Development of Lumbar Spine Injury Risk Prediction Tools for Physical and Virtual Human Surrogates in Frontal Motor Vehicle Crashes

Lumbar spine injuries have been consistently reported in epidemiology studies of frontal motor vehicle crashes (MVCs). Both prevalence and severity of lumbar spine injuries are suggested to increase with the introduction of reclined torso postures in future vehicles, as predicted by recent research with virtual human body models (HBMs) and post-mortem human subjects (PMHS). Such studies have shown that frontal MVCs elicit combined compression and flexion loading within the lumbar spine that is affected by several occupant and environment factors, but appropriate data on human injury response and tolerance in this loading mode is absent. Further, the corresponding tools for predicting injury risk do not exist for any human surrogate utilized in research, vehicle assessment, and certification (e.g., PMHS, HBMs, and crash test dummies, or ATDs). Together these surrogates produce a well-rounded view into injury occurrence and prediction, but each serve different and specific purposes. Thus, the goal of this dissertation was to develop surrogate-specific injury risk prediction tools (IRPTs) for the lumbar spine that predict injury risk in combined compression-flexion loading scenarios.

First, injury tolerance and response of the lumbar spine in this loading paradigm was quantified with component experimentation of PMHS representing a large array of ages, sizes, and both sexes. The developed human IRPTs described the pseudo-stress necessary to predict the risk for compression-flexion-based fractures for a diverse donor population, as well as the variation in characteristic response and stiffness. Second, the human IRPTs were translated to an HBM, creating HBM-specific lumbar spine IRPTs that incorporated differences in surrogate characteristics and loading behavior. Third, the effects of several occupant, restraint, and crash factors on HBM lumbar spine loading were investigated using component and sled test environments, while simultaneously building confidence in the developed HBM-specific IRPTs. Torso angle and speed had the largest effects on load metrics, and predicted injury risk reasonably complied with injury incidence from PMHS sled tests and field data. Lastly, an ATD was simulated in matched sled test conditions to understand the effects of the same critical factors on ATD lumbar spine loads and directly relate HBM and ATD loads. The matched simulations provided meaningful interpretations of spine loads relative to various frontal impact conditions, although IRPTs could not be translated from the HBM to the ATD due to mixed relationships between surrogates.

The outcomes from this dissertation provided robust and rigorous injury risk prediction for two human surrogates and bounds over which all surrogates were behaving reasonably. The intersection of knowledge gained from the PMHS, HBM, and ATD can provide a strong foundation for considering lumbar spine injury risk in future safety countermeasure design and linking human risk to an ATD criterion that can be implemented in vehicle assessment protocols. The culmination of this research is timely and has the potential to be particularly powerful within the automotive safety field given the advancement of human surrogates over the years, the desire to use the surrogates to predict injury risk, and the increased loading and injury risk within the lumbar spine for various occupant postures and vehicle compartment designs.