Abstracts Category : Other

Characterisation of vehicle infant restraint systems for assessment of adaptive restraint technologies

by Timothy Barnard

This thesis investigates the potential to use fast solving computational models in conjunction with injury prediction metrics for use in global optimisation schemes for the assessment of restraint system concepts. The issue has been addressed in the past through live and cadaver testing but issues arise due to ethical issues associated with their use and the required limits on maximum accelerations as well as issues related to their associated accuracy. Test dummy testing provides a means of acquiring accurate results without the associated difficulties but repeated testing is associated with high costs and is, therefore, not a feasible option for global optimisation techniques. Computational models allow for high iteration optimisation techniques if the correct modelling method is utilised. Further, differing accuracy models can be used in conjunction with each other in order to further establish accuracy and identify global minimisations of injury potential. These topics are discussed in detail in the literature review chapter in order to provide a better understanding on the area of research. This research is then used to develop a simplified computational model of an infant exposed to crash accelerations. The design of each body segment is then independently assessed and combined in to one model before the equations of motion describing their motion are derived and incorporated in to a python model. Validation of the computational model was conducted through comparison with actual test dummy results and injury potentials. The results of the simulation were representative of the test dummy results, however, some small error as a result of design choices were present. The pendulum design of the head, in conjunction with the high centre of gravity associated with infants, resulted in an over estimation in the prediction of injury to the head. The injury of the head, however, while higher than those experienced by the test dummy, were still within the maximum permissible values specified by associated technical standards. Finally, a case study assessing the potential to use injury potential as a platform for restraint stiffness optimisation was then performed. Difficulties in correctly simulating the head injury metric within the minimisation of injury method did arise but it was shown that with the addition of a method that adaptively selects the time envelope based on restraint stiffness would eliminate the issue as the actual calculation of the head injury metric was accurate. While the simulations prediction of particular injury metrics were overestimated, the model was able to be used as a basis for optimisation of restraint stiffness. The case study indicated that by reducing the total stiffness of the shoulder belt stiffness it is possible to reduce injury potential, thereby, proving that the model can be used for global optimisation schemes.