AbstractTraumatic injuries are evaluated using the Abbreviated Injury Scale (AIS), which is a risk to life measurement. Human computer models currently use stresses and strains to evaluate the critical points of injuries. This is problematic if these assessment values are below the cut-off stain thresholds as these current indicators bear no direct relation to AIS. This critical limitation prevents vehicle design and pedestrian safety protocols to use human computer models as the AIS levels cannot be predicted. This research proposes to use for the first time a power method, named Peak Virtual Power (PVP), on the Total Human Model for Safety (THUMS) to extract all the AIS levels of pedestrian trauma at organ and tissue level. This coding was created by calibrating the critical AIS for brain tissue and critical organs against a critical principal strain injury cut-off value. This was achieved by impacting a human model in critical impact directions observed in pedestrian accidents with speeds varying from 2.0m/s to 17.0m/s, covering the EuroNCAP test speed as well as the maximum impact speed provided in real-life pedestrian accident scenarios by the UK Police Force (UKPF). The AIS response was then scaled and bound using corridors using the relationship between AIS and probability of death, which was known to be a cubic. These unique trauma injury corridors were tested against four real-life pedestrian accidents which were reconstructed, and for which the Post Mortem information was available. The study concludes that the PVP method can predict pedestrian head trauma, and in some cases slightly under-estimate it by 1 AIS level, because of post-impact haemorrhage which cannot be captured using a Lagrangian solver. However, for other body organs there are significant differences between the estimates of trauma from PVP and the PM reports. This may be due to a) the head calibration using cylindrical impactors to mimic the impact contact between vehicle and pedestrian head, whereas the body contact against the bonnet may be more diffuse, and b) problems with the THUMS model for which the heart may be oversimplified.
For the first time, this work provides a foundation for the development of a numerical tool to predict AIS in vehicle design.
|Date of Award||Jun 2019|
|Supervisor||Christophe Bastien (Supervisor)|