Adjoint-based Shape Optimization of High-lift Airfoil using the NSU2D Unstructured Mesh Solver
Adjoint-based Shape Optimization of High-lift Airfoil using the NSU2D Unstructured Mesh Solver
复制标题
DOI：
10.2514/6.2014-0554
复制
发表时间：
2014-01
期刊：
影响因子：
--
通讯作者：
T. Nambu;D. Mavriplis;K. Mani
中科院分区：
文献类型：
--
作者：
T. Nambu;D. Mavriplis;K. Mani
采用离散伴随方法对一种高升力多段翼型进行了优化设计。非结构网格RANS求解器NSU 2D被用作流动和伴随求解器，并且LBFGSB（其是基于拟牛顿法的优化算法）被用于驱动形状优化。为了获得更大的设计空间，NSU 2D中的网格变形求解器通过在网格变形过程中重新计算网格刚度矩阵来修改。设计变量包括操纵参数，如襟翼和缝翼间隙，重叠和偏转，以及表面形状参数。优化了两种目标函数，即目标升力系数约束下的阻力系数和最大升力系数。在第一种情况下，优化形状减少了64个计数的阻力，在第二种情况下，优化形状的最大升力系数从基线值4.340增加到4.602。
A high-lift multi-element airfoil is optimized using a discrete adjoint method. The unstructured mesh RANS solver NSU2D is used as the flow and adjoint solver, and LBFGSB, which is an optimization algorithm based on a quasi-Newton method, is used for driving shape optimization. In order to achieve a larger design space, the mesh deformation solver in NSU2D is modified by re-computing the mesh stiffness matrix during the mesh deformation process. Design variables consist of rigging parameters such as flap and slat gap, overlap and deflections, as well as surface shape parameters. Two kinds of objective functions are optimized, a drag coefficient constrained by a target lift coefficient and a maximum lift coefficient. In the first case, the optimized shape reduces drag by 64 counts, and in the second case, the maximum lift coefficient of the optimized shape is increased from the baseline value of 4.340 to 4.602.