Studies Of Dynamic Mesh Technique And Its Application In Aeroelastic Numerical Simulation

Today, numerical simulation of aeroelasticity has already become a hot spot of the Computational Fluid Dynamics(CFD), and the generation of dynamic mesh is an essential technology about it. This thesis develops two hybrid dynamic mesh techniques based on traditional dynamic mesh techniques, and studies their applications in numerical simulation of aeroelasticity.Aiming at a complex multi-block structured grid, an efficient hybrid dynamic mesh technique is presented. Combine Radial Basis Functions(RBFs) method and Transfinte Interpolation(TFI) method together, this present method can be devided into two steps. At first, all block vertexes with known deformation are selected as control points for RBFs interpolation, and apply RBFs interpolation to get the grid deformation on block edges according to the deformation of control points. Then, an arc-length-based TFI is employed to efficiently calculate the grid deformation on blovk faces and inside ech block. Generaly, the number of block vertexes is very small, time cost in RBFs interpolation can be ignored, which makes this present dynamic mesh technique efficient, almost the same with TFI method. On the other hand, the deformability of this method is as strong as RBFs due to the deformation on block edges is calculated by RBFs.A novel hybrid dynamic mesh technique is developed for hybrid mesh which is based on RBFs and Delaunay graph mapping(DGM). Step one, a set of very coarse background grid is designed according to the computational mesh, the cell type is triangle for two-dimensional(2D) cases and tetrahedron for three-dimensional(3D) cases. Step two, select all boundary points of background grid as control points for RBFs interpolation. Step three, refine the background grid near geometry surface for ensuring it covers the same area with computational mesh. Step four, locat the computational mesh at the background grid, build the relationship of every computational mesh point with the cell in background grid which it belongs to. Step five, apply RBFs interpolation to move the background grid. Step six, the deformation of computational mesh can be easily worked out according to the deformation of background grid through the relationship between computational mesh and background grid. The application of RBFs interpolation makes the deformability of this hybrid dynamic mesh technique much stronger than DGM. Like RBFs-TFI hybrid dynamic mesh generation method, time used in RBFs interpolation is few because the background grid is very coarse, so the RBFs-DGM dynamic mesh technique is very efficient. In the whole moving mesh process, all information about computational mesh used is only the coordinate value of grid point, therefore, this present hybrid dynamic mesh technique has no limit for mesh type.Aiming at dynamic mesh, a parameterized mesh quality criteria is proposed for measuring the overall quality of mesh. This mesh quality criteria considers the contributions of the change of grid cell to the overall mesh quality in quality, volume and position when the mesh is moved. The effect of dynamic mesh to mesh quality is studied through steady numerical simulations, which lay the foundation for the study of the application of dynamic mesh technique in aeroelasticity numerical simulation. The numerical simulations demonstrate that, this mesh quality criteria can well reflect the change of mesh quality in dynamic mesh, about the same as the change of aerodynamic results. On the other hand, the RBFs-DGM dynamic mesh technique has good performance in efficiency and deformability, is an ideal dynamic mesh technique for aeroelasticity analysis.The present RBFs-DGM hybrid dynamic mesh technique is applied in 3D aeroelasticity problems. Based on the wing structural deformation measurement, a structural stiffness inversion technique is presented for high-aspect-ratio wing model in wind tunnel test. By simplifying the wing model to a single-beam model, the structure stiffness distribution can be decomposed to bending stiffness distribution and torsional stiffness distribution. Take the bending stiffness as an example, select a few appropriate basis functions for matching the bending stiffness, and calculates the deflection and torsional angle under the aerodynamic loading respectively. Through comparison of numerical results with experimental data, the weight of every basis function can be worked out by least square method. Then the bending flexibility distribution and torsional flexibility distribution are inverted. After that, the flexibility matrix of wing model can be extracted by finite element analysis software. Finally, the static aeroelastic correction problem in every state can be solved by applying the static aeroelastic calculation method. Compare to measuring the structural deformation in every test state, the present technique is more efficient, and it has good prospect in engineering applications.