Capturing and characterizing geometric complexities of metal additively manufactured parts using x-ray micro-computed tomography and multiscale curvature analyses

Abstract

This study explores a quantitative analysis of multiscale geometric characterization of laser melted parts made of Ti-6Al-4V to establish functional relations between geometric characterisations and formation, or process, parameters. A focus is here is on multiscale curvature as a characterization of local shape which, unlike conventional texture characterizations, does not require a datum. Surfaces were measured using microCT scanner which, unlike conventional optical and stylus methods, can measure surfaces of re-entrant features. Laser power and scanning speed were considered as crucial AM process parameters in this study. It was found that curvature performed well in discriminating surfaces by both crucial process parameters and their combination for all analysed scales ranging from 0.005 to 1.000 mm. Strong functional correlations were established using linear regression (R2 = 0.83) for the largest considered scale, while no correlations were observed for fine scales associated with the powder particle sizes. It was noticed that large scale curvature increases with laser power and decreases with the scanning speed. That showed that multiscale curvature can be an easy-to-implement and valuable analysis for complex, freeform, AM metal surfaces.