基于双值参数化方法且考虑制造约束的炭纤维复合材料铺层优化

Ply optimization of carbon fiber-reinforced plastics laminates based on a gradient-based optimization method with manufacturing constraints

基于梯度的优化方法对炭纤维复合材料层合板的铺层数量和顺序进行优化。优化问题中以铺层质量为目标,并以刚度和制造约束为约束。采用改进双值参数化方法对铺层的材料性能进行插值,并基于凸规划对偶算法对优化问题进行求解。为了适应凸规划对偶算法的特点,将关于铺层角度的制造约束表述为少量非线性约束。同时引入离散度约束和惩罚指数以消除优化结果中的中间变量。算例结果验证了该优化方法的有效性。

We optimized the number of plies and the stacking sequence of carbon fiber-reinforced plastics laminates using a gradi- ent-based optimization method. To attain industrial relevance and reduce the risk of failure and the cost, certain manufacturing con- straints were considered in the optimization. Manufacturing constraints were： 1 ） candidate fiber orientations were limited to a finite set, 2） the bottom layer must be solid; 3 ） an intermediate void was not allowed, and 4） two or three identical contiguous layers were allowed at most. In the optimization, the mass was minimized with stiffness and manufacturing constraints. In order to avoid introducing linear constraints to the optimization model, a material property interpolation scheme called a bi-value coding parameter- ization scheme was modified and the manufacturing constraints concerning fiber orientation were expressed as nonlinear constraints. Meanwhile, a dispersion constraint and penalty index were introduced to eliminate intermediate variables. The optimization problem was solved by a convex programming dual algorithm. Numerical examples validate the optimization method.