Artificial intelligence assisted calculation of interfacial heat transfer coefficient in low pressure die castings

TY - JOUR

T1 - Artificial intelligence assisted calculation of interfacial heat transfer coefficient in low pressure die castings

AU - Li, Zhongyao

AU - Wu, Xuelong

AU - Hou, Qinghuai

AU - Li, Xiang

AU - Wang, Wenbo

AU - Qiao, Haibo

AU - Ma, Xiaoying

AU - Feng, Shuwei

AU - Wang, Shihao

AU - Kong, Decai

AU - Miao, Yisheng

AU - Dou, Ruifeng

AU - Lang, Yuling

AU - Wang, Junsheng

N1 - Publisher Copyright: © 2025 Elsevier Ltd

PY - 2026/3

Y1 - 2026/3

N2 - The interfacial heat transfer coefficient (IHTC), or contact thermal resistance, characterizes the heat transfer characteristics between two contacting objects. During the casting process, the cooling of the molten metal in the mold cavity relies mostly on the heat transfer between the metal and the mold. Therefore, the IHTC between the mold and the casting is crucial for the accurate prediction of temperature distribution in a casting component. In this study, a new methodology has been developed to obtain the full sets of IHTC in complex castings like wheel hubs. Using limited experimental data and finite element simulation results, such artificial algorithms as XGBoost, SVR, and Transformer have been applied to establish the correlation between output temperature distribution and input IHTC. It has been found that the XGBoost model performed best, which was then used as the objective function in a non-dominated sorting genetic algorithm (NSGA II) optimization. Therefore, accurate simulations have been performed by applying specific IHTC to different boundaries, enabling successful validation by experimental data from thermocouple measurements.

AB - The interfacial heat transfer coefficient (IHTC), or contact thermal resistance, characterizes the heat transfer characteristics between two contacting objects. During the casting process, the cooling of the molten metal in the mold cavity relies mostly on the heat transfer between the metal and the mold. Therefore, the IHTC between the mold and the casting is crucial for the accurate prediction of temperature distribution in a casting component. In this study, a new methodology has been developed to obtain the full sets of IHTC in complex castings like wheel hubs. Using limited experimental data and finite element simulation results, such artificial algorithms as XGBoost, SVR, and Transformer have been applied to establish the correlation between output temperature distribution and input IHTC. It has been found that the XGBoost model performed best, which was then used as the objective function in a non-dominated sorting genetic algorithm (NSGA II) optimization. Therefore, accurate simulations have been performed by applying specific IHTC to different boundaries, enabling successful validation by experimental data from thermocouple measurements.

KW - Boundary condition

KW - Casting

KW - Interfacial heat transfer coefficient

KW - Machine learning

KW - Multi-objective optimization

KW - Transformer

UR - https://www.scopus.com/pages/publications/105026906677

U2 - 10.1016/j.icheatmasstransfer.2025.110422

DO - 10.1016/j.icheatmasstransfer.2025.110422

M3 - Article

AN - SCOPUS:105026906677

SN - 0735-1933

VL - 172

JO - International Communications in Heat and Mass Transfer

JF - International Communications in Heat and Mass Transfer

M1 - 110422

ER -