In the case of additively manufactured components made of aluminium, failure under cyclic loading is often caused by manufacturing defects such as pores or non-metallic inclusions. The size, location and frequency of these defects are directly influenced by the process parameters of the additive manufacturing process. An optimization of these parameters with regard to reduced porosity is therefore of great interest.
Within this work, the fatigue strength of four different test series on AlSi10Mg specimens produced by selective laser melting is determined. The parameter sets differ with regard to the process parameters volume rate and volume energy density. By fractographic examinations, the failure relevant defects are localized and the position and size of these defects are investigated in terms of their effect on the occurrence of failure. The high dispersion of the test results is reduced by adapted concepts of Murakami and Kobayashi through determination of a strain-based intensity factor K_ε with exact consideration of the respective pore situation. Despite serious differences in porosity of the four test series, a single significant Wöhler curve with reduced scatter can be determined for all experiments on the basis of this fracture mechanics-based parameter.
Statistical evaluations of the defect size and position on the fracture surface also provide information on which values of the volume rate and volume energy density are to be preferred for selective laser melting of AlSi10Mg with regard to high fatigue strength and low porosity.