Part consolidation (PC) is becoming a viable cost savings approach due to the increased design freedom associated with industry adoption of additive manufacturing. However, there is little research focused on mathematical approaches for assembly level design generation, with most work aimed at providing best practices for merging several parts into one. This article presents a novel topology optimization approach to PC that determines the ideal number of parts, their geometry, and optimal joining pattern, without bias towards the original assembly. Multiple layered design domains are created, and a joining domain that determines the connections between parts is introduced. A multiobjective problem statement optimizes the complex trade-off between compliance, support structure volume, surface area, and number of joints, to minimize the total cost of the final assembly. Design variable initialization and boundary condition placement are discussed for problems with multiple domains. Three test cases are presented and solved for a range of cost trade-offs to demonstrate optimized solutions as design objectives are varied.

中文翻译：

---

增材制造的零件整合：一种多层拓扑优化方法

由于与行业采用增材制造相关的设计自由度增加，零件整合 (PC) 正成为一种可行的成本节约方法。然而，很少有研究关注用于装配级设计生成的数学方法，大多数工作旨在提供将多个零件合并为一个的最佳实践。本文介绍了一种新颖的 PC 拓扑优化方法，该方法可确定理想的零件数量、它们的几何形状和最佳连接模式，而不会偏向原始组件。创建了多层设计域，并引入了确定零件之间连接的连接域。多目标问题陈述优化了顺应性、支撑结构体积、表面积和接头数量之间的复杂权衡，以最小化最终组装的总成本。针对多域问题讨论了设计变量初始化和边界条件放置。提出并解决了三个测试案例以进行一系列成本权衡，以展示随着设计目标的变化而优化的解决方案。