Optimizing thermoplastic pultrusion parameters for quality long fiber thermoplastic pellets in glass fiber-reinforced polypropylene

Abstract

A long fiber thermoplastic pellet (LFTP) is a pellet that contains discontinuous fibers reinforced in thermoplastic. In this experiment, LFTPs were produced using the thermoplastic pultrusion process. The thermoplastic pultrusion process is uncomplicated in terms of the machine device, and it can improve the impregnation quality of LFTP. Glass fiber was used as the reinforced fiber, and PP fiber was used as the matrix. This experiment studied the effect of molding temperature and pulling speed on impregnation. The die system contained six heating zones. In the first step, zones 3 and 4 (melting zone) were varied at 200, 210, 220, and 230°C. The melting zone that exhibited good results was selected for the next step. In the second step, zones 5 and 6 (heat reduction zone) were varied at 160°C–190°C and 150°C–180°C, respectively. After determining the optimal molding temperature, the pulling speed was studied, with speeds varied at 10, 20, 30, 40, and 50 cm/min. The impregnated quality of the LFTP was investigated through microstructure analysis. A data logger was used to record the temperature profile in the pultrusion die. Furthermore, the impregnation quality and microstructure of LFTP from thermoplastic pultrusion were compared to commercial LFTP. The results showed that un-impregnation decreased with an increase in the melting zone temperature. In the heat reduction zone, un-impregnation decreased with a decrease in temperature. For the pulling speed, un-impregnation decreased with a decrease in speed. The optimal die system had a melting zone temperature of 230°C and a heat reduction zone temperature of 160 and 150°C. The optimal pulling speed was 10 cm/min, resulting in an un-impregnation rate of 8.06%. Additionally, commercial LFTP had an un-impregnation rate of 13.73%.

Graphical Abstract

Introduction

Composite materials are currently being applied in many industries such as aerospace, automotive, civil infrastructure, and industrial parts. ¹ The composite material market is predicted to grow at a rate of 12% from 2021 to 2028, according to forecasts. ² Composite materials are formed by combining two or more materials with different properties. The matrix material bonds the reinforced fibers together. The reinforcement reinforces the composite material. Generally, matrix materials are ceramics, metal and polymer. ³ Polymer matrices are particularly popular in the composite market. ⁴ There are several different polymer matrices used in composite materials. The reinforcement can be classified with various forms such as flakes, particles, short fibers, long fibers, and continuous fibers. The properties of the composite depend on the form of reinforcement used. ⁵ Longer fiber forms generally result in higher mechanical properties, but they are more difficult to form. ⁶ There are many processes for fabricating composite materials depending on the form of reinforcement, such as filament winding, injection molding, hand lay-up, compression molding, and pultrusion. ⁷

One common process for composite fabrication is injection molding, which is popular and relatively easy. However, when short fibers are mixed with polymer for injection molding, the resulting composite has lower mechanical properties and poor fiber orientation. To improve orientation and mixing, short fiber thermoplastic pellets are prepared for injection molding and compression molding. However, short fiber thermoplastic pellets may not be sufficient for certain products. Therefore, long fiber thermoplastic pellets (LFTPs) have been developed to enhance mechanical properties and orientation. LFTPs are currently used in products fabricated by injection molding and compression molding in pellet form, with diameters of 2–4 mm and lengths of 6–25 mm. ⁶ LFTPs exhibit high strength, high modulus, and high impact strength. ⁸ The method for classifying between long fiber and short fiber is an aspect ratio. ⁹ Generally, LFTPs are fabricated through the hot melt impregnation process, which combines pultrusion and extrusion processes. In the pultrusion process, continuous fiber is pulled into an impregnation die along with molten thermoplastic from an extruder. The molten plastic impregnates the fiber, and the material is then cooled and cut to the required length.^6,10 The impregnated quality is a problem in the hot melt impregnation process because the high viscosity of the thermoplastic makes it difficult to impregnate into the fiber. Additionally, the extruder machine used in the hot melt impregnation process is expensive, resulting in high production costs.

In this work, there is an interest in fabricating LFTP using the thermoplastic pultrusion process because it can improve impregnation quality and reduce machine costs. The thermoplastic pultrusion is a process for forming continuous fiber-reinforced composites. Reinforced fiber and thermoplastic fiber are pulled through a hot pultrusion die with a constant cross-section shape. The plastic is then melted and impregnated into the fiber, followed by pulling it out of the die to form a unidirectional composite. According to literature reviews, continuous composites can be fabricated through the thermoplastic pultrusion process with an impregnation quality ranging from 80% to 93%.^11,12 The parameters of the pultrusion process include molding temperature, pulling speed, and filling ratio.^11,13,14 This research studied the effect of parameters on the impregnation of LFTP from thermoplastic pultrusion. The molding temperature was studied, including the melting zone and heat reduction zone, and pulling speeds were also studied. Glass fiber was chosen as the reinforced fiber due to its continuous fiber form, high mechanical properties, and low price. Polypropylene (PP) fiber was used as the matrix due to its low price, low density, low process temperature, and low water absorption. ¹⁵

Materials and methods

Materials

In this experiment, PP yarn (133 tex or 1200, a density of 946 kg/m³) was used as the matrix for LFTP. The melting temperature of the PP yarn was 178.8°C, as obtained from the differential scanning calorimetry (DSC) result, and a degradation temperature of 320°C was obtained from the thermal gravimetric analysis (TGA) result. Glass fiber yarn (1200 tex, a density of 2620 kg/m³) was used as the reinforced fiber in LFTP. The volume fraction of LFTP was determined using a filling ratio. ¹¹ The pellet length of commercial LFTP ranged from 10 to 12 mm. PP pellet (POLIMAXX 1100NK) was used for preparing a specimen for mechanical test.

Method

In this experiment, LFTP was prepared using thermoplastic pultrusion process as shown in Figure 1. The effect of molding temperature and pulling speed on un-impregnation was studied. The molding temperatures in the die system were varied as shown in Tables 1 and 2. The die system of zones 1 and 2 was fixed at 150 and 160°C, respectively, based on preliminary data from literature. ¹⁶ The melting zones in the die system (zones 3 and 4) were determined from DSC results and studied at temperatures of 200, 210, 220, and 230°C. The heat reduction zones (zones 5 and 6) in the die system were varied as shown in Table 2 to investigate resin impregnation before LFTP exited the die. The pulling speeds were varied at 10, 20, 30, 40, and 50 cm/min using the optimal temperature. The filling ratio was used to define the intermediate material before molding. Generally, in the thermoplastic pultrusion process, the filling ratio is designed as being more than 100% of the cross-sectional area of the die.^11,12 This allows for an excess of thermoplastic fiber beyond 100% to be impregnated or back flow phenomena at the taper zone of the die, thus improving the impregnated quality. Therefore, the intermediate material of LFTP was designed with a filling ratio of 104.73%, consisting of 91.76% PP and 12.97% glass fiber. The filling ratio was calculated based on the cross-sectional area of the intermediate materials and the die, as shown in equation (1). ¹¹ Meanwhile, the volume fraction of LFTP was at 100%, consisting of 87.03% PP and 12.97% glass fiber. Microstructure analysis was performed to evaluate the impregnated quality of LFTP. Furthermore, the impregnated quality of LFTP from the pultrusion process was compared to commercial LFTP.

$\text{Filling}\hspace{0.17em}\text{ratio}=\frac{\text{Cross}-\text{sectional}\hspace{0.17em}\text{area}\hspace{0.17em}\text{of}\hspace{0.17em}\text{materials}}{\text{Cross}-\text{sectional}\hspace{0.17em}\text{area}\hspace{0.17em}\text{of}\hspace{0.17em}\text{die}}$

(1)

Figure 1.

Schematic diagram of thermoplastic pultrusion process.

Table 1.

Melting zone of die system.

No.	Heat reduction zone temperature (°C)		Melting zone temperature (°C)		Start melting zone temperature (°C)		Preheat zone temperature (°C)
	Zone 6	Zone 5	Zone 4	Zone 3	Zone 2	Zone 1	Zone 3	Zone 2	Zone 1
1	150	160	200	200	160	150	100	100	100
2	150	160	210	210	160	150	100	100	100
3	150	160	220	220	160	150	100	100	100
4	150	160	230	230	160	150	100	100	100

Table 2.

Heat reduction zone of die system.

No.	Heat reduction zone temperature (°C)		Melting zone temperature (°C)		Start melting zone temperature (°C)		Preheat zone temperature (°C)
	Zone 6	Zone 5	Zone 4	Zone 3	Zone 2	Zone 1	Zone 3	Zone 2	Zone 1
1	150	160	230	230	160	150	100	100	100
2	160	170	230	230	160	150	100	100	100
3	170	180	230	230	160	150	100	100	100
4	180	190	230	230	160	150	100	100	100

Temperature profile measurement

The temperature profile was measured to confirm the effect of molding temperature on resin impregnation in the fiber. A Type K thermocouple was inserted into the intermediate material of LFTP, and it was pulled through a hot die along with the intermediate material until outside of the die. The temperatures of the die were recorded using a data logger from National Instruments (Model NI 9211) through the NI Signal Express program.

Microstructure analysis

The impregnation analysis of LFTP was conducted based on microstructure images. LFTP specimens were prepared using resin casting and subsequently polished using a polishing machine. LFTPs were cast with unsaturated polyester resin. Sandpapers ranging from No. 100 to 2000 and flannel fabric were used for polishing. Microstructures were investigated by using an Olympus microscope at 10 times magnification. Microstructure images were analyzed using the ImageJ program to calculate the value of impregnation. The un-impregnation values of LFTP were calculated based on the space area within the fiber bundle and the total fiber bundle area. Figure 2 shows the cross-section area of LFTP, with the black area indicating the un-impregnated region in the fiber bundle. The black region represents an area where PP resin has not impregnated, indicating a lack of impregnation by PP.

Figure 2.

Un-impregnated analysis by ImageJ program (a) Cross-section area of LFTP (b) Un-impregnated analysis of LFTP.

Mechanical properties

This investigation of mechanical properties involved conducting tensile, bending, and impact tests. Specimens for investigating the mechanical properties of LFTP were prepared using injection molding. LFTPs were mixed with PP until a glass fiber ratio of 15 wt% was achieved. The injection machine (LG model LGH50N) was set to 190°C for preparing the specimens. The universal testing machine (Hounsfield load 25 kN) was used for the tensile and bending tests. The tensile test was carried out according to ASTM D638 with a head speed of 10 mm/min, while the bending test was conducted following ASTM D790 with a head speed of 1.30 mm/min. An impact Izod machine was used for the impact test, following ASTM D256 with a hammer size of 2 J.

Results

In this experiment, the first group of samples was conducted with varying molding temperatures to determine the optimal temperature for further parameter studies. Different molding temperatures were set as shown in Table 1, with changes made only in zones 3 and 4. After selecting the best temperature, the impregnation of resin before LFTP exited the die was studied by adjusting the temperature in zones 5 and 6, as presented in Table 2. Once the best molding temperature was determined, the effect of different pulling speeds on the mechanical properties of the composites was also evaluated.

The effect of molding temperature

The effect of melting zone temperature on un-impregnation is shown in Table 3. It was observed that un-impregnation decreased with an increase in melting zone temperature, although there were slight differences. The un-impregnation percentages for melting zone temperatures of 200, 210, 220, and 230°C were 13.35%, 11.18%, 9.35%, and 8.03%, respectively. The results showed that un-impregnation decreased at high temperatures because the resin could be easily impregnated. These findings are consistent with the literature,^11–13,17 which explains that the viscosity of PP decreases with increasing temperature. Figure 3 shows the temperature profile of die system ¹² indicating that the temperature profile changes when different mold temperatures are set. Therefore, the temperature profile confirms that temperature changes occur within the die. While high molding temperatures can improve impregnation quality, it is important to avoid setting the molding temperature above 230°C, as this could cause the PP to melt and break at the taper zone before entering the constant cross-section die. The temperature profile results revealed that the temperature at the start of the melting zone increases with higher melting zone temperatures. Hence, it is recommended to maintain the melting zone within the range of 200 to 230°C. The temperature of the materials became higher than the set temperature due to friction between the die surface and the molten materials. This friction caused an increase in the viscosity of the molten polymer, leading to shear heating of the polymer. According to literature, ¹⁸ friction increases with higher viscosity of the materials and a greater pulling force. Furthermore, Kurihara ¹⁹ explained that the temperature of the molten material increases due to shear heating. Figure 4 shows microstructures of LFTP cross-sections, confirming a decrease in un-impregnation when the temperature of the melting zone was increased. More black regions appeared at lower melting zone temperatures. Therefore, the optimal temperature for the melting zone was determined to be 230°C, which aligned with the impregnation quality results.

Table 3.

The result of the effect of melting zone temperature on an un-impregnation.

No	Molding temperature (°C)	Un-impregnation (%)	Standard deviation (S.D.)
1.	200	13.35	5.92
2.	210	11.18	2.67
3.	220	9.35	4.12
4.	230	8.03	3.87

Figure 3.

Temperature profile of the material of LFTP.

Figure 4.

Cross-section of LFTP in different temperature of melting zone (a) melting zone at 200°C (b) melting zone at 210°C (c) melting zone at 220°C (d) melting zone at 230°C.

Similarly, the heat reduction zones of the die system, as shown in Table 2, were investigated using the same method as the melting temperature. The effect of the heat reduction zones on un-impregnation is shown in Table 4. It was observed that un-impregnation increased with higher molding temperatures. The un-impregnation percentages for heat reduction zones for No. 1, 2, 3, and 4 were 8.03%, 20.82%, 26.19%, and 28.80%, respectively. These results indicate that high temperatures in the heat reduction zone caused the resin to flow out of the glass fiber, adversely affecting the impregnation quality. The heat reduction zone is located near the exit of the die, which led to resin flowing out of the glass fiber at the outer edge of the die when the heat reduction zone was set at high temperatures.

Table 4.

The result of the effect of heat reduction zone on un-impregnation.

No.	Die system	Un-impregnation (%)	Standard deviation (S.D.)
1.	1	8.03	3.87
2.	2	20.82	8.56
3.	3	26.19	8.17
4.	4	28.80	5.69

Figure 5 confirmed the effect of different temperatures at the heat reduction zone. It can be observed that as the temperature of the heat reduction zone increased, there was a higher likelihood of resin flowing out of the fiber bundles. The un-impregnation was particularly high at heat reduction zone No. 4. While lower heat reduction zones demonstrated better impregnation, the temperature could not be lowered beyond heat reduction zone No. 1 due to the risk of increased viscosity and potential breakage of the LFTP. These results align with the literature, ²⁰ which explains that resin flow occurs at high temperatures and long molding times, as observed in compression molding. Consequently, a molding temperature of 230°C and heat reduction zone No. 1 were selected to investigate the effect of pulling speed.

Figure 5.

Temperature profile of heat reduction zone in die system.

Figure 6 shows the cross-section of LFTP at different temperatures of the heat reduction zone, indicating that the un-impregnated area increased with higher heat reduction zone temperatures. More black regions appeared at higher heat reduction zones. Therefore, the optimal temperature for the heat reduction zone was determined to be in the die system of No. 1. These results correspond to the impregnation quality results.

Figure 6.

Cross-section of LFTP in different temperatures of heat reduction zone (a) Die system 1 (b) Die system 2 (c) Die system 3 (d) Die system 4.

The effect of pulling speed

The optimal molding temperature was selected to study the effect of pulling speed. Pulling speeds were varied at 10, 20, 30, 40 and 50 cm/min. The effect of pulling speed on un-impregnation is shown in Table 5. It was observed that un-impregnation increased with higher pulling speeds. The un-impregnation percentages for pulling speeds of 10, 20, 30, 40, and 50 cm/min were 8.03%, 12.93%, 13.77%, 13.97%, and 16.24%, respectively. These results indicate that higher pulling speeds resulted in poorer impregnation quality, as the intermediate material was molded for a shorter duration. Consequently, the resin was unable to impregnate the glass fiber as effectively as at lower pulling speeds.

Table 5.

The result of the effect of pulling speed on un-impregnation.

No	Pulling speed (cm/min)	Un-impregnation (%)	Standard deviation (S.D.)
1.	10	8.03	3.87
2.	20	12.93	8.46
3.	30	13.77	9.71
4.	40	13.97	7.47
5.	50	16.24	9.90

Figure 7 confirms the impact of changing the temperature in the die. It can be observed that higher pulling speeds in the temperature profile result in shorter material residence time in the die. Moreover, the temperature in the temperature profile decreases at a slower rate with higher pulling speeds, indicated by the difference in slope on the graph. These results align with the literature,^11,12,17,21 which explain that increasing the pulling speed leads to higher un-impregnation due to insufficient time for the resin to impregnate the fiber.

Figure 7.

Temperature profile of pulling speed in die system.

Figure 8 shows the cross-section of LFTP at different pulling speed, showing an increase in the un-impregnated area with higher pulling speeds. More black regions appear at higher pulling speeds. Consequently, the optimal pulling speed was determined to be 10 cm/min, which aligns with the impregnation quality results.

Figure 8.

Cross-section of LFTP in different pulling speed (a) Pulling speed at 10 cm/min (b) Pulling speed at 20 cm/min (c) Pulling speed at 30 cm/min (d) Pulling speed at 40 cm/min (e) Pulling speed at 50 cm/min.

Therefore, the optimal conditions for achieving the best impregnation quality of LFTP are a melting zone temperature of 230°C, heat reduction zone in the die system of No. 1, and a pulling speed of 10 cm/min. LFTP produced through the pultrusion process was cut to a length of 10–12 mm. Figure 9 shows a comparison between LFTP from the pultrusion process and commercial LFTP. Moreover, the impregnation quality of LFTP from the pultrusion process was compared to commercial LFTP.

Figure 9.

(a) LFTP and (b) commercial LFTP.

The comparison of impregnation quality between LFTP and commercial LFTP is shown in Table 6. It was found that LFTP produced through the pultrusion process exhibited better impregnation quality compared to commercial LFTP, with un-impregnation percentages of 8.03% and 13.73% respectively. There is a possibility that the pultrusion process can improve impregnation quality by increasing the filling ratio. Additionally, the impregnation quality was confirmed through microstructure analysis, as shown in Figure 10. The microstructure results indicated that the black region in the commercial LFTP was greater than that in LFTP from the pultrusion process. Therefore, it is possible that LFTP produced through the pultrusion process possesses higher mechanical properties compared to commercial LFTP. It should be noted that impregnation quality directly affects mechanical properties.^11,17,20

Table 6.

The result of comparison between LFTP and commercial LFTP.

No.	Materials	Un-impregnation (%)	Standard deviation (S.D.)
1.	LFTP	8.03	3.87
2.	Commercial LFTP	13.73	2.70

Figure 10.

Cross-section between LFTP from pultrusion and commercial LFTP.

Mechanical properties of LFTP

Specimens of PP, LFTP, and commercial LFTP were prepared using the same process and the same ratio of glass fiber. The mechanical properties, including tensile strength, tensile modulus, flexural strength, flexural modulus, and impact strength, are presented in Table 7. These results indicate that the tensile strength of LFTP is slightly higher than that of commercial LFTP, while the tensile modulus has clearly increased. The flexural strength of LFTP is slightly lower than that of commercial LFTP, whereas the flexural modulus of LFTP has significantly increased. The impact strengths between LFTP and the commercial LFTP showed no significant differences. Based on these results, it is possible to attribute the increase in tensile modulus and flexural modulus to improvements in impregnation quality. Tensile modulus had increased by up to 35%, while flexural modulus had increased by up to 21%. Therefore, the mechanical properties of LFTP increased with a decrease in un-impregnation, aligning with the findings in the literature.^11,12,22

Table 7.

Mechanical properties of LFTP and commercial LFTP.

Materials	Tensile strength (MPa)	Tensile modulus (MPa)	Flexural strength (MPa)	Flexural modulus (MPa)	Impact strength (J/m)
PP virgin	23.37	1262.86	32.34	864.86	19.27
LFTP	38.64	2735.71	47.27	1885.71	56.47
Commercial LFTP	37.01	1765.83	48	1500.42	56.85

Conclusions

In this study, thermoplastic pultrusion was conducted to prepare LFTP with the aim of enhancing its impregnated quality before its intended use. The impregnated quality significantly affects the mechanical properties of the material. As a result, optimization of parameters was carried out. The molding temperature and pulling speed were identified as important parameters in the pultrusion process. The results revealed a decrease in un-impregnation with an increase in the molding temperature (melting zone). Conversely, the heat reduction zone temperature before exiting the die led to an increase in un-impregnation. The effect of pulling speed, un-impregnation increased with an increase of pulling speed. The comparison between LFTP and commercial LFTP indicated that LFTP produced through the pultrusion process exhibited superior impregnation quality compared to commercial LFTP. The mechanical properties comparison between LFTP and commercial LFTP also showed a significant increase in tensile modulus and flexural modulus of LFTP from pultrusion process. Tensile modulus had increased by up to 35%, while flexural modulus had increased by up to 21%. While tensile strength, flexural strength, and impact strength may not increase significantly, when considering the cost of the machinery, the thermoplastic pultrusion process becomes a viable alternative for producing LFTP.

Future work on LFTP will involve surface treatment to improve the interface between the matrix and reinforcement of LFTP before its use in an injection process.