In-depth Study of Volvo Cars in Rollover Accidents

Rollover is not the most common type of accident, but the consequences may be serious. Understanding of the injury mechanisms during the complex rollover movement is limited and requires further investigation. ln-depth investigations of 21 Volvo cars involved in rollover crashes were analysed with regards to cause and outcome. Detailed information was examined about the crash scenario, the roadside environment, the car, its occupants and injuries, if any occurred. Parameters such as the initiation of rollover, the rollover kinematics, the deformation of the car and the environmental interaction with the car were analysed with a view to determine a measure of rollover severity. As a complement, the results from the in-depth investigations were compared with statistical analyses of rollover cases in Volvo's accident database and the NASS database. The results from these databases support the findings from the in-depth study. Having this comprehensive knowledge about the crash performance and kinematics, no single parameter or combination of parameters was found appropriate for describing the rollover severity related to injury risk. On the other hand, conclusions are drawn regarding deformation characteristics, the in itiation of rollovers, the number of turns, the roll direction, where the rollover takes place, the body parts most often injured, and the most common injury mechanisms. ROLLOVER CRASHES ARE VERY COMPLEX in terms of vehicle motion, occupant kinematics, occupant contacts within the vehicle's interior and possible occupant contact with the road or ground surface. Keeping this complexity in mind, it is understandable that finding a way to determ ine rollover crash severity is not easy. For instance Terhune et al, 1 991 , did not manage to define good measures of rollover crash severity, while studies by Digges and Klisch, 1 99 1 , indicate vehicle speed as an important parameter. IRCOBI Conference Sitges (Spain), September 1999 297 Of the injury mechanisms discussed in the literature the mechanism of head and neck injuries is the most frequently mentioned. Neck injuries can be caused by the occupant hitting the roof: either by the occupant moving towards the roof, resulting in a diving-type injury or by the roof itself being crushed (Bahling 1 990, Orlowski 1 985). The extent of roof crush could be an indicator of the injury severity, but according to Rechnizer et al, 1 998, Mackay, 1 991 , Piziali et al, 1 998, roof crush is not a factor in determ ining injury severity. Another mechanism of injury is complete or partial ejection during the rollover. Rechnitzer et al, 1 996, report partial ejection of the head occurring, even if the occupant was belted, resulting in severe head-crushing injuries. There are only two standard test procedures for rollovers; one dynamic and one static. The dynamic test method is the US legal requirement FMVSS 208; describing a test which in itiates a rollover around the longitudinal axis of the car. The FMVSS 208 test method is mainly for testing partial ejection. The test method is widely used and is quite simplified for being a rollover, and there may be a great variability in number of turns, roof deformation and final position. Only the first roll could be made reproducible, according to Rechnizer et al, 1 998. The static test is for roof crush resistance, the FMVSS 216. The roof is deformed by a plate until a certain force is reached. This force has to be reached within 1 27 mm. Apart from these two rollover test methods several types of non standard rollover tests are used. For example, rollovers in itiated by soil, a ditch or a ramp (Cooperrider et al, 1 998, Wech and Ostmann 1 996, Winn 1 989, Brown 1 985). One objective of this study was to define rollover crash severity measures by identifying and correlating the occupant injury mechanisms with the analyses of accident material. Another aim was to categorise the car damage and find the relevance of existing rollover test methods.