Correlation Study on Different Bumper Impact Test Method and Predicted Results 2003-01-0211

Laboratory testing and computer aided engineering (CAE) techniques have long been employed in the development and optimization of automotive bumper systems. However, as more aggressive automotive styling and performance trends and mass and cost reduction objectives continue to push the limits of known bumper system construction, early optimization techniques will need to improve proportionately. The limitations of today's test and analysis techniques for providing accurate predictions can lead to excessive factors of safety (waste) and/or expensive rework during latter stages of vehicle development. Much work has been done to improve the capability of computer simulations for predicting impact behavior (1). However, some of the development limitations can also be attributed to an inability to achieve repeatable results or to accurately simulate the low speed impact behavior of a vehicle with universal test vehicles (UTV).

This paper will first describe a series of experiments aimed at understanding the differences in impact testing due to the test method and set-up. Three different labs replicated the same impact event, each using a different test cart and impact set-ups. The impacts consisted of a flat barrier, 30 degree angled barrier, longitudinal pendulum, angled (corner) pendulum and IIHS pole impact. Each impact type was run three times at each lab to determine a degree of statistical repeatability.

The paper will also address how these results compare to traditional Finite Element Modeling techniques used to predict bumper system performance. A better understanding of how different test methods can effect bumper system impact results and how these correlate to both FEA prediction and real vehicle performance can help reduce development cycle times and costs required to design bumper systems.