ABSTRACT Computational modelling of mould filling is a component of paramount importance for the design of gating systems of castings. The main methods are the body-fitted grid system and the Cartesian grid system. While the former comes with high accuracy but high calculation times, the latter is fast but less accurate. A hybrid method combining a three-dimensional Cartesian cut-cell method and a porous media approach has been further developed. The calculated filling patterns compared favourably with high-speed video images of sand mould filling during pouring of ductile iron, and with visual images recorded through an acrylic glass-water model device simulating the gravity die casting process. In addition, numerical experiments were conducted to demonstrate the improved filling accuracy of the developed hybrid method. To this goal, three different symmetry tests with and without sub-mesh were designed and evaluated. All tests convincingly demonstrated the computational time and accuracy advantages of the developed method.

中文翻译：

---

开发一种提高铸件充型计算模型精度的混合方法

摘要 模具填充的计算建模是铸件浇注系统设计中最重要的组成部分。主要方法有贴体网格系统和笛卡尔网格系统。前者精度高但计算时间长，后者速度快但精度低。进一步开发了一种结合三维笛卡尔切割单元法和多孔介质方法的混合方法。计算出的填充模式与球墨铸铁浇注过程中砂型填充的高速视频图像以及通过模拟重力压铸过程的丙烯酸玻璃水模型设备记录的视觉图像相比具有优势。此外，还进行了数值实验以证明所开发的混合方法提高了填充精度。为了这个目标，设计和评估了有和没有子网格的三种不同的对称性测试。所有测试都令人信服地证明了所开发方法的计算时间和准确性优势。