Objective: Frontal crashes in which the vehicle has poor structural engagement, such as small-overlap and oblique crashes, account for a large number of fatalities. These crash modes are characterized by large intrusion and vehicle yaw rotation. Results from previous studies have shown mixed results regarding the importance and effects of these parameters. The aim of this study was to evaluate how vehicle yaw rotation, instrument panel intrusion, and the time history of the pulse angle influence chest injury outcomes.
Method: This study was conducted using kinematic boundary conditions derived from physical crash tests, which were applied on a finite element simulation model of a vehicle interior including a finite element human model. By performing simulations with different levels of simplified boundary conditions and comparing the results to a simulation with a full set of boundary conditions, the influence of the simplifications was evaluated. The injury outcome measure compared between the simulations was the expected number of fractured ribs. The 3 simplifications simulated were (1) removal of vehicle yaw rotation, (2) removal of vehicle yaw rotation plus an assumption of a constant pulse angle between the x- and y-acceleration, and (3) removal of instrument panel intrusion.
The kinematic boundary conditions were collected from 120 physical tests performed at the Insurance Institute of Highway Safety; 77 were small-overlap tests, and 43 were moderate overlap tests. For each test, the full set of boundary conditions plus the 3 simplifications were simulated. Thus, a total of 480 simulations were performed.
Results: The yaw rotation influences occupant interaction with the frontal airbag. For the approximation without this kinematic boundary component, there was an average error in injury outcome of approximately 13% for the moderate overlap cases. Large instrument panel intrusion increases the risk of rib fracture in nearside small-overlap crashes. The mechanism underlying this increased fracture risk is a combination of increased airbag load and a more severe secondary impact to the side structure. Without the intrusion component, the injury risk was underestimated by 8% for the small-overlap crashes.
Conclusion: The approximation with least error was version 2; that is, a model assuming a constant pulse angle, including instrument panel intrusion but no vehicle yaw rotation. This approximation simulates a sled test with a buck mounted at an oblique angle. The average error for this approximation was as low as 2–4%.
Taylor & Francis, 2014. Vol. 15, no Supplement 1, S88-S95 p.