Sensitivity Analysis for Efficient Parameterizations of Automotive Composite Structures for Crash

A significant challenge in structural optimization of vehicle architectures for crash is handling of the typically large number of quantities defining the input and system parameters. The use of advanced composites materials increases this challenge by introducing more parameters and complex failure behavior. This paper introduces a novel design workflow that can help reduce the problem complexity for composite vehicle structures and provide for a more efficient design workflow. First a computationally efficient prognosis method is introduced to smooth the design space and reduce the number of required samples. Second a sensitivity analysis using the Sobol decomposition is introduced to provide a parameter importance hierarchy. Results show a reduction of 71.12% of required samples at the same time achieving a better quality design space. The sensitivity analysis results in a reduction of 22 to 8 parameters.      «

A significant challenge in structural optimization of vehicle architectures for crash is handling of the typically large number of quantities defining the input and system parameters. The use of advanced composites materials increases this challenge by introducing more parameters and complex failure behavior. This paper introduces a novel design workflow that can help reduce the problem complexity for composite vehicle structures and provide for a more efficient design workflow. First a computa...     »