PSI - Issue 38

Andreas Kempf et al. / Procedia Structural Integrity 38 (2022) 77–83 Author name / Structural Integrity Procedia 00 (2021) 000 – 000

82

6

per specimen was determined. Defects in L- PBF processed materials are usually distributed ‘regularly’ in the overall part volume. Hence, these analyses allow a comparative evaluation regarding the porosity between the specimens fabricated with the different L-PBF machines, since the presence of defects in the stress-critical surface area of the mechanically finished rotating bending test specimens becomes increasingly likely with an increasing porosity. The results are summarized in Fig. 5a and b emphasizing the lowest porosity for specimens processed with L-PBF machine B. Nearly 60% of these specimens failed due to a singular defect. The number of crack initiating defects per specimen is in the order of 2. In contrast, the manufacturing process with L-PBF machine C led to the highest porosity. Only 14% of the specimen showed a singular crack initiation. The number of crack initiating defects per specimen reaches a value of about 6. Additionally, the size of each defect was measured and the equivalent defect diameter was calculated. The cumulative defect size distribution in Fig. 5c highlights that specimens of L-PBF machine B contain the smallest defects, whereas those fabricated with L-PBF machine C the largest ones. Based on these results, the observed differences in the HCF performance between the three L-PBF machines become plausible. 6. Correlation of tensile and fatigue properties To estimate the fatigue strength of metal parts, empirical relationships with quasistatic strength properties from tensile tests can be found in literature. For aluminium alloys, the fatigue strength is in the order of 0.25 UTS to 0.35 UTS (Issler et al., 1997). That is valid up to an ultimate tensile strength of 325 MPa. A further increase in the ultimate tensile strength does not improve the HCF resistance (König et al., 2020). The ultimate tensile strength in the as-built condition of specimens processed with the three L-PBF machines are in range between 346 to 484 MPa so that the HCF lifetime should not differ. Interestingly, a comparison with the maximum stress amplitude of fatigue specimens who passed 10 7 cycles reveals an improvement of the HCF resistance with increasing ultimate strength (Fig. 6a). However, this apparent contradiction can be explained by the size of the pre-existing material defects which has been demonstrated in the section before. Furthermore, a connection between ultimate tensile strength and HCF resistance is refuted, if the T6 heat treated specimens are included in the discussion. On the one hand, the ultimate tensile strength of these specimens is lower than those in the as-built condition, but the heat treatment improves dramatically the HCF strength. On the other hand, the T6 heat treatment promotes a homogenization in the microstructure so that differences in ultimate tensile strength between the three L-PBF machines were removed. Nevertheless, due to the impact of the material defects, differences in the HCF resistance were obtained. Finally, these are also the reasons why the yield strength does not correlate with the fatigue properties (Fig. 6b). As already discussed in section 4, a general relationship between ductility and HCF properties seems not to be valid for L-PBF processed AlSi10Mg if different heat treatment conditions are considered. However, as depicted in Fig. 6c, the elongation at fracture correlates very well with the HCF resistance for the individual examined heat treatments. Fracture surface analyses in Fig. 5 revealed that an increasing defects size (which reduces the HCF strength) is accompanied with an increasing number of defects which causes an increased porosity and, thus, a decreased elongation a fracture. This can be seen as a possible verification of the observed relationship between the ‘global’ Fig. 6. Comparison of tensile and fatigue properties; (a) ultimate tensile strength vs. maximum stress amplitude at 10 7 cycles; (b) yield strength vs. maximum stress amplitude at 10 7 cycles; (c) elongation at fracture vs. maximum stress amplitude at 10 7 cycles; Values of specimens manufactured with the three different L-PBF machines in the as-built and T6 heat treatment condition.