Sensitivity Analysis of Virtual Crash Simulation Software Using Design of Experiments (DOE)

<div class="section abstract"><div class="htmlview paragraph">This paper is a continuation of a previous effort to evaluate the post-impact motion of vehicles with high rotational velocity within various vehicle dynamic simulation softwares. To complete this goal, this paper utilizes a design of experiments (DOE) method. The previous papers analyzed four vehicle dynamic simulation software programs; HVE (SIMON and EDSMAC4), PC-Crash and VCRware, and applied the DOE to determine the most sensitive factors present in each simulation software. This paper will include Virtual Crash into this methodology to better understand the significant variables present within this simulation model. This paper will follow a similar DOE to that which was conducted in the previous paper. A total of 32 trials were conducted which analyzed ten factors. Aerodynamics, a factor included in the previous DOE, was not included within this DOE because it does not exist within Virtual Crash. The same three response variables from the previous DOE were measured to determine the effect of the various factors. These response variables are the <i>x</i>-coordinate, <i>y</i>-coordinate and total rotational displacement of the vehicle at rest. The results show that, consistent with the previously tested vehicle dynamic simulation software programs, roadway drag was a significant factor in predicting all three response variables. There is also a sensitivity to the longitudinal speed, as well as multiple factor interactions in determining the y-coordinate. Multiple factors – weight, yaw inertia, CG height, roadway drag, longitudinal speed, and angular speed – were important for rotational displacement. Along with the DOE from the previous papers, this paper conducted a separate DOE analysis to evaluate two additional factors; roll inertia and pitch inertia. Five trials were run to evaluate the effect of these factors on the same three response variables from the previous DOE. The results show that the only significant factor was the pitch inertia in determining rotational displacement. It’s noted that pitch inertia was nearly significant in determining the y-coordinate but fell marginally short of the statistical threshold (0.005) used by the authors.</div></div>