PSI - Issue 34

5

Radomila Konecna et al. / Procedia Structural Integrity 34 (2021) 135–140 Author name / Structural Integrity Procedia 00 (2021) 000 – 000

139

Roughness of as-built surfaces: Fig. 3 quantifies the previous structural observations in terms of the longitudinal roughness measurements (averaged over multiple specimens) of the top surface of the three miniature specimens. All roughness parameters behave consistently for the three directions with the highest roughness determined for the Type C specimens, followed by the Type B specimens and with Type A- specimens being the smoothest. If the mean roughness value R a is considered, the values range from 12  m (Type C) to 6  m (Type B) to 5  m (Type A-). These values are in agreement with published data thus confirming the expert production process adopted. Roughness measurements of Fig. 3 and cross-sectional views of Fig. 2 provide useful insight clarify that the roughness of Type B depends essentially on contouring parameters (as they influence balling or partially melted particles). On the other hand, the significantly higher roughness of Type C depends on the surface being a cross section of a sequence of 50  m layers. The low roughness of type A- specimens is related to intersection of the longitudinal cross-section with the regularly spaced hatches and the melt parameters affecting free surface curvature. 3.2. Directional fatigue behavior. Fig. 4 shows the significant directionality of the high cycle fatigue behavior of the present as-built SLM AlSi10Mg without heat treatment and with as-built surfaces. The maximum stress versus number of cycles to crack initiation are used to plot the data. The fatigue strengths at 2 10 6 cycles range from about 190 MPa for Type A- specimens to about 115 MPa for the Type C specimens. The data trend of Type B specimens is similar to Type A- at high stresses but is close to Type C data at high cycles (about 140 MPa). The data are reasonably consistent with the roughness measurements, that is Type C specimens have the highest roughness and the lowest fatigue performance while Type B specimens show a relatively low roughness and intermediate fatigue strength. Type A- and Type B specimens have a similar roughness values, but their near surface structures are completely different.

Fig. 4. Directional fatigue behaviour of L-PBF AlSi10Mg with as-built surfaces (no post-fabrication heat treatment).

3.3. Residual stresses As the specimens were left in the as-built state (i.e., high temperature heat treatment), residual stresses are expected to contribute significantly to the fatigue behavior and to be directional in nature. As a first approximation, the contribution of residual stresses to fatigue crack initiation may be likened to a mean stress effect: tensile residual stresses have a negative effect on fatigue behavior while the opposite effect is expected when residual stresses are