PSI - Issue 34

Sanne van den Boom et al. / Procedia Structural Integrity 34 (2021) 87–92

91

Sanne van den Boom et al. / Structural Integrity Procedia 00 (2021) 000–000

5

0

2.5 %

1.5

1

2

0.5

Fig. 5: DIC (left) and numerical results (right) of the three designs. From top to bottom: design A - uniform infill , design B - optimized infill , design C - optimized design . Strains higher than 2.5% are shown in black in the numerical simulations.

Fig. 6: Failed specimens, from left to right: design A - uniform infill , design B - optimized infill , design C - optimized design .

Indeed, this is also the location where the ladder steps tend to fail (Figure 6). Whereas the design A - uniform infill specimen only showed a crack, for design B - optimized infill , the middle of the ladder step has broken o ff . This happened for two of the three optimized infill specimens. The other specimens also developed cracks at both corners, but stayed in one piece. For design C - optimized design , the middle of the ladder broke o ff in all the specimens. Figure 7a compares the mechanical behavior of the di ff erent types of ladder steps, by comparing the numerical and experimental force-displacement curves. It is clear that design B - optimized infill and design C - optimized design are sti ff er than design A - uniform infill . This is unsurprising, as they were both optimized for minimum compliance. Besides being sti ff er, both optimized versions also have a higher failure load, although they were not specifically optimized for this. Because the weights of the di ff erent types of specimens are not the same, in Figure 7b ratio between the maximum load for each specimen and its weight is also listed. This table shows that the design B - optimized infill specimens are lighter than the design A - uniform infill specimens while having a slightly higher average ultimate load. On average, the design A - uniform infill specimens carry 16 N per gram, the design B - optimized infill specimens carry 20 N per gram, and the design C - optimized design specimens carry 31.6 N per gram.

6. Discussion and conclusions

In this work we have performed a numerical analysis, infill optimization and design optimization of a highly loaded 3-D printed part. It was shown through homogenization of the infill at di ff erent densities that a RAMP material interpolation with a penalty factor of 3 is a good approximation of the infill properties. For the design optimization, the standard material interpolation function is used. In both cases, it is important to carefully select the density level