Calibration of Inherent Strain used in L-PBF Part-Scale Modelling using a Numerical–Experimental Distortion Fitting
Please login to view abstract download link
Metal additive manufacturing has emerged as a transformative manufacturing technology, enabling the production of complex engineering components with exceptional design freedom compared to conventional manufacturing processes. However, accurate prediction of part distortion remains a major challenge in laser-powder bed fusion (L-PBF) due to the complex thermo-mechanical phenomena involved. The inherent strain method offers an efficient alternative to full thermo-mechanical simulations, but its accuracy strongly depends on the proper calibration of the inherent strain values. In this work, a numerical–experimental framework is proposed for the calibration of inherent strain based on distortion fitting. A set of benchmark geometries was manufactured by L-PBF using controlled process parameters, and the resulting distortions were measured using high-resolution metrology. In parallel, a finite element model based on the inherent strain method was developed to simulate the L-PBF process for each geometry. Then, the inherent strain components were identified through an inverse fitting procedure that minimizes the discrepancy between numerical predictions and experimental distortion data. The results show that the calibrated inherent strain values strongly depend on the part geometry. Moreover, the cantilever geometry, commonly employed for calibration purposes, presents significant limitations, as the maximum distortion achievable after support removal is constrained by the elastic response of the structure. In addition, the calibrated inherent strain values are mesh-dependent, since the thickness of the deposited finite element layers is dictated by the mesh size, directly influencing the distortion generated during each layer deposition.