Based On The Response Surface Method For Aerodynamic Optimization Design

Aerodynamic optimization design technology based on CFD becomes a very active object in the CFD field. With the rapid advancement in computing power, more general design techniques called direct optimization method are suggested in recent years. It is likely that the design result would fall into just local optima without reliable gradient based on these traditional optimization methods. In recent years the response surface methodology(RSM) was applied in many design fields, because of its high efficient and practical. RSM is a collection of statistical and mathematical optimization techniques. RSM can approach the global optimum in design space based on analyzing a set of candidate design points. RSM can fit complex response relations if the response models are appropriate. In general the models are simple algebra equations. RSM can find a solution near the global optimum only based on the character of the cost functions. In this paper, an optimization design method based on RSM was developed in order to improve aerodynamic performance of the airfoils and wings.The main aspects of this paper are as follows:1. RSM was advanced for aerodynamic optimization design. A complete optimum design programme based on RSM was established.2. Aerodynamic inverse design of airfoils based on RSM by using Navier-Stokes equations was studied.3. Optimum aerodynamic design of transonic airfoils at single or multiple design-points, subject to specified constraints based on RSM by using Navier-Stokes equations was studied.4. Optimum aerodynamic design of transonic wings at single or multiple design-points, subject to specified constraints based on RSM by using Euler equations was studied.5. Optimum aerodynamic design of transonic wings subject to specified constraints based on RSM by using Navier-Stokes equations was studied.The optimum design results indicate that RSM can be successfully applied to improve aerodynamic performance of transonic airfoils and wings at single or multiple design-points, subject to specified constraints. This method is utility and valid, and will be very attractive for practical use.