Search Results

This paper describes a software system that automates the process of comparing experimental to simulation data. First, it plots all experimental and the numerical signals for direct comparison. A rating of each numerical signal is calculated by comparing peak amplitude values, timing of peak amplitudes, and signal shape information, and expressing the correlation as a scalar. Weighting factors can be applied to each score, which can be used to increase the importance of some, and decrease the importance of other signals. This allows for the evaluation of the model for a specific purpose or loading condition. A report describing the model validation can be generated automatically. The report contains a description of each test performed, a description of the corresponding simulation model, rating scores, and graphs for each output signal.

Side impact occupant safety has been evaluated and optimized using a MADYMO vehicle compartment model with prescribed structural motion. A CPU efficient approach is presented where reliability is optimized in three stages: Stochastic (random) analysis to calculate the scatter that could be expected for the initial design Optimization taking scatter into account by safety margins derived from the first stage Robustness verification To evaluate scatter resulting from dummy hardware variations a stochastic model of the ES-2 dummy was used. Variation of dummy & seat position and variation of airbag and door trim properties were included using the scatter database developed in the EC projects VITES and APROSYS. Two sets of 150 simulations were used for the scatter analysis (including 23 scatter parameters) and 120 simulations were necessary in the optimization (including 3 design parameters). The CPU per analysis was 2 hours and the total process took 9 days using up to 6 processors.

A computational model of a mid-size sport utility vehicle was developed using MADYMO. The model includes a detailed description of the suspension system and tire characteristics that incorporated the Delft-Tyre magic formula description. The model was correlated by simulating a vehicle suspension kinematics and compliance test. The correlated model was then used to simulate a J-turn vehicle dynamics test maneuver, a roll and non-roll ditch test, corkscrew ramp and a lateral trip test, the results of which are presented in this paper. The results indicate that MADYMO is able to reasonably predict the vehicle and occupant responses in these types of applications and is potentially suited as a tool to help setup a suite of vehicle configurations and test conditions for rollover sensor testing. A suspension system sensitivity study is presented for the laterally tripped non-roll event.