Sandwich structure T-joints are increasingly broadly applied in aviation and aerospace industries due to the need for lightweight design. This paper deals with the lightweight optimization of a typical adhesively bonded Nomex honeycomb-core sandwich T-joint in side bending load, considering the strength constraints. The optimization problem, with discrete and continuous design variables, is a compound optimization problem involving size optimization for the whole structure and stacking sequence optimization for multiple variable-thickness composite laminates. A self-adjusted parametric modeling with user-defined suppression process is proposed. An integrated combination of progressive damage model methodology, self-adjusted parametric modeling with user-defined suppression process and multi-island genetic algorithm is applied for the optimization problem. The optimization result showed 30.75% weight reduction compared to the original T-joint configuration. On the basis of history data, we investigate the correlations between design variables and concerned constraint variables.

由于需要轻巧的设计，三明治结构的T型接头在航空和航天工业中的应用越来越广泛。考虑到强度约束，本文针对侧弯载荷下典型的粘结性Nomex蜂窝芯三明治T形接头的轻量化优化进行了研究。具有离散和连续设计变量的优化问题是一个复合优化问题，涉及整个结构的尺寸优化和多层厚度可变复合材料层压板的堆叠顺序优化。提出了一种具有用户自定义抑制过程的自调整参数建模方法。渐进式损害模型方法论的综合结合，优化问题采用了用户自定义抑制过程和多岛遗传算法的自调整参数建模。优化结果显示，与原始T形接头相比，重量减轻了30.75％。基于历史数据，我们研究了设计变量与相关约束变量之间的相关性。