Casting CAE (Computer Aided Engineering) has recently become an indispensable tool for casting design and defect prediction. Casting filters are often installed in the flow path to control the melt flow rate in gating design, but the effectiveness of such filters must be validated through simulation. But, there is a limitation in using approximate models, such as Darcy’s law, which was used in the previous study, to analyze complex flow behavior in the casting filters. Recently, numerical analysis methods that closely approximate filter shapes are available. One approach is the SPH (Smoothed Particle Hydrodynamics) particle method. However, one drawback of this method lies in its significant computational costs.

  In this study, we investigated the relationship between particle size, calculation accuracy, and calculation cost in the particle-based SPH method to accurately and efficiently simulate the flow behavior in casting filters. Furthermore, we also evaluated the characteristics and practicality between the Eulerian lattice method (PM method) with Darcy’s law, which models the filter section in terms of permeability, and the SPH method. The results clarified that although the performance of the SPH particle method can be improved by optimizing the particle size, its computational costs are still more expensive than the other methods. This therefore remains a practical issue that needs to be addressed.

  In the present study, a method was proposed for automatically combining elements in the mold region to reduce the number of elements of mesh data divided into voxel meshes. Three types of evaluation methods for heat flux at the interface between fine and coarse elements were evaluated based on the strict solution of one-dimensional unsteady heat conduction. Solidification analyses simulating die casting of an oil pan were carried out to verify the reduction in the number of elements and calculation time. As a result, the following findings were obtained. A method for estimating the temperature of virtual fine elements inside a coarse element was proposed by considering both spatial distribution and conservation low. As for the evaluation method for heat flux at the interface between fine and coarse elements, the method using the estimated temperature of virtual fine elements within a coarse element was found to be the most accurate in the tested three methods. By defining appropriately sized merging-prohibited regions around casting and cooling elements and combining mold elements in other regions, the number of elements was significantly reduced with accuracy remaining nearly unchanged. For the case study simulating die casting of oil pan, the reduction ratio of number of elements was up to 93.3% and that of time for solidification analysis was up to 84.2%.

  This study introduces a metamodeling approach to predict the microstructure and mechanical properties of aluminum alloys during casting and heat treatment processes. Using a step-shaped specimen with varying cooling rates, the microstructure and mechanical properties in both as-cast and T6 heat-treated states were predicted based on physical models. Since thermodynamic calculations require significant computational time, Orthogonal Latin Hypercube Design (OLHD) sampling and Kriging-based metamodeling were employed to reduce computation time while maintaining predictive accuracy. The present method was applied to a step-shaped cast specimen, and predicted values were compared with experimental data, confirming that they were satisfactory.

  Since die casting simulations need to calculate complex multi-physics simulations, such as the interaction between solid and the flow of molten metal, a calculation method which can handle a wide range of calculation conditions is required. The MPH (Moving Particle Hydrodynamics) method, which is a particle method based on analytical mechanics, may be effective for the die casting simulations as it has high numerical stability. In this paper, the applicability of the MPH method to the die casting simulation was investigated. The fundamental models of heat transfer, phase change, and variable volume were introduced to the MPH method, and the calculations which simulate the die casting process with simple geometries were conducted. It was shown that the proposed method has the potential to simulate various phenomena in the die casting process in a unified manner.