Road traffic injuries are the second leading cause of death among 5-14 year-olds. Traumatic brain injuries are the most common severe injuries sustained by pediatric occupants and responsible for one third of all pediatric injury deaths. Literature shows that current pediatric Anthropomorphic Test Devices, and more specifically the Hybrid III 6-year-old (6YO), fail to predict the kinematics of the pediatric head and spine. The goal of this dissertation research is to provide corridors for the trajectories of the head and thoracic spine of a 6YO occupant in a 40 km/h frontal impact. The challenge is the absence of experimental data that can guide the development of these corridors at this speed.

To overcome the dearth of pediatric kinematic data in high-speed impacts, four different data sources were combined:​ pediatric and adult volunteers test at 9 km/h, cadaveric tests at 9 km/h and 40 km/h, animal surrogate tests at 9 km/h and 40 km/h and in vitro bending tests of sections of the pediatric and adult thoracic spine. The results from the 9 km/h volunteer tests showed that conventional methods that scale between pediatric and adult subjects underpredicted the forward excursion of the pediatric head by 42% (SAE method) and 49% (mass scaling). Two new methods predicting the displacement of pediatric occupants were developed within this dissertation. The first one assumed conservation of energy and underpredicted the excursion of the head by 29%. The second one was based on the use of a linear timeinvariant 2D model of the occupant. The values of the effective stiffness and damping joint parameters were obtained to minimize the error between the model and the observed pediatric displacements at 9 km/h. A quasilinear viscoelastic characterization of the bending behavior of the pediatric thoracic spine was used to relate the stiffness of the upper and middle thoracic spine regions and to reduce the number of unknown joint properties in the model. The model overpredicted the forward displacement of the head (5% error) and T1 (6% error). This model was then used to predict the trajectories of a 6YO in a 40 km/h frontal impact. The assumptions made regarding the time-invariant characteristic of the model as well as the loading environment at 40 km/h were checked against the animal-surrogate and cadaveric tests. The predictions of the sagittal trajectories of the pediatric head, T1 and T8 obtained from the simulation of the model were combined to produce corridors. The limitations of the method are discussed in the dissertation.