Seatbelts provide substantial benefits in rollover crashes, yet occupants still receive head and neck injuries from contacting the vehicle roof interior when the roof exterior strikes the ground. Prior research has evaluated rollover restraint performance utilizing anthropomorphic test devices (dummies), but little dynamic testing has been done with human volunteers to learn how they move during rollovers. In this study, the vertical excursion of the head of restrained dummies and human subjects was measured in a vehicle being rotated about its longitudinal roll axis at roll rates from 180 to 360 deg/sec and under static inversion conditions. The vehicle's restraint design was the commonly used 3-point seatbelt with continuous loop webbing and a sliding latch plate. This paper presents an analysis of the observed occupant motion and provides a comparison of dummy and human motion under similar test conditions. Thirty-five tests (18 static and 17 dynamic) were completed using two different sizes of dummies and human subjects in both nearside and farside roll directions. The research indicates that farside rollovers cause the restrained test subjects to have greater head excursion than nearside rollovers, and that static inversion testing underestimates head excursion for farside occupants. Human vertical head excursion of up to 200 mm was found at a roll rate of 220 deg/sec. Humans exhibit greater variability in head excursion in comparison to dummies. Transfer of seatbelt webbing through the latch plate did not correlate directly with differences in head excursion.
Abstract