Analysis and Comparison of Metrics to Assess Submarining Behavior with the Hybrid III 5th Percentile Dummy in the Rear Seat

Real-world data show that abdominal loading due to a poor pelvis-belt restraint interaction is one of the primary causes of injury in belted rear-seat occupants, highlighting the importance of being able to assess it in crash tests. This study analyzes the phenomenon of submarining using video, time histories, and statistical analysis of data from a Hybrid III 5th female dummy seated in the rear seat of passenger vehicles in moderate overlap frontal crash tests. This study also proposes different metrics that can be used for detecting submarining in full-scale crash tests. The results show that apart from the high-speed videos, when comparing time-series graphs of various metrics, using a combination of iliac and lap belt loads was the most reliable method for detecting submarining. Five metrics from the dynamic sensors (the maximum iliac moment, maximum iliac force drop in 1 ms, time for 80% drop from peak iliac force, maximum pelvis rotation, and lumbar shear force) were all statistically significant predictors of submarining, but the presence of a rear seat-belt pretensioner was not significant. Though no one metric was able to provide accurate submarining judgement in all the cases, a combination of these significant metrics can be used along with video to gain more confidence in submarining judgement.