Effect of Occupant and Restraint Variability in Reclined Positions on Submarining Probability in Frontal Car Crash Scenarios

Erik Brynskog^1*Jonas Östh^1,2Karl-Johan Larsson³Johan Iraeus¹

• ¹Department of Mechanics and Maritime Sciences, Chalmers University of Technology, Göteborg, Sweden
• ²Volvo Cars, Göteborg, Sweden
• ³Autoliv Research, Vårgårda, Sweden

In future autonomous vehicles, a greater seat back recline angle has been suggested to accommodate a more relaxed occupant position. Due to the reclined position, the pelvis rotates rearward resulting in less favorable in-crash pelvis to lap belt interaction. In a crash, this issue can increase the likelihood of the lap belt disengaging from the pelvis and instead loading the abdomen, i.e., submarining. Hence, to enable assessment of submarining prevention measures for reclined occupants in frontal car crash scenarios, it is motivated to enhance the understanding of pelvis to lap belt interaction.In this simulation study, the submarining outcome of a population of reclined 50%ile (in terms of height and weight) male occupants, subjected to restraint variability in a semi-rigid seat setup, was analyzed through finite element human body model (FE-HBM) simulations (n = 369). To account for the substantial individual variability associated with pelvic shape, a statistical shape model was utilized to predict a large set of random 50%ile male pelvises. Based on select measurements, a sub-sample was drawn from this set (n = 78) for inclusion in the analysis. The simulated submarining outcome and corresponding occupant/restraint parameters were used to generate a metamodel predicting probability of submarining. The results showed that random variations of a 50%ile male can be comparable with restraint design variability on submarining outcome for reclined occupants. Significant predictors included three from the occupant (pelvis angle, iliac spine hook angle, and H-Point forward/rearward position), and three from the restraint (buckle angle, seat friction, and seat pan angle). Non-significant predictors included occupant soft tissue thickness and fat stiffness, shoulder belt load limit force, and inclusion/exclusion of single lap belt pre-tensioning.In conclusion, this study implies that future vehicle safety ratings, using different versions of a 50%ile male FE-HBM, may be subject to variation in submarining outcome if harmonization of the target occupant anatomy/posture is not established. In addition, based on the response in a semi-rigid seat setup, this study also indicates that to achieve robust submarining protection for reclined occupants, the current legal requirements on buckle angle might need a shift towards more vertical angles.