将数值优化方法和CFD仿真结合,对某轿车前轮阻风板进行了优化.以前轮阻风板的高度和安装位置为变量,通过DOE拉丁试验方法选取了20组样本点.利用CFD仿真得到各样本点的响应值,由响应值建立Kriging近似模型.在验证了近似模型的可信度基础上,通过多岛遗传算法对近似模型进行全局寻优.优化结果表明:在不增大升力的前提下,前轮阻风板最多可将气动阻力由827.4 N降到792.4N,降幅为4.23%.若阻力、升力的权重分别设为0.85和0.15,气动阻力降为796.8 N,降幅为3.70%;气动升力降为382.6 N,降幅为6.55%.
The air dam skirts of the front-wheels of a certain car were optimized by combining data optimization method and CFD simulation.The height and the installation coordinate of the air dam skirts were chosen as the variables and 20 groups of sample data were adopted by the method of DOE.The response values corresponding to the sample were achieved by CFD simulation,and then Kriging model was established.The credibility of the model was verified and global optimization was carried out by multi-island genetic algorithm.The result showed that the aerodynamic drag could be reduced by 4.23% at the most,from 827.4 N to 792.4 N,based on the prerequisite that the aerodynamic lift wouldn't be increased.And when the weight was set at 0.85 and 0.15 separately,the aerodynamic drag could be decreased by 3.70% to 796.8N while the aerodynamic lift could be reduced by 6.55% to 382.6 N.