Evaluation of calibration accuracy in numerical simulation of laser powder bed fusion using a simplified thermomechanical model

Numerical simulations are essential in additive manufacturing for predicting distortions and stresses, thereby ensuring the reliability and performance of parts in industrial applications. However, traditional high-fidelity simulation methods often require significant computational resources. To overcome this limitation while maintaining accuracy, simplified approaches have been developed by incorporating experimental data through a preliminary calibration step. Nevertheless, certain variables involved in the calibration stage, specifically laser absorptivity and mesh size, can significantly influence both the accuracy and the efficiency of the simulation. This study investigates how these variables affect calibration accuracy using a simplified thermomechanical modeling approach. Simulations were performed and validated against experimental measurements. For this analysis, SS316L cantilever specimens oriented orthogonally were fabricated using the Laser Powder Bed Fusion process. The results indicated that lower laser absorptivity values led to better accuracy. Additionally, the model showed high sensitivity to mesh size, emphasizing the importance of selecting appropriate simulation parameters. Validation confirmed that the calibration errors aligned closely with simulation predictions. However, the results also suggested that the accuracy could be compromised with replication to a more complex geometry. These findings highlight the potential and limitations of this simplified approach for simulating large components and offer insights into how parameter selection affects model accuracy. Clarifying the assumptions of this model and explaining how the parameters influence its accuracy can be considered essential for future research aimed at improving this approach, as well as for supporting its dissemination in industrial applications.