Determination of Postmortem Human Subject Response to Vertical Drop Tower Testing and Initial Application for Total Human Model for Safety Finite Element Model Validation

Le résumé fourni par la source

Drop tower testing was conducted using 50th percentile male PMHS at 15G peak acceleration in a rigid seat, with a seat pan-to-seatback angle of 90°. Subjects were instrumented with 6DOF motion blocks at T1, T4, T12, L3, and S1 to capture detailed vertebral body kinematics. Pressure sensors were also placed throughout the lumbar spine to estimate force in the intervertebral discs from S1-L2. PMHS were restrained using a pilot torso harness attached to the seat at the shoulders and lap belt, both pretensioned to 89 N. Reaction forces were measured in the seat using six-axis loads under the seat pan. Final positioning of the occupant was documented using a FARO arm point probe and laser scanner. To recreate the experimental setup, CAD models of the experimental fixture were meshed using a commercial FE modeling software (Hypermesh) and imported into LS-Dyna for incorporation with the THUMS model. The belt routing tool in LS-PrePost v4.9.12 was used to develop the torso harness and shoulder and lap belts. Pre-simulation was performed to position the THUMS model in accordance with recorded FARO data, and the experimentally recorded seat vertical acceleration was assigned using Boundary_Prescribed_Motion. Finally, instrumentation locations were duplicated within the THUMS model to match the PMHS experimental setup. The THUMS model showed similar head kinematics compared to the experiment, which went first into extension, followed by flexion during the primary pulse. The torso of the model, however, experienced an increased flexion/compression response compared to PMHS. The peak reaction force in the simulated seat load cells measured 12.7 kN, which was within one standard deviation of the average normalized experimental values (average = 12.8 kN, standard deviation = 0.4 kN). The average load in the lumbar spine in the model was found to be 3.3 kN, which was lower than PMHS average by more than two standard deviations.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.