Multi-objective interface design for uniform assembly

Precision assembly is the final performance safeguard in the intelligent manufacturing stage of precision electromechanical systems. Uniform assembly, featuring uniform stress distribution in assembled parts, is of great significance in ensuring the accuracy and performance stability of precision electromechanical systems. Assembly interface is the fundamental existence in assembled systems. Targeting uniform assembly, this study leverages assembly interface's capacity in regulating stress distribution, and further extents it to simultaneously control the dynamic performance of assembled systems. Specifically, the multi-objective assembly interface stiffness design problem of concurrently reducing the stress peak at the assembly interface and increasing the natural frequency of the assembled structure is addressed. ε-constraint method and non-stationary multi-stage assignment penalty function method are utilized for modelling. The solver is an enhanced particle swarm optimization (PSO) method named MREOBL-PSO, which is developed by introducing and combing the opposition-based learning (OBL) and multi-dimensional random elite mutation strategy. Numerical simulations are experimentally verified. Applications of multi-objective assembly interface stiffness design to bolted flanges and aero-engine rotor systems are demonstrated.