Controlled rollover test methods have been developed where touchdown conditions of the vehicle are specified as test inputs. Rollover crash touchdown parameters can vary widely due to variations in road surface and topography, maneuvers, and vehicles. While vehicular accident reconstruction teams have performed steering induced rollover tests and reported on touchdown conditions in the literature, such kinematic parameters are only available for an extremely limited set of conditions and vehicles. Furthermore, information about the sensitivity of touchdown conditions to changes in vehicle and maneuvers is missing from the literature. Thus, the goals of this study were threefold:​ to develop and validate two vehicle models in ADAMS TM, use them to simulate common types of steering- induced soil-trip rollovers, and to evaluate how differences in maneuvers and vehicle type affect vehicle kinematics at touchdown.

First, vehicle inertia measurement tests, suspension tests, tire tests, bushing tests, and driving tests, including double lane change, J-turn, and fishhook, were performed using a sedan and a pickup truck. Next, vehicle models for each vehicle were built and validated with the experimental data. A straight highway was modeled following road design guidelines and a soil-tire interaction model was implemented. Analysis of NASS-CDS cases showed that rollover accidents occurred as a result of the vehicle leaving the roadway and either attempting to drive back onto the road (corrective) or continuing to steer from the road (non-corrective). Then specific cases exemplifying the corrective and non- corrective maneuvers were reconstructed with the two vehicle models to determine baseline driver inputs. Lastly, 120 Monte Carlo simulations were performed to compare vehicle kinematics and touchdown conditions of the two types of vehicles and maneuvers.

The two vehicle models showed good correlations with the static and dynamic test data. The median values of roll rates of the sedan were 290 deg/sec and 380 deg/sec in corrective and non-corrective maneuvers, respectively. The pickup truck showed lower roll rates in the same maneuvers (210 and 250 deg/sec, respectively). Touchdown roll angles were higher in the sedan (120 and 190 degrees) than in the pickup (103 and 104 degrees) and higher in the non-corrective maneuver for both vehicles. Vertical speeds at touchdown were about 2.6 m/s higher in the non-corrective maneuver than in the corrective maneuver.

The vehicle models were validated with results from component tests, static tests, and dynamic tests but no steering-induced rollover test data were available to validate the vehicle models. Subsequent to this study, steering-induced rollover tests will be performed to validate the models further and the soil model will be validated by testing the soil at the test site.

Despite these limitations, the methodology and results presented provide for the best available means to determine touchdown parameters for use in controlled rollover crash testing. The data presented show a substantial difference in touchdown conditions with respect to types of vehicles and maneuvers. Therefore, when a rollover test is performed, the test conditions should be carefully selected depending on types of vehicles or maneuvers to generate realistic outcome.