飞翼气动优化中参数化和网格变形技术

几何参数化和网格变形是飞行器气动外形数值优化迭代过程中的两个关键技术。基于非均匀有理B样条(NURBS)的自由型面变形(NFFD)技术对几何表示形式具有普适性,距离权函数(DWF)网格变形技术具有计算快速和网格拓扑无关性,两者广泛应用于曲面优化。基于NFFD技术改进了反求参数的Newton迭代算法,并通过提高物面附近网格刚度改进的距离权函数(IDWF)技术使其适用于更大程度的网格变形。还提出了改进后的参数化和网格变形两种技术并行计算的具体实现算法。结合离散伴随方法,使用参数化和网格变形技术,实现了由NACA0012初始翼型到飞翼标准翼型EH1590的反设计;针对某飞翼标模完成了单点全机升阻比优化,升阻比提高约18%。数值结果表明,建立的NFFD和IDWF动网格技术可满足飞翼气动外形优化参数化和快速网格变形的需求。

Parameterization and grid deformation techniques for flying-wing aerodynamic optimization

Geometry shape parameterization and grid deformation techniques are two key issues for aerodynamic shape optimization of aircraft by CFD. The non-uniform rational B-splines (NURBS)-based free-form deformation (NFFD), which is universal for arbitrary representation of geometry, and the dynamic grid technique based on distance weighted function (DWF), which is independent from grid topology and fast on calculation, are widely used for shape optimization. In this paper, the NFFD technique is introduced in detail and the Newton iteration algorithms for inverse local coordinates calculation are developed. The DWF is improved (IDWF) to satisfy larger range grid deformation by increasing the stiffness of grid cells near surface. The parallel implementation algorithms for both techniques are designed and introduced in detail. With an optimization solver based on gradient calculated by discrete adjoint method, the standard flying-wing airfoil EH1590 is inversely designed form the initial airfoil NACA0012, and the ratio of lift and drag is optimized and increased by 18% for a whole flying-wing aircraft in a single design state. The result indicates that for flying-wing aerodynamic shape optimization, the NFFD technique is sufficient for shape parameterization and IDWF technique is efficient for grid deformation.