旋翼干扰流场数值模拟中的新型Blend低速预处理方法

针对定常低速预处理在双时间步长内收敛性差、旋翼干扰流场数值模拟中非定常低速预处理自动退化为无预处理的问题，在非定常低速预处理方法的基础上，采用量级分析方法研究了对流项和黏性项量级，通过组合定常和非定常预处理特征值矩阵，建立了一种新型Blend低速预处理方法。通过圆柱非定常扰流、前飞直升机旋翼/机身干扰非定常流场和直升机/舰船干扰非定常流场的数值模拟，对各低速预处理方法的收敛性和预测精度进行了细致研究。结果表明，本文方法解决了在旋翼干扰流场数值模拟时非定常低速预处理自动退化为无预处理的问题，相比定常低速预处理大幅提高了收敛效率。

A new Blend low speed preconditioning method for numerical simulation of rotor interaction flowfield

Steady low speed preconditioning is of poor convergence in dual-time iterations, and unsteady low speed preconditioning automatically reverts back to no preconditioning in numerical simulation of rotor aerodynamic interaction. To overcome these problems, we conduct an order analysis to blend the steady and unsteady low speed preconditioning eigenvalues matrices based on the unsteady low speed preconditioning framework. A new Blend low speed preconditioning method is developed from altering the dissipation formulation using this matrix.The unsteady flow around the circular cylinder, unsteady rotor-fuselage aerodynamic interaction in the helicopter forward and coupling ship/rotor flowfields are simulated to validate the proposed method. The numerical results indicate that the proposed method can solve the problem of automatical reversion of unsteady low speed preconditioning back to no preconditioning, and can improve significantly convergence of steady low speed preconditioning.