Fast Distortion Prediction in Directed Energy Deposition Using Inversely-Identified Inherent Strains

Yun Chen,Yang Xu,Liang Hou,Wen Shao,Xiubing Jing,Yuan Li,Shuyuan Chen
DOI: https://doi.org/10.1080/09544828.2023.2193879
2023-01-01
Journal of Engineering Design
Abstract:Large thermal distortions due to the cyclic rapid melting and solidification mechanism in metal additive manufacturing (MAM) affect part manufacturing precision. The so-called inherent strain method is one of the most computationally efficient methods to predict distortion, but the current inherent strain methods extract inherent strains through the full-scale detailed thermo-mechanical model (TMM) with a long computation time. This work proposes an inverse parameter identification method to fast identify inherent strains based on the measured distortion results from a two-track-two-layer workpiece. The identified inherent strains are employed in a static mechanical analysis to efficiently predict distortion in MAM. To verify the proposed method, a multi-track-multi-layer workpiece and a square-shaped workpiece deposited by the directed energy deposition process are studied. The simulated distortion results demonstrate high simulation accuracy by comparing with the experimental results. In addition, comparisons with the TMM and the modified inherent strain method indicate that the inversely-identified inherent strains can improve the distortion simulation accuracy and reduce the simulation time, which is practical to be applied in industrial applications.
What problem does this paper attempt to address?