Optimization of Constraint Location, Orientation, and Quantity in Mechanical Assembly

Abstract

A mechanical assembly aims to remove 6 degreeoffreedom (DOF) motion between two or more parts using features such as fasteners, integral attachments, and mating surfaces, all of which act as constraints. The locations, orientations, and quantity of these constraints directly influence the effectiveness of a constraint configuration to eliminate DOF; therefore, constraint design decisions are crucial to the performance of a mechanical assembly. The design tool presented in this paper uses an analysis tool developed by the authors to explore a userspecified constraint design space and help the designer make informed decisions based on quantitative data so as to optimize constraint locations and orientations. The utility of the design tool is demonstrated with an assembly case study that contains both threaded fasteners and integral attachments. The results identify the opportunity for significant improvements by separately exploring individual design spaces associated with some constraints and further gains through a search of a multidimensional design space that leverages interaction effects between the location and orientation variables. The example also highlights how the tool can help identify nonintuitive solutions such as nonrectilinear, nonplanar parting lines. A tradeoff study demonstrates how the design tool can quantitatively aid in optimizing the total number of constraints. Adding constraints generally improves an assembly's performance at the expense of increased redundancy, which can cause lockedin stresses and assembly inaccuracies, so the design tools helps identify new/removable constraints that offer the greatest/least contribution to the overall part constraint configuration. Through these capabilities, this design tool provides useful data to optimize and understand mechanical assembly performance variables.