ARSRPE Conference Paper Database

Full crash test reconstruction and analysis of four real-world impacts

Franklyn, Melanie, Logan, David, Hillard, Peter, Fildes, Brian

Crashes - Analysis


Reconstruction of real-world crashes using computer software has been performed by previous researchers, as demonstrated by numerous studies in the literature. However, this method has a number of disadvantages, particularly with regard to the accuracy of the crash reconstruction software used to replicate real-world crashes. In the current research, full physical crash test reconstructions of four real-world crashes were conducted. For each case, a comparison of crash test damage between the real-world and crashed vehicles was performed, and injury parameters from the anthropomorphic test devices were evaluated against the actual injuries sustained by the real-world occupants. The results demonstrated that the crash damage profiles produced, despite some variation, were a reasonable replication of the real-world damage, with some over or underestimation of the damage in each case. Injury assessment functions from the anthropomorphic test devices correlated well with the real-world injuries in almost all instances. Hence, despite some degree of over or underestimation of crush in the crash test reconstructions, the results indicate that the acceleration experienced by the anthropomorphic test devices must have been similar to that experienced by the real-world occupants. This suggests that injury parameters measured by anthropomorphic test devices do not entirely depend on vehicle crush, but also on other crash parameters.