Fracture toughness enhancement of carbon fiber–reinforced polymer composites utilizing additive manufacturing fabrication

Abstract

The present work reports a novel approach to enhance the fracture resistance and notch sensitivity of carbon fiber-reinforced polymer composites utilizing additive manufacturing (3-D printing) fabrication. The 3-D printed composites utilize carbon fiber bundles to reinforce nylon/chopped fiber resin in a multilayered structure configuration. Single-edge (60°) notched samples were printed using Mark Two printer. Three reinforcement schemes were designed and used to manufacture the specimens. The focus was placed on selective reinforcement at the crack tip to arrest crack initiation. The mechanical properties, fracture toughness, and fracture behavior of the printed composites were evaluated. It was found that wrapping fiber around the notch effectively blunted the notch and redirected crack propagation away from the notch tip, thereby lengthening the crack path and leading to improved fracture resistance. It was also found that such improvement reaches a saturation level. Excessive notch reinforcement beyond optimal limit can reverse the gains in fracture resistance due to notch-targeted reinforcement. Examination of the fracture surface morphology of the printed composites reveals lack of fusion of the sizing of the individual continuous carbon fiber bundles and the lack of adhesion between the matrix layers (nylon/chopped fiber resin) and the adjacent carbon fiber bundle reinforcement. Damage to the fibers within the carbon bundle was also observed. Thus, a synergetic effect of the carbon fiber bundles reinforcement and the matrix requires more optimization to manufacture carbon-reinforced polymer composites using 3-D printing.

Keywords

3-D printed composites fracture resistance unidirectional reinforcement additive manufacturing surface morphology

Introduction

The ability to fabricate strong and durable lightweight parts for load-bearing applications has been the subject of extensive research. Such parts are of particular interest for aerospace, military, automotive, and clean energy applications. Although polymers have higher strength-to-weight ratio than metals, their ability to replace them for such applications is hindered by their inherently low mechanical and fracture resistance properties. To enhance those properties, two types fiber reinforcement based on fiber continuity have been used: short (chopped) and continuous. Compared to chopped fiber, continuous fiber reinforcement is more effective due to longer fibers and improved load transference from the matrix. Several techniques have been developed to fabricate continuous fiber polymer composites.¹ Molding is a common approach involving the manual or automated placement of fibers in a mold cavity with the subsequent infusion of liquid resin under pressure or vacuum. In pultrusion, commonly used to produce long prismatic parts, continuous fiber rovings drawn from fiber spools are impregnated with resin before introducing them into a preforming die and a heated die for curing. Filament winding is another technique whereby continuous fiber is wrapped or woven around mandrels to produce reinforced hollow structures. In spite of the effectiveness of those techniques, they have drawbacks in terms of the intensity of cost, time, and labor involved as well as the level of precision and control of fiber placement. Ideal reinforcement patterns selectively targeting critical stress raising features such as holes and fillets as well as external part walls cannot be freely created using those methods. Finally, in all the listed techniques, thermoset resin is commonly used as the matrix material due to its ease of flow and fiber impregnation compared to thermoplastic resin, which has the advantage of better impact resistance, weldability, and recyclability.²

Additive manufacturing (AM), commonly known as 3-D printing, offers a new approach to fabricate complex parts without the use of expensive complex molds or heavy machinery.³ Starting with a CAD model of the part, thin (fraction of millimeter thick) slices are extracted by a slicing software and a tool path is generated to create each slice (layer). As layers are built one top of the other, the complete part is gradually realized. Different techniques for creating the layers have been developed for polymers. The most widely used technique, commonly known as fused deposition modeling (FDM), involves the liquefaction and extrusion of thermoplastic filament feedstock from a small nozzle (typically 0.5 mm in diameter) followed by the deposition of the resulting bead along the predetermined tool path necessary to create the layer.⁴ The tightly packed beads fuse together upon contact creating a solid. Typical thermoplastics used in traditional polymer fabrication such as acrylonitrile butadiene styrene (ABS), nylon, and polycarbonate (PC) can be processed using FDM, making it the most versatile polymer AM process. Although the FDM process produces the strongest parts compared to other polymer AM processes, FDM parts remain inherently weak compared to their counterparts produced by traditional thermoplastic fabrication processes such as plastic injection molding.⁵

To enhance the mechanical properties and durability of FDM polymeric parts, fiber reinforcement strategies used in traditional (as opposed to additive) composite manufacturing are currently being investigated. By combining the advantages of 3-D printing and fiber reinforcement technologies, this approach to composite manufacturing can be transformative in terms of cost, time, new material systems as well as mechanical properties and overall performance. Zhong et al. reinforced ABS filaments with chopped glass fiber and noticed an increase in bonding between layers which was attributed to fiber bridging across neighboring layers.⁶ Ning et al.⁷ used chopped carbon fiber to reinforce ABS filaments and measured a peak improvement in strength of 27% at 7.5 wt% fiber loading. Tekinalp et al.⁸ studied the effect of adding short (0.2–0.4 mm) carbon fiber coated with epoxy-based sizing to ABS copolymer filaments.⁸ They studied the effect of fiber loading and orientation on the tensile strength and modulus and reported an improvement of about 115% and 700%, respectively, over the neat ABS. They also found that the extrusion process in FDM resulted in improved fiber alignment compared to that observed in compression-molded samples.

For continuous fiber reinforcement of FDM-fabricated polymeric parts, two main approaches are being considered. The first involves the simultaneous extrusion of a thermoplastic filament and a central continuous fiber strand from a single nozzle resulting in coaxial plastic/fiber bead. Using carbon fiber to reinforce polylactic acid (PLA), Matsuzaki et al.⁹ achieved about 6-fold improvement in tensile modulus and 4.4-fold improvement in tensile strength using this approach. They also examined the reinforcement due to twisted yarn of natural jute fiber, but the improvement over neat PLA was not significant. Fiber pullout was the main failure mechanism reported in Matsuzaki et al.’s work, indicating poor adhesion between the fiber and the PLA matrix. The poor carbon fiber adhesion to PLA matrix was also observed by Li et al.¹⁰ To improve adhesion, they modified the carbon fiber through the application of PLA sizing agent to the fiber. They examined the tensile strength, flexural strength as well as the storage modulus, loss tangent, and glass transition. The carbon fiber modified with sizing resulted in 13.8%, 164%, and 351% additional tensile strength, flexural strength, and storage modulus, respectively, over the original carbon fiber (with no sizing). Another issue with the coaxial fiber reinforcement noted by Brooks et al.¹¹ is the stress concentration effect at the central fiber surface and the off-center fiber shifting in the deposited bead, Li et al.¹⁰ The second approach involves the independent deposition of the thermoplastic matrix and the fiber reinforcement using two separate nozzles. The thermoplastic filament is similar to standard FDM filaments and the fiber reinforcement filament consists of a bundle of carbon fibers held together by a sizing agent. A specialized nozzle is used to heat and deposit the fiber as it follows a prescribed path to produce the desired fiber reinforcement layout. This approach allows for more flexibility in designing the fiber reinforcement scheme compared to the previous approach. From a practical point of view, an advantage of this approach is the maturity and commercial availability of the equipment and materials needed through Markforged Inc., (Watertown, MA).¹² Using this technology, Dickson et al.¹³ compared the enhancement of tensile and flexural strengths of the propriety nylon blend of Markforged due to continuous fiber reinforcement using Kevlar, carbon fiber, and fiber glass filaments supplied by the same company. They measured 6.3-fold enhancement in tensile strength (passing that of aerospace-grade aluminum) and 5-fold enhancement in flexural strength over the neat nylon, with carbon fiber offering the highest improvement. They indicated that the strength enhancement saturates with the increase in fiber content beyond about 25% due to the accompanying increase in porosity and the inherent weak fiber/matrix interface. Their examination of the fracture surfaces indicated that fiber pullout occurred frequently implying weak fiber matrix bonding. Using the same proprietary nylon filament and technology of Markforged, van derer Klift et al.¹⁴ measured ninefold increase in tensile strength, albeit with large variability reaching 22% standard deviation. Melenka et al.¹⁵ examined the effect of fiber volume fraction on the elastic modulus of reinforced nylon using Kevlar. They measured a modulus of 1767, 6920, and 9001 MPa for 4.04%, 8.08%, and 10.1% volume fraction, respectively, compared to the neat’s 0.94 GPa as specified by the supplier.

In the current work, we study the fracture resistance enhancement due to continuous fiber reinforcement of thermoplastic polymers utilizing AM. Fiber reinforcement is selectively placed within a part at critical locations where failure can most likely occur to improve fracture resistance. We believe that our novel fabrication procedures are more effective, and potentially more economical, for reinforcing critical structural components compared to traditional reinforcement strategies. To prove the concept, we study the effect of different reinforcement strategies on the fracture resistance and notch sensitivity of unidirectional carbon fiber-reinforced nylon composite.

Materials and experimental work

Materials

Two materials were used to prepare our composite samples. The matrix material was Onyx, a proprietary nylon blend loaded with chopped carbon fiber for improved strength and heat deflection resistance. It was supplied by Markforged Inc.¹² in the form of spooled filament 1.75 mm in diameter. The reinforcement material, also supplied by Markforged Inc., was a filament about 0.35 mm in diameter consisting of a bundle of continuous carbon fibers held together by a sizing agent. The complete specifications of the reinforcement filament are not disclosed by the supplier; however, Dickson et al.¹³ counted 1000 fibers in the bundle, each about 0.008 mm in diameter. Table 1 presents the tensile properties of Onyx and the carbon fiber reinforcement measured using 3-D printed samples, as reported by the supplier. Each layer of the carbon fiber test samples was completely filled with unidirectional fiber parallel to the tensile axis using the reinforcement filament described above. The Onyx test samples consisted of completely filled layers with beads of undisclosed infill orientation. Carbon fiber samples and Onyx samples were reportedly tested in compliance with ASTM D3039 and ASTM D638 standards, respectively.

Table 1.

Tensile properties of Onyx and carbon fiber reinforcement.¹²

	Onyx	Carbon fiber
Tensile modulus (GPa)	1.4	54
Tensile strength (MPa)	30	700
Tensile strain at break	0.58	0.015

Description of samples

Two types of tensile samples, un-notched and single-edge 60° notched, were used and are shown in Figure 1. In all cases, the notch was created through the printing process. Matrix samples and reinforced samples using three different continuous reinforcement schemes, referred to as A, B, and C, were studied (the schemes are explained in the next subsection). Table 2 describes the six sets of samples, which were fabricated and tested. For each set, four replicates were tested. The un-notched samples were used to establish the tensile properties of the matrix in order to evaluate the effect of reinforcement. The notched samples were used to determine the effect of the different reinforcement schemes on fracture resistance.

Table 2.

Description of the different sets of samples studied.

Set ID	Description	Composition	Reinforcement scheme
1	Un-notched matrix	Onyx only	None
2	Un-notched—reinforcement A	Onyx/continuous carbon fiber	A
3	Notched—matrix	Onyx only	None
4	Notched—reinforcement A	Onyx/continuous carbon fiber	A
5	Notched—reinforcement B	Onyx/continuous carbon fiber	B
6	Notched—reinforcement C	Onyx/continuous carbon fiber	C

Sample fabrication and reinforcement schemes

All samples tested in this work were fabricated using Mark Two 3-D printer from Markforged Inc.,¹² which is based on the FDM process described in the Introduction. Mark Two has two dedicated nozzles, one for Onyx filament and the other for the continuous fiber reinforcement filament. During printing, the Onyx filament is liquefied and extruded from a 0.5-mm diameter nozzle, while the fiber filament is heated to liquefy the sizing to infuse the fibers and provide adhesion to the matrix. The construction of any layer starts by creating its external contour using Onyx, followed by infilling the interior with Onyx and/or fiber reinforcement as needed. The orientation of the infill beads with respect to the loading axis of the sample has a significant impact of the mechanical response and will be described below.

All our samples were constructed using a two-layer repeating sequence with each layer having a thickness of 0.125 mm. In the case of matrix samples, both layers were fully filled with Onyx with the infill bead orientation parallel to the longitudinal direction of the sample (also the tensile loading direction) in one layer and transverse to it in the next layer. For the reinforced samples, one layer was fully filled with Onyx (matrix) with infill bead orientation transverse to the longitudinal direction of the sample, while the other layer (reinforcement layer) was filled with continuous fiber strands according to schemes A, B, or C, as specified in Table 2. For reinforcement A, straight bundles of carbon fiber were laid along the longitudinal direction of the sample without targeted reinforcement of the notch in the case of the notched samples (Table 3). This scheme is similar to unidirectional continuous fiber reinforcement produced by traditional reinforcement technique. For reinforcement B, one fiber bundle (approximately 0.9 mm wide and containing 1000 fibers) followed the contour of the notch, wrapping around it and providing targeted reinforcement, while the remaining area of the layer was filled with longitudinal fiber bundles as in reinforcement A. Finally, reinforcement C was similar to reinforcement B, except that three, instead of one, fiber bundles were used to reinforce the notch. Table 3 clarifies the different reinforcement schemes. The fourth column in the table shows the actual tool path generated by the slicing/pathing software and followed by the nozzle during the actual printing process. The outermost path shows the Onyx contour path, as explained earlier, while all other lines show the fiber routing path. In all cases, the reinforcement ratio (volume of reinforcement filament to the total volume) was about 36%.

Table 3.

Summary of the characteristics of the different reinforcement schemes.

Type of samples	Reinforcement Scheme	Reinforcement configuration
Un-notched	Reinforcement A	Unidirectional
Notched	Reinforcement A	Unidirectional
Notched	Reinforcement B	Notch wrapping—one reinforcement bundle
Notched	Reinforcement C	Notch wrapping—three reinforcement bundles

It is emphasized that this study is focused on the effect of different continuous fiber reinforcement schemes embedded in the otherwise Onyx matrix. As such the chopped fiber reinforcement in Onyx is not a variable and contributes only to the baseline properties of the matrix.

Results and discussion

Microstructure analysis of the 3D-printed composite

As it was explained previously, the composite matrix is Onyx (chopped carbon fiber-reinforced nylon) deposited in layers with the infill beads perpendicular to the loading direction of the tested specimens, while the unidirectional carbon fiber reinforcement was deposited between the matrix layers along the loading direction. The scanning electron microscopic (SEM) micrographs in Figure 2(a) is taken from a cut section of the as-printed sample perpendicular to the reinforcement direction. Figure 2(a) shows the edges of the matrix layers and the reinforcement layers. The vertical matrix layers are clearly shown in Figure 2(a). At higher magnification, Figure 2(b), the matrix layer clearly displays the imbedded chopped fibers aligned with the Onyx bead. Some of these fibers are dislodged from the matrix and oriented in the vertical direction along the matrix layer. A broken continuous fiber reinforcement bundle is shown in the left side of Figure 2(b). The matrix edge is clearly identified adjacent to the fibers bundle indicating lack of real wetting between the carbon fiber bundle and the matrix material. Dickson et al.¹³ and van der Klift et al.¹⁴ made similar observations about this weak fiber/matrix bonding.

Stress–strain behavior of the 3-D printed composite

Tensile tests were performed on two sets of un-notched samples: matrix (Onyx with no continuous reinforcement) and matrix reinforced with continuous fiber reinforcement scheme A (sets 1 and 2, Table 2). As mentioned earlier, the reinforcement ratio was estimated at 36%. Representative stress–strain curves of both sets are shown in Figure 3. Also shown is the carbon fiber behavior generated using the tensile strength and modulus reported by the supplier (Table 1), assuming linear behavior. The measured values for strength and modulus of the Onyx matrix are comparable to those of the supplier, Table 1. The discrepancy between the measured strain-to-break (average 0.14 mm/mm) and the supplier reported value (0.58 mm/mm; Table 1) is most likely attributed to the difference in the construction of the 3D-printed samples. As explained earlier, 50% of the layers in our matrix samples had the infill beads oriented along the longitudinal axis of the sample, while the rest had beads deposited transverse to it, with the latter layers contributing poorly to ductility. As the supplier does not disclose the construction of their sample, it is believed that the bead orientation in their samples alternate between +45 and −45° to the loading direction from one layer to the next, which is expected to improve ductility. Ignoring the contribution of the matrix to the overall strength (36.5 MPa compared to the 700 MPa strength of the carbon fiber reinforcement), the predicted strength of the 36% reinforced samples should be about 252 MPa, which is comparable to the average measured strength as shown in Figure 3.

Residual strength–strain behavior of 3-D printed composite

3-D printing of composites can allow for tailoring the reinforcement of load-bearing structures. This is done by selectively reinforcing critical sections or locations where crack initiation is likely to start from. To prove this concept, we studied the reinforcement of the crack tip of single-edge 60° notched specimens. Three different reinforcement schemes, A, B and C, as detailed in Table 3 were used. Four sets of notched samples, sets 3–6, Table 2, were loaded in tension until fracture to determine the residual strength (the strength of the notched sample). Figure 4 shows representative residual tensile stress–strain curves of the four sets. The stress (hence the residual strength) was calculated by dividing the load by the full cross-sectional area away from the notch.

The first observation to make is that the residual stiffness (stiffness of the notched sample as determined from the stress–strain curve of the notched sample) was largest for reinforcement A followed by that of reinforcement B, and lastly C (Figures 4 and 5(a). This trend can be rationalized by noting that reinforcement A consisted of seven straight fiber bundles running along the longitudinal direction of the sample between the notch tip and the other side of the sample, while the corresponding numbers in the case of reinforcements B and C were six and four, respectively, see schematics in column 4 of Table 3. Those fiber bundles are more effective at carrying the tensile load than the fibers wrapping around the notch (none, one and three in the case of reinforcements A, B, and C, respectively). Hence, the larger the number of straight fiber bundles ahead of the notch, the higher the stiffness for the same cross-sectional area of specimen.

On the other hand, the fiber bundles wrapping around the notch are more effective at arresting crack initiation from the notch tip, that is, blunting the notch, which is expected to increase fracture resistance and the residual strength. Figure 5(b) supports this conclusion by noting that reinforcement B resulted in higher residual strength than that of reinforcement A. To help explain this result, Figure 6 shows the fracture pattern exhibited by the samples with different reinforcement schemes. The dashed lines overlaid on the fractured samples pictures traces the inner boundary of the fiber bundle wrapped around the notch. The arrows overlaid on the schematics on the left of the figure indicate the crack initiation site (at arrow tail) and its propagation direction. In the case of reinforcement, A, which is typical to traditional fiber reinforcement methods, the crack starts at the notch tip where stress is highest and cuts through the fiber ahead of the notch taking the shortest path to the opposite side of the sample, as expected (Figure 6(a)). On the contrary, in the case of reinforcement B, where continuous fiber was wrapped around the notch, the crack initiation site was on the other side of the sample far away from the notch tip, indicating that such reinforcement strategy effectively blunted the notch (Figure 6(b)). Furthermore, the wrapped fiber controlled the direction of crack propagation, redirecting its path along the inner boundary of the fiber away from the notch, thus increasing the crack path length and fracture energy.

The same phenomena, notch blunting and redirecting crack propagation, were observed in the case of reinforcement C that has three fiber bundles ahead of the crack tip (Figure 6(c)). The fact that the residual strength of reinforcement C is lower than that of B suggests that the notch blunting effect saturates and that an optimal degree of notch reinforcement exists. Excessive fiber reinforcement of the notch beyond what is needed to blunt it comes at the expense of the straight fibers, which are more effective at resisting the tensile load than the curved fibers wrapping around the notch. Hence, excessive notch reinforcement, as suggested by the case of reinforcement C, results in lower fracture resistance.

Fracture behavior

The fracture behavior of the 3-D printed composites without notch tip reinforcement (reinforcement A in Table 3 and Figure 6(a)) was examined in view of Figure 7, stitched SEM images of a fractured notched specimen. The crack propagated is from right to left in Figure 7. It is seen from the area ahead of the crack tip that the crack started by rupturing the Onyx matrix layers (nylon blend loaded with chopped carbon fiber). Then the crack propagated causing the rupture of the reinforcement layers consisting of bundles of continuous carbon fibers. The macro fracture surface morphology of the 3-D printed composites under investigations reveals the lack of fusion of the individual continuous carbon fiber bundles together. This is clear from the space that exists between the bundles. Thus, a synergetic effect of the carbon fiber bundles reinforcement is not fully utilized. In addition, there is a lack of adhesion between the matrix layers and the adjacent carbon fiber bundle reinforcement layers.

The lack of fusion between the carbon fiber bundles and the lack of adhesion of these bundles to the Onyx matrix are further examined using SEM. High-magnification micrographs of the fracture surface are shown in Figure 8(a) to (d). The micrograph in Figure 8(a) displays two horizontal layers of the Onyx matrix with carbon fiber bundles sandwiched between them. The matrix layers display ductile fracture as evident from the ridges formation in both Figure 8(a) and (b).

The lack of fusion of the fiber bundles together is also displayed in Figure 8(a). This has created voids between the fiber bundles and hence porosity between the matrix layers. This can probably be addressed by tightening the reinforcement together during printing and optimize the temperature of the fiber bundles during printing.

Another issue observed from Figure 8(c) and (d) related to the continuous fiber reinforcement usage in 3-D printed composite is the crushing of the individual fibers within some reinforcement bundles. This can reduce the effectiveness of the reinforcement. A plausible explanation of this damage to the fibers could be due to the displacement mismatch between the matrix and the fiber bundle during deposition. If, for example, the reinforcement fibers bundle has any slack during the deposition, a quasi-static loading can cause such damage as evident in Figure 8(d).

Conclusions

3-D printed continuous carbon fiber-reinforced polymer composites have been studied. The 3-D printed composites utilized unidirectional carbon fiber bundles to reinforce nylon/chopped fiber resin in a multilayered structure configuration. The aim of the current study was to explore 3-D printing in the manufacturing of load-bearing structures with inherent reinforcement of simulated defects such as side notch. It was found that wrapping fiber around the notch, enabled by 3-D printing technology, effectively blunted the notch tip causing cracks to initiate away from the notch tip, lengthening the crack path and leading to improved fracture resistance.

Examination of the fracture surface morphology of the printed composites reveals lack of fusion of the sizing of the individual continuous carbon fiber bundles together. Thus, a synergetic effect of the carbon fiber bundles reinforcement is lacking. Increasing the packing density and optimizing the deposition temperature of the fiber bundles can enhance the fusion and hence the effectiveness of the reinforcement.

It was also found that there is a lack of adhesion between the matrix layers (nylon/chopped fiber resin) and the adjacent carbon fiber bundle reinforcement. As both the sizing of the carbon fiber bundles and the matrix material are thermoplastic, an optimization of the deposition temperature during printing may enhance the adhesion leading to a denser composite.

Damage to the fibers within the carbon bundle during deposition can occur leading to reduction in the effectiveness of the reinforcement. It is believed that this is due to the displacement mismatch (slack) between the fiber bundle and the matrix during deposition. This can be addressed by optimizing the tension on the deposited carbon bundle on the matrix.

Footnotes

Acknowledgements

The support of Ms. Telisa Toliver, Vice-President, Business Development and Ms. Niccole Boswell, Diversity Portfolio Manager-University Affairs, Chevron Corporation is greatly appreciated.

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 Chevron Corporation through the Tuskegee University-Chevron Additive Manufacturing Lab.

References

Mallick

. Fiber-reinforced composites: materials, manufacturing, and design. Boca Raton: CRC press, 2007.

Epple

Bonten

. Production of continuous fiber thermoplastic composites by in-situ pultrusion. AIP Conf Proc 2014; 1593: 454–457.

Gibson

Rosen

Stucker

. Additive manufacturing technologies: 3D printing, rapid prototyping, and direct digital manufacturing, 2nd ed. New York: Springer, 2015.

Chua

Leong

Lim

. Rapid prototyping: principles and applications (with companion CD-ROM), 3rd ed. Singapore: World Scientific, 2010.

Shofner

Lozano

Rodríguez-Macías

, et al. Nanofiber-reinforced polymers prepared by fused deposition modeling. J Appl Polym Sci 2003; 89: 3081–3090.

Zhong

Zhang

, et al. Short fiber reinforced composites for fused deposition modeling. Mater Sci Eng A 2001; 301: 125–130.

Ning

Cong

Qiu

, et al. Additive manufacturing of carbon fiber reinforced thermoplastic composites using fused deposition modeling. Compos B Eng 2015; 80: 369–378.

Tekinalp

Kunc

Velez-Garcia

, et al. Highly oriented carbon fiber-polymer composites via additive manufacturing. Compos Sci Technol 2014; 105: 144–150.

Matsuzaki

Ueda

Namiki

, et al. Three-dimensional printing of continuous-fiber composites by in-nozzle impregnation. Sci Rep 2016; 6: 1–7.

10.

Liu

. Rapid prototyping of continuous carbon fiber reinforced polylactic acid composites by 3D printing. J Mater Process Technol 2016; 238: 218–225.

11.

Brooks

Tyas

Molony

. Tensile and fatigue failure of 3D printed parts with continuous fibre reinforcement. Int J Rapid Manuf 2017; 6: 97.

12.

Markforged Inc., Watertown, MA, USA. High Strength 3D Printing. https://markforged.com (accessed 14 August 2018).

13.

Dickson

Barry

McDonnell

, et al. Fabrication of continuous carbon, glass and Kevlar fibre reinforced polymer composites using additive manufacturing. Addit Manuf 2017; 16: 146–152.

14.

der Klift

Koga

Todoroki

, et al. 3D printing of continuous carbon fibre reinforced thermo-plastic (CFRTP) tensile test specimens. Open J Compos Mater 2016; 06: 18–27.

15.

Melenka

Cheung

BKO

Schofield

, et al. Evaluation and prediction of the tensile properties of continuous fiber-reinforced 3D printed structures. Compos Struct 2016; 153: 866–875.