Garcia-Michelena Pablo, Ruiz-Reina Emilio, Gordo-Burgoa Olaia, Herrero-Dorca Nuria, Chamorro Xabier
Mechanical and Manufacturing Department, Mondragon University, 20500 Mondragon, Spain.
Institute Carlos I for Theoretical and Computational Physics (iC1), Department of Applied Physics II, University of Malaga, 29071 Malaga, Spain.
Materials (Basel). 2025 Jan 5;18(1):199. doi: 10.3390/ma18010199.
This study investigates fixed and moving mesh methodologies for modeling liquid metal-free surface deformation during the induction melting process. The numerical method employs robust coupling of magnetic fields with the hydrodynamics of the turbulent stirring of liquid metal. Free surface tracking is implemented using the fixed mesh level set (LS) and the moving mesh arbitrary Lagrangian-Eulerian (ALE) formulation. The model's geometry and operating parameters are designed to replicate a semi-industrial induction melting furnace. Six case studies are analyzed under varying melt masses and coil power levels, with validation performed by comparing experimentally measured free surface profiles and magnetic field distributions. The melt's stirring velocity and recirculation patterns are also examined. The comparative analysis determines an improved performance of the ALE method, convergence, and computational efficiency. Experimental validation confirms that the ALE method reproduces the free surface shape more precisely, avoiding unrealistic topological changes observed in LS simulations. The ALE method faces numerical convergence difficulties for high-power and low-mass filling cases due to mesh element distortion. The proposed ALE-based simulation procedure is a potential numerical optimization tool for enhancing induction melting processes, offering scalable and robust solutions for industrial applications.
本研究探讨了用于模拟感应熔炼过程中液态金属自由表面变形的固定网格和移动网格方法。该数值方法采用了磁场与液态金属湍流搅拌流体动力学的稳健耦合。自由表面跟踪通过固定网格水平集(LS)和移动网格任意拉格朗日-欧拉(ALE)公式实现。该模型的几何形状和操作参数旨在复制一个半工业感应熔炼炉。在不同的熔体质量和线圈功率水平下分析了六个案例研究,并通过比较实验测量的自由表面轮廓和磁场分布进行了验证。还研究了熔体的搅拌速度和再循环模式。对比分析确定了ALE方法在收敛性和计算效率方面的改进性能。实验验证证实,ALE方法能更精确地再现自由表面形状,避免了在LS模拟中观察到的不切实际的拓扑变化。由于网格单元变形,ALE方法在高功率和低质量填充情况下面临数值收敛困难。所提出的基于ALE的模拟程序是一种潜在的数值优化工具,可用于改进感应熔炼过程,为工业应用提供可扩展且稳健的解决方案。