Assessment of a dummy model in crash simulations using rating methods

Abstract

With fast development in high performance computer area and modeling techniques, design of injury countermeasures based on numerical simulations could make the process more cost-efficient. To reduce the risks and uncertainties associated with the use of predictive models for design decision making, an increased attention should be paid to the validation process of human dummy models. The main goal of this study was to assess a numerical THOR dummy model in frontal crash environment using kinetic and kinematic test data and various validation metrics. A THOR-NT dummy positioned on a rigid planar seat and restrained by a standard 3-point shoulder and lap belt system was subjected to two frontal crash sled tests with “initial velocity 40 km/h”. In the numerical analysis, the dummy model was positioned in a test setup model using the same procedure employed in testing. The dynamic and kinematic data of dummy model were compared with corresponding test data using validation metrics implemented in two rating systems. The overall dummy kinematics, the 3D displacement time histories of certain dummy body regions, and the interaction forces of dummy with the sled and the belt obtained in simulation were in reasonable agreement with the corresponding test data. Generally, the rating scores showed to be sensitive to the validation metrics and to the weighting factors used in the assessment process. The current average score of THOR model positioned it in the “fair-to-acceptable” range, which would recommend it for use in impact simulations. The results of this study and the validation methodology may be used for further refinement of the current THOR model, in setting up future dummy tests, the assessment of other human dummy models, and the improvement of model rating methods.