PSI - Issue 61

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

153

6

Author name / Structural Integrity Procedia 00 (2024) 000–000

(b) Difference Between Factor Combinations - Layout Only ABS PLA

(a) Difference Between Factor Combinations - Layout + Nozzle ABS PLA

PC

PC

Weakest combination

0.8 Gyroid 0.6 Gyroid 0.6 Concentric

Weakest combination

Gyroid

Gyroid

Gyroid

0.6 R45 0.6 R90

R45 Average R90 Average Gyroid Average

1.15 1.35 1.01 1.61 1.46 1.16

1.22 1.19 1.47 1.36 1.15 1.03 1.51 1

1.72 1.50 1.45 1.62 1.61 1.54 2.08 1

1.30 1.25

1.27 1.16

1.11 1.04

0.6 Gyroid

1

1

1

0.6 Concentric

Concentric Average Difference with strongest combination

1.58

1.47

1.03

0.8 R45 0.8 R90

58%

47%

11%

0.8 Gyroid

1

0.8 Concentric

1.56

Difference with strongest combination

61%

51% 108%

Fig. 5. (a) Comparison chart between the weakest and strongest cases and (b) comparison chart only considering layout.

in the form of gaps and voids. There has been very little discussion on this topic in the published literature, but it is clearly an issue that must be watched for when selecting process parameters. This phenomenon is discussed further in the crack pattern analysis. This result was very consistent across samples, as shown in Figure 4b. This was not noted for either of the other materials, so it appears to be a particular risk for PC.

4.3. Crack Pattern Analysis

Qualitative analysis of the observed crack patterns also provides insight into the e ff ects of the element layout. Figure 6 shows examples of each crack path relative to material and element layout. For the R45 samples the cracks followed the raster, occasionally experiencing a 90 ◦ turn to follow the raster of a di ff erent layer. The R90 samples all had straight cracks from the notch tip toward the edge. The gyroid samples showed behavior similar to R45, except that the cracks were far more jagged and irregular. Of all the geometries, the concentrically printed samples were the only ones that showed specific dependence on the material. As shown in Figure 6 for concentric cases, the ABS crack dissipated sideways into the structure (guided by the contour), the PLA crack tended to split into branches and cause crazing, and the PC samples remained sti ff until the crack hit a contour and then failed in fast fracture. These observations were consistent across multiple replications, including the widely-varying concentric PC samples.

R45

R90

Crack

Crack

Crack

Crack

Crack

Crack

PC

PLA

PLA

ABS

ABS

PC

Concentric

Gyroid

Crack

Crack

Crack

Crack

Crack

Crack

PC

PLA

ABS

ABS

PLA

PC

Fig. 6. Typical observed crack patterns

Further examining the layout of the 0.6mm concentric PC samples (Figure 4b and Figure 6), it appears likely that the specific combination of nozzle size and layout built in a weak spot in the structure that encouraged crack growth. Coupled with the brittleness of FFF-processed PC, this explains the variability in fracture load not noted with other