基于非结构化网格下TVD格式的一种r因子算法

A r-factor algorithm for TVD schemes on unstructured grids

  • 摘要: 针对对流主导问题的数值模拟,提出了一种改进的r因子算法,以实现TVD(total variational diminishing)格式在非结构化网格中的扩展。新算法将面中心到辅助逆风节点的距离考虑在内,以实现更合理的远逆风节点位置预测;通过包含更多逆风信息的插值方法和采取考虑实际点与辅助点之间的物理值差异的方法来提高TVD格式的精度;并避免引入复杂的向量计算过程以提高效率。非结构化网格下对流主导问题的基准测试结果表明,相比于现有的其他算法,新r因子算法在精度和效率方面具有更好的性能。

     

    Abstract: To realize the implementation of TVD schemes on unstructured grids, an improved r-factor algorithm is proposed for the numerical simulation of the convection-dominated problems. The distance from the face center to the auxiliary upwind node is considered in the new r-factor algorithm, which achieves more reasonable prediction of the far-upwind node position. The accuracy of TVD schemes on unstructured grids is improved by invoking an interpolation scheme that contains more upwind information and considering the physical deviation between the actual nodes and the auxiliary nodes. Meanwhile, the computational efficiency is improved by avoiding the introduction of complex vector computation procedures in the new algorithm. Benchmark testing results of the convection-dominated problems on unstructured grids indicate that the new r-factor algorithm outperforms existing ones in terms of both accuracy and efficiency.

     

/

返回文章
返回