Arndt, M., Stevens, D., and Arndt, S., "Comparison of Linear Variable Deceleration Rate Rollover Reconstruction to Steer-Induced Rollover Tests," SAE Technical Paper 2012-01-0469, 2012, doi:10.4271/2012-01-0469.
A variable deceleration rate approach to rollover crash reconstruction was proposed in 2009 by Rose and Beauchamp. A detailed description of Rose and Beauchamp's method was outlined in 2010. The method used a Linearly Variable Deceleration Rate (LVDR) as a function of roll distance. Improvements in responses as a function of time was demonstrated by Rose and Beauchamp using test data from two 208 dolly rollover tests; however, they noted that additional validation work using steering-induced rollover tests would be desirable. This paper provides additional validation of the LVDR model using the steer-induced rollover test data reported in 2011 by Stevens et al. The Over-The-Ground Speed (OGS) and recorded roll rate results from the five steer-induced rollover tests reported by Stevens' in 2011 were compared to reconstructed speed and roll rates as a function of time using the 2010 Rose and Beauchamp LVDR method. Using an appropriate range of average drag factors, the LVDR method produced agreement with the measured results of the Stevens rollover tests. Comparisons showed agreement with the predicted rollover duration, the shape of the roll rate curve and the maximum roll rate. Calculated roll rates were high if the calculated roll duration was low. Low roll durations were found associated with a Constant Drag Factors (CDF) method and the LVDR method utilizing high average drag factors. Greater roll rate uncertainty occurred in roll segments that have long airborne duration and/or high speed change. The LVDR method significantly improves prediction of speed and roll rate time history compared to a CDF method.