| Along with the increase of hypersonic aircraft flight speed, the contradiction of the lift-drag ratioand aerodynamic heating is becoming more and more serious. It gets more important to findappropriate design process to comprise this contradiction. By far, MDO methods have been takenmainly to solve these problems. The design of traditional supersonic aircraft is based on separateduncoupling patterns, in which work on each discipline is independent and different disciplinedestinations are chosen for designs and optimizations. This method is capable to solve some problemand satisfy the requirement in designs though. Ignoring the coupling effects of performances in eachdiscipline and design parameters, some design allowance is wasted and on fixed constraint, satisfyingresults can’t be got in design process. According to genetic algorithm and definition of Paretooptimum solution, a multi-objective genetic optimization algorithm based on Pareto front isconstructed. Solving partial differential equation is involved in the optimization process, and largepopulation iterations may add large calculation cost. Therefore, aerodynamic thermal numericalmodel based on multinomial response front and Kriging response front is constructed as agent. Anaircraft is designed whose configuration resembles X-33’s. Wing is taken as main optimization object,whose optimization parameters are the length of exposed wing, backward swap and diameter ofleading edge of the wing. The optimum parameters are lift-drag ratio and density of the thermal flowfrom leading edge. Taking sample point by Latin hypercube experimental method, aerodynamicthermal characteristics at sample points are calculated through CFD method. After the results arecalculated, polynomial response front and Kriging response front are constructed respectively. Thencomprehensive optimization on aerodynamic and thermal is processed based on response front model.At last, a discussion about the feasibility of the method is made. |
PDF Full Text Request