Optimization System Development Of Automobile Side Impact Simulation Based On Experiment Data

As the improving of road safety regulations and car crash test standard, more investment are made by auotomobile manufactors on passive safety development and tests, meanwhile, higher reliability of side impact simulation is demanded. The complicate structure of automobile results in complicate model of simulation and further decrease of simulation precision, repeated simulations are needed in one project. These factors above have a serious impact on the efficiency of new car development. What's more, the dead an injury caused by side impact significantly higher than other crash cases in traffic accidents. It is urgent to improve the side protection performance. In summary, there are positive significance both for engineering and reality to study optimizing mathods of side impact simulation and.The paper explored optimization mathods of the car side impact simulation. Firstly, PSM(Prescribed Structure Motion) mathods was used to convert the simulation from whole vehicle environment to sub-system emvironment, which greatly reduced the data exchange and calculating scale. Secondly, sensors data obtained in the crash experiment were used to settle the problem of the decreasing precision caused by accumulated bias in a series of simulations. Thirdly, a method of parametric structural deformation was put forward to transform the traditional one-way feedback(structural deformation to restraint system) to bi-diractional feedback, making it possible for targeted adjustment of structural deformation data according to the restraint system simulation result and decrease of simulation time.An optimization system of automobile side impact simulation based on experiment data was developed to integrate the above optimization mathods and bridge the data between the automotive structural crash performance and restraint system integration optimization. The system first finish the data exchange with LS-DYNA and MADYMO, second, develop a fitting algorithm support up to 10 sensors according to the structural deformation features and configuration of sensors, and avoid deformation distortion by partition optimizing method. Carry out the parametric structural deformation method by defining the feature points on X-axis and Z-axis direction. Further more, the system provides auxiliary functions such as node displacement curve plot and customized nodes output, considering the practical need of engineer. The paper end with an example applying the optimization system to verify liability and practicability of the system.