Experimental Analysis of Fiber Reinforcement Rings’ Effect on Tensile and Flexural Properties of Onyx™–Kevlar® Composites Manufactured by Continuous Fiber Reinforcement

Additive manufacturing of composite materials is progressing in the world of 3D printing technologies; composite materials allow the combination of the physical and mechanical properties of two or more constituents to create a new material that meets the required properties of several applications. In this research, the impact of adding Kevlar® reinforcement rings on the tensile and flexural properties of the Onyx™ (nylon with carbon fibers) matrix was analyzed. Parameters such as infill type, infill density and fiber volume percentage were controlled to determine the mechanical response in tensile and flexural tests of the additive manufactured composites. The tested composites showed an increment of four times the tensile modulus and 1.4 times the flexural modulus of pure Onyx™ matrix when compared with that of the Onyx™–Kevlar®. The experimental measurements demonstrated that Kevlar® reinforcement rings can increase the tensile and flexural modulus of Onyx™–Kevlar® composites using low fiber volume percentages (lower than 19% in both samples) and 50% of rectangular infill density. However, the appearance of some defects, such as delamination, was observed and should be further analyzed to obtain products that are errorless and can be reliable for real functions as in automotive or aeronautical industries.


Introduction
Composite materials have revolutionized the world of manufacturing processes with the ability of blending the mechanical and physical properties of the constituents, creating a new material [1]. Composite materials consist of two or more constituents, a soft and continuous matrix and a discontinuous and hard reinforcement. The matrix is responsible for distributing the loads applied along the reinforcements, and reinforcements withstand the loads [1,2]. The selection of constituents and the distribution gives the material specific mechanical properties such as tensile modulus, flexural modulus and impact resistance, among others [2].
Cofaru et al. [60] analyzed the tensile properties of Onyx™ composites reinforced with carbon, glass and Kevlar ® fibers; the lowest strain was obtained using carbon fibers, followed by Kevlar ® and finally glass fibers, and the highest elastic modulus was obtained with carbon, followed by Kevlar ® and finally glass fibers. Papa et al. [26] analyzed Onyx™ reinforced with different carbon fiber directions, obtaining a higher response to forces using unidirectional fibers at 0 • , in comparison with the 0 • /90 • arrangement. Ansari and Kamil [61] also tested the tensile properties of Onyx™ composites reinforced with carbon, glass and Kevlar ® fibers; the results showed that the fiber angle is the one that is responsible for the ultimate tensile strength behavior, and the highest tensile was obtained using unidirectional isotropic direction. Ojha et al. [63] observed that adding fibers would increase elastic modulus and stiffness but adding more than eight layers of fibers would result in a decrease in elastic modulus and stiffness.
The presented studies worked with different combinations and analyzed the mechanical response of a particular CFRC printing configuration because Markforged Inc. only reported the constituent materials properties separately. The capability of adding the desired proportions of matrix and fiber opens the possibility of varying printing parameters and obtaining the desired mechanical behavior of a CFRC product.
The use of an internal geometry in AM products results in lightweight products that can maintain certain mechanical properties, translating to material optimization. The impact of only using reinforcement rings in combination with rectangular infill geometry has not been studied profoundly. The presented studies use samples with solid or triangular infill and fiber reinforcement infill using different fiber orientations with their respective reinforcement rings. Therefore, this research focuses on the study of the impact of using rectangular infill geometry and only the Kevlar ® reinforcement rings, reducing the percentage of fiber (lower than 19% of fiber volume) and material utilized to create the Onyx™-Kevlar ® samples to obtain the mechanical properties and analyze the effect of adding the Kevlar ® reinforcement rings to Onyx™, for comparison with the obtained properties with pure Onyx™ results published by Markforged [64].

Materials
The CFRC samples were fabricated with Onyx™ as the thermoplastic matrix constituent and Kevlar ® as the fiber reinforcement constituent. Onyx™ and Kevlar ® fiber filaments have a diameter of 1.75 mm and 0.3 mm, respectively, and they were supplied by Markforged Inc., Watertown, MA, USA. The mechanical properties of Onyx™ and Kevlar ® filaments are described in Table 1. The samples were manufactured in a Mark Two™ 3D printer by Markforged Inc., Watertown, MA, USA. This printer works with a dual nozzle, as seen in Figure 1. The first nozzle deposits the thermoplastic matrix, and the second nozzle deposits the fiber reinforcement in the specified zones, substituting the infill geometry selected for the piece. The printer requires a special slicer software called "Eiger™", by Markforged ® , Inc. In this software, the printing parameters to achieve the 3D-printed samples are set.

Property
Onyx™ The samples were manufactured in a Mark Two™ 3D printer by Markforged Inc., Watertown, MA, USA. This printer works with a dual nozzle, as seen in Figure 1. The first nozzle deposits the thermoplastic matrix, and the second nozzle deposits the fiber reinforcement in the specified zones, substituting the infill geometry selected for the piece. The printer requires a special slicer software called "Eiger™", by Markforged ® , Inc. In this software, the printing parameters to achieve the 3D-printed samples are set.  The testing method, the test standard, fiber orientation, infill geometry, fiber volume and reinforcement rings in the samples used in this research are summarized in Table 2. The samples used for tensile and bending tests are presented in Figure 3. The testing method, the test standard, fiber orientation, infill geometry, fiber volume and reinforcement rings in the samples used in this research are summarized in Table 2. The samples used for tensile and bending tests are presented in Figure 3.   The testing method, the test standard, fiber orientation, infill geometry, fiber and reinforcement rings in the samples used in this research are summarized in T The samples used for tensile and bending tests are presented in Figure 3.

Mechanical Analysis
To characterize the mechanical properties of the Onyx™-Kevlar ® composite, tensile and flexural tests were performed in an MTS Insight machine by MTS Systems Corporation, Eden Prairie, MN, USA and an MTS-634.25 axial extensometer with 50 mm (2 in.) Gage Length was used to measure the strain in the specimens. Tensile tests were conducted following ASTM D3039 standards and flexural tests following ASTM D790 standards. Five repetitions of each mechanical test were performed to obtain reliable and repeatable results, as marked in ASTM D3039 and ASTM D790 standards. For tensile tests, the samples were placed between a pair of clamps, presented in Figure 4a. These clamps have a hydraulic system that allows the user to adjust the holding pressure to ensure the samples are clamping; the clamps were configured to apply a pressure of 150 psi to hold the tensile samples in place. To ensure a holding distance that would allow correct measurements in the tensile tests, prior trials were made to have the separation of the clamps that would secure the pieces in place and that would allow the rupture in a mid-zone of the CFRC samples; this distance was 80 mm between clamps. Then, a tensile load of 100 KN at a rate of 2 mm/min (as mentioned in ASTM D3039) was applied to the samples until failure at room temperature. where represents the position of the ℎ load and displacement values to be substituted in Equation (5).

Dehumidification Process
Mechanical characterization results can be affected by the humidity absorption of nylon polymer and interlayer zone imperfections [48]. Onyx™ is made of nylon; it absorbs humidity from ambient [66,67], which is why it is stored in a dry-box in the whole manufacturing process [50,53,59,63]. The printed samples require a dehumidification that is achieved with a dry thermal process in an electrical furnace to eliminate the humidity absorbed during the printing process [42,48,50]. The electrical furnace used was TE-M20AT from Terlab, Escondido, CA, USA. The samples were heated at 70 °C for 4 h to eliminate the humidity in Onyx™. Some researchers have stored the samples in sealed bags to prevent changes caused by the environment [60]; in this research, the treated samples were stored in a desiccator to reduce the mechanical test variation results produced by moisture in the samples [67].

Density Analysis
Onyx™-Kevlar ® composite samples were subjected to a density analysis in a Metter Toledo XS104 Analytical Balance at room temperature. To determine the real density of the samples, standard ASTM D792 was followed using the water displacement method. The density of the tested samples was calculated to determine the specific tensile and flexural modulus considering the fiber volume percentage and infill density of the samples.

Tensile Properties
The tensile properties of the five 3D-printed composite samples are represented in Figure 5 where the strain-stress curves are plotted. Figure 5 also displays a picture of the samples used and the failure mechanism at . The samples' designation is KFT, where KFT means Kevlar ® Fiber Tensile, followed by the sample test number from one to five. The tensile test allowed us to characterize the response of the composite to the load applied in the axial direction of the samples. The results of the tensile test were analyzed using the Hooke's general law to obtain the elastic modulus of the composite material by using Equation (1): where σ max is the maximum stress applied to the sample, E T is the elastic modulus of the composite material and ε is the elastic strain at σ max , which was obtained using Equation (2), The E T of the Onyx™-Kevlar ® composite can be determined by solving for the slope in the linear section of the stress-strain curve using Equation (3), where i represents the position of the i th stress and strain values to be substituted in Equation (3). The flexural test selected was a three-point bending test, as seen in Figure 4b. Before the tests were conducted, the samples were measured in order to obtain the values for the geometrical parameters used in Equation (4). The samples were placed in two supports that had a separation of 50 mm. The separation between supports was calculated by multiplying the samples' width times 16, as mentioned in ASTM D695. Then, a load of 1 KN at a rate of For bending test results, the flexural modulus E f can be obtained using Equation (4) [65].
where L is the span length, b is the sample width and d is the beam thickness. The values of m were obtained by finding the slope of the straight-line segment of load-displacement curves, using Equation (5) [65].
where i represents the position of the i th load and displacement values to be substituted in Equation (5).

Dehumidification Process
Mechanical characterization results can be affected by the humidity absorption of nylon polymer and interlayer zone imperfections [48]. Onyx™ is made of nylon; it absorbs humidity from ambient [66,67], which is why it is stored in a dry-box in the whole manufacturing process [50,53,59,63]. The printed samples require a dehumidification that is achieved with a dry thermal process in an electrical furnace to eliminate the humidity absorbed during the printing process [42,48,50]. The electrical furnace used was TE-M20AT from Terlab, Escondido, CA, USA. The samples were heated at 70 • C for 4 h to eliminate the humidity in Onyx™. Some researchers have stored the samples in sealed bags to prevent changes caused by the environment [60]; in this research, the treated samples were stored in a desiccator to reduce the mechanical test variation results produced by moisture in the samples [67].

Density Analysis
Onyx™-Kevlar ® composite samples were subjected to a density analysis in a Metter Toledo XS104 Analytical Balance at room temperature. To determine the real density of the samples, standard ASTM D792 was followed using the water displacement method. The density of the tested samples was calculated to determine the specific tensile and flexural modulus considering the fiber volume percentage and infill density of the samples.

Tensile Properties
The tensile properties of the five 3D-printed composite samples are represented in Figure 5 where the strain-stress curves are plotted. Figure 5 also displays a picture of the samples used and the failure mechanism at σ max . The samples' designation is KFT, where KFT means Kevlar ® Fiber Tensile, followed by the sample test number from one to five.
The results of the tensile test, summarized in Table 3, are the individual elastic properties of Onyx™-Kevlar ® composites tested in this research and the mean values.
The Kevlar ® reinforcement rings showed an increment in tensile response. All the samples presented higher values, with a mean value of 9.57 GPa in tensile modulus, in comparison with the 2.4 GPa tensile modulus of pure Onyx™ published by Markforged [64]. This increment is the result of the arrangement of the fibers unidirectionally aligned parallel to the same axis to which the tensile force was applied, as mentioned in [45,50,53,61]. The unidirectional fibers withstand higher tensile stress in comparison with different orientation of reinforcement fibers. This is due to the distribution of stresses along the axis on which the fibers are deposited, having minimal or no significant shear stresses to affect tensile performance. The results of the tensile test, summarized in Table 3, are the individual elastic properties of Onyx™-Kevlar ® composites tested in this research and the mean values. The Kevlar ® reinforcement rings showed an increment in tensile response. All the samples presented higher values, with a mean value of 9.57 GPa in tensile modulus, in comparison with the 2.4 GPa tensile modulus of pure Onyx™ published by Markforged [64]. This increment is the result of the arrangement of the fibers unidirectionally aligned parallel to the same axis to which the tensile force was applied, as mentioned in [45,50,53,61]. The unidirectional fibers withstand higher tensile stress in comparison with different orientation of reinforcement fibers. This is due to the distribution of stresses along the axis on which the fibers are deposited, having minimal or no significant shear stresses to affect tensile performance.
The 3D printing process has natural variations in the printed products. For example, the KFT5 sample showed a lower maximum strain compared with the other tensile samples; this can be attributed to defects in the manufacture of the part, since it is known that the technologies for creating CFRC are not yet reliable enough to have a product that always fulfills the same physical and mechanical characteristics due to printing defects that can lead to premature failure [29,56,68].  The 3D printing process has natural variations in the printed products. For example, the KFT5 sample showed a lower maximum strain compared with the other tensile samples; this can be attributed to defects in the manufacture of the part, since it is known that the technologies for creating CFRC are not yet reliable enough to have a product that always fulfills the same physical and mechanical characteristics due to printing defects that can lead to premature failure [29,56,68].
The obtained E T was compared with other research found in the literature; the different CFRC samples tested and the obtained E T are presented in Table 4. The obtained E T was achieved with a reduced fiber volume in comparison with other studies [39,44,50,53,60,61,63], who obtained their maximum results using 25.8% to 60% of fiber volume in comparison with the 18.77% used in this research.

Flexural Properties
The flexural test results are shown in Figure 6 plotted in a load-displacement curve, with a picture of the samples used and the deformation after maximum displacement was reached. The samples' designation is KFB, where KFB means Kevlar ® Fiber Bending, followed by the sample test number from one to five.
The obtained was compared with other research found in the literature; the different CFRC samples tested and the obtained are presented in Table 4. The obtained was achieved with a reduced fiber volume in comparison with other studies [39,44,50,53,60,61,63], who obtained their maximum results using 25.8% to 60% of fiber volume in comparison with the 18.77% used in this research.

Flexural Properties
The flexural test results are shown in Figure 6 plotted in a load-displacement curve, with a picture of the samples used and the deformation after maximum displacement was reached. The samples' designation is KFB, where KFB means Kevlar ® Fiber Bending, followed by the sample test number from one to five. The results of the flexural test, summarized in Table 5, are the individual flexural properties of Onyx™-Kevlar ® composites tested in this research and their mean values. The results of the flexural test, summarized in Table 5, are the individual flexural properties of Onyx™-Kevlar ® composites tested in this research and their mean values.
The Kevlar ® reinforcement rings also have a positive increasing effect on the flexural modulus of the Onyx™-Kevlar ® composite. The flexural modulus increases from 3.0 GPa for pure Onyx™ [64] to 4.11 GPa in mean for Onyx™ with Kevlar ® reinforcement rings. This increment is also due to the direction of the fibers, since the fibers along the length of the samples help to distribute the stresses along the sample geometry, having a distribution from the center of the sample, where the load is applied, towards the supports of the three point bending test. The obtained E F , indicated with a rectangle in Figure 6, was compared with other research found in the literature. The different CFRC samples tested and the obtained E F are presented in Table 6. The bending properties of Onyx™-Kevlar ® are not as high as Onyx™-carbon products, where the E T is higher than 24 GPa, but it was demonstrated that the addition of Kevlar ® reinforcement rings can improve the mechanical response of CFRC products to bending forces. Table 6. Flexural modulus comparison with other CFRC research.

Printing Defects
After experimental tests, the samples were inspected; different type of failures and repetitive printing defects in tensile and flexural samples were found. In Figure 7, the different types of failures presented in tensile test samples can be seen. The failure modes were compared with the ASTM D3039 failure modes atlas to determine a qualitative standardized comparison. The samples presented in the matrix present a premature failure due to delamination, as in [52], leaving all the load to the Kevlar ® fibers until they also reach and overpass the σ max of fibers.
In Figure 8, the different types of defects and failure mechanisms in flexural samples are shown. In these samples, fracture occurs in the same direction that the load was applied. In the samples KFB2, KFB3 and KFB4, delamination was present after maximum deflection was reached. In the samples KFB2 and KFB4, the fibers came out of the matrix and were exposed to the environment due to poor bonding between the matrix and fibers.
The Kevlar ® fibers in the tensile test suffer pull-out from the matrix, as seen in Figure 7a,b, due to an interlaminar failure, as mentioned in [46]. The failures and defects presented in Figures 7 and 8 show that the matrix and fibers have poor bonding; this can be the result of air bubbles or voids in the samples, as mentioned in [41,52,53,70]. The delamination presented in Figure 8 is described in the literature as poor adhesion of the different layers of matrix and reinforcement, an effect that is typically present in the FFF and CFRC processes [41]. The defects and failure mechanisms that were observed in this research are similar to the ones reported in [38,39,56].
CFRC samples can be affected by the presence of voids, lack of molding conditions and low impregnation causing a low adhesion between the matrix and fibers, resulting in lower mechanical properties compared with traditionally manufactured composites [34,71,72]. Voids can be reduced using the postprocessing technique presented in [68], which consists of placing the 3D-printed composite into a mold and applying heat and pressure to it, to reduce the voids between matrix filaments and matrix-fiber bonding. The poor matrix interface will cause a reduction in mechanical performance [70]. In Figure 8, the different types of defects and failure mechanisms in flexural samples are shown. In these samples, fracture occurs in the same direction that the load was applied. In the samples KFB2, KFB3 and KFB4, delamination was present after maximum deflection was reached. In the samples KFB2 and KFB4, the fibers came out of the matrix and were exposed to the environment due to poor bonding between the matrix and fibers. The Kevlar ® fibers in the tensile test suffer pull-out from the matrix, as seen in Figure  7a,b, due to an interlaminar failure, as mentioned in [46]. The failures and defects presented in Figures 7 and 8 show that the matrix and fibers have poor bonding; this can be the result of air bubbles or voids in the samples, as mentioned in [41,52,53,70]. The delamination presented in Figure 8 is described in the literature as poor adhesion of the different layers of matrix and reinforcement, an effect that is typically present in the FFF and CFRC processes [41]. The defects and failure mechanisms that were observed in this research are similar to the ones reported in [38,39,56].  In Figure 8, the different types of defects and failure mechanisms in flexural samples are shown. In these samples, fracture occurs in the same direction that the load was applied. In the samples KFB2, KFB3 and KFB4, delamination was present after maximum deflection was reached. In the samples KFB2 and KFB4, the fibers came out of the matrix and were exposed to the environment due to poor bonding between the matrix and fibers. The Kevlar ® fibers in the tensile test suffer pull-out from the matrix, as seen in Figure  7a,b, due to an interlaminar failure, as mentioned in [46]. The failures and defects presented in Figures 7 and 8 show that the matrix and fibers have poor bonding; this can be the result of air bubbles or voids in the samples, as mentioned in [41,52,53,70]. The delamination presented in Figure 8 is described in the literature as poor adhesion of the different layers of matrix and reinforcement, an effect that is typically present in the FFF and CFRC processes [41]. The defects and failure mechanisms that were observed in this research are similar to the ones reported in [38,39,56]. Another effect that can impact the mechanical response in the samples used in this research is the fiber waviness, a phenomenon that creates variations in the out-of-plane zones where the continuous fiber is deposited, altering the distribution of loads along the fibers in the different layers of the 3D-printed product [73,74].

Density Test
Three representative samples of the five used in tensile and flexural tests were subjected to density analysis to determine the specific elastic and flexural modulus of the samples considering matrix, reinforcement and infill selected for the Onyx™-Kevlar ® composite. The results of these density analyses were used to calculate the specific elastic modulus and the specific flexural modulus of the Onyx™-Kevlar ® composites. The mean density value for tensile samples was 1.028 g/cm 3 and the mean density value for bending samples was 1.033 g/cm 3 . These results gave a specific elastic modulus of 9.31 GPa·cm 3 /g and a specific flexural modulus of 3.98 GPa·cm 3 /g. The variation in den-sity can be attributed to variation in the printing process as over or under extrusion of the matrix materials, even though the samples were printed using the same equipment.

Conclusions
Tensile and bending tests were carried out to determine the effect of fiber reinforcement rings on the tensile and flexural properties of Onyx™-Kevlar ® 3D-printed composites.

•
The addition of 18.77% of fiber volume as Kevlar ® reinforcement rings to Onyx™ with 50% of rectangular infill resulted in an increment of almost 400% of pure Onyx™ elastic modulus. A value of 9.57 GPa was obtained with lower fiber volume compared with other studies that go from 1.3 GPa to 4.95 GPa using 25.8% to 60% of fiber volume.

•
The addition of 17.25% of fiber volume as Kevlar ® reinforcement rings to Onyx™ with 50% of rectangular infill resulted in an increment of almost 140% of pure Onyx™ flexural modulus. A flexural modulus of (4.11 GPa) was obtained, and similar results (4.73 GPa) were obtained using 19% of fiber volume and triangular infill. • A specific elastic modulus of 9.31 GPa·cm 3 /g was obtained and a specific flexural modulus of 3.98 GPa·cm 3 /g was obtained for Onyx™-Kevlar ® CFRC.
After experimental tests, some defects were registered; the effect of defects on composite printed parts requires further investigation to improve surface quality of the printing samples. In the sample fracture mechanism, a weak adhesion in the interlayer zone of the matrix and reinforcement was observed, so further investigations are needed to control the interlayer adhesion. The presence of the described defects in the CFRC samples leads to the conclusion that further analysis needs to be conducted in the FFF technology for composite materials, since the defects decrease the reliability of CFRC products and thus diminish its early use in the industry as a replacement for traditionally manufactured composite materials.
With the presented results, it can be concluded that the reinforcement rings occupied a low volume in the 3D-printed product geometry. These reinforcement rings can be used in industrial applications such as aeronautics or the automotive field, where lightweight structures are designed to withstand the mechanical stresses to which products in these fields are subjected. However, more research is required in the study of reinforcement rings, considering more levels in the analysis, by analyzing the continuous addition of reinforcement rings used with the aim of optimizing the materials used but fulfilling the required mechanical properties for certain applications.
The printing parameters that the Mark Two printer have are limited for the user to modify them; thus, more options to vary in printing parameters can help in the improvement of 3D-printed composite materials and achieve similar mechanical properties of traditionally made composites.