Multi-Objective Design Optimization Applied to Light Weighting a GM Equinox Rear Cradle 2007-01-1724

A redesigned 2005 Chevrolet Equinox rear cradle was optimized using metamodel techniques coupled with large scale finite element simulations. The rear cradle was redesigned to accommodate an electric drive system to convert the equinox to a hybrid as part of the General Motors sponsored student project Challenge X. The design optimization uses metamodels to represent the finite element responses of the cradle under multiple loading situations. The metamodels provide an optimization technique that reduces design time and the number of required simulations compared to the traditional, iterative type simulation based optimization. The result of the optimization was a 12% weight savings over the initial design with a higher confidence in safety. The optimization results were confirmed by an extra set of finite element simulations outside of those performed for the optimization process.