Evaluation of new GFRTP and CFRTP using epsilon caprolactam as matrix fabricated with VaRTM

2.1 Matrix and reinforcement

The matrix of FRTP was obtained from anionic ring opening polymerizing for a monomer of ε-caprolactam with sodium salt of ε-caprolactam as a catalyst and hexa-methylene diisocyanate as an activator. The reinforcements were twilled fabrics of carbon fibers and plain woven textiles of glass fibers. The thickness of the carbon fabrics was 0.22 mm, and their fabric densities were 12.5 tows/25 mm in both directions of the warp and the waft. The thickness of the glass textiles was 0.21 mm, and their fabric densities were 20 tows/25 mm in both directions of the warp and the waft.

The carbon fabrics were washed with acetone in order to remove the carboxylic acid of the binding agent before the installation into the lower metallic die because this agent gave undesirable effects on the hardening of ε-caprolactam.

In contrast, the surface treatment of glass fabrics was performed with a silane coupling agent.

2.2 Molding method

2.2.1 Vacuum-assisted resin transfer molding

The catalyst for the anionic polymerization of ε-caprolactam may lose its catalytic function owing to the humidity in the air, and polymerization may fail as a result. Therefore, to fabricate FRTP with ε-caprolactam, it may be necessary to use sealed (air tight) molding methods that can control the humidity in the system, e.g., resin transfer molding, vacuum-assisted resin transfer molding (VaRTM) and infusion molding using a vacuum bag. This study selected the VaRTM method in which a simple vacuum pump was used as a resin infusion system for fabricating two kinds of FRTPs. Figure 1 shows a schematic view of the VaRTM molding system.

Figure 1

Schematic view of VaRTM for FRTP.

At first, the fabrics were piled on the lower metallic die in a closed space bounded by the upper metallic die (sealing the mold) and they were heated at specified temperatures of 120°C, 140°C, 160°C and 200°C. At the same time, the interior of the mold dies was decompressed by a vacuum pump at up to -90% gauge pressure. After that, the monomer solution of ε-caprolactam was mixed with the catalyst and the activator at 110°C, and the mixed monomer solution was poured into the mold dies maintained at the specified temperature. After the mixed monomer was poured, the mold and the solution were kept heated for some curing time at the specified temperature. Finally, the FRTP plate composed of PA6 as the matrix and the reinforcement fabrics was removed from the mold without cooling.

The obtained FRTP plate had neither voids nor unfilled parts because ε-caprolactam had a very low viscosity before the polymerization. Figure 2 shows the fabricated GFRTP and CFRTP. Hereafter, they are called I-GFRTP and I-CFRTP, respectively, and the resin plate fabricated with VaRTM is called I-PA6. All of the dimensions were the same, and their lengths, widths and thicknesses were 710, 630 and 3 mm, respectively. They had a flange of 50 mm in depth and 50 mm in width around their four sides. The number of reinforcement fabrics was 15 for the GFRTP and 13 for the CFRTP, and the volume fractions were 42% for the GFRTP and 49% for the CFRTP, respectively.

Figure 2

Photographs of fabricated FRTP plates: (A) I-CFRTP and (B) I-GFRTP.

3.1 Measurements of heat of fusion and degree of crystallinity

In order to evaluate the crystallinity, the I-PA6 plate and the matrix of I-GFRTP and those of the C-PA6 and the C-GFRTP were measured. In the case of I-PA6 and I-GFRTP, their values were measured twice, namely, the heat of fusions for the I-PA6 and I-GFRTP specimens after the first molding was immediately measured (first heating). After remelting (over 250°C) and then cooling, the heat of fusions for the specimens was measured (second heating) in order to examine their values of crystallinity after the polymerization.

Each resin sample of 10 mg was placed in an aluminum pan, and a differential scanning calorimeter was used to evaluate its thermal properties in an atmosphere of nitrogen supplied at a rate of 40 ml/min.

The temperature of the samples was raised from room temperature to 160°C, 180°C, 200°C, 220°C, 240°C and 250°C at a rate of 20°C/min, and the heat of fusion was measured for the crystals formed upon the polymerization (first heating).

Next, in order to evaluate the heat of fusion in the samples of I-PA6, I-GFRTP and I-CFRTP after polymerization and to compare them with the results of C-PA6, C-GFRTP and C-CFRTP, the samples were heated again up to each fabrication temperature at a rate of 20°C/min to measure the heat of fusion for the crystals of the samples (second heating). With the heat of fusion obtained from the first and the second heating, the degrees of crystallinity (DC) for each sample were calculated by the following equation:

where ΔH_m is the measured heat of fusion and ΔHm100% is the theoretical value of the heat of fusion for a polymer of 100% crystallinity. The theoretical value of 188 J/g for PA6, reported by Dole and Wunderlich [13], was used here.

3.2 Measurements of unreacted monomer and water absorption

Since the monomer of in situ polymerizing ε-caprolactam left the unreacted state after the polymerization dissolved quickly in water, the percentage of the unreacted portion of the monomer in the matrices of I-PA6, I-GFRTP and I-CFRTP, and their water absorptions were measured using the following procedure. Just after the molding, a rectangular piece of size 10 mm×60 mm was cut from the respective plates of 3 mm in thickness. Each piece was dried at 60°C for 24 h in a depressurized chamber and its weight (W₀) was measured precisely. Then each was soaked in hot water at 80°C for 72 h and then its weight (W₂) was measured precisely. Finally, each was again dried at 60°C for 72 h in a vacuum chamber and then its weight (W₁) was measured. With these weight values and Eqs. (2) and (3), the percentage of the unreacted portion of the monomer (W_u) and the water absorption (W_a) were calculated.

3.3 Three-point bending test

In order to obtain the bending strength and modulus of each specimen, namely, I-PA6 and C-PA6, I-GFRTP and C-GFRTP, and I-CFRTP and C-CFRTP, a three-point bending test was performed based on the JIS K 7017 standard. Each specimen had a dimension of 3 mm in thickness, 15 mm in width and 100 mm in length. The span length was 80 mm. Five specimens were tested for each case.

3.4 Izod impact test

The Izod impact test was performed based on the JIS K 7110 standard in order to evaluate the impact resistance of each specimen. The specimens were subjected to an impact with a hammer hitting in the normal direction to the surface (edgewise impact test). This surface had a width of 3 mm, a height of 80 mm, a depth of 10 mm and a notch of 2.54 mm deep.

3.5 SEM observation

In order to evaluate the adhesion between the fibers and the resin, the fracture surface of the cross sections after the bending test was observed with a scanning electron microscope. These observations were performed for the I-GFRTP and the C-GFRTP, and for the I-CFRTP and the C-CFRTP. After cutting the fractured specimens into pieces, they were polished using a waterproof abrasive paper and were finished with a silica polishing solution. The observation was performed under a vacuum atmosphere.

4.1 Gel time and weight of average molecular weight

Figures 3 and 4 show the relations of the molding temperature to the gel time of the in situ polymerizing ε-caprolactam and the average molecular weight. In Figure 3, it can be seen that the heating accelerated the polymerization and the time to polymerization was within 2 min at a temperature of 140°C. This fact leads to a quicker fabrication time of FRTPs with in situ polymerizing ε-caprolactam. Since the average-weight molecular weight showed the maximum at around 160°C, the polymerization of ε-caprolactam was finished within a few minutes at this temperature.

Figure 3

Relation of gel time to temperature.

Figure 4

Relation of weight average molecular weight of polyamide 6 to polymerization temperature.

4.2 Degree of crystallinity

Figure 5 shows how the degree of crystallinity was affected by the molding temperature for the I-PA6 and I-GFRTP, and for the C-PA6 and C-GFRTP. In this figure, the first heating values were the degree of crystallinity of the I-PA6 and I-GFRTP measured immediately after they were molded at temperatures of 140°C, 160°C, 180°C and 200°C, respectively, and the second heating values were measured after they were cooled and solidified following remelting of the post-polymerization. It was found that the degrees of crystallinity just after the molding (first heating) were higher than the other values.

Figure 5

Effect of molding temperature on the degree of crystallinity for the first and second heating for I-PA6 and I-GFRTP.

Generally, crystallization occurs in two stages, namely, the generation of crystal nuclei and the growth of crystal. It is known that a higher temperature leads to greater difficulty in generating crystal nuclei. At a temperature of 140°C, the crystallization showed the highest value because the crystallization that occurred in the polymerization process grew faster owing to the presence of much smaller sizes of nucleus. In contrast, the values of crystallization became smaller at higher temperatures because the crystal nuclei with a smaller size did not grow well. The values of crystallinity after the second heating were smaller than those after the first heating because the I-PA6 and the matrix of I-GFRTP were already highly polymerized and their crystal growth speeds were also already small such that they were almost the same at each molding temperature.

If the results of the I-PA6 and the I-GFRTP are compared, the degree of crystallinity seems almost the same, which suggests that it is not affected by the presence of glass fiber reinforcements.

For C-PA6 and C-GFRTP, the degree of crystallinity was a constant value of around 23% for both the first heating and the second heating. This is probably because the C-PA6 resin and the C-GFRTP matrix were already highly polymerized at the molding process whereby the molecular motion did not differ from the first and second heating processes, namely, no change in the crystal growth speed.

In Figure 6, the case of I-CFRTP showed the same tendency as that of I-GFRTP for the change in molding temperature. The crystallization of I-CFRTP (the first heating) changed from 40% to 25% as the molding temperature increased. The value at 140°C (first heating) was larger than that of C-CFRTP, but the values after the second heating were almost the same as those of C-CFRTP.

Figure 6

Effect of molding temperature on the degree of crystallinity for the first and second heating for I-CFRTP and C-CFRTP.

4.3 Unreacted monomer and water absorption

For the cases of the I-PA6, I-GFRTP, C-PA6 and C-GFRTP, Figures 7 and 8 show the percentage of the unreacted monomer and the water absorption, respectively. From these figures, it can be seen that both the contents of the unreacted monomer and the water absorption were smaller over the molding temperature range between 140°C and 160°C for both I-PA6 and I-GFRTP, and their values were almost the same as the C-PA6 and the I-GFRTP. However, both the contents of the unreacted monomer and the water absorption increased at the molding temperature of 200°C. From these results, there was an appropriate range of polymerization temperature for ε-caprolactam. That is, the matrix resin molded at temperatures of 140°C to 160°C seemed to be highly polymerized such that the contents of the unreacted monomer and water absorption had decreased. For the results of the I-PA6 and the I-GFRTP, there was no difference in the tendency for the contents of the unreacted monomer and the water absorption. This indicates that the glass fibers used as the reinforcement rarely affect the polymerization of the I-PA6 and I-GFRTP.

Figure 7

Content of unreacted monomer of I-PA6 and C-PA6 and of I-GFRTP and C-GFRTP.

Figure 8

Water absorption of I-PA6 and C-PA6 and of I-GFRTP and C-GFRTP.

The water absorption of C-PA6 and C-GFRTP was approximately the same as that of I-PA6 and I-GFRTP molded at 140°C–160°C. This indicates that I-PA6 and I-GFRTP have approximately the same water absorption as C-PA6 and C-GFRTP made from already polymerized polyamide 6.

In Figure 9, the unreacted monomer of the I-CFRTP showed the lowest content at temperatures between 140°C and 160°C, and this value was almost the same as that of C-CFRTP. Figure 10 shows that the water absorption of I-CFRTP at 140°C and 160°C was almost the same as that of C-CFRTP. The optimum molding condition for water absorption was between 140°C and 160°C, which was the same as that of the unreacted monomer.

Figure 9

Content of uncreated monomer of I-CFRTP and C-CFRTP.

Figure 10

Water absorption of I-CFRTP and C-CFRTP.

The range of optimum molding temperature, which was not correlated much with the contents of unreacted monomer and the water absorption, was between 140°C and 160°C for in situ polymerizing ε-caprolactam.

4.4 Bending strength and modulus

Figures 11 and 12 show the bending test result for the I-PA6 and the C-PA6, and for the I-GFRTP and the C-GFRTP, respectively. These figures mean that both the bending strength and the modulus of the I-PA6 and the I-GFRTP showed the highest values when the molding temperature was 140°C–160°C. This dependency of the bending strength and modulus on the molding temperature seemed to be correlated to the lowest unreacted monomer contained in the matrix.

Figure 11

Results of the three-point bending tests for I-PA6 and C-PA6.

Figure 12

Results of the three-point bending tests for I-GFRTP and C-GFRTP.

In contrast, for the C-PA6 that was prepared for the purpose of comparison against the I-PA6, both the bending strength and the modulus were approximately the same as those of I-PA6 molded at 140°C–160°C. Likewise, C-GFRTP was approximately the same as I-GFRTP molded at 140°C–160°C in terms of both the bending strength and the modulus.

In the case of carbon fabric, the bending strength and modulus of I-CFRTP showed the highest values at the temperature range between 140°C and 160°C, which were the same as those of I-GFRTP in Figures 13 and 14. However, the strength of I-CFRTP was larger than that of C-CFRTP with hot-press molding; the reason for this will be explained in the following scanning electron microscopy (SEM) result.

Figure 13

Bending strength of I-GFRTP and C-CFRTP for molding temperature.

Figure 14

Bending modulus of I-GFRTP and C-CFRTP for molding temperature.

4.5 Impact strength

The impact tests of I-PA6, I-GFRTP and I-CFRTP were performed for the specimens fabricated only at 140°C, and this temperature was the optimum one for the unreacted monomer, water absorption and also the bending properties. Table 1 lists the results of the five specimens for each of the test plates. The symbols “C” and “H” represent the complete brake accompanying a separation of specimen and the partial brake with a hinged style of specimen after the impact, respectively. In the I-PA6, the average value was 2.83 kJ/m² and all of the specimens showed a failure mode of “C.” In the impact test of I-CFRTP, the no. 3 specimen showed a failure mode of “H” and this value was the maximum value of 26.6 kJ/m² among the I-CFRTP specimens. The average value was less than half that of I-GFRTP.

4.6 SEM results

The cross-sectional views of I-GFRTP and C-GFRTP molded at various temperatures are shown in Figure 15. Much of the resin around the glass fibers in the specimen fabricated at 200°C did not remain, but the I-GFRTP fabricated at 140°C and 160°C and the C-GFRTP had a homogeneous impregnation of the matrix over the entire surface of the glass fiber reinforcement without defects such as voids. The monomer of ε-caprolactam used as the matrix of I-GFRTP was well impregnated into the fiber reinforcement because it was still in the state of the monomer with a lower viscosity, so that satisfactory impregnation could be obtained even in the case of using a long-fiber fabric that contains normally a larger amount of fiber reinforcement. Therefore, the in situ polymerizable FRTPs using a monomer of ε-caprolactam as the matrix did not need high-temperature and high-pressure molding systems such as the sheet stacking method, since they could be produced using simpler molding systems with lower energy consumption.

Figure 15

SEM observation of I-CFRTP and C-CFRTP.

From Figure 16, the interface strength of I-CFRTP at 140°C was larger than that of C-CFRTP because much of the resin remained around the fibers. In contrast, the resin around the surface of I-CFRTP at 200°C hardly remained. So the bending properties of I-CFRTP at 200°C became the lowest as shown in Figure 13.

Figure 16

SEM observation of I-CFRTP.