The relation between impact severity and risk of injury is a fundamental issue in terms of comparing vehicles and occupant protection systems. Normally, such risk functions would have to be based on reconstruction of crashes, limiting the possibility to generate risk functions down to individual car models. In this study, an alternative way to derive risk functions was developed and used. In the present method, risk functions were derived using matched pairs of crashes, varying mass relations in a controlled way, and generating risk versus relative change of velocity. The data used was police reported crashes in Sweden during 1994-2000. The results show that there are major differences in injury risk functions between individual car models. The results are of major importance for the development of car model safety rating and for the evaluation of new car safety technology. The method is also of importance in understanding possible scenarios of sub optimisation. In the development of vehicles there is a risk of concentration on certain crash severities instead of looking at the overall performance.
Abstract