An integrated computational materials engineering framework for process-structure-property mapping in laser powder bed fusion

Fabien Briffod; Phuangphaga Daram; Masahiro Kusano; Makoto Watanabe

doi:10.1016/j.matdes.2025.115097

Article An integrated computational materials engineering framework for process-structure-property mapping in laser powder bed fusion

Fabien Briffod ; Phuangphaga Daram ; Masahiro Kusano ; Makoto Watanabe

Download file (9.16 MB)
Download as zip (8.28 MB)

Collection

Citation

Fabien Briffod, Phuangphaga Daram, Masahiro Kusano, Makoto Watanabe. An integrated computational materials engineering framework for process-structure-property mapping in laser powder bed fusion. Materials & Design. 2025, 260 (), 115097. https://doi.org/10.1016/j.matdes.2025.115097

(BibTeX)

Description:

(abstract)

This work presents a comprehensive, experimentally validated integrated computational materials engineering framework for mapping the process-structure-property relationships in laser powder bed fusion (L-PBF) of Hastelloy-X alloys. The framework couples heat transfer, cellular automata (CA) solidification, and crystal plasticity finite elements (CPFE) simulations within one workflow. The heat transfer model was calibrated using single-track experiments and Bayesian inference to accurately capture melt pool geometry and the transition from conduction to keyhole melting. The CA model, driven by thermal simulation data, successfully reproduced key microstructural features, including the equiaxed-to-columnar grain transition and the formation of a strong crystallographic texture. The mechanical behavior was then predicted by CPFE simulations on representative volume elements extracted from the CA microstructures, revealing a direct correlation between crystallographic texture and macroscopic mechanical properties. The framework was applied to the mapping of the (P,v) process space, identifying distinct regions based on defect formation, microstructure and mechanical response. This validated approach offers a robust and efficient alternative to experimental trial-and-error identification of optimal process window, paving the way for data-driven optimization of L-PBF processes.

Rights: