Developing Lumbar Spine Fracture Injury Risk Functions for Frontal Impact Anthropomorphic Test Devices using Paired Human Body Model Simulations

Abstract

Thoracolumbar fractures from car crashes have gained importance over the past halfcentury. While fatal injuries have declined due to vehicle structural- and infrastructural improvements, the rate of moderate and more severe thoracolumbar fractures has remained steady or slightly increased. Recent research has introduced a new Finite Element (FE) lumbar spine model and a corresponding Injury Risk Function (IRF), integrated into the SAFER Human Body Model (SHBM) to enhance fracture risk prediction capabilities in car crash simulations. This thesis aimed to develop new lumbar spine injury prediction functions for current 50th percentile male frontal impact Anthropomorphic Test Devices (ATD), Hybrid III and the Test device for Human Occupant Restraint (THOR) using paired simulations with SHBM to find potential Injury Criteria (IC) in the ATDs. This to enable evaluations of lumbar spine fracture risks in simulated crashes at an early stage in the development of car restraints using models of ATDs and in real crash tests. Occupant models were positioned on the passenger side in a FE model of a mediumsized SUV front occupant compartment with seat back angles of 20°, 25°, and 30° using primer and ANSA. Paired simulations were run for all models using LS-DYNA. Data sampling employed a full factorial method, varying simulations with factors like impact pulse severity, belt forces, and seat pan stiffness. The study focused on common accident scenarios for moderate spine injuries: full frontal and run-off road impacts. Data analysis using linear regression identified correlations between ATD measures and SHBM peak lumbar strain. After data post-processing, Peak Pelvis Z-Acceleration (Peak Pelvis AZ) for both ATD’s and Peak T12 AZ for THOR showed the highest correlation. To improve the correlation between peak lumbar strain and ATD IC, the Dynamic Response Index (DRI) was assessed. The DRI measures maximum dynamic spinal compression using a simplified mass-spring-damper model. By using pelvis AZ as input, an optimized DRI was achieved, establishing it as a viable IC. IRFs for the ATDs were generated by substituting SHBM’s peak lumbar strain with linearly regressed strains from ATD ICs for specific loading scenarios. Bootstrapping established a 95% confidence interval to account for simulation and SHBM IRF variations. Verification in a medium-sized SUV’s rear seat mid-position showed predicted risks for Hybrid III within 30% of SHBM risk, while THOR’s differed by up to 40%. Generally, ATD IRFs underestimated fracture risk, except for the Hybrid III in the most severe crash pulse. In summary, Peak Pelvis AZ is recommended as IC for both Hybrid III and THOR. DRI can also be used for improved correlation but needs additional processing. The new IRFs provide a basis for fracture risk prediction using Hybrid III and THOR in frontal and run-off road crashes.