PSI - Issue 61

Albert E. Patterson et al. / Procedia Structural Integrity 61 (2024) 148–155 Author name / Structural Integrity Procedia 00 (2024) 000–000

149

2

and fuses discrete elements of material to form the part layers [1–4]. Therefore, the layout pattern of the elements can be expected to greatly a ff ect the fracture toughness of the materials as it determines the available crack paths in the fundamentally structured material. In addition, the part shell (layer outline) will have an e ff ect the fracture whenever a crack begins at or reaches a layer boundary as seen with printed notches; this consideration is an important driver for selecting part geometry and notching method. The near-net-shape design freedom provided by ST-AM processes such as FFF allows the development of man ufacturable metamaterials, smart materials, functionally-graded materials, lattice structures, accurate topology opti mization solutions, and other di ffi cult and sensitive design problems. While significant manufacturability constraints do exist, the fundamental structured nature of ST-AM process o ff ers many interesting design opportunities. These constraints, the material behavior during processing, the e ff ects of voids and other defects, and the elastic-plastic be havior of the AM polymer materials are all well understood and easily interpreted and applied during design. This is widely known in the existing design and AM literature, but is especially clear in the studies that dealt with the design of structured materials such as those by Allum et al . [5], Peng et al . [6], Lenti [7], Gardan et al . [8, 9], Lanzillotti et al . [10, 11], Djouda et al . [12], McLouth et al . [13], Patterson et al . [1–3, 14, 15], and Hart et al . [16]. All of these works either focused on automated design, manual layouts, or on a single material. To further explore the problem, the present study examines three materials simultaneously under the same con ditions and geometries. The focus of this work is on parametric design using only the existing well-known material layouts and notching methods. This work contributes to conversation on AM-enabled structured materials by pro viding a comparative dataset, by examining the e ff ect of notching method, and by making conclusions about several materials within the same study. This work will complement the existing works, helping to improve the existing knowledge base on the design of structured materials using FFF. To determine the e ff ect of the notching method on standard ASTM D5045 compact tension samples (W = 30 mm, B = 6 mm) of each material (acrylonitrile butadiene styrene (ABS), polylactic acid (PLA), and polycarbonate (PC)) and select a method for the layout experiments in the section, four common techniques were compared for each material (Figure 1a-d). Printed and machined notches were used, each with and without pre-cracking treatments. The factor combinations and results (fracture load) are shown in Table 1 with each treatment tested three times for a total of 36 tests. The collected data was statistically analyzed using ANOVA (Minitab V.20), which found that both material selection ( p < 0 . 001, F = 194 . 29) and notching style ( p < 0 . 001, F = 18 . 29) had a statistically significant impact on the results. Therefore, while the e ff ect was not as high as material selection, the notching method certainly had a significant e ff ect and must be considered. Machining was done with a milling machine using a 20 ◦ engraving bit with a 0.1 mm cutting tip and 3.175 mm shaft (Figure 1e-f). Pre-cracking was done using the method from [17]. Testing was done using a desktop MTS tensile testing machine with compact tension clevis attachments and a 2kN load cell. The samples were printed in a flat orientation using a Prusa i4 machine (Sunhokey Electronics Co., Ltd., Shenzhen, China) with a heated glass bed and inside of an enclosure. All filament was sourced from Hatchbox (Pomona, California, USA). The extrusion temperatures were 235 ◦ C (ABS), 210 ◦ C (PLA), and 250 ◦ C (PC), while the bed temperatures were 70 ◦ C (ABS and PC) and 60 ◦ C (PLA). Print speed was 50 mm / s in all cases. The largest variation in fracture load recorded is about 15%, which is relatively small compared with the variability observed in the literature for some materials simply from arranging the material elements [7–12, 14, 18–21]. For the non-pre-cracked cases, the di ff erence can easily be explained by notch geometry (as seen in Figure 1a-d). ASTM D5045 recommends always machining the samples but does not give a standard machining method or tool for this. It was observed when looking at the samples after testing that all had clean breaks with single cracks except a few of the printed notches which were not pre-cracked. In these cases (such as the one shown in Figure 1g), there were multiple cracks or small areas of shell separation from the infill. These e ff ects likely explain the major di ff erences between the printed and print + pre-crack samples. This e ff ect was not observed in the printed notches which had also been pre-cracked; it was concluded that the shell in the printed notches was the main contributor. Examining the crack tips (Figure 1a-d), the shell around the crack tip is in tension and helps to make the sample more sti ff during testing. Therefore, the stress concentration provided by the pre-crack and the extra stress concentration combined 2. Notching Method Analysis