面向非结构混合网格高精度阻力预测的梯度求解方法

针对非结构混合网格的特点，通过改进传统的Green-Gauss梯度求解方法，提出了一种可提高非结构混合网格黏性计算精度的节点型Green-Gauss梯度求解方法。利用改进后的方法，完成了DLR-F4翼身组合体算例的计算和对比分析。改进后的梯度求解方法残差收敛更好，下降量级更多，阻力系数和试验吻合更好，激波区域压力分布和分离区域流场细节的模拟更精确，说明改进后的梯度求解方法有效提高了程序的鲁棒性和阻力预测精度，验证了方法的有效性。采用改进后的方法对第5届AIAA阻力预测研讨会的通用研究模型（CRM）进行了详细的模拟分析，结果表明：改进的梯度求解方法更加适用于非结构混合网格的黏性计算，计算精准度达到国际同类CFD软件水平，进一步验证了改进方法的可靠性。

Gradient calculation method of unstructured mixed grids for improving drag prediction accuracy

Based on the characteristics of the unstructured hybrid grid, a new gradient calculation method named node-based Green-Gauss is presented by improving the original Green-Gauss method. To validate the new method, a numerical simulation of the DLR-F4 wing-body configuration is performed. The numerical results are analyzed and compared with that of the original Green-Gauss method. The results indicate that better residual convergence and more order of magnitude decrease can be obtained with the new method, the drag coefficients show a good agreement with experiment results, and the pressure distributions in the shock wave and separation region can be computed more accurately. It is found that the new method has excellent drag prediction accuracy and more robustness. Moreover, the Common Research Model (CRM) adopted in the 5th AIAA Drag Prediction Workshop is studied by using the method proposed. The results demonstrate that the new method is much more suitable for computation of viscosity of the unstructured hybrid grid, and the accuracy reaches the same level of famous CFD software. All of this further verify the drag prediction accuracy of the new method.