A comparbon of optimization techniques for solving the coupling suspension problem

With the tremendous computational power available to designers today, the temptation to include "everything" in our analyses grows greater by the day. As designers, there is a desire for a perfect understanding of how a product will behave under a multitude of conditions. This desire is restrained by the time constraints on product development. The need to produce items "Faster, Better, and Cheaper" dictates a timeline that may not allow for both highly detailed analyses and the consideration of many different possible design points. Research in the area of system reduction has shown the potential for considerable computational savings with a minimal sacrifice in solution accuracy. This trade off is based on the selection of what system analysis couplings to include, and which may be ignored for a time and included later. This paper compares multiple "black box" improvements to solving the coupling suspension problem (CSP), as well as an interactive solution technique that allows a designer more control over the problem at hand. The comparisons are based on the different techniques' abilities to exploit good regions of the design space, the amount of designer control, and computational impact A technique that allows a designer full control over the problem topology through visualization is also demonstrated.