Replicating Real World Rollover Crash Injuries

This paper details the injuries occurring in real world trip-over only rollover crashes, for seat belted and contained occupants, and assesses whether these injuries can be replicated using a Jordan Rollover System (JRS) crash test rig recently installed at Crashlab in Sydney. This research forms part of the Dynamic Rollover Occupant Protection (DROP) project funded by the Australian Research Council and industry collaborators to develop a dynamic rollover crash test protocol that can assess a vehicle’s rollover crashworthiness. Australian National Coroners Information System (NCIS) fatality data and US NASS-CDS serious injury data of seat belted occupants involved in single vehicle pure rollover crashes ranging over the period of 2000 to 2010, were investigated. AIS3+ head and thorax injuries and AIS2+ spinal injuries were analysed to determine rollover injury characteristics and to determine possible test conditions under which they occur. Publicly available dynamic rollover crash tests carried out by other researchers were also analysed to determine their capability of replicating these real world injuries. Serious head injuries (SHI), serious neck/spine injuries (SSI) and serious thorax injuries (STI) were found to be distributed in roughly equal proportions, most occurring independently of each other, indicating different injury causal mechanisms. A significant portion of these injuries occurred where there was minimal or no roof crush involvement. Investigations of other researcher’s crash test results show dynamic rollover crash test rigs, crash test protocols and anthropomorphic test devices (ATD) have not, in general, been able to replicate ATD loadings consistent with these real world injuries repeatedly in a manner similar to frontal or side impact crash test protocols. The dynamic test conditions, measurement systems (possible ATD) and criteria required to consistently replicate vehicle damage and a particular injury mode (SHI, STI and SSI) using the JRS are discussed. It was concluded that to date it appears that current test protocols are not capable of consistently replicating the injuries identified in real world rollover crashes. Addressing roof crush alone via quasi-static testing will not mitigate all real world rollover injuries in typical trip-over only rollover crashes. A more advanced dynamic rollover crash test protocol must be developed that is more representative of the real world crashes and be capable of consistently replicating SHI, STI and SSI. It may be possible using the JRS test rig albeit the rig may need to be modified to tolerate much heavier impacts and a suitable rollover ATD may need to be developed. Until such time that the real world injuries observed in strong roof vehicles can be replicated repeatedly in a realistic manner, research on the development of an appropriate crash test protocol and ATD will need to continue.