Calibrating Inherent Strain for Additive Manufacturing: An investigation of different subscale geometries

Abstract

Additive manufacturing (AM) is gaining popularity over the years and process simulation has become a crucial step to effectively assess the AM process. Due to the high temperature gradients, residual stress and deformation that resides in the final product, predicting residual stress, and deformation during simulation is a crucial step to estimate the producibility of the geometry. This thesis work aims to identify whether the modified inherent strain method can replace the computationally costly thermo-mechanical simulation method to assess the deformation for the deposition of material onto large scale geometries. In this thesis, a multi-scale approach is used to establish a workflow for the inherent strain method were two different scales of models are used for the simulation, i.e., a sub-scale model and a large-scale model. The thesis aims to study the geometric influence of the sub-scale model for prediction and calibration of the inherent strain tensors with respect to deformation. In this study, eight different subscale models with different configurations were simulated and calibrated. The calibrated inherent strain tensors of the subscale model were compared with the sensitivity analysis of the large-scale model to identify the best fit sub-scale model, taking the first step towards establishing a viable workflow for the inherent strain.