Background
During crashes, vehicle occupants may experience a wide variety of injuries that often correspond to their location within
the vehicle, their age and gender, and type of vehicle and crash. Current finite-element models that are used to assess the level
of injuries employ only 60,000 to 100,000 elements and require 12 hours of computation to assess vehicle structural components.
Occupant models mostly represent the “50% adult male” (i.e., average-size male). However, models representative of
the “95% adult male” (larger in size than 95% of adult male population), “5% adult female” (smaller in
size than 95% of adult female population), child, and infant vehicle occupants are required for a complete assessment of crash
scenarios.
Role of High-Performance Computing
High-fidelity models, such as THUMS (Total Human Model for Safety), present detailed anatomical components of the human body.
The highest level of biomechanical modeling uses the anatomically correct structures of vehicle occupants to assess trauma to
the brain, internal organs, joints, and skeletal system during crashes. High-fidelity modeling can add 80,000+ elements per occupant
and assess up to four passengers per vehicle, for a total of 320,000 elements. This information can be used to develop and
improve mechanisms that can increase the levels of passenger safety protection for all sizes of occupants within the vehicles. |