泡沫铝填充薄壁金属圆管吸能特性的预测及优化方法

Prediction and Optimization Method of Energy Absorption Characteristics of Thin-walled Metal Circular Pipe Filled with Aluminum Foam

为提高泡沫铝填充薄壁金属圆管的吸能特性并进行高效率结构设计,基于均匀拉丁超立方试验设计方法和径向基函数神经网络,构建轴向冲击下泡沫铝填充铝管吸能特性的近似模型,并将近似模型嵌入遗传算法中实现构件的结构优化。研究结果表明:基于径向基函数神经网络构建的近似模型的拟合优度大于0.99、均方根误差小于0.08,且近似模型的计算耗时仅为数值计算的0.56%;近似模型在保证较高精度的同时可大幅提高计算效率;对泡沫铝填充铝管的结构优化发现,通过增大圆管半径和壁厚、减小高度可使泡沫铝填充铝管的平均压缩力最大化,反之,可使泡沫铝填充铝管的峰值压缩力最小化;基于遗传算法的多目标优化显著提高了泡沫铝填充铝管的吸能特性;该研究成果可为泡沫铝填充薄壁金属圆管的快速设计和优化提供参考。

In order to improve the energy absorption characteristics of aluminum foam-filled thin-walled metal circular tube and realize high-efficiency structural design,an approximate model of the energy absorption characteristics of aluminum foam-filled aluminum tube under axial impact is established based on the uniform Latin hypercube design method and radial basis neural network,and it is embedded into the genetic algorithm to realize the structural optimization of components.The results show that the goodness of fit of the approximate model based on the radial basis neural network is greater than 0.99,the root mean square error is less than 0.08,and the calculation time of the approximate model is only 0.56%of the numerical calculation.The approximate model can greatly improve the calculation efficiency while ensuring higher accuracy.Through the optimization of the structure of aluminum foam-filled aluminum tube,it is found that the average compression force of aluminum foam-filled aluminum tube can be maximized by increasing the radius and wall thickness of the circular tube and reducing the height,and on the contrary,its peak compression force can be minimized.The multi-objective optimization based on the genetic algorithm significantly improves the energy absorption characteristics of aluminum foam-filled aluminum tube.The research results can provide a reference for the rapid design and optimization of aluminum foam-filled thin-walled metal circular tube.