Citation

Boyd PL. Proc. Int. Tech. Conf. Enhanced Safety Vehicles 2005; 2005: 10p.

Copyright

(Copyright © 2005, In public domain, Publisher National Highway Traffic Safety Administration)

DOI

unavailable

PMID

unavailable

Abstract

Starting in the 2004 model year, the National Highway Traffic Safety Administration (NHTSA) improved the rollover resistance ratings in its New Car Assessment Program (NCAP) consumer information by adding a dynamic maneuver test. NHTSA had provided rollover resistance ratings in the 2001 – 2003 model years based solely on the Static Stability Factor (SSF) measurement of vehicles. The ratings express the risk of a vehicle rolling over in the event of a single vehicle crash, the type of crash in which most rollovers occur. The SSF, which is determined by a vehicle’s center of gravity height and track width, had proved to be a powerful predictor of rollover risk based on a linear regression study of rollover rates of 100 vehicle models in 224,000 single vehicle crashes (R-squared = 0.88). The TREAD Act required NHTSA to change its rollover resistance ratings to use a dynamic maneuver test, and the 2004 and later NCAP rollover resistance ratings use both SSF and a dynamic maneuver. This paper describes the development of the risk prediction model used for present rating system. Twenty-five vehicles were tested using two highly objective automated steering maneuvers (J-turn and Fishhook) at two levels of passenger loading. A logistic regression risk model was developed based on the rollover outcomes of 86,000 single-vehicle crashes involving the make/models that were tested. The vehicles were characterized by their SSF measurements and binary variables indicating whether or not they had tipped up during the maneuver tests. It was found that the Fishhook test in the heavy (5 passenger equivalent) load was the most useful maneuver test for predicting rollover risk. The relative predictive powers of the SSF measurement and the Fishhook test were established by a logistic regression model operating on the rollover outcomes of real-world crash data. This model was used to predict the rollover rates of vehicles in the 2004 and 2005 NCAP program based on their SSF measurements and Fishhook maneuver test performance. The information in this paper first appeared in NHTSA’s Federal Register notice that established the NCAP rollover resistance rating system for model year 2004.