Case Study of Vehicle Maneuvers Leading to Rollovers: Need for a Vehicle Test Simulating Off-Road Excursions, Recovery and Handling

The sequence of events leading to rollovers was determined from 63 in-depth investigated cases in the NASS-CDS database from 1995-1999. The sequence was evaluated by vehicle maneuvers, vehicle stability, surface type, road and shoulder transition condition, posted and estimated speeds, vehicle type and driver injury severity. The cases were divided into 3 categories defining the precipitating event leading to a rollover: 1.) negotiating a curve at usually too high a speed, 2.) drifting off the road, or 3.) avoiding an obstacle in the traveling lane.

47% of the precipitating events were negotiating a curve at too high a speed prior to the sequence leading to a rollover, 27% were drifting off the road, and 25% left the road after avoiding an obstacle, such as an animal or vehicle in the traveling lane. 90% left the roadway with 66% having four wheels off the road and 34% two wheel off the road. The transition from a road surface to the shoulder often involved a drop of approximately 50 mm onto grass, gravel or dirt from an asphalt or concrete road surface. In 45% of the cases, the driver attempts a recovery to the road and steers. When the front wheel regains the road surface, the vehicle abruptly crosses the road and often departs it on the opposite side and then rolls over. In 45% of the cases, the driver attempts to keep the vehicle on the shoulder, but drifts further off road until it leaves the roadway and rolls over. In 10% of the cases, the precipitating event happens on the road, leading to a loss of control and rollover.

This study prioritizes the need to carry out tests that simulate a vehicle leaving the road and having at least two wheels on the shoulder after negotiating a curve at high speed, drifting off the road or leaving the road by obstacle avoidance. The vast majority of rollovers involve the vehicle departing the road, so handling tests need to include a transition from a road surface to a shoulder where there may be a drop off of approximately 50 mm and an attempt to recover back onto the road with various degrees of steering or an attempt to maintain a straight heading during braking under a varying coefficient of friction road condition. Since current tests are carried out on a flat road, a new suite of handling maneuvers should be considered where the wheels leave the road surface to a simulated shoulder. Dry conditions should first be captured, followed by wet conditions.