Head Injuries in Car-to-Pedestrian Accidents - Investigation of Head Impact Dynamics, Injury Mechanisms and Countermeasures

Doktorsavhandling, 2010

The aim of this study was to investigate head impact dynamics and to establish risk functions for head injuries by mathematical reconstructions of real-world car-to-pedestrian accidents. Furthermore, an attempt was made to develop an approach for the design and evaluation of head-protective devices. Child and adult pedestrian accidents from the German In-Depth Accident Study (GIDAS) database were analyzed and selected for reconstruction. These accident cases contained detailed information concerning the car, pedestrian and environment in the pre-crash, crash and post-crash phases. Two approaches, multibody system (MBS) dynamics, implemented by the MADYMO program; and the finite element method (FEM), implemented by the LS-DYNA program, were used for this study. The child pedestrian accidents were reconstructed using corresponding MBS pedestrian and car models. The GEBOD program was used to generate the mass/inertial properties and characteristic dimensions of various body segments, based upon anthropometric data of the victims. The MBS car models were developed according to the geometry of the cars involved in the accidents. The contact properties of the car structures were obtained from impactor tests. Initial conditions in the reconstruction, such as car impact speed, pedestrian position and posture, were set up according to accident information. The adult pedestrian accidents were first reconstructed using the MBS approach to reproduce pedestrian kinematics. A validated finite element (FE) head model with detailed anatomy structure was then used to reconstruct the head-to-windshield impacts. The initial impact conditions – including head impact velocity, head impact location, and head orientation – were defined according to the results from the MBS reconstructions. Brain injury parameters, such as intracranial stress and pressure, were calculated and correlated with the injury outcomes using a logistic regression model. An FE dummy model was developed as a substitute for a mechanical pedestrian dummy, and was validated on both component and full-scale levels across an extensive range of tests. The FE dummy model was used for the design and optimization of a pedestrian airbag system. The accident analysis demonstrated that the head was the most vulnerable body region in pedestrian accidents. Its injury severity depends on car impact speed and head impact location. In child pedestrian accidents, head impact locations were mainly on the hood top. In adult pedestrian accidents, the MAIS 2 head injuries were frequently caused by impacts with the windshield centre, while impacts with the windshield edge and A-pillar were associated with a higher risk of AIS 3+ head injuries. The accident reconstructions using MBS models showed that head impact conditions depend on the shape and stiffness of the car front, car impact speed, and pedestrian stature. In general, the head impact speed is proportional to the car impact speed and is lower than the maximum head speed. The head impact angle could be influenced by several factors, such as pedestrian height, hood edge height, hood angle and car impact speed. The head impact timing increases as a function of pedestrian height and is inversely proportional to car impact speed. The HIC is an important criterion in predicting head injury risk in pedestrian accidents, with a tolerance level that could vary considerably among individuals. At a HIC value of 700, the corresponding AIS 2+ head injury risk of a child pedestrian varies between 40% and 68%. The calculated brain injury parameters using the FE head model show a good correlation with brain injury severity. Critical values of the parameters were determined for AIS 3+ brain injuries, which are 256 kPa of coup pressure, 152 kPa of countercoup pressure, 14.8 kPa of Von Mises stress, and 7.9 kPa of shear stress. Compared with headform impactors, a pedestrian dummy could render a more biofidelic response as to head impact kinematics. The optimized airbag system could reduce the HIC value of the FE dummy head from 2,960 to 602 at the car impact speed of 40 km/h. The results also indicate that the protection performance could be influenced by design parameters such as ignition timing and the presence of a vent hole on the airbag. This study demonstrates that accident reconstructions could help to promote knowledge of head injury mechanisms in pedestrian accidents. This knowledge could be utilized, with proper mathematical tools such as the FE pedestrian dummy model, to develop safety countermeasures and thereby enhance pedestrian head protection.