Vehicle Drift Analysis using Functional Simulation and Statistical Tools 2013-36-0325
The lateral drift of a vehicle while driving straight ahead has been a major concern of the OEMs on the South America market. Due to its natural way to evaluate and also to different types of roads on the market (roads with different types of bank angle), a vehicle with slight tendency to drift will certainly be a reason of customer complaint. Since such a vehicle dynamics property is very sensible to small variation of some parameters, such as road bank angle, alignment setting, etc, sometimes the subjective evaluation tends to become worthless due to the parameters control. In this scenario, simulation becomes important.
As it is a quite difficult subject, since there is big influence of small parameters variation, the usage of statistical approach allow to obtain better understanding of the phenomenon.
This work presents a statistical approach for simulation based on DOE analysis and Monte Carlo method. As an exercise of the method a functional CAE vehicle dynamics model was specific created to evaluate the likelihood to meet a given target under variation of certain parameters. The DOE was firstly carried out in order to rank the major impacting parameters and, finally, the Monte Carlo analysis was carried out with the main five parameters identified.
The method has proven to be a powerful design tool. An advantage of the method which could be verified is its outcome, not normally obtained from conventional deterministic CAE simulations: the likelihood of achieving a specific target and the design robustness inputting variation.