Child Safety in Car Crashes

Abstract: Traffic related trauma is the main cause of child fatality and injury in Europe and North America, with the large majority of fatalities and injuries occurring to children in cars. Specifically, head injury is a common outcome in injury causing car accidents. Furthermore, injury statistics show that the injury rate increases at the graduation from one type of restraint system to another. In Sweden, the graduation from a rear facing restraint to the forward facing booster child restraint has been identified as the first critical graduation. It occurs as the children reach the approximate weight of 18 kg. This size is similar to the anthropomorphic test devices (ATD) that correspond to 3-year-old children, and the two with most measurement capabilities are the HIII 3YO and the Q3. These ATDs are used in order to improve restraint systems for children, and they need to emulate real child crash behaviour adequately in order to properly inform proposals for improvements. The Q3 is the newest and it is under consideration to be included in European requirements. Thus, both further improvements of the restraint systems for 3-year-old children and additional evaluation of the Q3 are needed. Therefore, this thesis presents two studies: one that investigates which restraint system characteristics are beneficial for forward facing 3-year-olds in frontal crashes; the other study evaluates the head kinematics of the Q3 math model and a modified version of it, by comparing the math models’ responses to the assumed kinematics of 3-year-old children in real accidents. The restraint system study was done through mathematical simulations with the MADYMO model of the Q3 ATD. Nine parameters were studied: six related to the seat belt (such as pretensioner, load limiter and belt anchor positions); two parameters were related to the seat (stiffness and pitch angle); and one related to the foot support. The parameters that had the greatest effect on the dummy responses were: belt anchor positions, retractor pretensioner and retractor load limiter. The conclusions were that: the upper belt anchor should be positioned rearward of the dummy so that the belt encloses the dummy’s shoulder; the lap belt anchors should be positioned to make the lap belt angle as horizontal as possible without inducing submarining; and the seat belt should have pretensioner and load limiter functions. However, these recommendations need to be balanced with the recommendations for other occupant sizes and ages before implementation into any vehicle. In the second study, the Q3 math model and the modified Q3 math model head kinematics were evaluated through mathematical reconstructions of real accidents. Head x, y and z displacements were compared with the assumed displacements of the children in the accidents, and differences were identified. The modifications to the Q3 math model consisted of an increase in the bend and twist flexibility of the thorax, and an update of some anthropometric measures to new data of weights and lengths of body parts. The simulation results showed that in straight frontal crashes, the Q3 math model predicted the head displacement adequately. In oblique frontal crashes, the rigid torso of the Q3 math model appeared to obstruct the expected head displacements. The modified version correlated better to the assumed head kinematics of the children in the oblique frontal crashes. Its performance in straight frontal crashes was adequate. Future updates of the Q3 ATD and math model should include consideration of these findings. Finally, despite the identified lacks of the Q3 math model, the findings of the second study do not alter the conclusions of the first study; the results are valid also for the modified Q3 math model. The two studies are thus important contributions to the continuous work of mitigating traffic accident induced injury and fatality to children.