Research On Aerodynamic-thermal-structural Multidisciplinary Analysis And Efficient Optimization Strategies

High-speed flight vehicle is animportant direction in future development of flightvehicles, and the aerodynamic-thermal-structuralcoupled problem isone of the researchfocuses in this field. Due to the complexity of the aerodynamic-thermal-structuralcoupledproblem, it is very important and interesting for high-speed flight vehicles design toreasonably simplify in terms of actual situations and establish a correctaerodynamic-thermal-structural multidisciplinary analysis model, involving aerodynamicheating analysis, transient heat transfer analysis, thermal stress analysis, thermal modeanalysis and thermal flutter analysis. Since computationally expensive high fidelitynumerical simulation modelshave been widely employed in each discipline,it is difficult toefficiently achieve the optimal solution ofthe aerodynamic-thermal-structural coupledproblem without decreasing precision.The modularized aerodynamic-thermal-structural multidisciplinary analysis model ofhigh-speed flight vehicle is established andthe efficientoptimization strategiesusingtheadaptive metamodel based on fuzzy clustering are proposed.Theaerodynamic-thermal-structural multidisciplinary integrated design optimizationfor F104lifting surface isperformed.More detailed work of this dissertation is presented as follows:1)According to the background and significance of this subject, the state-of-the-artinthe aerodynamic-thermal-structuralcoupled analysis method and the multidisciplinarydesign optimization strategy is summarized and discussed, which provides references forthe further study.2) Themathematical model of the coupled aerodynamic-thermal-structural problem issummarized.The theoretical foundations of severalsub-analysis modelsin thismultidisciplinary coupled problemare introduced.Besides,the classification of theaerodynamic-thermal-structural coupled problems according to differentcouplingrelationships is introduced, and then, based on the classification,the analysisformulationsare simplifiedin accordance withtheflight conditions. 3) In aerodynamic heating analysis with non-zeroangle of attack, precision ofempiricalalgorithm for leeward is relatively low, whileCFDsimulation is rathertime-consuming. Thus, an aerodynamic heating analysis method using varying complexmodelsis proposed, which ensures theprediction precision for leeward and greatly reducesthe computation cost. Considering strong coupling relationships, complicated coupledaerodynamic-thermal-structural multidisciplinary problem can be simplified byusingone-way solution and loose coupling methods. Disciplinary analysis modelsincluding theaerodynamic heating empiricalalgorithm, the transient heat transfer, the aerodynamicforce-aerodynamic heating CFD calculation, the thermal stress, the thermal modal and thethermal flutter are built. Furthermore,an integrated simulation framework of modularizedaerodynamic-thermal-structural multidisciplinary problem is established. Based on designstructure matrix, the mathematical model of thismultidisciplinary design optimization isformulated.4) Aiming at the time-consuming problem in high-speed flightvehiclesaerodynamic-thermal-structuralintegratedanalysis, a metamodel-based globaloptimization using fuzzy clustering for design space reduction (FCR) is proposed. In FCR,fuzzy c-mean method is improved to cluster even distribution approaching data points inhigh dimensional, and automatic space reductionmethod based on fuzzy clustering isproposed to gradually shrinkthe design space, and in reduced design space metamodels areconstantly renewed by sequential sampling to achieve the global optimal solution. FCRoptimization strategy based on Lagrange multiplier method is proposed to deal withexpensive constrained optimization. Comparative study shows that FCR has better globaloptimizing ability and efficiency,which canefficiently resolve the multidisciplinary designoptimization problem with high fidelity analysis models.5) Based on integrated simulation frameworkforaerodynamic-thermal-structuralintegratedanalysis,each disciplinary analysis model forF104lifting surfaceisestablished. And then,integrated analysis for the entireflight trajectoryis carried out.At the same time, mathematical model for the optimization problemof thelifting surface is established and design optimization forthis lifting surface is performed by using FCR optimization strategy based on the Lagrange multiplier method. Results showthat the feasible optimal solutions satisfyingthe constraint conditions can be obtained withlimitedcomputational cost.6) Sincethere may exist multiple local optima in aerodynamic-thermal-structuralmultidisciplinary design optimization, two kinds of optimizationmethods for multipleoptima problems are proposed based on FCR optimization strategy. Optima validationcriteria and matching criteria are developed to filter local optima during theoptimizationprocess. The numerical examples with multiple optima are used to validate the performanceof the algorithms. The optimization results show that the proposedmethodspossessfavorable capability for searching multiple optima, higheroptimization efficiencyand betternumerical stability.