Efficient predictive model of part distortion and residual stress in selective laser melting

Selective laser melting (SLM) is a promising technology to manufacture functional (end-use) metal parts with complex geometry directly from CAD data. The process induced high tensile residual stress and part distortion due to the non-uniform heat input during a SLM process would detrimentally affect the part performance. However, it is extremely challenging to predict distortion of a practical SLMed part if each single track is taken into account by using the conventional modeling methods The complex multiphysics phenomenon such as fluid flow in the melt pool, phase transformation during cooling, and resulted anisotropic properties further complicate this issue. In this study, a temperature-thread multiscale modeling approach has been developed to effectively predict residual stress and part distortion of a twin cantilever. An equivalent body heat flux has been proposed from the microscale laser scan model and imported as the temperature-thread to the subsequent mesoscale layer hatch model. The hatched layer is then heated up by the equivalent body heat flux and used as a basic unit to build up the macroscale part in a layer by layer fashion. The thermal history and residual stress fields of the twin cantilever during the SLM process were simulated. The predicted cantilever distortion agrees with the measured data with a reasonable accuracy. (C) 2017 Elsevier B.V. All rights reserved.