基于定涡黏性假设的连续伴随优化方法研究

Study on Optimization Design of Continuous Adjoint Method based on the Constant Eddy Viscosity Assumption

基于网格节点位置坐标变分并结合通量雅克比矩阵技术,完成了定涡黏性假设下带湍流模型连续伴随系统的建立,极大降低了RANS方程下复杂伴随系统的推导难度,简化了伴随系统表达式,节约了计算资源。同时采用定涡黏性假设耦合SpaIart-Allmaras湍流模型,实现了二维叶栅壁面压力反设计的数值验证,针对变量梯度、优化过程及结果与全湍流系统进行了分析对比;并以通道内熵增为优化目标完成了某亚音速叶栅基于该假设的气动优化及叶栅吹风试验,试验结果验证了优化设计结果的可靠性。

The technique of the variation in the grid node coordinates combining with Jacobian Matrices are introduced to the establishment of the turbulent continuous adjoint method with the constant eddy viscosity assumption.This method reduces the derivation difficulty for the complex adjoint system of the RANS equations,which simplifies the formulas and saves the cost.Meanwhile,the constant eddy viscosity assumption and the Spalart-Allmaras turbulence model are employed to achieve the numerical validation for inverse design of 2D blade pressure distribution,and the gradients,optimization process and results are compared with the fully turbulent adjoint system.And with the objective of the entropy generation,a subsonic blade is tested with the numerical optimization and experiment of wind tunnel to demonstrate the reliability of the optimal results.