A statistical approach to determine data requirements for part porosity characterization in laser powder bed fusion additive manufacturing

Abstract

A major factor in determining the fatigue life of fracture-critical parts is the effect of process-induced porosity. Prediction of critical pore size in different process regimes of a laser powder bed fusion (L-PBF) processed part could provide invaluable information for process development and qualification and certification efforts. However, the amount of data required to accurately populate the pore size distribution to predict the critical pore size is still unknown. To address this gap, the present study utilizes extreme value statistics to determine the data required to characterize porosity in the L-PBF additively manufactured parts fabricated using different processing conditions. 2D cross-sectional porosity data obtained via optical microscopy was used as an example to demonstrate the statistical modeling approach. The statistical modeling described here can also be applied to other manufacturing processes and other types of data such as 3D porosity measurements, grain size, and inclusions.