Safety rating of vehicles can be defined and measured in different ways. One important and difficult question to answer is; what is safety? What injuries should be prevented, fatalities, severe injuries, all injuries or injuries leading to permanent disability? Some rating systems may also look for rating accident involvement. There are several rating systems developed and used in most parts of the world. Most of them are intended to measure crashworthiness, however defined in different ways. The presented method is developed to consider as many of the confounding factors as possible, without losing too much of the possibility to generalise the results. The aim is to measure the risk of a fatal or disabling injury without influence of driver behaviour or exposure, i.e. to isolate the crashworthiness of the car from other factors. Many of the confounding factors are therefore related to impact condition and severity. The most complicated issue to handle is probably the normalisation of impact severity. While the normal way to generate measurements of impact severity is to either reconstruct every individual case or use on-board crash recorders, such methods are not suitable for the amount of crashes needed for crashworthiness rating on individual car models. Indirect measures must therefore be used. Two methods have been used: indirect parameters predicting impact severity such as speed limit and the relative outcome in two vehicle crashes. In the presented method, the latter is used, often called matched-pairs or paired comparison. The technique has been further developed and used by Folksam. Apart from measuring crashworthiness of different car models, other objectives with this study were to demonstrate how aggressiveness, impact severity and mass differences may influence the rating results and to show some trends and developments in crashworthiness of cars during recent years. (Author/publisher)
Abstract