As part of a comprehensive plan to reduce the risk of death and serious injury in rollover crashes, the National Highway Traffic Safety Administration (NHTSA) has a program to characterize restraint system response in rollovers. A rollover restraint tester (RRT) is utilized to produce a 180 degree roll followed by a simulated roof-to-ground impact. Recognizing the unpredictability of the real world rollover phenomenon, this test provides a repeatable and consistent dynamic environment for suitable lab evaluation. Similar NHTSA research during the mid-1990s demonstrated an excursion reduction of up to 75% when an inflatable belt was compared to the standard three-point belt with a 50th percentile (50th ) male dummy [Rains, 1998].

Technologies being considered include integrated seat systems, pyrotechnic and electric resetable pretensioners, four-point belt systems, and inflatable belts. High speed video data are collected and analyzed to examine occupant head excursion throughout the tests and are presented for discussion. The RRT has demonstrated to be repeatable;​ however, there are some concerns about the real world relevancy of the RRT dynamics in the absence of a lateral component. The RRT does not have a mechanical component for lateral motion that is typical in some real world rollover events.

This research attempts to determine if reducing occupant excursion during a rollover event is possible by utilizing the RRT. Results presented at the 20th ESV conference demonstrated that excursion characteristics can be affected with the implementation of advanced restraints in the 50th percentile male dummy [Sword, 2007]. This paper presents expanded research with the 50th percentile male dummy and also includes the 5th percentile (5th) female and 95th percentile (95th) male dummies.

When compared to a baseline 3-point restraint, advanced restraints utilizing pretensioning and other technology reduced excursion of all the dummies in both the Y and Z directions, where the Y direction is lateral motion and the Z direction is vertical motion. The current production technologies, pyrotechnic and motorized retractors, were able to reduce Y and Z excursion in RRT tests, by up to 66% and 60%, respectively. The advanced restraints, inflatable belts and 4-pt belts, reduced excursion in the Y and Z directions up to 80% and 86%, respectively.