Effect of Flax/PP pellets’ morphology on the mechanical properties of injection-molded parts

Abstract

Flax and polypropylene (PP) pultrudates of 4.76 mm in diameter were produced using a multi-die pultrusion system. The flax content was 50 vol.%. Pultrusion conditions were varied to produce four porosity content between 4% and 22%. The pultrudates were pelletized into lengths of 6 and 15 mm, and injection-molded with pure PP to reduce the part flax content to 25 wt.%. Results showed that well consolidated pultrudates, having porosities up to 8%, were more resilient to the pelletizing process such that the pellets remained structurally intact after the process. These pellets went smoothly through the hopper into the injection molding machine screw. Pellets with void content higher than 8% lost integrity during pelletizing leading to uniform mixing. Exposed fibers segregated from the polymer in the hopper leading to ununiform mixing. The highest mechanical properties improvement compared to pure PP was using 15 mm pellets that had a prorosity of 8%. The tensile modulus doubled at 4652 ± 113 MPa. The impact strength increased by almost five times at 10.5 ± 0.7 $k J / m^{2}$ . 15 mm pellets provided 1.7 times the impact strength of the 6 mm pellets. These improvement are attributed to low flax fiber thermal degradation and improved fiber dispersion.

Keywords

Pultrusion flax thermoplastic long fiber thermoplastics injection

Introduction

Inorganic fibers and fillers such as glass fibers, talc and mica are widely used to reinforce the thermoplastic polymers in the automotive industry.¹ Yet, the production and disposal of such materials have an adverse impact on the environment. Due to rising environmental awareness, new environmental laws and regulations were introduced to help reduce the inorganic waste problem.² These new environmental laws and regulations are putting great pressure on the automotive industry to replace some of these inorganic fillers with biodegradable alternatives. In fact, the U.S Department of Energy’s Vehicle Technologies Office (VTO) introduced a multidimensional program to promote automotive Lightweight Materials (LM) to ensure energy security through renewables. VTO argued that the replacement of glass fibers with natural fibers will allow lighter parts to be used while increasing the percentage of renewables within the vehicle.³ It is estimated that a 10% weight reduction could reduce fuel consumption by 6–8%⁴ since lighter objects need less energy to get transported compared to heavier ones. Therefore, lighter materials provide an opportunity to improve vehicle fuel economy and mitigate greenhouse gas (GHG) emissions. In response, many auto-makers are increasing the use of natural fiber-reinforced composites. For example, Audi uses natural fiber-reinforced composites to produce boot liners, hat racks, seat backs, back door panels and spare tire-linings for their new cars. Like the automotive industry, many other industries such as medical and pharmaceutical, protective clothing, food packaging and hygiene clothing are finding more ways to utilize natural fiber reinforced composites for both structural and non-structural applications.⁵ For example, Bcomp ltd. produces winter and summer sports equipment using only natural fiber-reinforced composites. GAF materials corp. uses natural fiber-reinforced PP to make composite deckings.^5,6

Flax fibers are becoming more popular in industrial applications due to their low environmental impact and light weight. In comparison to other natural fibers, flax fibers can be processed at higher temperatures due to its high cellulose content.⁷ Natural fiber thermal degradation normally occurs when cellulose is exposed to a temperature range above 200°C. Lignin, on the other hand, starts to degrade when the temperature approaches 160°C. Flax fibers are made of low lignin content (2%) and comparatively high cellulose content (64%).⁸ This makes flax more resistive to thermal degradation compared to other natural fibers. In addition, flax is also preferred due to its higher tensile properties, which can be attributed to the high cellulose content as well. Flax fibers have the highest Young’s modulus values amongst natural fibers, comparable to e-glass fibers.⁹ Therefore, flax-reinforced composites are a great choice for sustainable industrial applications.

There is normally a very limited interaction between hydrophilic fibers such as flax fibers and hydrophobic matrices such as PP resulting in a reduced interfacial strength as well as increased moisture absorption. Different kinds of treatment can be used to improve the interaction including physical and chemical treatments.^10,11 Of all chemical methods to improve the interaction in natural fiber composite, maleic anhydride polypropylene (MAPP) is recognized to be the most popular treatment. It has been proven to yield twice the strength as achieved when used with silane treatment.¹²

Thermoplastic polymers have a very high melt viscosity compared to thermoset polymers. This high viscosity makes it very difficult to impregnate fibers and mold the final part in a one-step process.¹³ Thus, an intermediate form, in the form of bulk charge or pre-impregnated pellets, is introduced prior to the final part production which is normally done using compression or injection molding process. Fibre-reinforced thermoplastic composite materials are often categorized based on their fiber reinforcement lengths. For instance, long fibre-reinforced thermoplastics (LFT) are composite materials consisting of thermoplastic polymer and discontinuous reinforcement fibres whose length-to-diameter aspect ratio is larger than the critical aspect ratio above which fiber rupture is favored over fiber/matrix debonding.¹³ LFT offers superior mechanical properties compared to short fibre-reinforced thermoplastics (SFT). LFT production has experienced fast growth in the last two decades. A market survey published in January 2017 projected that the LFT global market value will grow from USD 3.28 billion to USD 5.55 billion by 2022.¹³ Several thermoplastic polymers were utilized as LFT polymers including PP,^14,15 Polyamide 9 (PA6),¹⁶Polylactic acid (PLA),¹⁷ Polymethylmethacrylate (PMMA),¹⁸ and high density polyethylene (HDPE).¹⁹

The desire to reduce processing steps and shipping costs for LFT intermediate form resulted in the implementation of Direct LFT (D-LFT) in the 90’s.^13,20D-LFT works by combining the injection or compression molding station with an inline-compounding of fibers and polymers.^21,22 The polymer is fed into an extruder, often a twin screw extruder, where it is melted and mixed with fillers, pigments and additives in order to achieve a homogeneous molten mixture. It is then mixed with continuous fiber reinforcement strands. Glass fibers are typically used as reinforcements for such applications. This material is then extruded and cut into a charge, and then transferred to a cold mold where it is compressed into the desired size of the final component.²³ The polymer only passes through a single melt history. This minimizes thermal degradation and improves mechanical properties.²⁴ Additionally, this process can be used to produce both SFT and LFT.²⁵ Another LFT manufacturing technique is referred to as the direct strand deposition.²⁶ The direct strand deposition was developed to produce LFT charges that are placed into the mold for compression molding but using a single-screw extruder. However, this process completely depends on the single-screw plasticisation for both fiber dispersion and impregnation. Ren et al.²⁷ faced impregnation challenges when they directly mixed the dry fibers with PP in a single screw-extruder. They tested the effect of changing the fiber weight fraction from 10 to 40% on impregnation and dispersion levels. They observed that fiber length and dispersion decreased as fiber content increased. Fiber-to-fiber interactions were thought to increase with adding fiber content. Low impregnation was also observed at higher fiber content, and it was recommended to reduce the fiber weight fraction.

The adoption of the D-LFT process is costly. It was estimated that the adoption cost can be offset once the annual production volume is increased to 30,000 units and above.²⁸ In addition, the D-LFT process can be complicated to implement since it forces the part manufacturer to handle purchasing the dry fibers and the polymer at the same time. Offering ready-to-use LFT pellets gives part manufacturers the flexibility to choose the process that they wish to implement. It simply allows them to place the LFT pellets in the hopper for injection or bulk molding compound.

The second intermediate form, pellets, can be used to make SFT and LFT. SFT pellets are usually produced by extrusion compounding using short chopped fibers that are introduced into the extruder together along with the polymer. Commonly, 4 mm-long chopped fibers are used, and the fiber length is subsequently shortened to less than 0.5 mm in the pellet while passing through the extrusion screw.²⁹ Thus, due to the limitation of short fibers, SFT pellets add inadequate reinforcing capabilities to the final composite material. On the other hand, using LFT pellets was proven to improve impact/stiffness as well as creep, fatigue and high temperature properties.³⁰ LFT pellets are often produced using the thermoplastic composites (TPC) pultrusion process. During the TPC pultrusion process, yarn precursors which contain the polymer and reinforcement fibers are wound into bobbins that are placed in a creel. The polymer and reinforcement fibers are then pulled together through a number of dies that are heated to liquefy the polymer fibers. In order to fully impregnate the reinforcement fibers, the yarns that enter the dies contain an excess of polymer fibers. This excess polymer, consequently, creates a backflow which helps raise the impregnation pressure.^31–34 This technique does not damage the reinforcing fibers and the fibers typically remain quasi-continuous after they have been processed. The pultrudates are then set to cool off and are cut into pellets of desired length. The LFT fiber length is determined by the cut length of the pellets.

Vaidya et al.³⁵ produced 12 mm Glass/PP LFT pellets using a TPC pultrusion process. These pellets were blended with inorganic pigment-colored compound pellets in order to study the processing and performance of inorganic-pigmented colored LFT in comparison with unpigmented LFT. Their work proved that LFT can be colored with inorganic pigments and still retain good mechanical properties. Batrus et al.³⁶ designed and fabricated E-glass/PP LFT bus seat using extrusion-compression molding. The LFT pellets were produced using a TPC pultrusion process. The LFT seat design acheived 43% weight savings and 18% total production cost savings over typical seats that contain a circumferential steel frame. Yet, these two research papers did not specify the pellets’ void contents nor analyze the impact of LFT pellets’ void contents on the mechanical properties of the final injected part. The TPC pultrusion conditions such as pulling speeds, die geometry and set temperatures were not reported. Furthermore, inorganic reinforcement fibers were used in both studies.

Hartness et al.³⁷ produced Glass/PP LFT pellets using a TPC pultrusion process. The glass contents in the LFT pellets were varied between 10 and 70% by weight. It was observed that the TPC pultrusion process led to a high level of impregnation although the LFT pellets’ void contents were not specified in their research. Several other researchers have implemented TPC pultrusion to produce LFT pellets as well. It was observed that the fibers attained high levels of impregnation with PP polymer using TPC pultrusion.^38–40 Additionally, LFT pellets can also be produced by similar processes such as wire coating,⁴¹ solid state polymerization⁴² and commingling.⁴³

Achieving low void contents in the pultrudates at relatively high pulling speeds using TPC pultrusion can be challenging as there seems to be a trade-off between increasing the pulling speed and decreasing the void contents due to the high melt viscosities of the thermoplastic polymers.⁴⁴ In fact, Lapointe et al.⁴⁴ observed a significant increase in void contents in C/PEEK pultrudates when the TPC pultrusion speed was increased from 50 to 100 mm/min. The void contents increased by over 85%. Wiedmer et al.⁴⁵ pultruded Carbon/Polymide-12 (C/PA-12) hybrid yarn using different pulling speeds. The void content tripled when the pulling speed increased from 100 to 400 mm/minute. Ghaedsharaf et al.⁴⁶ pultruded Carbon/Polyetherimide (C/PEI) biaxial braid which was made of commingled yarns. The void content significantly increased when the pulling speed was raised from 50 to 150 mm/minute. They hinted that the reason for the void increase was possibly due to the polymer flow time which became too short to fully penetrate the carbon fiber bundles when the pulling speed was raised to 150 mm/minute. Thus, all these findings indicated that increasing pulling speed, while keeping the same pultrusion tooling geometry, could lead to a significant increase in the void content.

There are very few published papers that studied the effect of intermediate forms on the performance of the end part of the Flax/PP composites.^47–49 Barkoula et al.⁴⁹ studied Flax/PP SFT and LFT pellets to be used for injection molding. They compared the reinforcing capabilities for both the SFT and LFT pellets using impact and tensile tests. Producing the Flax/PP pellets was done by an extruder for SFT pellets and by TPC pultrusion for the LFT pellets. They observed that the LFT pellets resulted in a significantly higher impact strength in the injection-molded parts. Yet, Barkoula et al.⁴⁹ did not discuss the influence of the LFT pellets’ void contents on the manufacturability of injection molded parts.

Having a low void content requirement in LFT pellets may restrict the pulling speed of TPC pultrusion which may lead to limiting the use of LFT pellets in industrial applications. Thus, it is crucial to quantify the LFT pellets’ void content that does not jeopardize the mechanical properties of the injected final parts. On the other hand, low void contents in the pultrudates may result in LFT pellets being structurally intact. In a previous study, Flax/PP LFT pellets having low void contents using TPC pultrusion were produced.¹⁵ The low-void LFT pellets were able to remain structurally intact during the pelletizing process and experienced a smooth injection molding process. However, the investigation was limited to well impregnated LFT pellets. To better understand the influence of the LFT pellets’ void contents in injected molded parts, we produced pultrudates using four different void contents. These pultrudates were pelletized into two different lengths, 6 mm and 15 mm. The LFT pellets were then added, with pure PP pellets, into a master mixture before being added to the injection molding hopper to produce injection molded parts. Finally, the injected molded parts were characterized by tensile and Charpy impact tests to evaluate the mechanical strength. In the injection molded specimens, the fiber/polymer impregnation was characterized using scanning electron microscopy (SEM).

Experimental

Materials

Pure PP (Guangzhou Chemical Fiber Co.) in the fiber form (finess of 100 tex) and in the pellet form were used for the pultrusion and injection molding. The melting point of the PP was 166.5 $℃$ as obtained by DSC at heating rate of 10 $℃ / m i n$ . Flax yarns (Flax Low Twist Roving, Safilin) having a size of 200 tex and a solid density of 1.53 g/ $c m^{2}$ were used.

Flax thermal characterization

Thermogravimetric analyses (TGA) of flax fibers were conducted (Q50, TA Instruments) with a temperature ramp-up of 2°C/min until 300°C under air atmosphere (40 mL/minute). However, during the TPC experiments, the total residence time at the processing temperature of 220°C is dependent of pultrusion speed. At low speed (50 mm/minute), flax fibers spent 3.2 min inside the heating dies. While at the 250, 500 and 1000 mm/minute pultrusion experiments, flax fibers spent 0.64, 0.32 and 0.16 min respectively. To better quantify the flax degradation during the TPC pultrusion at different pultrusion speeds, TGA of flax fibers was conducted by reproducing the thermal conditions during the pultrusion experiments. Three flax samples of approximately 1g were dried for 48 h at 60°C before the TGA tests. All TGA experiments featured 40 mL/minute air flow. First, the flax fibers were exposed to 100°C to reproduce the preheating stage, followed by a thermal ramp at a 20 °C/minute to 220°C to reach the pultrusion dies’ temperature. Finally, the flax sample was exposed to an isothermal environment for 3.2 min, the same duration as the travel time in the pultrusion dies at 50 mm/min.

Pultrusion apparatus

Figure 1(a) presents the schematic for the pultrusion system used in this study. The pultrusion apparatus contained a creel, a guide, a preheater, four heating dies, a cooling die, and a pulling system. The creel allowed for individual tension adjustment for each of the bobbins. The tension was set at 1 N per bobbin holding the reinforcement fibers. No tension was applied to the bobbins that were holding the polymer fibers. The preheater was a metallic tube, 300 mm-long and 16.2 mm in diameter, wrapped with a strip heater. Figure 1(b) and (c) present detailed dimensional description of the system. The final diameter decreases from die 1 to 4. All four heating dies had a 31.75 mm conical cavity tapered at $5 °$ relative to the horizontal axis followed by a straight 6.35 mm (0.25 inch) cylindrical cavity. The diameter of each die’s cylindrical exit was 5.13, 5.00, 4.90 and 4.76 mm respectively. The white triangles in Figure 1 denote the location of thermocouples in the heating dies. All heated dies (Die 1 to 4) were heated to 220°C. Additionally, the red circles represent the location of the heating cartridges whereas the blue circles represent the location of the air-cooling channels. A thin-walled tube was protruding from the last heated pultrusion die toward the cooling die. The tube length exposed to the air was 27.6 mm. The tube was inserted into a 3 mm-deep hole at the cooling die’s inlet. The thickness of the thin wall was 1.59 mm in order to reduce the thermal interference between the fourth heating die and the cooling die. The cooling die was set to 60°C. The pultrusion speed was controlled by a puller throughout the process. The flax fiber volume content was 50% by using 68 flax yarns. The speeds and filling ratios ( $R_{f}$ ) were varied to produce the LFT pellets at various void contents. The $R_{f}$ is calculated as:

R_{f} = \frac{A_{i n}}{A_{D i e}}

(1)

Where

A_{i n}

is the material area entering the die and

A_{D i e}

is the die exit area. The nominal

R_{f}

in the first die uses the flax and PP material areas pulled from the creel. These areas are obtained by dividing the flax and PP yarn finenesses (in Tex, or g/1000m) by their corresponding densities (in g/cc). The other dies nominal

R_{f}

are calculated using the same flax area as in the first die. For the PP area, in case of under filling (

R_{f} < 100 %

), all the PP is assumed to follow the pultrudate. In case of over filling (

R_{f} > 100 %

), the PP area was considered to be the non-flax area from the previous die exit area. Table 1 shows the used filling ratios, pultrusion speeds and set temperatures to produce LFT pellets having target void contents of 4%, 8%, 15% and 22%. To obtain void contents of 4% and 8% at pultrusion speeds of 50 mm/min and 250 mm/min, all four dies had

R_{f} > 100 %

by using 99 PP yarns. To obtain void contents of 15% and 22% at higher speeds of 500 mm/min and 1000 mm/min, only 74 PP yarns were mounted on the creel. This led to the first three dies having a

R_{f} < 100 %

. Only the last die was in an overfilling situation at

R_{f} > 105 %

Table 1.

Filling ratios, pultrusion speeds and set temperatures used to produce pultrudates with different void contents.

Target void content (%)	Pultrusion speed (mm/min)	Die filling ratios
Target void content (%)	Pultrusion speed (mm/min)	Die 1 (%)	Die 2 (%)	Die 3 (%)	Die 4 (%)
4	50	106	105	104	106
8	250	106	105	104	106
15	500	90	95	99	105
22	1000	90	95	99	105

Morphological characterization

The pultruded rods impregnation morphology was characterized by a microscopic observation of the cross-section. An injected coupon was polished and then analyzed using a digital microscope; (Keyence, VHX 7000). Full ring was selected for lighting type. Photos of 150 and 300 X magnification were taken of the injected coupon. Void content was measured according to ASTM2734-09 Method C. The fracture surfaces of specimens were examined in a SEM (TM300, Hitachi).

Injection molding

After production, Flax/PP pultrudates were cut into 6 and 15 mm long pellets using a manual paper trimmer (Model #118, Dahle). The LFT pellets were oven-dried for 24 h at 60 $℃$ right before the injection molding process. The Flax/PP LFT pellets were then blended with pure PP pellets in order to lower the fiber weight fraction. The mixture contained 60 wt. % PP pellets and 40 wt. % PP flax PP LFT pellets. According to this mix, the flax weight content in the injected coupons was 25%. The mixture was then added directly to the hopper of the injection molding machine (SE50S, Sumitomo).

Table 2 displays the process set points for compounding Flax/PP in the injection molding machine. The maximum temperature in the injection molding process was kept around 200°C in order to minimize the degradation of the flax fibers. The injection velocity used for testing was 65 mm/s. The naming format follows the order: void content-LFT length. Table 3 displays the void contents, set points, and LFT pellets’ lengths for the injection molding experiments. Each injection was repeated three times. One coupon was randomly selected from each injection for mechanical testing.

Table 2.

Injection molding machine process set points for compounding Flax/polypropylene.

Zone	1	2	3	4	Cooling
Temperature	200°C	200°C	200°C	200°C	25°C
Inject. velocity	65 mm/s
Cooling time	30 s

Table 3.

Void contents set points and LFT pellets’ lengths for the injection molding experiments.

Experiment	Void content (%)	LFT pellet’s length (mm)
4%–6 mm	4	6
4%–15 mm	4	15
8%–6 mm	8	6
8%–15 mm	8	15
15%–6 mm	15	6
15%–15 mm	15	15
22%–6 mm	22	6
22%–15 mm	22	15

Mechanical Characterization

Tensile tests were performed according to ASTM D638-44 in a universal testing machine (Model, Brand). The size of the tensile specimen was 165 x 13 x 19 mm according to ASTM D638. The Charpy impact test specimens were tested according to ASTM D6100.

Results and discussions

Thermogravimetric analysis of the flax fiber

Figure 2 shows a typical TGA plot of flax fibres during a temperature ramp-up at 2°C/min under air atmosphere. This figure shows the flax mass loss and mass rate of change with respect to temperature. As the temperature rises, and before reaching 200°C, the flax mass decreases slightly due to water evaporation. Beyond 200°C, the flax weight continues to decrease at a much higher rate. Results from this test are consistent with previous TGA of flax fibre found in literature ^50,51: as thermal degradation is triggered above 200°C, mass loss starts to increase. This finding implies that the thermal degradation during the pultrusion experiments happens inside the heating die at which the flax is exposed to a temperature above 200°C. To verify if substantial degradation could occur in the time span of manufacturing, TGA experiments were conducted. The TGA experiments resembled the thermal conditions for all pultrusion experiments. Figure 3(a) shows the temperature evolution and mass loss monitored using flax fibre weight percentage. Three samples were used to reproduce flax thermal history in the pultrusion heating die. In the first 30 min, the temperature was set at 100 $℃$ in order to remove mass associated to the humidity content. This drying was followed by a thermal ramp-up of 20 °C/minute to 220°C, and finally an isothermal environment for 3.2 min. Figure 3(b) shows the last 3.2 min of the TGA experiment. This time interval corresponds to the time that the flax fiber spent inside the heating dies during pultrusion at 50 mm/minute. Table 4 and Figure 3(b) show the weight percentage change that occurred after exposing flax fiber to 220°C for a certain number of minutes. Pultrusion experiments that were done at low speeds resulted in a larger mass loss compared to the other higher speeds experiments. The LFT pellets produced at 50, 250, 500 and 1000 mm/minute experienced a weight loss of 0.42% $\pm$ 0.029%, 0.093% ± 0.006%, 0.036% ± 0.002% and 0.016% ± 0.001%. The LFT pellets produced at the lowest pulling speed were significantly more degraded than the other speeds. The LFT pellets produced at 250, 500, 1000 mm/minute were only 22%, 9% and 4% degraded compared to the LFT pellets produced at 50 mm/minute. This calculation was done by dividing the resulting weight loss of the pellets produced at each pulling speed by weight loss of the pellets produced at 50 mm/minute. This finding indicates that flax fibers in the LFT pellets that had the lowest void content were more degraded compared to the other LFT pellets.

Table 4.

Weight percentage change that occurred after exposing flax fiber to 220°C.

Pult. Speed	50 mm/min	250 mm/min	500 mm/min	1000 mm/m
Exposure time (min)	3.2	0.64	0.32	0.16
Weight change	−0.420 ± 0.029%	−0.093 ± 0.006%	−0.036 ± 0.002%	−0.016 ± 0.001%

Pultrudate consolidation

The pultrusion of Flax/PP experiments were conducted without any fiber breakage event. Figure 4 shows cross sections with respect to the pultrusion process speed. Figure 4(a) and (b) show cross-section images of LFT pellets produced at 50 and 250 mm/minute respectively. The cross-section images show round shapes with very few voids. The edges do not look damaged and the polymer had fully melted during the pultrusion process. Additionally, the two top insets show that the polymer fibers located at the center of the rod fully melted during the pultrusion process and impregnated the flax fiber even at the centre, which is the last place in the pultrudate to melt.

Figure 4(c) shows a cross-section image of LFT pellets produced at 500 mm/minute. The cross-section image shows less of a round shape with some voids. Additionally, the inset shows that the polymer fibers located at the center of the rod did not fully melt during the pultrusion process. However, the middle inset shows one flax fiber inside an area which seems to be fully impregnated near the outer surface. This finding was expected since the outer surface is heated first during pultrusion. Figure 4(d) shows a cross-section image of LFT pellets produced at 1000 mm/minute. The cross-section image shows an arbitrary shape with a significant amount of void. Additionally, the polymer did not fully melt during the pultrusion process. The arbitrary shape is attributed to the lack of flax fiber wetting by the PP polymer. Poor load transfer between the two phases resulted in a deconsolidation state and difficulty to retain the circular shape. Moreover, similarly to the LFT pellets produced at 500 mm/minute, the middle-enlarged view close to the rod edge also shows a flax fiber inside an area where there seems to be full impregnation near the outer surface. Obviously, increasing the pulling speed results in a reduction of the exposure duration at elevated temperatures. The cross-section images in Figure 4 clearly indicate that this short exposure duration reduces the time that the polymer is in a melted stated and able to flow. For 500 mm/min and 1000 mm/min, that exposure was not long enough to melt the polymer at the center of the pultrudate.

Table 5 shows pultruded rods’ measured porosities. The void content characterization indicated that 50 mm/min led to a 4% void content, 250 mm/min to an 8% void content, 500 mm/min to a 15% void, and 1000 mm/min led to the highest void content of 22%. As expected, higher pultrusion speed led to low impregnation quality. This is attributed to two factors. First, as seen in the microscopic images at high speed, the temperature within the pultrudate was not reached leading to un-melted polymer regions. Second, at high speed, the time at which the polymer was under pressure in the conical regions of the dies was short, which led to a shorter impregnation time. LFT pellets produced at 500 mm/minute and above had un-melted PP fibers since the processing temperature did not reach the core of the pultrudates at the heating dies.

Table 5.

Pultruded measured porosities.

Target void content (%)	Pultrusion speed (mm/min)	Measured void content (%)
4	50	4.3 ± 1.0
8	250	8.4 ± 1.2
15	500	15.1 ± 2.5
22	1000	22.4 ± 2.8

Pelletizing quality

These pultrudates were then cut into two different lengths; 6 and 15 mm. Figure 5 shows the appearance of LFT pellets cut at 6 and 15 mm for the four void contents, i.e. 4%, 8%, 15% and 22%. The top four images represent the 6 mm LFT pellets and the lower four images represent the 15 mm LFT pellets. Figure 5(a) and (b) represent the LFT pellets with the lowest void content which was measured to be 4.3 $\pm$ 1.0%. The LFT pellets with the two lengths look structurally intact. The pelletizing process did not seem to significantly damage the LFT pellets. This is attributed to the impregnation quality of the rod indicated by low void content. Figure 5(c) and (d) represent the LFT pellets which had a slightly higher void content, which was measured to be 8.4 $\pm$ 1.2%. Yet, the LFT pellets still look structurally intact. The increase in void content did not seem to cause any damage to the LFT pellets. Figure 5 (e) and (f) present the LFT pellets which had a higher void content and was measured to be 15.1 $\pm$ 2.5%. Here the LFT pellets seem to be damaged from the pelletizing process. In fact, some of the 6 mm LFT pellets have lost their cylindrical shape. Others cracked along their length and opened through their center. The 15 mm length LFT pellets at the same 15% void content have also lost their cylindrical shape. Very few were cracked and opened. This indicates that the 6 mm LFT pellets might be more prone to damage than the 15 mm LFT pellets at such void content. Figure 5(g) and (h) represent the LFT pellets which had the highest void content, which was measured to be 22.4 $\pm$ 2.8%. Here most LFT pellets appeared damaged from the pelletizing process, for both LFT lengths. The damage is again a rupture along the length of the pellet, opening it to expose its center. Moreover, un-melted PP fibers at the center of the cracked and opened pellets were observed. This confirms that high pultrusion speed prevented the core of the pultrudate to reach melting temperature. The damage which occured during the pelletizing process is attributed to the high void content that reduces the transverse strength of the pultrudate. This is combined with the fact that un-melted fibers cannot provide transverse structural integrity to the pellet since they cannot bond the flax fibers together. Finally, it must be noted that the pelletizing was done using a hand activated paper trimmer. The cutting action of the moving knife in such device is relatively gentle compared to industrialized pelletizers. Therefore, it is suspected that poorly impregnated pultrudates will not lead to repeatable pellet morphologies when using industrial pelletizers.

Observation of the manufacturability of the injection process

During injection molding, the LFT pellets with void content of 4% and 8% went smoothly through the hopper and the screw of the injection molding machine without any manual intervention as seen in a supplementary video, which can be viewed at https://youtu.be/0kOILBqbq0U. The remaining LFT pellets with higher void contents, 15% and 22%, required a manual intervention by using a plastic stick to push the mixture down toward the screw (see video above). Figure 6 shows image of the 15%-6 mm master mixture. The broken LFT pellets resulted in a tangled fiber cluster which did not pass down smoothly though the hopper. The pure PP pellets went down smoothly carried by their own weight, but the tangled fiber cluster needed manual intervention, using a pushing tool, to go down towards the screw of the injection molding machine. Without the manual intervention, the tangled fiber cluster would get larger and only the pure PP pellets would pass towards the screw. The high void LFT pellets were observed to bend and entangle with each other forming fiber clusters, obstructing uniform mixing. The diameter of the clusters became larger than the screw inlet, which blocked them from reaching the injection molding screw.

Figure 7(a) shows the dog-bone coupons after the injection molding. Specimens made with pure PP and with Flax/PP LFT pellets. The pure PP dog-bone coupons were fully transparent. All Flax/PP dog-bone coupons had very similar appearances. The dog-bone coupons made out of the high void LFT pellets led to the same injected coupons’ appearance, being opaque and a dark brown color. The fiber distribution in the dog-bone coupons appeared to be evenly and randomly oriented. However, there were some areas in the dog-bone coupons that were a light brown colour. These areas are attributed to bundles of less impregnated flax fibers. Those bundles are not favoured since they indicate inefficient integration between the polymer and the fiber. Yet, this is a typical appearance for biocomposites produced using LFT pellets and injection molding. According to Chaitanya et al.⁵² it is typical that the injection molding process causes the long fibers to bend and twist during processing, leading to the formation of fiber clusters within the produced biocomposites. Figure 7(b) Microscopic images of the injected coupons produced using LFT pellets having 22% void content; the highest void content. No void can be observed in the injected copouns. The white lines are traces of flax fibers pullouts which were caused by the polishing process and the low chemical interaction between flax fibers and PP. This finding indicates that irrespective of the LFT void content, the injection molded coupons did not have any void.

Mechanical properties of the injected parts

Figure 8 presents the impact strength of the injection molded parts as function of void contents and LFT pellets' lengths. The specimens were made with (a) 6 mm and (b) 15 mm LFT pellets. The quasi-continuous LFT pellets resulted in an overall increase in impact strength. The pure PP injection molded parts had an impact strength of 2 kJ/m². All the other injection molded parts had higher impact strength than the pure PP. The LFT pellets with lowest void content, 4%-6 mm and 4%-15 mm, resulted in an impact strength of 5367 ± 218 and 7469 ± 551 $J / m^{2}$ respectively. Here, the 15 mm LFT size caused a 40% improvement in the impact strength. The LFT pellets with the second lowest void content, 8%-6 mm and 8%-15 mm, resulted in an impact strength of 6173 ± 213 and 10,515 ± 730 $J / m^{2}$ respectively. Here also, the 15 mm LFT size caused a significant improvement in the impact strength. In fact, specimen 8%-15 mm had the highest impact strength value compared to all other specimens. This value is almost five times (4.6x) as high as the pure PP impact strength. Considering only the 15 mm LFT pellets, those with around 8% void content resulted in coupons with higher impact strengths than the coupons made using the 4%-15 mm. The LFT pellets with 4% void content were produced using a much lower pultrusion speed of 50 mm/minute. As mentioned above in Thermogravimetric analysis of the flax fiber, the LFT pellets with 4% void content experienced higher levels of thermal degradation which possibly lowered the impact strength of the injection molded specimens. On the other hand, LFT pellets with around 8% void contents were produced at a significantly higher speed, 250 mm/min and thus had experienced less thermal degradation.

Overall, the impact strengths for the 15% and 22% void contents were very comparable to the specimens produced with lower void content LFT pellets. However, the specimens produced using high void content LFT pellets required manual intervention during the injection molding process. Thus, the manual intervention was mostly responsible for improving the mixing quality in the hopper of the injection molding machine. In addition, the specimens produced using LFT pellets with having higher void content had wider standard deviations. This high standard deviation occurred possibly due to the inconsistent mixing during the injection molding experiments.

Figure 9 presents the tensile properties of injection molded parts as a function LFT pellet’s void content and LFT pellet’s length; (a) presents the modulus strength for injection molded parts made using 6 mm LFT pellets and (b) using 15 mm LFT pellets. It should be noted that in all cases the addition of flax fibers led to an increase in stiffness compared to the pure PP sample. The 8%-15 mm specimens had the highest tensile modulus average of 4652 ± 113 MPa, a value which was more than twice as high as the pure PP tensile modulus of 1817 ± 106 MPa. For the specimens made using shorter pellets, shown in Figure 9(a), the modulus slightly increases between specimen 4%-6 mm to 8%-6 mm, then slightly increases again to 15%-6 mm and then stabilizing at approximately the same value. However, the difference in modulus values between 8%-6 mm, 15%-6 mm and 22%-6 mm seems insignificant, and the pooled average modulus of the three conditions is 3764 ± 192 MPa. This indicates that the fiber content and orientation is similar in all cases. Furthermore, the fiber degradation for specimen 4%-6 mm was the highest and specimen 8%-6 mm was the second highest. As seen in Table 4, the 4% void content LFT pellets, which were produced at 50 mm/minute, experienced significantly higher thermal degradation. This explains the lowest modulus measured for 4% void content LFT pellet injected coupons.

For the longer pellets shown in Figure 9(b), the 8%-15 mm had the highest tensile modulus average of 4652 ± 113 MPa. The difference observed with the 4%-15 mm pellets is striking. The modulus value of the 4%-15 mm pellets is 2837 ± 182 MPa, that is, 39% lower than the 8%-15 pellets. This difference is attributed to the thermal degradation due to longer thermal exposure during pultrusion. This may have weakened the flax fibers and resulted in shortening during injection molding. Further studies are needed to verify this explanation. The lower modulus values for 15%-15 mm and 22%-15 mm are attributed to the lower load transfer caused by the bundling and improper mixing of dry flax fibers during injection molding. Thus, the inadequate flax fiber distribution was likely the cause behind the drop in modulus.

In comparing Figure 9(a) and (b), it is observed that, except for the 8%-15 mm, the coupons made using shorter flax fibers resulted in a higher tensile modulus than the longer flax fiber reinforced biocomposites. This is counter-intuitive since longer fibers provide higher surface for stress transfer, which results in higher modulus. The 15%-15 mm and 22%-15 mm modulus values, of 3744 ± 81 MPa and 3283 ± 221 MPa respectively, are equivalent or lower than the pooled average modulus of the 8%-6 mm, 15%-6 mm and the 22%-6 mm coupons. In this case, the lower modulus values are attributed to fiber entanglement due to the fiber clustering during injection molding. An increase in the modulus can be generally attributed to an improved dispersion of the fibers within the biocomposites enabling better stress transfer between fibers and the polymer.^53–55 Chaitanya et al.⁵² observed that long fibers in LFT biocomposites bend and entangle with each other forming fiber clusters, thus impeding uniform dispersion and orientation of fibers within the biocomposite. Therefore, it is likely that the manual intervention to mix the unimpregnated fibers in the injection machine hopper was insufficient to properly mix and detangle the fibers in the injected coupons.

Figure 9(c) presents the maximum tensile strength of the injection molded parts as a function of void content and length. As can be seen in the figure, there is no significant difference between specimens in the maximum tensile strength. All results are around the value of 30 MPa. Also, it is observed that the addition of flax fibers in the PP did not improve the maximum tensile strength compared to the pure PP dog-bone coupons. This finding was expected since there is normally a very low chemical interaction between the hydrophilic fibers, such as flax fibers, and hydrophobic polymers, such as PP, resulting in a low interfacial strength.⁵⁶ Adhesion between the polymer and the reinforcement fiber is an important characteristic to improve the mechanical strength of a composite part.⁵⁷ Yet, it is worth mentioning that the dog-bone coupons did not contain any coupling agent to improve adhesion which could improve the tensile properties of these samples. This result confirms similar findings by Barkoula et al.⁴⁹ Typically, MAPP can be used as the coupling agent to improve wettability for natural fiber/PP composites.^58–61

Morphological analysis of the injection molded specimens

Figure 10 shows the SEM images of an impact fractured region. Typical SEM images are represented in (a) & (b), showing smooth surfaces at the flax fibers as well as fiber pull-outs after the impact fracture. The smooth surfaces indicate that the flax fibers had weak adhesion and chemical interaction with the PP polymer. The low chemical interaction between the flax fibers and PP was the cause behind making the maximum tensile strengths unchanged relative to pure PP as seen in Figure 9(c). Figure 10(c) 8%-6 mm & (d) 8%-15 mm show paired specimen which had the same void content but was produced using a different LFT pellet length. As can be seen in Figure 10(d), the SEM image shows relatively longer fiber compared to 8%-6 mm specimen, even after the impact fracture. The longer fibers could explain the improvement of the impact strength of the 15 mm LFT pellets compared to the 6 mm LFT pellets.

Conclusions

This study investigated the effects of LFT pellets’ void contents during the pelletizing process and on the mechanical properties of the injection-molded parts. The fiber volume content of the pultrudates was 50% and four void content points were tested between 4% and 22%. The pultrudates were pelletized into two different lengths of 6 and 15 mm. Three sets of dog-bone coupons were injection-molded using a combination between the void contents and the LFT pellets' lengths. The LFT pellet’s void content was found to have an influence on the LFT pelletizing quality as well as on the mechanical properties of the injection-molded parts. Flax/PP rods pultruded at 50 and 250 mm/min had void contents up to 8%. These rods were cut into well consolidated pellets. These structurally integral pellets were injection-molded without any issues. Flax/PP rods produced at 500 mm/min and 1000 mm/min led to void contents above 15%. The pelletizing process of these poorly impregnated rods was very damaging, resulting in broken pellets with exposed dry fibers. Manual intervention to allow the LFT pellets and dry flax fibers to sink in the injection molding machine was required. The mechanical properties of the injection-molded parts using high void LFT pellets resulted in less consistent values attributed to inadequate flax fiber distribution. The mechanical properties of the injection-molded parts using low-void LFT pellets resulted in more consistent values. The highest recorded impact strength was 10,515 ± 730 J/m2, corresponding to an experiment that was done using 15 mm LFT pellets with around 8% void content. These pellets were cut from rods pultruded at 250 mm/min. The LFT void content and length did not have any influence on the maximum tensile strengths of the injection-molded parts. All SEM images showed that the exposed fibers had smooth surfaces, indicating very low chemical interaction between the flax fibers and PP. Irrespective of the LFT void content, the injection molded coupons did not have any voids. Finally, further work will need to be conducted on the parameters of the injection molding machine such as using different injection speeds and pressure values to confirm the effects of Flax/PP LFT pellets’ morphology on the mechanical properties of injection-molded parts. This will help applicable industries, which make LFT components, to choose the appropriate void contents and injection molding parameters. Also, future work will characterize, discuss and improve the seemingly poor interfacial properties between flax and PP and thoroughly investigate the blending of flax-PP compatibilizing agent with Flax/PP LFT pellets.

Footnotes

Acknowledgements

The authors would like to thank Danielle Szydlowski for her help with pelletizing pultrudates. We would like also to thank Bombardier Aerospace, Pultrusion Technique, Saudi Arabia’s Saline Water Conversion Corporation and Ministry of Education (KSP12014087), NSERC (CRDPJ488387-15) and Prima Quebec (R10-009) for financing this research project.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the PRIMA Quebec (R10-009), Natural Sciences and Engineering Research Council of Canada (CRDPJ488387-15), Saudi Arabia’s Saline Water Conversion Corporation and Ministry of Education (KSP12014087).

Data availability

The raw/processed data required to reproduce these findings cannot be shared at this time due to technical or time limitations

References

Fogorasi

Barbu

. The potential of natural fibres for automotive sector-review. In: IOP conference series: materials science and engineering. IOP Publishing; 2017, p. 012044.

Zhu

Njuguna

, et al. Recent development of flax fibres and their reinforced composites based on different polymeric matrices. Materials 2013; 6: 5171–5198.

Pervaiz

Panthapulakkal

Sain

, et al. Emerging trends in automotive lightweighting through novel composite materials. Mater Sciences Appl 2016; 7: 26–38.

Cheah

. Cars on a Diet: The Material and Energy Impacts of Passenger Vehicle Weight Reduction in the US. Massachusetts, MA: Massachusetts Institute of Technology, 2010.

Ngo

T-D

. Natural Fibers for Sustainable Bio-Composites. Natural and Artificial Fiber-Reinforced Composites as Renewable Sources, 2018, pp. 107–126.

Bispo

SJL

Freire

RCS

Aquino

EMFD

. Mechanical properties analysis of polypropylene biocomposites reinforced with curaua fiber. Mater Res 2015; 18: 833–837.

Angelov

Wiedmer

Evstatiev

, et al. Pultrusion of a flax/polypropylene yarn. Composites A: Appl Sci Manufacturing 2007; 38: 1431–1438.

Baley

Kervoëlen

Lan

, et al. Flax/PP manufacture by automated fibre placement (AFP). Mater Des 2016; 94: 207–213.

Van de Velde

Kiekens

. Thermoplastic pultrusion of natural fibre reinforced composites. Compos Structures 2001; 54: 355–360.

10.

Bénard

Fois

Grisel

. Roughness and fibre reinforcement effect onto wettability of composite surfaces. Appl Surf Sci 2007; 253: 4753–4758, DOI: 10.1016/j.apsusc.2006.10.049.

11.

Sinha

Panigrahi

. Effect of plasma treatment on structure, wettability of jute fiber and flexural strength of its composite. J Compos Mater 2009; 43: 1791–1802. DOI: 10.1177/0021998309338078.

12.

Bera

Alagirusamy

Das

. A study on interfacial properties of jute-PP composites. J Reinforced Plastics Composites 2010; 29: 3155–3161. DOI: 10.1177/0731684410369723.

13.

Ning

Hassen

, et al. A review of Long fibre thermoplastic (LFT) composites. Int Mater Rev 2020; 65: 164–188.

14.

Gottgetreu

. Injection compression molding of long glass fiber reinforced polypropylene. SAE Transactions 1998; 107: 611–618.

15.

Alsinani

Laberge

. Effect of length variation of flax/polypropylene pellets produced by multi-die pultrusion on the quality of injection molded product. In: Automotive Composites Conference & Exhibition. ACCE; 2018.

16.

Rohan

McDonough

Ugresic

, et al. Mechanical study of direct long fiber thermoplastic carbon/polyamide 6 and its relations to processing parameters. In: Proc 14th-Annual SPE Automotive Composites Conference & Exhibition, Novi, MI, 9–11 September, 2015.

17.

Wang

Uawongsuwan

Yang

, et al. Effect of molding condition and pellets material on the weld property of injection molded jute/polylactic acid. Polym Eng Sci 2013; 53: 1657–1666.

18.

Xie

Yang

. Interface and mechanical properties of poly (methyl methacrylate)–fiber composites. J Applied Polymer Science 2004; 93: 2478–2483.

19.

Garcia‐Rejon

Meddad

Turcott

, et al. Extrusion blow molding of long fiber reinforced polyolefins. Polym Eng Sci 2002; 42: 346–364.

20.

Fan

Liu

Whitfield

, et al. Effects of processing parameters on the thermal and mechanical properties of LFT-D-ECM glass fiber/polyamide 6 composites. SPE Automotive 2016; 73(1): 44–45.

21.

Bondy

Altenhof

. Low velocity impact testing of direct/inline compounded carbon fibre/polyamide-6 long fibre thermoplastic. Int J Impact Eng 2018; 111: 66–76.

22.

Weidenmann

Grigo

Brylka

, et al. Time temperature dependence of long fiber reinforced polypropylene manufactured by direct long fiber thermoplastic process. Int J Mater Metallurgical Eng 2016; 11: 13–18.

23.

Perez

Osswald

Goris

. Study on the fiber properties of a LFT strand. SPE ACCE 2013; 2: 1115–1126.

24.

Hawley

Jones

. In-line compounding of long-fiber thermoplastics for injection molding. J Thermoplastic Compos Mater 2005; 18: 459–464.

25.

Kumar

Alwekar

Kunc

, et al. High-performance molded composites using additively manufactured preforms with controlled fiber and pore morphology. Additive Manufacturing 2021; 37: 101733.

26.

Brast

Michaeli

. Processing of long-fibre reinforced thermoplastics using the direct strand-deposition process. Plastics, Additives and Compounding 2001; 3: 22–29.

27.

Ren

Dai

. Fiber dispersion and breakage in deep screw channel during processing of long fiber-reinforced polypropylene. Fibers Polym 2014; 15: 1507–1516.

28.

Leaversuch

. In-Line compounding of long-glass/PP gains strength in automotive molding. Plastics Technology 2001; 9: 39–40.

29.

Kim

Park

. Effect of fiber length on mechanical properties of injection molded long-fiber-reinforced thermoplastics. Macromolecular Res 2020; 28: 433–444.

30.

Lee

. Handbook of Composite Reinforcements. New Jersey, NJ: John Wiley & Sons, 1996.

31.

Carlsson

Åström

. Experimental investigation of pultrusion of glass fibre reinforced polypropylene composites. Composites Part A: Appl Sci Manufacturing 1998; 29: 585–593.

32.

Alsinani

Ghaedsharaf

Lebel

. Effect of cooling temperature on deconsolidation and pulling forces in a thermoplastic pultrusion process. Composites B: Eng 2021; 219: 108889.

33.

Ghaedsharaf

Brunel

J-E

Lebel

. Multiscale numerical simulation of the forming process of biaxial braids during thermoplastic braid-trusion: predicting 3D and internal geometry and fiber orientation distribution. Composites Part A: Appl Sci Manufacturing 2021; 150: 106637.

34.

Absi

Alsinani

Lebel

. Carbon fiber reinforced poly (ether ether ketone) rivets for fastening composite structures. Compos Structures 2021; 280: 114877.

35.

Vaidya

Thattaiparthasarthy

Pillay

, et al. Colored inorganic-pigmented long-fiber thermoplastics. J Thermoplastic Compos Mater 2015; 28: 1627–1643.

36.

Bartus

Vaidya

Ulven

. Design and development of a long fiber thermoplastic bus seat. J Thermoplastic Compos Mater 2006; 19: 131–154.

37.

Hartness

Husman

Koenig

, et al. The characterization of low cost fiber reinforced thermoplastic composites produced by the DRIFT™ process. Composites Part A: Appl Sci Manufacturing 2001; 32: 1155–1160.

38.

Tan

Wang

. Preparation, microstructures, and properties of long-glass-fiber-reinforced thermoplastic composites based on polycarbonate/poly (butylene terephthalate) alloys. J Reinforced Plastics Composites 2015; 34: 1804–1820.

39.

Friedrich

. Carbon fiber reinforced thermoplastic composites for future automotive applications. In: AIP Conference Proceedings. Melville, NY: AIP Publishing LLC; 2016, p. 020001.

40.

Friedrich

Almajid

. Manufacturing aspects of advanced polymer composites for automotive applications. Appl Compos Mater 2013; 20: 107–128.

41.

Bockstedt

Skarlupka

. An Overview of Long Fiber Reinforced Thermoplastics. In: American Chemical Society (ACS) national meeting, Anaheim, CA, 2–6 Apr 1995.

42.

Han

Liu

. Preparation and mechanical properties of long glass fiber reinforced PA6 composites prepared by a novel process. Macromolecular Mater Eng 2005; 290: 688–694.

43.

Bayerl

Mitschang

. Rheological and mechanical behavior of long‐polymer‐fiber reinforced thermoplastic pellets. J Appl Polym Sci 2014; 131(1): 39716.

44.

Lapointe

Laberge Lebel

. Fiber damage and impregnation during multi‐die vacuum assisted pultrusion of carbon/PEEK hybrid yarns. Polym Composites 2019; 40: E1015–E1028.

45.

Wiedmer

Manolesos

. An experimental study of the pultrusion of carbon fiber-polyamide 12 yarn. J Thermoplastic Composite Materials 2006; 19: 97–112.

46.

Ghaedsharaf

Brunel

J-E

Laberge Lebel

. Thermoplastic composite rod manufacturing using biaxial braid-trusion. In: Proceedings of the ECCM18–18th European Conference on Composite Materials. Athens, Greece, 24–28 Jun 2018.

47.

Bos

Müssig

van den Oever

. Mechanical properties of short-flax-fibre reinforced compounds. Composites Part A: Appl Sci Manufacturing 2006; 37: 1591–1604.

48.

Nyström

Joffe

Långström

. Microstructure and strength of injection molded natural fiber composites. J Reinforced Plastics Composites 2007; 26: 579–599.

49.

Barkoula

Garkhail

Peijs

. Effect of compounding and injection molding on the mechanical properties of flax fiber polypropylene composites. J Reinforced Plastics Composites 2010; 29: 1366–1385.

50.

Barneto

Vila

Ariza

, et al. Thermogravimetric measurement of amorphous cellulose content in flax fibre and flax pulp. Cellulose 2011; 18: 17–31.

51.

Arias

Heuzey

M-C

Huneault

. Thermomechanical and crystallization behavior of polylactide-based flax fiber biocomposites. Cellulose 2013; 20: 439–452.

52.

Chaitanya

Singh

. Processing of PLA/sisal fiber biocomposites using direct-and extrusion-injection molding. Mater Manufacturing Process 2017; 32: 468–474.

53.

Bourmaud

Pimbert

. Investigations on mechanical properties of poly (propylene) and poly (lactic acid) reinforced by miscanthus fibers. Composites Part A: Appl Sci Manufacturing 2008; 39: 1444–1454.

54.

Bledzki

Jaszkiewicz

Scherzer

. Mechanical properties of PLA composites with man-made cellulose and abaca fibres. Composites A: Appl Science Manufacturing 2009; 40: 404–412.

55.

Yan

, et al. Fabrication and characterization of fully biodegradable natural fiber-reinforced poly (lactic acid) composites. Composites Part B: Eng 2014; 56: 717–723.

56.

Hoa

. Principles of the Manufacturing of Composite Materials. Lancaster, PA: DEStech Publications, Inc, 2009.

57.

X-F

Dzenis

. Droplet on a fiber: geometrical shape and contact angle. Acta Mechanica 2006; 185: 215–225.

58.

Arbelaiz

Fernandez

Ramos

, et al. Mechanical properties of short flax fibre bundle/polypropylene composites: Influence of matrix/fibre modification, fibre content, water uptake and recycling. Composites Science Technology 2005; 65: 1582–1592.

59.

Bledzki

Fink

Specht

. Unidirectional hemp and flax EP‐and PP‐composites: Influence of defined fiber treatments. J Appl Polym Sci 2004; 93: 2150–2156.

60.

Gassan

. A study of fibre and interface parameters affecting the fatigue behaviour of natural fibre composites. Composites Part A: Applied Science Manufacturing 2002; 33: 369–374.

61.

Kannan

Cheng

. Influence of laminate lay-up, hole size and coupling agent on the open hole tensile properties of flax yarn reinforced polypropylene laminates. Composites Part B: Eng 2014; 57: 80–85.