Manufacturing method of carbon and glass fabric composites with dispersed nanofibers using vacuum-assisted resin transfer molding

2.1 Fiber reinforcement and matrix

The carbon and glass fiber used were CF-3327 and GF-T132, respectively, with a plain weave type of fabric (Hankuk Fiber Co., Miryang, Kyungnam, Korea). Fabric weights of the carbon and glass fiber were 205 and 339 g/m², respectively. The reinforcing fibers were plain woven, and the resin fluidity was good because the path of the resin was built by the space where the warp and fill yarn intersected.

The epoxy (YD-128, Kukdo Chem. Co., Seoul, Korea) to hardener (KBH-1089, Kukdo Chem. Co., Seoul, Korea) ratio of the mixed resin was 90:100 by weight percentage. A hardener based on acid anhydride was used. Epoxy resin with an epoxy equivalent weight (EEW) of 184–190 g/eq, viscosity of 11,500–13,500 cps at 25°C, and specific gravity of 1.17 was obtained.

2.2 Carbon nanofibers

Vapor-grown carbon fiber (VGCF), a promising nanofiller, has very good mechanical properties and low production cost. Because a nanofiber has the form of a nearly straight line, without twisting and bending, CNF is suitable as a high strength material. The physical and mechanical properties of the VGCFs (Showa Denko K.K., Tokyo, Japan) used and a scanning electron microscope (SEM) image are shown in Table 1.

Table 1

Properties and SEM photograph of VGCF.

VGCF (Showa Denko, Japan)
Fiber diameter (nm)	150
Fiber length (μm)	10–20
Aspect ratio	10–500
Real density (g/cm3)	2.0
Bulk density (g/cm3)	0.04
Specific surface area (m2/g)	13
Young’s modulus (GPa)	273–760
Tensile stress (MPa)	2700–3500

3.1 Permeability coefficient of the reinforcement

For the permeability coefficient tests, a reinforcement fiber was inserted into the mold and the resin filled the voids in the reinforcing fiber. Typically, the flow physics in porous media are modeled using Darcy’s law (35), given by

where V¯ and P are the volume averaged velocity and pressure, respectively; K is the permeability of the porous medium; and μ is the fluid viscosity.

As shown in Equation 1, the resin flow velocity is influenced by the permeability, viscosity, and pressure gradient of the injection process. The technique for improving the flow velocity was to shorten the resin filling distance by installing a multiple injection gate or by reducing the resin viscosity. The resin velocity was measured at the flow tip along with the time to fill the mold in a two-dimensional radial direction. The permeability of the reinforcement was calculated by Equation 1 using these measurements. This is an effective method for determining the permeability even when the reinforcing fibers are anisotropic. The upper plate was covered with a vacuum film to observe the resin impregnation process inside the reinforcement. After laminating the reinforcing fiber on the mold, vacuum was maintained in the fiber layers using a vacuum bag. The resin reservoir was under atmospheric pressure, and a manometer was attached to the injection gate to measure the exact pressure inside the mold. A video of the changing surface of the resin was taken.

In the permeability test, the flow material used was oil (350 CS, ShinEtsu, Tokyo, Japan). The laminated five-ply fiber fabric had a dimension of 100×100 mm. Usually, the permeability coefficient can be determined by the fabric shape. The permeability coefficient is one of the dependent variables for determining the flow rate of the resin, which numerically expresses the air gap between the fibers. The average coefficient (K) is 1.66E-11 m². Generally, this coefficient has a value in the range of 10^-10–10^-13 for fabrics, and the material used in this research had a permeability coefficient value in the effective range (36).

3.2 Volume fraction of the fiber

In this research, based on the MIL-HDBK-17-1F standard, the laminating ply of the reinforced fiber was determined so that the volume content was 50% (37). The volume of the fiber, V_f, can be calculated by the following equation:

where FAW is the fiber areal weight, n is the number of ply, t is the laminating thickness, ρ_f is the density of the reinforced fiber, and k is the conversion factor. In this research, the area density and the density of the carbon fiber (CF-3327) were 0.0205 and 1.5 g/cm³, respectively. Those of the glass fiber (GF-T132) were 0.0339 and 2.6 g/cm³, respectively. Table 2 shows the number of laminating ply according to thickness.

3.3 Viscosity of the resin

The viscosity of a resin is an important factor affecting the diffusion flow rate into a reinforcing material. The flow rate is inversely proportional to viscosity. In a general RTM process for composites, because the reinforcement content is low (<20 wt.%), the flow resistance is also reduced and a full impregnation of the resin is possible. However, because impregnation is limited by higher resin viscosity, setting the injection temperature was very important in this study to ensure the dispersion of CNFs in the epoxy resin for this research. The transition of the resin viscosity was measured using a rheometer (Gemini 200 Rheometer, Bohlin Instruments Ltd., UK) for the optimization of the molding process.

After a gap of 500 μm between the top and the bottom plates of the rotator in the viscosity measurement was set, the temperature was increased and the change rate of shear stress was converted to the viscosity value. The temperature of the resin was raised to 100°C at a rate of 20°C/min, and the chemical activity was investigated. Figure 1 shows the results of the viscosity test.

Figure 1

Results of viscosity in dynamic and isothermal test.

In order to confirm the viscosity behavior during the early stage of injection, dynamic injection was repeated six times – heating and cooling between temperatures of 30°C and 100°C. Initially, the viscosity at high temperature was lower than that at low temperature, but the change rate of the viscosity over time gradually increased. Based on the results of the dynamic injection and static scanning, it was determined that an injection temperature of 60°C was the most suitable for a total process time of about 3 h.

3.4 Glass transition temperature

A thermal analyzer (TMA/SDTA840, Mettler Toledo Co., Greifensee, Switzerland) was used to measure the glass transition temperature (T_g) of the resin (YD-128/KBH-1089) used. The composite specimen for T_g measurement had a dimension of 8×8×8 mm³, and the temperature was increased at a rate of 1°C/min from 20°C to 200°C. In Figure 2, graph 1 shows the probe displacement measured by the TMA equipment and graph 2 is the differential of the displacement curve. T_g was determined to be 138.22°C from graph 2.

Figure 2

Result of thermal analysis for glass transition temperature.

4.1 CNF Dispersion process in epoxy

Figure 3 shows a schematic of the CNF dispersion process. CNFs were dispersed in an ethanol solvent by ultrasonication for 10 min. First, the ethanol and the CNFs were mixed by using an ultrasonic dispersion device and this solution was blended with the epoxy resin. The compound was again mechanically mixed for 30 min. Residual solvent was eliminated using an agitator at 80°C for 1 h. The test specimen was prepared by curing an epoxy (YD-128) with dispersed CNFs and a hardener (KBH1089). The degree of CNF cohesion when dispersed in a composite was investigated by optical microscopy, as shown in Figure 4. The dispersion of CNFs was found to be almost uniform with no obvious aggregates.

Figure 3

Schematic of ultrasonic dispersion method.

Figure 4

SEM photographs for CNF dispersion in composites.

4.2 Molding process of VARTM

The reinforced composite had anisotropic mechanical properties, depending on the kind of fibers used as reinforcement and fiber orientation. The design of composites can be varied unlike metals, and their mechanical properties can be controlled by designing their structures by changing either the fiber orientation or the lamination method or both.

The VARTM process can be divided into the preforming step, resin filling step, and hardening step. In the performing step, the cutting and lamination of the reinforcement are conducted. The hardening step includes controlling the temperature and pressure and inducing chemical reaction, which cause the resin to cure. Resin filling is the core process for impregnating the resin with fiber. This process is affected by the permeability coefficient of the reinforcement, resin viscosity, hardening property of the matrix, and the vacuum level. A flow media (80HT, Airtech International, Inc., USA) was used so that the CNFs could smoothly penetrate the reinforcing material under vacuum.

Table 3 summarizes the various composites according to the manufacturing factors of the matrix (epoxy) and fiber (carbon/glass), thickness (4–10 mm), laminating orientation (0°/45°/90°), and CNF content (0.1–0.5 wt.%). The peel ply was laminated on the top and the bottom plates, and then the flow media and reinforcement fiber were placed on the inside. First, a carbon or glass reinforcement fabric piece was laminated. Next, the flow media and mold release were placed. Last, these materials were covered with a double vacuum bag so that the laminated material was tightly sealed. After these mold preparation steps, a liquid resin was injected by vacuum pump through the entrance of the vacuum bag and past the laminated fibers.

After evaporating the excess chemicals from the resin mix, the laminated mold was kept under vacuum and put into an oven at 60°C. The mixed resin was transferred inside of the laminating mold. The flow direction was suitably designed so that the resin injected into the laminated plate was able to spread through the whole lamination along the flow media. A pressure of 0.9 atm was used at the inlet port in order to remove the resin-rich layer after the resin infusion was completed. The material was hardened in an oven at 80°C for 30 min and then at 120°C for 120 min with a full resin charge. The temperature was raised after the first 30 min to minimize internal stress. Thus, the integrated composite material was finally obtained through this series of processes. Figure 5 shows the whole curing cycle conducted in this research.

Figure 5

Curing cycle for VARTM process.

4.3 Integrity analysis of the composites

The integrity of the composite samples was evaluated by investigating the arrangement of pores and fibers. The specimens were prepared by the process described below, and optical analysis was performed.

1. cold mounting: mounting by polyester resin

2. rough polishing: using abrasive paper #400–#2400

3. micropolishing: using diamond suspension (6, 3, 1, 1/4 μm)

4. sample observation: using ×50 and ×200 magnification with an optical microscope (Nikon-EPIPHOT200, Japan)

5. calculation of the porosity: determining the number of pores and mean size

The following representative images were obtained. For the 0°/90° type of laminates with fiber weave, the perpendicular (warp) and the horizontal (weft) bundles of fibers were woven by turns and had a regular arrangement, as shown in Figure 6. Pores as small as 10 μm were found throughout the sample. Figure 6 exhibits a mean pore size of about 45 μm and a porosity of about 3.6×10^-5%. These data suggest that the integrated composite was produced with low porosity.

Figure 6

Optical scan of 0+90° specimens (×200).

As shown in Figure 7, for the laminate with a 90°/45° fiber orientation, the woven fabrics were stacked in layers, especially with a flat array with an elliptical tilt of 45°. Overall, the arrangement inside the composite samples was observed to be homogeneous. The mean size of the pores was about 50 μm, and the porosity was about 3.7×10^-5%. The manufactured composites differed in fiber array orientation according to the laminating angle. However, the quality of the manufactured composites achieved was good, exhibiting a desirable array of fibers and low porosity in each case.

Figure 7

Optical scan of 90+45° specimens (×200).