压气机角区分离是一类典型的复杂分离流动。目前常用的线性涡黏性模型预测得到的结果常与实验结果有较大差别。使用雷诺应力模型的计算或大涡模拟等可以得到良好的预测结果,但受计算量和稳定性限制,目前尚未得到广泛应用。本文探索了一种形式简单的新型非线性涡黏模型在压气机角区分离流动预测中的可行性和有效性。计算结果表明:该非线性模型有效地改进了精算精度,压力分布和分离起始位置等的预测值得到了明显改进。同时,使用该模型时的额外计算量小于5%且计算的收敛性没有明显恶化。本文结果表明该非线性涡黏模型在叶轮机械复杂流动中的应用值得进一步探讨。
In this work, the compressor corner separation is predicted with a simple nonlinear eddy-viscosity model, the Quadratic Constitutive Relation (QCR) model. Compared to the linear Boussinesq model where the turbulent stress is assumed to linear with the strain, in this model a correction tensor term is added to resolve the turbulent anisotropy. Numerical results show that improved pressure distribution is obtained at various incidents and the numerical results agrees better with the experiment. The flow details in the corner separation region, including the anisotropic turbulent stress distributions and the velocity distributions in the corner region are discussed for better understanding of the flow physics that results in the improved prediction. QCR model is proved to have very low cost and be able to improve the prediction of complex compressor corner separation.