结构化网格快速生成工具的开发及其在冶金模拟仿真中的应用

Development of the fast structured mesh generation tool and its application in metallurgical simulation

  • 摘要: 结构化网格在模拟计算高精度、复杂问题时具有更高的计算准确度和更好的收敛性,但其划分难度大且耗时较长,如何快速生成高质量结构化网格是进行CFD(Computational fluid dynamics)等模拟计算领域的前沿问题. 根据冶金容器的几何结构特征,引入了参数一体化建模+网格划分的理念,利用Python脚本快速批处理数据和开源CFD软件OpenFOAM的优势,成功开发了适用于冶金反应器的网格快速生成工具. 以钢包–中间包–结晶器及其附属连接装置为例,进行了案例分析测试. 结果表明,该工具软件生成的结构化网格在扭曲度和正交性等方面质量较高,可以满足各类模拟仿真的需求. 该工具在解决冶金容器几何结构更新和网格分布调整方面相比于传统前处理方法可以在几十秒内快速生成高质量结构化网格. 此外,开发了便于用户使用的图形用户界面,界面中仅需输入关键的几何结构和网格参数,即可对冶金容器快速便捷地生成相应的结构化网格,有效提升了前处理的简易度和效率. 这一工具软件有望应用于冶金数字化和智能化转型发展的多种场景,并为其提供有效的软件支撑.

     

    Abstract: A structured grid enables higher computational fidelity and better convergence in the calculation of high-precision complex simulation problems, particularly for high-speed flow field or multiphase flow. However, its generation is usually difficult, labor-intensive, and time-consuming; thus, how to quickly generate a quality structured grid is a frontier topic in various simulation scenarios, such as computational fluid dynamics (CFD). Metallurgical vessels need to be served at high temperatures, and their inner cavity are filled with refractory materials. The structure is usually relatively regular. According to the geometric structural characteristics of metallurgical vessels, parametric modeling and mesh generation have been adopted to successfully develop a fast mesh generation tool suitable for metallurgical reactors by taking advantage of batch data processing of Python Script and open-source CFD software OpenFOAM. Following the analysis of the internal topological logical relationship of the target geometry structure, it was divided into several basic geometric units, such as torus, cones, and quadrilateral columns, and their geometric parameters and topological parameters were obtained. The shape and spatial position of the basic geometric structure were determined by its geometric control parameters. The code was written according to the topological relation of the target geometry structure, and the structured grid of basic geometry units was gradually stacked into the target geometry structure. Thousands of coordinate information can be output and sorted using Python Script according to the predetermined logic in a short time. After script execution, a dictionary file called blockMeshDict required by blockMesh can be generated and then converted into a .msh file. With ladle, tundish, mold, and its auxiliary connecting device as examples, a mesh-generation tool is developed and examined. Consequently, the generated grid had high quality considering its checking distortion and orthogonality, which can meet the needs of various simulations. This tool exhibits remarkable advantages over traditional pretreatment methods in solving the geometric structure update and mesh adjustment of metallurgical vessels (e.g., changing the position of the retaining wall in tundish and the position of the bottom blowing plugs in the ladle) and can quickly generate high-quality structured meshes in several or tens of seconds. A user-friendly graphical user interface was also developed, which only needed key geometric structure and mesh parameters, and the corresponding structured mesh can be generated conveniently. The program is continuously being optimized to provide more visual and interactive performance. It can effectively improve the simplicity and efficiency of preprocessing. This tool is expected to be applied to digitalization and intelligent transformation for the metallurgical industry and provide effective support in industrial software.

     

/

返回文章
返回