Investigation and analysis of glass fabric/PVC composite laminates processing parameters

1 Introduction

Due to the high strength-to-weight ratio of polymer-based composite products, different industries such as vehicle and aerospace are very interested in the use of these materials. The high strength-to-weight ratio of composites results in lighter components and reduces the energy consumption in the transportation industries [1]. Polymer-based composites are divided into two main groups, including thermoset and thermoplastic matrices. Thermoplastic matrix composites are more attractive for industries than thermosets due to their advanced properties such as higher environmental resistance, reformability at elevated temperatures, higher fracture toughness, recyclability, higher impact strength, and unlimited storage life [2], [3], [4]. However, the production and usage of thermoplastic composites are limited due to manufacturing difficulties and high production costs. Viscosity of thermoplastic polymers is higher than that of thermosets, and it makes the impregnation and wetting process very difficult [5].

The matrix/fiber impregnation and bonding quality has an important role on the integrity and mechanical properties of the composite materials and determines the load and stress distribution quality in the product [6]. For this, different studies have been carried out to improve the impregnation quality of thermoplastic composites. Ye et al. [7] investigated the impregnation of polyether ether ketone (PEEK)/carbon fiber composites and studied the effects of processing time, temperature, and pressure on the mechanical properties of the composites. Mayer et al. [8] investigated the macro and micro impregnation of carbon and glass fibers with molten nylon6.6 thermoplastic in a continuous manufacturing process. Han et al. [9] improved the impregnation quality and mechanical properties of polypropylene/carbon fiber composites by sizing the fibers using dissolved polypropylene. Ferreira et al. [10] studied the mechanical and physical properties of polypropylene matrix composites reinforced with different size glass fibers. Jonoobi et al. [11] investigated the mechanical properties and impregnation quality of cellulose nano-fibers/bio-based thermoplastics composites.

Cyclic butylene terephthalate (CBT) is a thermoplastic oligomer with a very low viscosity at elevated temperatures (0.15–0.03 Pa.s), which recently is employed in different studies to overcome the impregnation difficulties of thermoplastic composites [3], [12]. However, the brittleness of polymerized CBT matrix is a weak point, and a lot of researchers have tried to increase its toughness and ductility [13], [14]. Polyvinyl chloride (PVC) is another thermoplastic that was used with natural fibers and reinforcements such as kenef, wood, and sisal [15], [16], [17] to achieve a high-quality impregnated composite products. However, there is no report on the impregnation and production of PVC-based composites reinforced with synthetic fibers; therefore, investigation in this field seems necessary.

PVC is a commercial polymer that is widely produced and employed in different industries due to its excellent chemical, mechanical, and physical properties such as high resistance to corrosive conditions and high abrasion strength [18]. Due to the amorphous structure of PVC, it softens gradually by heating without any specific melting temperature; this feature helps PVC-based composite sheets to be formed easily into final products at elevated temperatures [19]. Also, PVC has high adhesion strength and surface wettability due to its polar functional groups [20]. Therefore, PVC can make strong bonding with the reinforcing fibers to produce high-quality composites. Despite the mentioned benefits, impregnation of the reinforcing fibers with PVC matrix is restricted significantly due to its high viscosity and low thermal stability at elevated temperatures [21].

The main target of this research is the study of PVC/fiberglass composites impregnation quality and determination of the proper processing parameters condition. For this, PVC/woven fiberglass composites were produced and the effects of most important processing parameters including processing time, temperature, and pressure on the flexural strength and modulus of the products were investigated. Impregnation and bonding quality between the fibers and matrix and failure mechanisms of the products also were studied.

2 Materials and methods

2.1 Materials

Plain weave E-glass fabrics with a surface density of 200 g/m² were used as the reinforcing phase, and commercial PVC films with a thickness of 0.2 mm, density of 1276 kg/m³, and glass transition temperature (Tg) of 73.4°C were used as the matrix. Figure 1 shows the differential scanning calorimetry (DSC) curve of the used PVC.

Figure 1:

DSC curve of the used PVC (glass transition temperature: 73.4°C, beginning of degradation is at 270°C, and pick of degradation is at 296°C).

2.2 Composite laminate preparation

The composite laminates were produced using the film stacking procedure. For this, 10 woven fiberglass layers with [0/90]₁₀ layup were alternatively placed between 11 PVC films in a 100-mm-diameter die. During the processing time, the PVC films soften and penetrate between the fibers, which results in consolidated composite laminates with a thickness of approximately 3 mm. The processing pressure is applied from the beginning of the process, and it is maintained constant throughout the processing time until cooling the samples to under Tg of the PVC. The temperature and pressure cycles of the process are shown in Figure 2.

Figure 2:

Processing pressure and temperature cycle of the production process (heating rate is 7°C/min and cooling rate is 5°C/min).

3 Design of experiments

Design of experiments was done using a five-level central composite response surface methodology with an alpha value of 2. Three parameters including processing time, pressure, and temperature are dominant parameters to improve impregnation and mechanical strength of thermoplastic composite laminates. Therefore, these three parameters were investigated in five levels using 18 different processing conditions. Table 1 shows the levels of the processing parameters.

Also, three-level full factorial design of experiments was used to accurate the optimization of processing time and temperature.

4 Characterizations

Inter-laminar strength is an important mechanical characteristic of composite laminates, which represents impregnation quality of the products. Low strength inter-laminar bonding results in delamination of the laminates without fiber fracture. Three-point bending is a common test method to evaluate inter-laminar properties and impregnation quality of composite laminates [22], [23], [24]. Therefore, this test method was used in this work with three repetitions to evaluate the products.

Three specimens for each processing condition were prepared and tested according to ASTM D790-07 standard. The tests were carried out on a computer-controlled press machine with a cross head speed of 5 mm/min. Figure 3 shows the schematic of specimen preparation and three-point bending test.

Figure 3:

Schematic of the composite laminate production process and three-point bending test.

Flexural strength and modulus of the specimens could be determined using the three-point bending test results and the linear elastic beam bending theory [24]. According to this theory, flexural strength of a beam with rectangle cross-section is calculated using Eq. 1 [25].

where σ_f is the flexural strength (MPa); F is the applied load at the failure of the specimen (N); and L, w, t are support span, width, and thickness of specimen, respectively (mm).

Equation (1) is applied to the materials with linear stress-strain behavior and failure strains <0.05.

The slope of the linear part of load-displacement curves is used to calculate the flexural modulus according to Eq. 2.

where L, w, and t are the same as for Eq. 1, and m is the slope of the linear part of load-displacement curves.

Due to non-ideal linearity of load-displacement curves, determination of the slope is a problem in the flexural modulus measuring. In this work, a quartic curve is fitted to the load-displacement data and the tangent of the fitted curve at inflection point is considered as the linear part of the load-displacement curve.

Density and void volume fraction of the specimens are also measured. In order to measure the density, the dimensions of the specimens are measured accurately using a 0.001-mm resolution micrometer and their weight is measured using a 0.01-g resolution scale. The average of three repetitions is reported as the total density of the processed material.

The measured weight of the specimens (ω_tot.) is the sum of fibers weight (ω_f) and matrix weight (ω_m), in which the fibers weights are calculated regarding the surface density of the fabrics and the matrix weight is calculated using Eq. 3.

The measured volume of the specimens (V_tot.) is as in Eq. 4.

where v_f, v_m, and v_v are fibers, matrix, and void volume of the specimens, respectively.

Density of the used fibers and matrix are known. So, the volume fraction of the void is calculated according to Eq. 5.

where ρ_f and ρ_f are density of the fibers and matrix, respectively.

The failure mechanisms including fibers fracture and inter-laminar sliding were investigated using an optical microscope, and the effects of the processing parameters on the failure mechanisms were studied. Also, the impregnation and bonding between the fibers and matrix were evaluated using a scanning electron microscopy (SEM). Finally, the effect of processing temperature on the PVC matrix structure and degradation was analyzed using Fourier transform infrared spectroscopy (FTIR) test.

5 Results and discussion

The produced samples have a thickness between 2.9 and 3.1 mm; fibers volume fraction, between 24% and 28%; and void volume fraction, between 1.2% and 35.7%. Also, the density of the samples is between 1158 and 1625 kg/m³.

The void volume fraction of the samples is relatively low and diminishes by increasing the processing temperature. However, for some of the samples, high volume fraction of void with a low density is observed that shows decomposition of the polymeric matrix.

5.1 Load-displacement curves

The results of three-point bending tests show different deflection behavior for the samples due to the variation in the impregnation and bonding strength between the fibers and matrix. For the specimens with high flexural strength, the load drops off sharply after failure, while for the low strength specimens, the load reduces slightly after the failure. The results of the load-displacement curve for some of the samples are shown in Figure 4.

Figure 4:

(A) Load-displacement curves of some samples, (B) the used three-point bending test setup.

These behaviors are because of different failure mechanisms of the samples. Fiber fracture occurs for high strength samples, while in the low strength sample, inter-laminar strength is low and the laminates slide over each other before the fibers fracture. In the following section, failure mechanisms of these samples are discussed in detail.

5.2 Flexural strength

Flexural strength and modulus of the specimens are calculated according to Eqs. 1 and 2, and the average of three repetitions is presented in Table 2. The results show a high sensitivity to the processing time at 220°C; therefore, experiments 17 and 18 were exerted to improve time effect analyzing.

Table 2:

Average flexural strength and modulus of the PVC/glass fiber composite laminates.

	Processing temperature	Processing time	Processing pressure	Density (kg/m3)	Fiber volume fraction (%)	Void volume fraction (%)	Flexural strength (MPa)	Flexural modulus (GPa)
1	1	−1	−1	1571.1	27.7	5.2	153.6	8.4
2	1	1	−1	1257.0	28.3	29.7	28.2	4.8
3	0	0	2	1602.4	26.9	1.9	136.0	9.4
4	0	0	0	1572.7	25.8	3.2	149.0	10.6
5	0	2	0	1593.1	26.6	2.3	151.1	10.8
6	−1	−1	1	1533.2	25.5	5.9	99.2	7.4
7	−2	0	0	1479.2	24.2	8.8	54.2	3.0
8	1	1	1	1251.2	28.2	30.8	33.6	4.9
9	0	0	−2	1553.5	25.5	4.4	145.8	8.3
10	1	−1	1	1598.3	26.7	2.0	147.2	9.7
11	−1	1	1	1555.9	25.0	3.6	120.4	8.7
12	2	0	0	1158.1	25.9	35.7	35.5	5.5
13	0	0	0	1582.8	26.4	3.0	138.0	10.1
14	−1	1	−1	1542.4	25.1	4.8	106.5	7.6
15	−1	−1	−1	1565.6	25.9	3.7	110.0	7.1
16	0	−2	0	1580.5	26.3	3.0	153.6	10.4
17a	1	−2	−2	1625.4	28.0	1.2	170.7	9.7
18a	1	−2	1	1616.8	28.3	2.2	189.0	10.3

1. ^aDenotes auxiliary experiments.

ANOVA is carried out to determine the effective processing parameters on the flexural strength. The results show that the main effects of time and temperature and the interaction effect of temperature-time are significant. Temperature has the maximum effect on the flexural strength.

For the processing temperatures below 220°C, increase of temperature increases the flexural strength, and the other two parameters affect the strength slightly. For samples produced at 220°C, the effect of processing time is excessive and any variation in the processing time affects the strength significantly. Figure 5 shows the effect of temperature on the flexural strength of the samples in different conditions of processing time and pressure.

Figure 5:

The effect of processing temperature on the flexural strength of the samples (N indicates the number of measurements, and range denotes results variation), the 200°C produced samples have the least dependency on the other parameters.

Due to the reduction of PVC matrix viscosity, impregnation and strength of the samples get better by increasing the processing temperature up to 220°C. However, degradation of PVC matrix at more elevated temperatures weakens the bonding between fibers and matrix, and consequently, strength diminishes extremely. Although the effect of processing time is not considerable at low temperatures, the processing time has a great effect at 220°C and increase of the processing time at 220°C reduces the strength excessively. This is because of the PVC matrix degradation at a long processing time. According to DSC results, degradation of PVC films begins at 270°C in the ambient atmospheric condition, while by keeping the sample at the temperature of 240°C, degradation occurs during 30 min, and by reducing the temperature to 220°C, the degradation occurs during 40 min. This trend shows the severe role of the processing time on the used PVC films thermal and degradation behavior. The effect of processing time on the strength of the samples differs at different temperatures. The interaction effects of processing time and temperatures on the flexural strength of the samples are shown in Figure 6.

Figure 6:

The interaction effects of time-temperature on the flexural strength of samples.

As seen in Figure 6, at low temperature (180°C), increase of processing time increases strength of the samples slightly. Penetration of softened PVC between the fibers is a time-dependent procedure, and by increasing the processing time, the impregnation gets better. The positive effect of time on the samples’ strength is reduced by elevating the temperature. At the temperature of 220°C, time has an inverse role, and the flexural strength of the samples shows a dramatic reduction by increasing the processing time up to 40 min due to PVC degradation.

According to ANOVA, the processing pressure does not have a significant effect on the strength. However, increase of the pressure at low temperatures improves the strength a little because the matrix viscosity is high and pressure helps the impregnation. Also, according to the experiments carried out at pressures below 0.5 MPa, the products do not have a proper impregnation and the inter-laminar strength is very weak. Therefore, pressure of at least 0.5 MPa is required to produce properly consolidated products.

5.3 Flexural modulus

For the structural and semi-structural applications, dimensional stability and stiffness are two important features that mostly depend on the elastic properties of products [26]. Low elastic modulus of polymeric composites is a weak point that restricts their usage. For the produced samples, the maximum flexural modulus of about 12 GPa is achieved, which is lower in comparison with metals. ANOVA for flexural modulus shows that temperature has the maximum effect, as it was seen for flexural strength. Figure 7 shows the effect of processing temperature on the flexural modulus of the samples.

Figure 7:

The effect of processing temperature on the flexural modulus of the samples (N indicates the number of measurements, and range denotes the results variation).

Like the flexural strength, increase of the processing temperature increases the flexural modulus, too. The highest flexural modulus is achieved at the processing temperature of 200°C, and above that, the modulus is approximately constant except for the degraded samples, in which a dramatic reduction happens. This trend proves that inter-laminar sliding occurs during the bending test for the samples with processing temperature below 200°C. However, for the samples produced above 200°C, the sliding does not occur and deflection and modulus of the samples are determined by the fibers and matrix properties according to the classic mechanics of laminates.

According to ANOVA results, the interaction effect of processing temperature and time on the flexural modulus is also significant, which is presented in Figure 8.

Figure 8:

The interaction effect of processing time and temperature on the flexural modulus of the samples.

Processing time has a positive role on the flexural modulus at low temperatures because of the impregnation improvement. However, by elevating the temperature, the effect of time reduces and at 220°C, processing time has an inverse effect on the modulus due to matrix degradation.

5.5 Failure mechanisms

Failure of the samples occurs in two main mechanisms including fibers fracture and inter-laminar sliding. At low-temperature-produced samples, inter-laminar sliding is the main failure mechanism, and by increasing the processing temperature, fracture of the reinforcing fibers appears. Failure mechanisms of some samples are shown in Figure 9.

Figure 9:

Optical microscope images of the samples failure mechanisms (inter-laminar sliding, fibers fracture, and matrix degradation are the main failure mechanisms).

As seen in Figure 9, inter-laminar sliding is the failure mechanism of samples produced at 160°C (Figure 9A); the weak bonding between the fibers and matrix causes delamination before the fibers fracture. By increasing the processing temperature, the impregnation quality and bonding strength between fibers and matrix get better and, as a result, fibers tolerate more loads. For samples produced at 180°C and 200°C, both inter-laminar sliding and fibers fracture occur simultaneously (Figure 9B and C). However, for the samples produced at 220°C during 10 min, fracture of the fibers is the only failure mechanism without any inter-laminar sliding (Figure 9D). But for the sample produced at 220°C during 40 min, samples fall to pieces by loading because the adhesiveness and integrity of PVC matrix are reduced due to degradation (Figure 9E and F).

5.6 FTIR results of degraded matrices

The microscopic images of the degraded samples show separation between the fibers and matrix without a proper stickiness between them; this demonstrates the loss of matrix cohesion to create a strong bonding with fibers. In order to analyze the transformations that occurred on PVC matrix, FTIR tests with wavelength of 400 to 40001cm were done on untreated PVC film and degraded PVC matrix (produced at 240°C during 30 min). The results are presented in Figure 10.

Figure 10:

FTIR spectra of untreated PVC film and degraded PVC matrix.

As seen in Figure 10, the FTIR spectra of the degraded matrix have none of the untreated PVC film spectra peaks; therefore, all of the functional groups of PVC and its additives are lost. In fact, due to PVC metamorphosis, the composite does not have a proper polymeric matrix to keep the fibers and transform the loads through them. So, the composite smashes easily by loading.

5.7 Impregnation of fiber/matrix

The variation of mechanical properties in the composites mostly depends on the impregnation quality and penetration of matrix between fibers. In order to explain the effect of parameters on the mechanical properties, SEM images of samples in different processing conditions are presented in Figure 11.

Figure 11:

SEM images of fiber/matrix impregnation of the composite laminates produced at different conditions.

As seen in Figure 11, the penetration of matrix between fibers does not occur in the sample with processing temperature of 160°C (Figure 11A); therefore, the fibers and PVC films are bonded together as separate layers. Although the PVC films are softened at 160°C to stick to the fibers and make a consolidated sheet, the viscosity is not appropriate to make a homogeneous product. As a result, the laminates slide over each other during loading due to weak bonding between them. Figure 11B shows a slight improvement in the impregnation of 180°C processed samples; however, the viscosity is yet high to wet all the fibers. Figure 11C exhibits an SEM image of 200°C processed sample, as it is seen, the PVC penetrates between the fibers properly. However, the impregnation in the condition of 220°C and 10 min is great and the matrix surrounded all the fibers completely (Figure 11D); this indicates the appropriate reduction of PVC matrix viscosity at 220°C to wet the fibers. Although 220°C seems suitable temperature to achieve a high-quality impregnation, more increase of the processing time or temperature causes dramatic reduction in the mechanical properties due to matrix decomposition. This trend proves the inverse effect of processing time (at temperatures above 220°C) on the mechanical properties of PVC matrix composites. By increasing the processing time, PVC matrix is degraded gradually and the bonding weakens. Figure 11E and F presents SEM images of the degraded samples. The followings could be seen for the degraded samples.

• The trace of fibers on the matrix is clear, which implies the low viscosity and good wettability of PVC at these temperatures. However, because of matrix decomposition, the bonding between fibers and matrix is not created appropriately.

• The chlorine gas arising from PVC decomposition is trapped between fibers and matrix interface and creates small cavities. These cavities reduce bonding surface and results in reduction of bonding strength between fibers and matrix.

6 Flexural strength optimization

Unlike tensile tests, shear stresses are created between laminates in the bending-based tests, which results in delamination and inter-laminar sliding. Therefore, flexural strength of composite laminates is mostly less than tensile strength [27]. Creation of shear stresses in three-point bending test makes it a suitable test method to evaluate inter-laminar strength and impregnation quality of composite laminates. Therefore, in the current study, the ANOVA model of flexural strength is maximized constrained in boundaries of experiments in order to find out the proper processing conditions of PVC/fiberglass composite laminates. According to the optimization results, maximum flexural strength of 192 MPa achieves in the processing temperature of 220°C, pressure of 1.6 MPa, and processing time of 10 min.

Although the optimum condition of processing pressure occurs in the interval of levels, the optimum condition of processing time occurs in the lower boundary while reduction of time improves the strength. Also, the optimum temperature seems not to be suitable because the only specimen produced above the optimum temperature is degraded. Therefore, in order to find out the optimum processing time and temperature, nine new processing conditions at three levels of temperature including 220°C, 230°C, and 240°C and three levels of processing time including 0, 5, and 10 min are considered according to full factorial design of experiments (processing pressure is assumed to be fixed at 1.6 MPa). Flexural strength of the products is measured with three repetitions, and the effect of parameters on the strength is presented in Figure 12.

Figure 12:

The effect of processing time and temperature on the flexural strength of PVC/fiberglass composites in constant processing pressure of 1.6 MPa.

As seen in Figure 12, increase of processing time until 5 min and processing temperature up to 230°C improves the product strength. By optimizing the surface fitted on the results, the maximum flexural strength of 229 MPa is achieved at processing temperature of 231°C and processing time of 5.2 min.

Although the flexural strength of these nine samples differs significantly, their flexural modulus is approximately 10 GPa. This trend demonstrates that the flexural modulus does not depend on the specimens’ strength and impregnation quality above a critical strength of the products. In order to find out the critical strength, regression analysis is done between specimens’ flexural strength and modulus.

7 Regression analysis between flexural modulus and strength

The flexural strength versus flexural modulus data of 72 valid tests (the results of three degraded samples are eliminated) are plotted in Figure 13.

Figure 13:

Flexural strength versus flexural modulus plot of PVC/fiberglass composite specimens.

As seen in Figure 13, the data can be divided into two parts. The first part is for strengths below 155 MPa, and the second part is for the strengths more than 155 MPa. With the increase of strength up to 155 MPa, the modulus also increases linearly up to 10 GPa while more increase of strengths does not affect the modulus.

This trend demonstrates that inter-laminar sliding has a dominant role on the deflections of low strength samples, while for high strength samples, strong bonding between the laminates prevents the inter-laminar sliding and the deflections are obtained just through the mechanical properties of fibers and matrix.

Therefore, the strength of 155 MPa, which is approximately for the samples produced at 200°C, acts as a transition point that shifts dominant failures mechanism from inter-laminar sliding to fibers fracture. These results are in a good agreement with the optical microscope images (Figure 7) and also SEM images (in Figure 11C, good impregnation of 200°C produced samples is seen).

8 Conclusion

PVC/fiberglass composite laminates were produced and characterized successfully through this study. The effects of three processing parameters including temperature, time, and pressure on the products quality were evaluated using mechanical tests and microscopic images. Some of the important achievements are summarized below:

• Processing temperature is the most effective parameter of PVC/fiberglass composite laminates production. Increase of the processing temperature up to 230°C improves the products quality; however, more increase of the temperature results in PVC matrix degradation and mechanical properties reduction.

• Processing time has an enhancing effect on the mechanical properties and impregnation quality at low processing temperatures; however, this enhancing effect is limited to a few minutes at high temperatures due to PVC matrix degradation.

• Processing pressure has the least effect on the products property, and a slight increase of mechanical properties by pressure increase is seen at low temperatures.

• Inter-laminar sliding is the dominant failure mechanism of low-temperature-produced samples, while fibers fracture is the failure mechanism of high-temperature-produced composite laminates.

• The maximum flexural strength of 229 MPa with a flexural modulus of 10 GPa is achieved for PVC/fiberglass composite laminates in the optimum production condition (processing temperature of 231°C, time of 5.2 min, and pressure of 1.6 MPa).

Eventually, it was concluded that PVC polymer can be used as a suitable matrix to produce high-quality thermoplastic composite laminates economically in a short time.