摘要
基于常规建模方法在汽车覆盖件曲面逆向造型中由数据点云很难构造出合适的A级曲面问题。文章在结合国内外学者得出的算法基础上,结合神经网络的优化提出了一种新的曲面造型方法,即NURBS曲面算法蒙皮造型方法。并在CFD仿真分析软件STAR-CCM中对整个车身进行外流场的数值模拟,通过对常规造型和蒙皮造型的数值模拟计算结果进行对比知:后者比前者在阻力系数上降低4.21%;升力系数降低1.69%;俯仰系数降低0.09%。该数据表明蒙皮造型方法能有效降低整车风阻确实提高了整车的动力性和其适用性。
Arming at the difficult problem associated with reconstructing a fitting and smooth of the a surface using points cloud in the reverse designing of an automobile curve surface was researched on the base of the method of normal modeling. Concluding on the basis of the algorithm of domestic and foreign scholars and combining with a neural network optimization, it puts forward a new method of surface modeling. This method is NURBS surface algorithm for skinning surface modeling method. And in the CFD simulation analysis software STAR-CCM, the whole car body was carried out the numerical simulation of flow field. Through conventional styling and skinning surface modeling numerical simulation results showing that: the latter than the former in the drag coefficient to reduce 4.21%; lift coeffwient decreases by 1.69%; pitch coefficient reduces 0.09%. The data shows the skinning surface modeling method can effectively reduce the vehicle drag indeed and improve the vehicle's power and its applicability.
出处
《机械设计与制造》
北大核心
2013年第10期103-105,109,共4页
Machinery Design & Manufacture
基金
陕西省科技攻关项目技持(2011K03-G20)
关键词
汽车覆盖件
点云数据
曲面重建
神经网络
蒙皮理论
CFD仿真分析
Automobile Curve Surface
Data Points Cloud
Surface Reconstruction
Neural Network
Skin Theory
CFD Simulation Analysis