High Efficient Multidisciplinary Design Optimization Of Static Aeroelasticity For Large-scale Structure

With the lightweight requirement of modern aircraft structure,it brings larger wing flexibility and more significant static aeroelastic effect.The traditional engineering structure design method which usually focus on the structure strength and stiffness can’t consider the static aeroelastic effect such as lift effectiveness,aileron effectiveness and aerodynamic center location changes dircetly.With this disadvantage,the structure designed by the traditional method is hard to meet the need of different disciplines.With the modern aircraft structure complexity and the need of digging deeper structure potential increased,the number of the structue optimization variables and the scale of various disciplines model become even larger.The large-scale aeroelastic-structure MDO method face challenges such as low computation efficiency,large computation capacity and derivative get difficulties when dealt with this large-scale structure optimization problem.These challenges limit the further fine optimization of the aircraft structure design.In this thesis,we intend to propose high precision and efficiency static aeroelastic performance and static aeroelastic design sensitivity methods.Combing with large-scale structure optimization method,these two methods will solve the problems of large-scale aeroelastic-structure MDO faced.The main parts of this thesis show as follow:At first,for imporving the efficiency and accuracy of the elastic aerodynamic load calculation and sloving the conflict that the Computational Fluid Dynamics(CFD)method takes too much time to calculate the aerodymamic load,a fine sectional correction panel method is proposed.The method which based on the rigid wing CFD data at different attack angles correct the subsonic panel method and get corrected matrixs.Using these corrected matrixs and downwansh segmented way,the method can calucate the aerodynamic load after the wing deformation in an efficiency way and keep high accuracy as well as CFD result.In order to improve the accuracy of the fine sectional correction panel method further,a panel mesh optimization method is also proposed.This method use CFD data,ISGHIT software platform and fine sectional correction panel method to get the optimal mesh.The aerodynamic load result calculated under the optimal mesh is closest to the CFD result.To improve the data transfer accuracy and efficiency of the aerodynamic and structure model coupling interface,the Radial Basis Function(RBF)method which performance high numerical accuracy and flexibility is adopted.Rules about selecting compact radius and surface interpolation nodes are set up to improve the computation efficiency of the RBF method too.Secondly,the definition of aeroleastic performance used in engineering is simple,can not fully reflect the impact of static aeroelastic effect and handling stability characteristics of the aircraft.By using the precise definition of the static aeroelastic performance and the complex-step derivative approximation,the method of calculating static aeroelastic performance such as lift effectiveness,aileron effectiveness and the chord location of aerodynamic center change rate is proposed after using the fine sectional correction panel method.Especially,an inner-outer iteration method is proposed to calculate the aileron effectiveness.This method can get the aileron effectiveness and the aircraft steady roll speed at the same time.It will be more fully reflect the wing flexibility and aileron angle impact to the aircraft mobility.In order to give designers more reference information,the thesis also presents an elastic lift compensation algorithm and an elastic roll speed compensation algorithm which will get the attack angle and aileron angle of the elastic wing to achieve a given load and a given roll speed.In addition,the thesis also proposes methods to estimate the wing divergence speed and reversal speed under a low accuracy.Thirdly,the gradient-based optimization algorithm is proven to be an effective method during solving large-scale structure optimization problem.In order to give this optimization algorithm support,the method of calculating the static aeroelastic design sensitivity —derivative of the structure optimization variables with respect to static aeroelastic performance —is proposed after using the fine sectional correction panel method.To satisfy the modular organization requirement of the of multidisciplinary optimization program,a high efficient and standardized reading/writing static aeroelastic program is proposed.By adopting OPENMP technique,the computation efficiency of this program can be improved greatly.Taken m6 wing static aerodynamic problem,CFD and Nastran results as examples,it proven that the methods and program present in this thesis can get the correct result and keep high efficiency.Finally,the principles and characteristics of the aeroelastic-structure MDO in this thesis and the constraints that used in engineering commonly are described.In particular,.MASS file is proposed to solve the problem of different central mass load in different load case.The large-scale structure optimization program developed by our research group is also described,including optimization algorithm,reduction techniques of design variables,constraint screening technology,programm organization and parallel processing.Base on this program,the static aeroleastic performance constraints are integrated,forming a large-scale aeroelastic-structure MDO program.On the basis of the preliminary work in the thesis,a flying wing UAV structure optimization with up to 699 variables,four limit load cases,two aerodynamic conditions and up to ten kinds of constraints is presented and checked.The results show that the large-scale aeroelastic-structure MDO method presented in this thesis has high computational efficiency.The optimization results reduce the structural weight by 15.66% while the constraints keep in line with engineering setting.In addition,the multidisciplinary optimization result under five different combinations of restrictions show the influence of the different constraints on the structure weight which will helps engineers to master and understand the design law of this structure type.