Research On The Aerodynamic Shape Optimization Design Of Vehicle Based On Free Form Derformation And Surrogate-based Optimization

Hypersonic glide vehicle(HGV)has some advantages of high speed,strong maneuverability and great range.This thesis focuses on the aerodynamic shape design of the HGV,which facing some challenges,such as the complexity of shape modeling,the time-consuming of numerical calculation,the nonlinearity of iterative optimization,and the feasibility of engineering application.Based on the free form deformation(FFD)method and surrogate-based optimization(SBO),the thesis has optimized the windward side of HGV,and achieved an optimal shape by designing the stability component and control component.Shape parametric modeling is the beginning of aerodynamic shape design.Some limitations need to be reduced when the FFD method is applied to parametric modeling of a HGV,such as the large computational complexity,low operational efficiency,and geometric discontinuity.In order to solve these problems,the FFD control lattice is simplified;the dimensions of the control lattice are reduced from 3 to 2,and the displacement changes of the control points are limited to one direction.The boundary of the traditional control lattice is transformed into the border of windward side.Take the non-uniform rational B-splines(NURBS)as the basis function,this thesis presents an Edge-NFFD method.The applicability of the Edge-NFFD method has been validated by modeling three typical HGV shapes(lifting body,blended wing body and waverider body),and all results show this method will provide better support for shape parametric modeling of the HGVs.A conflict exists between accuracy and efficiency when predicting aerodynamic performance of the HGV.Multi-fidelity calculation method can balance the accuracy and efficiency,which will provide the aerodynamic data to build the Co-Kriging surrogate model.The high-fidelity calculation method is based on the Navier-Stokes equations and turbulence model.The low-fidelity calculation method is based on the Euler equation and friction-based viscosity correction.Meanwhile the computational mesh generation is also a time-consuming process.By coupling the mesh deformation with shape parametric modeling,the geometric model and computational mesh can be updated synchronously.This thesis develops an improved elasticity-based mesh deformation method based on the robust parameters.The robust and validity of this improved elasticity-based method have been validated by several test cases.Based on the classical Kriging surrogate model,the Co-Kriging surrogate model with multi-fidelity is presented by expanding the structure of sample data.Compared with classical Kriging surrogate model,the Co-Kriging surrogate model has been verified with less time-consuming and higher accuracy.On this basis,an infilling sampling criterion of expected improvement(EI)is introduced to develop the framework of a SBO algorithm.In order to optimize the function of max(EI),the multi-objective evolutionary algorithm based on decomposition(MOEA/D)is employed.Based on the SBO,the maximum L/D of a base lifting body is optimized by manipulating its shape parameters of Edge-NFFD method.During the optimization,the internal loading volume is taken as the design constraint.The optimal shape has a larger L/D than base shape,while satisfying the internal loading constraint.The aerodynamic shape design of the HGV is a complicated system engineering.Firstly,the single/double vertical tails are designed to improve the stability of the HGV.Then,considering the current HGV does not have control ability,two symmetrical Flaps are designed as the control components.Finally,the performance evaluation of the optimal HGV is carried out from some aspects,such as aerodynamic force,aerodynamic heat,stability,controllability and loading volume,etc.The optimal shape has a good overall performance in multiple evaluated states.At the same time,this thesis points out the existing limitations of the optimal shape,which provides a guidance for its engineering application.This thesis is based on the future development of the HGV and researched some key methods in aerodynamic shape design,which including the shape parametric modeling,aerodynamic performance calculation,multi-fidelity surrogate-based optimization and overall design constraints.These methods are applied to a typical HGV shape,and the optimal results are better.The research work in this thesis is of great significance to the development of HGVs in the future.