Multiple criteria structural topology optimization for automotive structure design.

Practical structural design problems, involving the layout of material and/or stiffeners in a given design space while meeting static displacement and modal dynamic constraints, are some of the most challenging tasks for automotive design engineers. This problem is also encountered in the sign-off stage of product development cycle, when automotive chassis components prove out dictates a need for a surrogate body structure. This surrogate body is required to meet the static and dynamic target stiffness of the actual prototype body. The key to developing a dedicated prototype structure lies in the development of an optimization method, which can be applied to the design and manufacture of a surrogate body with static displacement and modal dynamic characteristics of the actual prototype body. In this research, the above optimization problem was formulated using the concept of mutual potential energy together with a mean value representation of the eigenvalues. Multi-objective functions were formulated using ratio operators to avoid scaling problems. Similar to the actual physical build of surrogate body structure, lumped masses were added to the optimization process. The optimality criterion proposed by Kikuchi and Ma for eigenvalue problems was extended for the above multi-objective optimization. This new method was applied and practically validated for various design examples, including an automotive surrogate body design problem by targeting the static and dynamic characteristics of the actual prototype body. The multiple criteria structural topology optimization problem was posed in terms of a frame ground structure with the design variables as the cross sectional depths of the individual frame members. Target displacements and eigenvalues were expressed as a weighted objective function using expressions for mutual mean compliance, mean compliance and mean eigenvalue function. The surrogate body structure was fabricated and practically validated by modal and time domain based laboratory testing. The performance of the realized surrogate body is significantly better than that obtained with a non-representative body structure that emanated from existing design practices. The major contribution of this research is the development and validation of an optimization method, which can be applied to the design and manufacture of a surrogate body structure with target static displacement and modal dynamic characteristics of the actual prototype body, along with experimental validation of the methodology on an industrially relevant automotive structure.PhDApplied SciencesAutomotive engineeringMechanical engineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/132462/2/9963886.pd