一种用于分离流动的网格自适应算法

对于可压缩粘性流动,提出利用流场速度的紊乱度作为指示变量进行网格自适应.Jameson中心格式有限体积法、五步Runge-Kutta时间推进法/双时间推进法求解定常/非定常N-S方程.基于雷诺平均N-S方程模拟紊流,选用SA一方程模型.在数值求解二维静态失速和动态失速问题过程中,加入网格自适应算法,提高数值模拟对流动分离特性的捕捉和分辨能力.算例结果表明在流场发生失速后,运用本文的自适应算法能够在增加少量网格单元的情况下明显提高计算精度.

Grid adaption technique for separation flow

Based on the compressible viscous flow,disorder degree of the flow velocity is proposed to be the indicator variable for a grid adaption.Jameson cell-centred scheme of Finite Volume Method and five-stage Runge-Kutta approach/dual time-stepping approach are used to solve steady/unsteady N-S equations.Spalart-Allmaras One-Equation is employed to the simulations of turbulent flows.In order to improve the capability of catching flow separation characteristics,the grid adaption method is used during the numerical simulation of the two-dimensional static stall and dynamic stall.The example results indicate that the grid adaption method can obviously improve the calculation accuracy in the case of increasing a small amount of grid cells.