A possible statistical biofidelity ATD evaluation scheme

ISO/TR9790 has been in existence for some years for evaluating the biofidelity of side impact ATDs (dummies) (ISO/TR 9790, 1999). NHTSA recently generated a new method for creating biofidelity corridors. The method was different from the ISO method, by incorporating statistics and the time relationships into the evaluation equation and automating the process (Maltese et al., 2002). Although both the ISO and NHTSA methods exhibit a number of strengths, they also have weaknesses. This paper attempts to build on these two methods to develop an ATD assessment method which offers added objectivity and is based on a statistical process. Improvements were explored in several key areas to address the existing numeric issues (Hsu et al., 2005). This process bases itself on the statistical correlations between the post-mortem human subject (PMHS) data. Detailed validation of the scheme is performed using PMHSs as “pseudo dummies”. A simple formula is proposed for ranking the biofidelity of the dummy, resulting in a score from 0-10, with 10 being the best. INTRODUCTION ince the 1950’s various mechanical human surrogates, or Anthropomorphic Test Devices (ATD), have been used for assessing the potential for injury in vehicle crash tests. These ATD crash test dummies have become more sophisticated, complex, and potentially more human-like through the years, but still provide only very limited estimations of what might occur in a real life crash. In order to improve this prediction, efforts have been made through the years to make ATDs more biofidelic. However, an omnidirectional dummy has not been developed. Instead, dummies have been created for each type of impact, resulting in a variety of different ATDs in frontal, side, and rear impacts. For some of these impact types, a whole family of ATD sizes has been developed. S Each of these ATDs has its own set of performance requirements, calibration procedures, and response corridors that have been developed in an attempt to make the dummy better mimic a human, as well as ensuring repeatability of responses. Many attempts have been made to determine the level of bio-fidelity. Tests have been performed using PMHSs to gather information on what injury response would be in certain