Grid self-adaptive method based on error estimation

Abstract

The invention relates to a grid self-adaptive method based on error estimation, comprising the following steps: 1) determining a target function; 2) obtaining a flow field solution and an adjoint solution based on the current grids of the target function; 3) constructing fine grids corresponding to the current grids; 4) interpolating the corresponding quantity of the current grids in the fine grids, and constructing an interpolation operator from a current grid solution to a fine grid solution; 5) calculating the residual amount of the flow field on the fine grids; and 6) carrying out error estimation and fairing, if the error is more than a set error, marking the current grids for fining, and returning to the step 2) after the fining is finished, and if the error is less than the set error, avoiding fining on the current grids. By utilizing the grid self-adaptive method, the error of the target function can be estimated according to the specified target function as well as a flow field control equation and the solutions of an adjoint equation thereof under the specific flow condition, self-adaptive local grid refinement is carried out according to the error, and then the calculation of the target function with higher accuracy is realized on the self-adaptive grids.

Description

technical field [0001] The invention relates to a grid adaptive method based on error estimation. Background technique [0002] Credibility research has been a hot topic in CFD in recent years. The grid has always been one of the most important sources of CFD uncertainty. This problem has also plagued CFD workers, focusing on how to reduce the numerical dispersion error caused by the grid. People have come up with various methods, such as gradient-based adaptive methods, and methods to improve grid quality, such as smoothing technology, density box technology and grid topology research, etc., to improve grid quality and improve calculation results. But the results are not always so ideal, or the wrong results are obtained, or the results have improved but are still far from meeting the requirements. For example, the grid adaptation based on Euler's gradient (pressure or density), and the grid adaptation based on entropy or Mach number for viscous flow calculations sometimes...

AI Technical Summary

Problems solved by technology

[0002] Credibility research has been a hot topic in CFD in recent years. The grid has always been one of the most important sources of CFD uncertainty. This problem has also plagued CFD workers, focusing on how to reduce the numerical dispersion error caused by the grid. People have come up with various methods, such as gradient-based adaptive methods, and methods to improve grid quality, such as smoothing technology, density box technology, and grid topology research, etc., to improve grid quality and improve calculation results. But the results are not always ideal, or get the wrong results, or the results have improved but are still far from meeting the requirements

Images