Effects of a new filling technique on the mechanical properties of ABS specimens manufactured by fused deposition modeling

The spread of 3D printing in many different fields has become eminent. This paper aims to improve the mechanical properties of parts printed by fused deposition modeling technique. Acrylonitrile butadiene styrene (ABS) specimens are printed with custom printing parameters. These parameters give a tensile strength that is 86% of the injection- molded ABS strength, and give one of the best recorded results for 100% infill printed ABS tensile specimens. Furthermore, a post filling technique has been studied. Specimens are printed with inner voids and different densities using slicing software. Void shape is precisely selected to conform to the filling process. High-strength, low-cost thermoset resin is injected through specimens to fill those voids. A tensile test has been performed after the full curing of the resin. A morphology analysis is done. Using this technique strength to printed weight ratio is improved by 151% and the cost is reduced by 51%.


Introduction
Additive manufacturing (AM) is the best choice to join the fourth industrial revolution, as it delivers complex and customized products with great freedom in design in a very short time, reducing the production cost, and shortening the development cycle [1][2][3][4]. Stereo-lithography (SLA) is the first technique that AM started with, which is a liquid-based raw material technique. Later on, many techniques were invented varying in the raw material from liquid, to powder, and solid [5]. All AM techniques use three-dimensional computer-aided designs (CAD) to produce -layer by layer -three-dimensional products by selectively adding materials, with variety in the materials being used. Fused deposition modeling (FDM), solid-based AM technique, is the most common and widely used technique due to its simplicity and low price as it utilizes low-cost polymers only [5,6]. FDM feeds thermoplastic filament into a heating chamber until it reaches a semi-liquid state, extrudes it through a nozzle above its melting point, then deposits it on the printing platform in specified positions according to the G-code generated by the machine's software [7]. FDM printed parts have been in use in many fields like rapid prototyping, building, healthcare devices and tools, motor drives, and aerospace components [8]. But due to the nature of the FDM technique, it experiences some drawbacks; one of them is the limitation in mechanical properties. This limitation is due to the weak and non-homogenous bonding between the corresponding layers building up the part and also between the roads building the layer itself. Also, since the cooling and heating rates in FDM are unstable, there would be uncontrolled shrink and defects in the microstructure [5,9].
Many studies have been performed to overcome the mentioned problems, covering different materials that are currently used in FDM (ABS, PLA, PA, NYLON, etc.). Many studies have considered printing parameters as important indices for product quality. They studied the influence of printing orientation, speed, raster angle, infill pattern, nozzle temperature, and layer height on the mechanical properties of the printed parts [10][11][12]. Table 1 summarizes the studies done by researchers on 3D printed ABS parts applying different printing parameters.
Other studies adopted the potential of improving physical and mechanical properties of the raw material itself. Nanoparticles were prepared and mixed with the printing filament to produce a new one [9,[21][22][23]. Some researches focused on improving the overall mechanical properties of the 3D printed part by decreasing its anisotropy or increasing its isotropic behavior [24,25]. They replaced the thermoplastic polymers that result in poor chemical and thermal properties with two-component epoxy resin [24], epoxy resin has good chemical and physical properties and gives high-strength bonding, with light weight structures withstanding high static loads [26][27][28]. Filippova et al. [25] impregnated ABS 3D printed samples in epoxy resin compound with different hardeners. 142% and 133% were achieved as an improvement in ultimate tensile strength depending on the hardener type. Instead of impregnation, Belter and Dollar [29] injected epoxy resin through ABS flexural specimens. This resulted in improving the overall part stiffness and strength up to 25% and 45%, respectively. Moreover, Jiang et al. [24] used epoxy as the main raw material in FDM technique to get the best use of its mechanical properties and print products that can be used in harsh conditions. Tensile specimens were printed of a mixture of only epoxy and CNT giving a tensile strength of 55 MPa.
In this paper, ABS tensile specimens are printed with printing parameters based on the survey presented in Table 1. A comparison between the results of many previous studies is held to decide on the best parameters to print with. Also, using these printing parameters, other tensile specimens are printed to test a new filling technique. This technique is a post processing technique that aims to improve the mechanical properties and reduce cost and weight. Hence, the parts are printed with inner voids which reduce the part's weight, printing time, and in turn printing cost. Afterwards, these voids are injected with low-cost, less dense resin which possesses higher mechanical properties. To make the part sparse inside, voids could be designed and located precisely during design stage by CAD software, or by slicing software which manages this process through infill density parameter [29]. In this study, the parts are made sparse by slicing software, three infill densities are printed (20%, 40%, and 60%) and injected with epoxy resin that is 91% cheaper than the ABS material. The infill pattern chosen is gyroid pattern in order to let the resin spread through the entire part. Tensile tests and morphology analysis are performed to investigate the improvement of this technique.

Materials and samples
In this study, ABS (acrylonitrile butadiene styrene) had been used as a filament material; it was obtained from  Table 2 shows the resin properties. Tensile specimens were designed according to ASTM638 type I [31], and printed on Ultimaker 3 with nozzle diameter 0.4 mm and Cura 4.8 slicing software in ideaspace company as shown in (Fig. 1).

Printing 100% infill specimens
In this study, 100% infill ABS tensile specimens were printed with printing parameters that were precisely selected based on the literature in Table 1; the selected parameters are believed to give the best results as they were selected based on the previous experimental work done on this material regarding the results obtained in every study. Table 3 summarizes the parameters used in the present study. After printing, tensile test was performed, and all tensile tests in this research had been performed on German Zwick testing machine Z010.

Filling technique
In this technique, tensile specimens were printed with inner voids to be injected with epoxy resin. ABS specimens were sparse inside using the slicing software. The software controls how dense the part is by infill percentage parameter. Three percentages were printed 20%, 40%, and 60% with a pattern that allowed the resin to flow through the whole specimen, and two holes of 0.4 mm for injection and venting were drilled in the specimens (Fig. 2).
To identify the mechanical properties of the epoxy resin being used, tensile test was done on deco-Pox039 Epoxy resin, two components A and B were mixed together under the predefined percentages, then the mix was poured into a custom-made silicone rubber mold in (Fig. 3a) and was let for 7 days to complete the full curing as written in the product data sheet; the test was done following ASTM638 type I.
After testing the epoxy specimens and confirming its properties, it was approved to be used as an injection material. Epoxy injection within the ABS printed tensile specimens was done using a syringe. The syringe was immersed inside the specimen through one of the holes and was sealed very well. The other hole was for venting. The syringe used was G16 to suit the highly viscous epoxy. The injection process was in vertical orientation, the syringe was immersed at the bottom, and the direction of the injection was against gravity  to make sure the epoxy filled the whole specimen by reaching the top. Figure 4 shows the direction of injection. Figure 5 represents sections at different locations in a specimen to make sure that epoxy flowed through the whole specimen. A tensile test was done -after 7 days from the injection of specimens, the full curing time -on German Zwick testing machine Z010, with 10 KN calibrated load cell, and crosshead speed of 5 mm min −1 . At least three samples for each case were tested. Morphology analysis on the fractured specimens had been done, as the fracture surfaces were observed via scanning electron microscope (SEM, JEOL JSM_5400LV). All the surfaces were gold-sputtered before observation, and the observation was at an acceleration voltage of 15 kV.

Results of epoxy resin test
The mechanical behavior of deco-Pox039 Epoxy resin is shown in (Fig. 6); the tensile test was done on four specimens to take the average and give the most accurate result. The epoxy used in this paper has a resin of aromatic bisphenol (bisphenol-A) and a hardener of amino methylamine and phenyl methylamine. The ultimate tensile strength obtained was 44.8 ± 1.1 MPa, and elongation at break was of 10.3 ± 0.03%. This result is higher than the tensile strength obtained by Filippova et al. [25], and  Figure 7 represents the result of the tensile test on 100% infill printed ABS specimens with the parameters in Table 3. Three specimens were tested, elongation at break was 7.4%, and tensile strength was 36.3 ± 0.81 MPa. This result is considered one of the best results obtained for 100% infill ABS tensile specimens. Figure 8 shows a comparison between the result of the present study, and most of the results obtained in the studies mentioned in Table 1. The variation in results  ensures the influence of printing parameters on the mechanical properties of the printed part. Table 4 compares the printing parameters used in the present study to those used by T. Coogan and Kazmer [16] as they got the closest results to this study's results. Both studies have very close parameters. The main difference is in layer height parameter as 0.2 mm layer height adopted in this research is believed to be the best height fulfilling the best boding between adjacent roads and consecutive layers [35,36].

Results of filling ABS test
The effect of epoxy resin injection within ABS printed parts is represented in (Fig. 9). All printing parameters, injection materials, and methods were controlled; the variables were the infill density of ABS while printing, and in turn the injected volume of epoxy.  As can be seen, the results are very close, average tensile strength for 20% ABS infill-80% epoxy injection (20B-80E) is 24.77 MPa, for 40% ABS infill-60% epoxy injection (40B-60E) is 27.03 MPa, and 21.8 MPa for 60% ABS infill-40% epoxy injection (60B-40E). Although the strength of 40% infill is marked the best, by analyzing the weight and corresponding printing time, (20B-80E) gave the optimum strength to printed weight ratio Table 5. Furthermore, the preference of 20% infill ABS injected with epoxy to the pure 100% infill printed ABS was investigated, and it was concluded that by reducing 56% of the part weight − in 20% infill -, the product mechanical strength reaches 68% of the printed part with 100% infill, and the strength to printed weight ratio is increased by 151% (Table 5). Strength to printed weight ratio is analyzed in this study rather than the common strength to weight ratio, because the injected epoxy resin has the same density of ABS filament used -1.1 g cm −3 . Some other thermoplastic filament materials like PLA which possesses higher density (1.17-1.24 g cm −3 ) [11]   results when injected with similar epoxy types. Results of the previous tests indicate adequate adhesion between epoxy and ABS. Since the effective surface is considered the main parameter to give best adhesion, and one of the techniques to enhance the effective surface is roughening it but to an acceptable level. As if the surface is more rough the adhesion stress would decrease [37]. And hence one of the themes of FDM process is the moderate surface finish [38,39], so that would fulfill good adhesion between epoxy and ABS. Figures 10 and 11 represent sections in 20%, 40%, 60%, and 100% infill specimens after injection and testing.

Morphology analysis
Results for morphology analysis are shown in Fig. 12; it illustrates the bonding between the adjacent roads of ABS filament in the same layer, and shows the bonding between the consecutive ABS layers. The first remark is that road cross-sections are not circular anymore and that is due to the compression induced from the layers on one another. The bonding between adjacent roads is on more than 50% of its height, when notice the lines in Fig. 12a, it could be concluded that the bonding height is between 90 and 170 μm. The gaps between consecutive layers are represented by lines in Fig. 12b, and the gap height is between 18 to 21 μm. The trans-granular cleavage shown in each road cross-section indicates the brittle fracture mode of the material (Fig. 12c). Figure 13 shows SEM micrographs of an overview for three sections in (20B-80E), (40B-60E), and (60B-40E) specimens (a, b, and c), respectively. In the micrographs of the injected ABS specimens, the contact surfaces between epoxy and ABS could be marked indicating super adhesion. Small discrete surface areas inside ABS specimens that were designed by printing software helped to make it fully composite as the epoxy was immersed through the whole part (Fig. 14), representing the three coated sections before being examined in the microscope. By going deep in (20B-80E) specimen (Fig. 15), there are three levels of the ABS road as the mechanical fracture leaves unpaved surface in Fig. 15b; in Fig. 15c, there is a perfect bond between epoxy and ABS with a vital defect in epoxy resin mix indicating some air bubbles; Fig. 15d, e indicate the presence of more air bubbles; beach marks appeared to ensure the brittle fracture; also, different levels of fracture in both epoxy and ABS appeared.
The same phenomena were recognized in Figs. 16 and 17 representing (40B-60E) and (60B-40E) specimens, respectively. Figure 16b is a magnified image of Fig. 16a where the brittle fracture is obvious, and also, it refers to discontinuity in injection. Figure 16d is a magnified image of Fig. 16c, and it shows 50 and 35 μm air bubbles in epoxy mix (Fig. 17a) representing the fracture; it was just before a relatively large air bubble, and also, there is another one that collapsed by the fracture; Fig. 17b represents epoxy that covers some sheared filaments.

Conclusions
Parts printed by FDM technique have some limitations in the mechanical properties. In this study, it was proofed that printing parameters highly affect the tensile strength. ABS tensile specimens were printed by FDM technique and evaluated according to ASTM638 giving one of the best recorded results, with an ultimate strength of 36.3 MPa. Also, this work was devoted to study the effect a new compositing technique on mechanical properties. Parts printed with different infill densities (20%, 40% and 60%) and injected with epoxy. This technique resulted in an improvement in strength to printed weight ratio by 151% in case of 20% infill and 129% in case of 40% infill specimens, but there was not any significant improvement in case of 60% infill specimens. The improvement in properties is directly proportional to the volumes of the injected material and ABS cavities. As in parts with 20% infill, large cavities allowed the viscous epoxy to fill the entire volume. On the contrary, in 60% infill, cavities were relatively small which justifies the results obtained. Morphology analysis showed good adhesion between epoxy and ABS cavities, with the presence of some air bubbles. Based on the present findings, parts printed with the adopted printing parameters and composited by the presented filling technique could be applied in a wide range of applications.
Author contribution All authors contributed to the study conception and design. Material preparation and data collection were performed by Heba Hussam. Data analysis was performed by Heba Hussam and Yasser Abdelrhman. The first draft of the manuscript was written by Heba Hussam. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB).

Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.