Statistical Analysis of Porosity and Process Parameter Relationships in Metal Additive Manufacturing

Abstract

Metal additive manufacturing is susceptible to formation of porosities in the build. Numerous process and material parameters have been reported to impact formation and evolution of different types of porosities in metallic additively manufactured parts. Conducting extensive experiments is a typical approach to investigate the effects of process and material parameters to formation of porosities. In this work, a computational framework is proposed to use two-point correlations to capture statistical distribution of porosities in metallic additively manufactured parts. The proposed scheme is robustly built based on mathematics of correlation functions, by combining the overall effects of all the parameters impacting porosity formation, into a coefficient term in a Fourier-type expansion of two-point correlation statistics. The functionality of the coefficients with respect to scanning speed (i.e. a process parameter) in selective laser melting of 304L stainless steel is derived numerically and consistency of the trend in change of the coefficient values compared to porosity volume fraction is confirmed. The formalism outlined and tested in this work provides a robust framework for prediction of porosity volume fraction as a function of process parameters for metal additive manufacturing.