Brain injuries, resulting from frontal impacts, were simulated with a FE model of the brain, in which kinematics data of the head are used as input. Estimations of three forms of brain injury are predicted using three criteria. The evaluation of the model's injury criteria is presented in this study and compared against current criteria using different sled tests configurations. The input data of the FE model came from tests on fresh cadaver: three series at 50kph with a 22Gs peak deceleration pulse (one with airbag and a 4kN load-limited belt, two without steering wheel and with a 4 and 6kN load-limited belt, respectively). A series at 30kph with lower deceleration pulse (15Gs), and only a 4kN belt was also performed. The head kinematics was measured at the centre of gravity using multi-accelerometers. The influence of impact conditions is evaluated comparing tests with identical restraint devices and reducing the velocity and deceleration. The HIC decreases strongly for a lower crash severity. The magnitude of the linear acceleration is similar, but the shapes of the curves are different. The rotational accelerations are also reduced but not the angular velocity.The change of the severity does not appear clearly on the Gambit value, because this criterion depends on both the linear and angular accelerations and, there is a lack of correlation between these two variables. There is also a great dispersion between the tests for the three brain injury criteria without any clear trend. The influence of restraint systems is assessed at 50kph using three different restraint systems. It appears that the HIC and Gambit values are lower with the 6kN belt than for the 4kN belt. The highest values of these criteria are found for the tests with airbag, but the two other series were performed without steering wheel (i.e. without head contact). The kinematics data showed the same trend. However, the angular accelerations and velocities do not show significant differences between the tests with and without airbag using the same belt. No trends can be found for the influence of the restraint systems on the RMDM values with respect to velocity, which were quite identical, nor on the DDM values which were scattered. The greatest CSDM values, corresponding to about 40% of the probability of DAI are obtained for the test with a 6kN belt. All the other test conditions do not show CSDM values above 20% except for one test with airbag where all the kinematics data were found high, and the lowest values (corresponding to less than 10% of probability) were found for the test with only the 4kN shoulder belt. The results presented here are preliminary observations from simulations in order to assess the traumatic brain injury risk in frontal car crash. Some attempt of description of injury mechanisms are given, but more detailed analyses are necessary and are planned for future work on more cases including the comparison with direct impact (impactor test in different directions). For the covering abstract see ITRD E141807.
Samenvatting