Each year, in the United States, about 220,000 passenger vehicles are involved in rollover crashes. The number of occupants in these vehicles is about 355,000 per year. Of these, about 224,000 are injured or killed; 9,800 fatalities, 14,100 seriously to critically injured, and 200,400 moderately or lightly injured survivors. The injured occupants incur about 789,000 injuries, aproximately 3.5 injuries per occupant. The comprehensive cost of these injuries is about 40 billion annually. As a class, rollover crashes constitute about 2.4% of the crashes but cause 33% of the injury costs.
This paper applies on-the-road crash data and computer modeling to clarify opportunities for reducing rollover crash casualties. Countermeasures to reduce rollover casualties include increased belt use, and technoloogical interventions to prevent ejection and reduce the severity of body contacts with the vehicle interior.
The vehicle speed immediately prior to the rollover is a pivotal predictor of rollover crash severity. The number of vehicle quarter turns, the vehicle damage, and the roll rate are also predictors, and are strongly related to the initial vehicle speed. Computer simulations described in this paper suggest that the severity of the tripping acceleration may also have a strong influence on occupant ejections and injuries.