Coordinate transformation induced errors of finite difference method

被引：0
|
作者：
Liu J. [1 ]
Wei Y. [1 ]
Han F. [1 ]
机构：
[1] School of Aeronautics and Astronautics, Dalian University of Technology, Dalian
来源：
Hangkong Xuebao/Acta Aeronautica et Astronautica Sinica | 2021年 / 42卷 / 06期
基金：
中国国家自然科学基金;
关键词：
Coordinate transformation; Discrete equivalence equation; Finite difference method; Geometric conservation law; Geometrically induced errors;
D O I：
10.7527/S1000-6893.2020.24397
中图分类号：
学科分类号：
摘要：
Coordinate transformation is required when the finite difference method is applied to the mesh with complex geometries, and the errors induced by the coordinate transformation are often introduced in this process. These errors are proved to be inevitable in the uniform flow field calculation with uniform grids in cylindrical coordinate systems, even if the transformation function of the physical coordinates to the calculated coordinates is continuously derivable, or the coordinate transformation coefficients in the calculation process are calculated by the accurate analytical formula, or the grid is completely orthogonal and fully smooth. Theoretical analysis shows the mechanism of the coordinate transformation induced errors: when the conservative Euler equation is transformed from the Cartesian coordinate system to the body fitted coordinate system, a source term is added. Currently, scholars usually use the geometric conservation law to construct a method based on coordinate transformation coefficients, which are matched with the format of the finite difference, to eliminate the source term. In this work, we introduce a new algorithm that processes the direct discretization from the discrete equivalent functions including the source term. Based on the above new algorithm, error analysis is carried out for the reconstruction process of MUSCL format under non-equidistant grid conditions. Theoretical derivation shows that the influence coefficient of the non-equidistant interpolation formula needs to be considered in reconstruction, only when the variables are transformed into the computational space for MUSCL reconstruction can the interpolation accuracy be guaranteed under uniform grid. Our theoretical analysis and numerical experiments have proven that this algorithm will not introduce coordinate transformation errors to the uniform flow field calculations. © 2021, Beihang University Aerospace Knowledge Press. All right reserved.
引用
下载
收藏
相关论文
共 21 条
  • [1] TRULIO J G, TRIGGER K R., Numerical solution of the one-dimensional Lagrangian hydrodynamic equations: UCRL-6267, (1961)
  • [2] STEGER J., Implicit finite difference simulation of flow about arbitrary geometries with application to airfoils, 10th Fluid and Plasmadynamics Conference, (1977)
  • [3] STEGER J L., Implicit finite-difference simulation of flow about arbitrary two-dimensional geometries, AIAA Journal, 16, 7, pp. 679-686, (1978)
  • [4] PULLIAM T, STEGER J., On implicit finite-difference simulations of three-dimensional flow, 16th Aerospace Sciences Meeting, (1978)
  • [5] HINDMAN R., Geometrically induced errors and their relationship to the form of the governing equations and the treatment of generalized mappings, 5th Computational Fluid Dynamics Conference, pp. 81-1008, (1981)
  • [6] HINDMAN R G., Generalized coordinate forms of governing fluid equations and associated geometrically induced errors, AIAA Journal, 20, 10, pp. 1359-1367, (1982)
  • [7] VIVIAND H, GHAZZI W., Numerical solution of the compressible Navier-Stokes equations at high Reynolds numbers with applications to the blunt body problem, Proceedings of the 5th International Conference on Numerical Methods in Fluid Dynamics, pp. 434-439, (1976)
  • [8] THOMAS P, LOMBARD C., The Geometric Conservation Law-A link between finite-difference and finite-volume methods of flow computation on moving grids, 11th Fluid and Plasma Dynamics Conference, (1978)
  • [9] THOMAS P D, LOMBARD C K., Geometric conservation law and its application to flow computations on moving grids, AIAA Journal, 17, 10, pp. 1030-1037, (1979)
  • [10] GAITONDE D, VISBAL M., Further development of a Navier-Stokes solution procedure based on higher-order formulas, 37th Aerospace Sciences Meeting and Exhibit, (1999)
123