Voxelization-based high-efficiency mesh generation method for parallel CFD code GASFLOW-MPI

Autor: Yabing Li, Thomas Jordan, Fujiang Yu, Andreas G. Class, Han Zhang, Jianjun Xiao
Rok vydání: 2018
Předmět:
Zdroj: Annals of Nuclear Energy. 117:277-289
ISSN: 0306-4549
DOI: 10.1016/j.anucene.2018.03.045
Popis: Hydrogen safety analysis is an important issue in the field of the nuclear severe accident. Fast, robust and accurate simulation of the complex hydrogen behavior in the nuclear containment is the first but critical step for hydrogen risk mitigation. GASFLOW-MPI is a widely used CFD numerical tool which provides the fast and reliable prediction, because of its parallel computational capability and well-validated models. However, due to its manual input-card based mesh generation, the whole meshing process is slow and lacks user friendliness. Therefore, a fast automatic mesh generation module is of importance and profit for practical industrial applications. Nowadays, Computer Aid Design (CAD) has become the formal standard for project design and preview. In this work, a mesh generation module is developed for GASFLOW-MPI, which directly uses the CAD file as the input for automatic mesh generation. The module exploits a voxelization-based method, which seeks to generate a Cartesian mesh by tracing rays directed into the geometry. The complexity of the mesh generation algorithm is also analyzed. Three models, including a toy problem, a steam generator compartment model, and a complex full scale reactor containment model are used to validate the new developed automated mesh generation module. The results show that meshes are generated at fast construction speed and well match the original CAD models.
Databáze: OpenAIRE