TY - JOUR
T1 - Solute pumping and trapping in laser powder bed fusion
T2 - A multiphysics numerical study of nonequilibrium microstructure evolution in AlSi10Mg
AU - Liu, Yufan
AU - Lian, Yanping
AU - Li, Ming jian
AU - Chen, Jiawei
AU - Xiong, Feiyu
AU - Lei, Liming
N1 - Publisher Copyright:
© 2026 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
PY - 2026/10
Y1 - 2026/10
N2 - The nonequilibrium microstructure in laser powder bed fusion (LPBF) largely dictates the final mechanical properties, yet understanding its formation remains a challenge due to the complex interplay between melt convection and solute redistribution. In this study, we propose a 3D multiphysics numerical framework for LPBF AlSi10Mg that integrates a thermo-fluid flow model, a solute transport model, and a grain structure evolution model. The thermo-fluid flow model is solved using the finite volume method (FVM) with a volume of fluid algorithm, providing the high-fidelity molten pool flow and temperature fields for the other two models. A two-way coupling scheme is proposed for the FVM and the cellular automaton method to simultaneously solve the solute transport, grain nucleation and growth, resolving both convection-driven solute pumping and the interface-kinetics-driven solute trapping. The key novelty lies in the synergistic coupling of these effects within a fully 3D molten pool, which has not been achieved in prior 2D or 3D models. The framework is validated against experimental data from two single-track LPBF AlSi10Mg cases documented in the literature. The predicted molten pool dimensions agree well with the experimental measurements, with a maximum relative error of 7.5%. The predicted Si concentration ranges from 1.39 wt% to 9.3 wt%, capturing the experimental range of 3.36–7.92 wt%; the predicted primary dendrite arm spacing ranges from 0.3 μm to 0.6 μm, close to the experimental range of 0.4–0.9 μm. The predicted average grain sizes for the two single-track cases show relative errors of 10.2% and 9.4%, respectively. Mechanistically, the simulations uniquely reveal a competitive-coexistent feedback loop between solute pumping and solute trapping. This coupled interaction governs the non-uniform solute distribution, the columnar-to-equiaxed transition, and the selection of solidification patterns. These findings provide new and generic insights into the process-microstructure relationships in LPBF, with implications beyond the specific AlSi10Mg.
AB - The nonequilibrium microstructure in laser powder bed fusion (LPBF) largely dictates the final mechanical properties, yet understanding its formation remains a challenge due to the complex interplay between melt convection and solute redistribution. In this study, we propose a 3D multiphysics numerical framework for LPBF AlSi10Mg that integrates a thermo-fluid flow model, a solute transport model, and a grain structure evolution model. The thermo-fluid flow model is solved using the finite volume method (FVM) with a volume of fluid algorithm, providing the high-fidelity molten pool flow and temperature fields for the other two models. A two-way coupling scheme is proposed for the FVM and the cellular automaton method to simultaneously solve the solute transport, grain nucleation and growth, resolving both convection-driven solute pumping and the interface-kinetics-driven solute trapping. The key novelty lies in the synergistic coupling of these effects within a fully 3D molten pool, which has not been achieved in prior 2D or 3D models. The framework is validated against experimental data from two single-track LPBF AlSi10Mg cases documented in the literature. The predicted molten pool dimensions agree well with the experimental measurements, with a maximum relative error of 7.5%. The predicted Si concentration ranges from 1.39 wt% to 9.3 wt%, capturing the experimental range of 3.36–7.92 wt%; the predicted primary dendrite arm spacing ranges from 0.3 μm to 0.6 μm, close to the experimental range of 0.4–0.9 μm. The predicted average grain sizes for the two single-track cases show relative errors of 10.2% and 9.4%, respectively. Mechanistically, the simulations uniquely reveal a competitive-coexistent feedback loop between solute pumping and solute trapping. This coupled interaction governs the non-uniform solute distribution, the columnar-to-equiaxed transition, and the selection of solidification patterns. These findings provide new and generic insights into the process-microstructure relationships in LPBF, with implications beyond the specific AlSi10Mg.
KW - Additive manufacturing
KW - Cellular automaton method
KW - Nonequilibrium solidification
KW - Solute pumping
KW - Solute trapping
UR - https://www.scopus.com/pages/publications/105047646789
U2 - 10.1016/j.jmatprotec.2026.119480
DO - 10.1016/j.jmatprotec.2026.119480
M3 - Article
AN - SCOPUS:105047646789
SN - 0924-0136
VL - 356
JO - Journal of Materials Processing Technology
JF - Journal of Materials Processing Technology
M1 - 119480
ER -