Revealing mechanisms of residual stress development in additive manufacturing via digital image correlation

Abstract

The severe thermal gradients associated with selective laser melting (SLM) additive manufacturing (AM) generate large residual stresses (RS) that geometrically distort and otherwise alter the performance of printed parts. Despite broad research interest in this field, it has remained challenging to measure warpage in general as well as RS distributions in situ, which has obfuscated the mechanisms of stress formation during the printing process. In pursuit of this goal, we have developed a non-destructive framework for RS measurement in SLM parts using three-dimensional digital image correlation (3D-DIC) to capture in situ surface distortion. A two-dimensional analytical model was developed to convert DIC surface curvature measurements to estimates of in-plane residual stresses. Experimental validation using stainless steel 316 L “inverted-cone” parts demonstrated that residual stress varied across the surface of the printed part, and strongly interacted with the component geometry. The 3D-DIC based RS measurements were validated by X-ray diffraction (XRD), with an average error of 6% between measured and analytically derived stresses. Systematic variation in RS was attributed to the sector-based laser raster strategy, which was supported by complementary finite element calculations. Calculations showed that the heterogeneous RS distribution in the parts emerged from the sequential re-heating and cooling of the new surface, and changed dynamically between layers. The unique DIC based RS methodology brings substantial benefits over alternatively proposed in situ AM RS measurements, and should facilitate enhanced process optimization and understanding leading towards AM part qualification.