Injury Prediction Methods for Highway Barrier Impact Using Occupant Dynamics Models

摘要

A computer program is developed to predict injuries to occupants when a vehicle strikes a highway barrier. The program is intended to eventually improve upon existing methods, which include crash testing using dummies and measurement of the vehicle's acceleration. The overall program includes portions to successively compute the vehicle's motion while striking a barrier, the occupant's motion given the vehicle's motion, and the occupant's injuries given his motion within the passenger compartment. BARRIER VII combined with the Rollover Vaulting Algorithm (RVA) and HVOSM constitute the vehicle-barrier dynamics options, UCIN and CVS are the occupant dynamics options used, and an injury prediction program is developed from the Kornhauser Sensitivity Curve. The program's output is checked against crash test results and barrier impact accident findings, and it shows reasonable agreement. A parametric study gives reasonable results for variations in occupant conditions, a barrier impact conditions and car size. UCIN's integration routine was replaced and a new human body joint model was developed in order to alleviate computational problems. Recommendations include further UCIN improvements, improved joint modeling, further validation, and additional investigation of injury mechanisms and tolerance values. The report includes discussions of occupant dynamics and occupant injury modeling.