × 30 cm carbon fabric. Composite series were different in the combination of fabric pre-treatment and preparation method. For clarity, the scheme of composites preparation with the sample marking system is shown in Figure 3. Twelve source pieces of carbon fabric for preparation of the first and second composite plates were kept in the air (CA) until the composite preparation. The remaining twelve pieces of fabric were placed into an oven at 300 • C for one hour to remove the epoxy layer from the surface of the carbon fibers (CO). Then, all fabrics were impregnated in a conventional manner with the A-matrix using a paint roller. Six impregnated CA fabric pieces were used for composite plate by classic method (CAC). The CAC plate was prepared by stacking impregnated fabrics one by one. The other six pieces of fabric were used for preparation of composite plate by prepreg method (CAP). In this case, the impregnated carbon fabrics were individually placed between two pieces of plastic foil (two pieces for every fabric) to prepare the prepregs. Prepregs were stored in a freezer at −18 • C. After seven days, the prepregs were taken out of the freezer, stripped of plastic foil, and used for preparation of composites by stacking one by one to get the CAP composite plate as in the case of CAC plate. Twelve impregnated carbon CO fabrics were used for preparation of the COC plate by a classic method and the COP plate by the prepreg method in the same way as CAC and CAP composite plates. Every prepared composite plate was placed between two pieces of peel-ply fabric, wrapped in a plastic foil, compressed at 440 kPa for one hour and then cured in the oven at 65 • C for 3 h. After this time, the plates were unwrapped from the plastic foil and peel-ply fabric and finally cured for 28 days at laboratory conditions. The fabric mass fraction of the plates is presented in Table 2. Then, all fabrics were impregnated in a conventional manner with the A-matrix using a paint roller. Six impregnated CA fabric pieces were used for composite plate by classic method (CAC). The CAC plate was prepared by stacking impregnated fabrics one by one. The other six pieces of fabric were used for preparation of composite plate by prepreg method (CAP). In this case, the impregnated carbon fabrics were individually placed between two pieces of plastic foil (two pieces for every fabric) to prepare the prepregs. Prepregs were stored in a freezer at -18 °C. After seven days, the prepregs were taken out of the freezer, stripped of plastic foil, and used for preparation of composites by stacking one by one to get the CAP composite plate as in the case of CAC plate. Twelve impregnated carbon CO fabrics were used for preparation of the COC plate by a classic method and the COP plate by the prepreg method in the same way as CAC and CAP composite plates. Every prepared composite plate was placed between two pieces of peel-ply fabric, wrapped in a plastic foil, compressed at 440 kPa for one hour and then cured in the oven at 65 °C for 3 hours. After this time, the plates were unwrapped from the plastic foil and peel-ply fabric and finally cured for 28 days at laboratory conditions. The fabric mass fraction of the plates is presented in Table 2.  Four prepared composite plates were cut into 250x25 mm samples ( Figure 4) by water jets. The obtained samples were kept at a laboratory temperature (LT) or treated with temperatures of 400, 500, and 600 °C for one hour. The treated temperatures were added to the sample names (CAC-LT, COP-400, etc.). All samples were tested for tensile strength and modulus of elasticity using the universal testing machine LabTest 6.200 (maximum load of the sensor 200 kN) at a loading speed of 2 mm/min. (LaborTech, s.r.o., Opava, Czech Republic) complying with ASTM 3039 ( Figure 5). The ends of the samples were reinforced with epoxy resin coating and covered with sandpaper to protect the composite surface from sharp grips. Prepared composite samples were studied by a scanning electron microscope.  Four prepared composite plates were cut into 250 × 25 mm samples ( Figure 4) by water jets. The obtained samples were kept at a laboratory temperature (LT) or treated with temperatures of 400, 500, and 600 • C for one hour. The treated temperatures were added to the sample names (CAC-LT, COP-400, etc.). All samples were tested for tensile strength and modulus of elasticity using the universal testing machine LabTest 6.200 (maximum load of the sensor 200 kN) at a loading speed of 2 mm/min. (LaborTech, s.r.o., Opava, Czech Republic) complying with ASTM 3039 ( Figure 5). The ends of the samples were reinforced with epoxy resin coating and covered with sandpaper to protect the composite surface from sharp grips. Prepared composite samples were studied by a scanning electron microscope.     Table 3. High content of Al2O3 and SiO2 with lower content of K2O and Na2O (points S_1, S_2, S_3) indicate presence of metakaolin, and the increased SiO2 content (S_4) induces undissolved SiO2 particle, probably covered by a thin layer of dissolved components. The bounded area (S_5) contains all components of the A-matrix in a proportion corresponding to the amount of material preparation.      Table 3. High content of Al2O3 and SiO2 with lower content of K2O and Na2O (points S_1, S_2, S_3) indicate presence of metakaolin, and the increased SiO2 content (S_4) induces undissolved SiO2 particle, probably covered by a thin layer of dissolved components. The bounded area (S_5) contains all components of the A-matrix in a proportion corresponding to the amount of material preparation.    Table 3. High content of Al 2 O 3 and SiO 2 with lower content of K 2 O and Na 2 O (points S_1, S_2, S_3) indicate presence of metakaolin, and the increased SiO 2 content (S_4) induces undissolved SiO 2 particle, probably covered by a thin layer of dissolved components. The bounded area (S_5) contains all components of the A-matrix in a proportion corresponding to the amount of material preparation.    Table 3. High content of Al2O3 and SiO2 with lower content of K2O and Na2O (points S_1, S_2, S_3) indicate presence of metakaolin, and the increased SiO2 content (S_4) induces undissolved SiO2 particle, probably covered by a thin layer of dissolved components. The bounded area (S_5) contains all components of the A-matrix in a proportion corresponding to the amount of material preparation.

Carbon Fabric
The carbon fiber surface studied by SEM is presented in Figure 8. Figure 8a shows the surface of the fibers not treated with elevated temperature, and surface of the fibers treated at 300 °C for one hour is showed in Figure 8b. The SEM photographs indicate that the differences in surface between these fibers are minimal and that heat treatment of the carbon fabric probably did not adversely affect the surface. Both types of fibers were subjected to the tensile strength test to compare the tensile properties. The obtained values (Table 4) confirmed that the exposure to temperature of 300 °C did not lead up to the deterioration of fibers' tensile properties.

Carbon Fabric
The carbon fiber surface studied by SEM is presented in Figure 8. Figure 8a shows the surface of the fibers not treated with elevated temperature, and surface of the fibers treated at 300 • C for one hour is showed in Figure 8b. The SEM photographs indicate that the differences in surface between these fibers are minimal and that heat treatment of the carbon fabric probably did not adversely affect the surface. Both types of fibers were subjected to the tensile strength test to compare the tensile properties. The obtained values (Table 4) confirmed that the exposure to temperature of 300 • C did not lead up to the deterioration of fibers' tensile properties.

Carbon Fabric
The carbon fiber surface studied by SEM is presented in Figure 8. Figure 8a shows the surface of the fibers not treated with elevated temperature, and surface of the fibers treated at 300 °C for one hour is showed in Figure 8b. The SEM photographs indicate that the differences in surface between these fibers are minimal and that heat treatment of the carbon fabric probably did not adversely affect the surface. Both types of fibers were subjected to the tensile strength test to compare the tensile properties. The obtained values (Table 4) confirmed that the exposure to temperature of 300 °C did not lead up to the deterioration of fibers' tensile properties.   Figure 9 presents the morphology of prepared composites. These SEM photos of the sample cross sections confirm that the carbon fiber distribution is similar in all composite samples, and the inorganic matrix surrounds the individual fibers of the fabrics independently of the thermal treatment of the fabrics or the preparation method. The uniform fiber saturation is one of the reasons for the high strength of composites.   Figure 9 presents the morphology of prepared composites. These SEM photos of the sample cross sections confirm that the carbon fiber distribution is similar in all composite samples, and the inorganic matrix surrounds the individual fibers of the fabrics independently of the thermal treatment of the fabrics or the preparation method. The uniform fiber saturation is one of the reasons for the high strength of composites. The fiber distribution of the fabric in the prepared composite is shown in the Figure 10. The comparison of the composite without heat treatment COC-LT (a) and after the heat treatment COC-500 (b) can be seen here. The embrittlement of the whole sample matrix is more significant with increasing the temperature, the fiber-bonding matrix cracking, crumbling and moving away from the fiber surface. This leads to a decrease in the tensile strength of the composites. The fiber distribution of the fabric in the prepared composite is shown in the Figure 10. The comparison of the composite without heat treatment COC-LT (a) and after the heat treatment COC-500 (b) can be seen here. The embrittlement of the whole sample matrix is more significant with increasing the temperature, the fiber-bonding matrix cracking, crumbling and moving away from the fiber surface. This leads to a decrease in the tensile strength of the composites.   Figure 9 presents the morphology of prepared composites. These SEM photos of the sample cross sections confirm that the carbon fiber distribution is similar in all composite samples, and the inorganic matrix surrounds the individual fibers of the fabrics independently of the thermal treatment of the fabrics or the preparation method. The uniform fiber saturation is one of the reasons for the high strength of composites. The fiber distribution of the fabric in the prepared composite is shown in the Figure 10. The comparison of the composite without heat treatment COC-LT (a) and after the heat treatment COC-500 (b) can be seen here. The embrittlement of the whole sample matrix is more significant with increasing the temperature, the fiber-bonding matrix cracking, crumbling and moving away from the fiber surface. This leads to a decrease in the tensile strength of the composites.

Tensile Properties of Composites
The influence of the composites' preparation method on their tensile properties was studied on four series of samples CAC, CAP, COC, and CAP composite plates. Changes in tensile properties due to heat treatment of the fabrics and use of two methods of plate preparation and finally influence of high temperature treatment on composite samples were observed. Tensile strength, strain, and Young's modulus were recorded for temperatures ranging from laboratory temperature to 600 • C on 30-day-old samples. The average values of tensile properties of composites are summarized in Figures 11 and 12. high temperature treatment on composite samples were observed. Tensile strength, strain, and Young's modulus were recorded for temperatures ranging from laboratory temperature to 600 °C on 30-day-old samples. The average values of tensile properties of composites are summarized in Figure  11 and Figure 12.
As expected [21][22][23], the composite samples cured at laboratory temperature in all four series (CAC-LT, CAP-LT, COC-LT, COP-LT) had the highest tensile strength and the strength decreased with increasing cure temperature. In Figure 11, we can see a significant difference in tensile strength (up to 63 MPa) between composites made of heat-treated fabrics (COC-LT, COP-LT) and composites made from fabrics with no heat treatment (CAC-LT, CAP-LT). The decrease in tensile strength could be due to the interaction of the alkaline matrix with the temperature exposed fabrics. This was visible only for samples cured at laboratory temperature. In case of the samples cured at temperatures 400-600 °C, the differences diminished. Each of these composite samples was treated by high temperature, so damages were similar. The above described facts will be examined in the following study. The measured values correlate with Krystek et al. [23].   high temperature treatment on composite samples were observed. Tensile strength, strain, and Young's modulus were recorded for temperatures ranging from laboratory temperature to 600 °C on 30-day-old samples. The average values of tensile properties of composites are summarized in Figure  11 and Figure 12.
As expected [21][22][23], the composite samples cured at laboratory temperature in all four series (CAC-LT, CAP-LT, COC-LT, COP-LT) had the highest tensile strength and the strength decreased with increasing cure temperature. In Figure 11, we can see a significant difference in tensile strength (up to 63 MPa) between composites made of heat-treated fabrics (COC-LT, COP-LT) and composites made from fabrics with no heat treatment (CAC-LT, CAP-LT). The decrease in tensile strength could be due to the interaction of the alkaline matrix with the temperature exposed fabrics. This was visible only for samples cured at laboratory temperature. In case of the samples cured at temperatures 400-600 °C, the differences diminished. Each of these composite samples was treated by high temperature, so damages were similar. The above described facts will be examined in the following study. The measured values correlate with Krystek et al. [23].   As expected [21][22][23], the composite samples cured at laboratory temperature in all four series (CAC-LT, CAP-LT, COC-LT, COP-LT) had the highest tensile strength and the strength decreased with increasing cure temperature. In Figure 11, we can see a significant difference in tensile strength (up to 63 MPa) between composites made of heat-treated fabrics (COC-LT, COP-LT) and composites made from fabrics with no heat treatment (CAC-LT, CAP-LT). The decrease in tensile strength could be due to the interaction of the alkaline matrix with the temperature exposed fabrics. This was visible only for samples cured at laboratory temperature. In case of the samples cured at temperatures 400-600 • C, the differences diminished. Each of these composite samples was treated by high temperature, so damages were similar. The above described facts will be examined in the following study. The measured values correlate with Krystek et al. [23].
The COP-600 plate had the lowest tensile strength. Removal of the organic sizing from the fabric in combination with the prepreg method proved to be the composite with the lowest tensile strength. The seven-day long exposure to the alkaline matrix on the temperature treated fibers probably led to the interaction with the fibers and the strength decreased compared to the other plates affected by 600 • C. The typical load vs. crosshead displacement profile of prepared composite is showed in Figure 13. The COP-600 plate had the lowest tensile strength. Removal of the organic sizing from the fabric in combination with the prepreg method proved to be the composite with the lowest tensile strength. The seven-day long exposure to the alkaline matrix on the temperature treated fibers probably led to the interaction with the fibers and the strength decreased compared to the other plates affected by 600 °C. The typical load vs. crosshead displacement profile of prepared composite is showed in Figure  13. The influence of the preparation method on the strength of the composites was not significant. The strength differences were predominantly within the standard deviation. While the classic method is faster, the prepreg method is advantageous in the industrial sphere, where it is necessary to divide production processes into fabric lamination and composite production. The final product can be shaped and layered from the prepreg prepared by the process described in this article to 30 days from the preparation of the matrix.
The behavior of the Young's modulus is illustrated in Figure 12. Samples cured at laboratory temperature show the highest Young's modulus. In contrast, samples treated at 600 °C showed the lowest values. In this case, the modulus was c. 1/3 compared to the modulus of samples cured at laboratory temperature. The measured values of four composite series are very similar; the differences are within the standard deviation.

Conclusions
In this investigation, the carbon fabrics reinforced aluminosilicate matrix composites were prepared by a simple classic method and by the prepreg method. Effects of heat treatment of fabric, high temperature treatment of the composite, and the preparation method on the mechanical properties and morphology of the composites were studied. Results lead to these conclusions: • All four types of composites showed homogenous microstructure and carbon fabric was well infiltrated by the inorganic aluminosilicate matrix independent of the fiber treatment or preparation method. • The highest tensile strength was seen in samples prepared without fiber heat treatment, with classic lay-up samples exhibiting a strength of 336 ± 19 MPa and prepreg prepared samples exhibiting a strength of 339 ± 9 MPa. • The composites lost high tensile strength with increasing curing temperature, but they retained 30-40 % of their original strength at 600 °C. • Significant decrease in tensile strength of samples with heat treated fabric. Therefore, removal of the organic sizing by elevated temperature did not show any positive effects. • The method of preparation of composite had no significant effect on the tensile strength or Young's modulus of the samples. The prepreg method of composite preparation is, in terms of tensile properties, a good substitute for classic composite preparation. The influence of the preparation method on the strength of the composites was not significant. The strength differences were predominantly within the standard deviation. While the classic method is faster, the prepreg method is advantageous in the industrial sphere, where it is necessary to divide production processes into fabric lamination and composite production. The final product can be shaped and layered from the prepreg prepared by the process described in this article to 30 days from the preparation of the matrix.
The behavior of the Young's modulus is illustrated in Figure 12. Samples cured at laboratory temperature show the highest Young's modulus. In contrast, samples treated at 600 • C showed the lowest values. In this case, the modulus was c. 1/3 compared to the modulus of samples cured at laboratory temperature. The measured values of four composite series are very similar; the differences are within the standard deviation.

Conclusions
In this investigation, the carbon fabrics reinforced aluminosilicate matrix composites were prepared by a simple classic method and by the prepreg method. Effects of heat treatment of fabric, high temperature treatment of the composite, and the preparation method on the mechanical properties and morphology of the composites were studied. Results lead to these conclusions:

•
All four types of composites showed homogenous microstructure and carbon fabric was well infiltrated by the inorganic aluminosilicate matrix independent of the fiber treatment or preparation method.

•
The highest tensile strength was seen in samples prepared without fiber heat treatment, with classic lay-up samples exhibiting a strength of 336 ± 19 MPa and prepreg prepared samples exhibiting a strength of 339 ± 9 MPa.

•
The composites lost high tensile strength with increasing curing temperature, but they retained 30-40 % of their original strength at 600 • C. • Significant decrease in tensile strength of samples with heat treated fabric. Therefore, removal of the organic sizing by elevated temperature did not show any positive effects.

•
The method of preparation of composite had no significant effect on the tensile strength or Young's modulus of the samples. The prepreg method of composite preparation is, in terms of tensile properties, a good substitute for classic composite preparation.