Parameterization and Validation of Road and Driver Behavior Models for CarMaker Simulations and Transmission HIL-Rig

The thesis analyzes whether the results from simulation and powertrain rig tests correlate to the actual loads of a vehicle driving in a real transport task. The rig used in the research uses the same software as the simulations, called CarMaker, but when conducting tests in the rig the models of transmission and engine are removed. Instead the actual components are used as Hardware-In-the-Loop (HIL). To be able to get good results from the simulation and rig tests, a representative model of the physical road had to be developed. The driver behavior model had to be parametrized in a way that resembles a real driver. The road used in the thesis is a pre-defined city route in Göteborg, used to resemble normal driving situations which include large amount of accelerations and decelerations, since city driving has the most influence on the durability of the transmission. The road model is built up by using x and y coordinates which describes the physical roads’ choice of path, surface friction and road attributes as speed limits and stop signs. The road model does not mimic the road on a micro but rather a macro scale, meaning that local abnormalities like potholes and stones are not modeled, the limitations are further explained in Section 1.4. The driver model used in the thesis is included in the CarMaker software, therefore the task was to parametrize the model to ensure that it was good representation of actual drivers. The focus when developing driver models was to investigate a defensive, normal and aggressive driver type. In order to shorten lead times and improve the economy of product development, testing and verification of products should be conducted earlier in the development process. An effective way to allow this is by the use of Computer Aided Engineering (CAE) tools. In order to be able to trust the results from the CAE tools, the models used must be trustworthy, which was the main focus of interest during the project. During the thesis, a good correlation of the road model was achieved, both in terms of choice of path and the ability to perform test on the road equally in vehicle, rig and simulation testing. It turned out to be hard to represent a real driver using the driver behaviour model included in the CarMaker software. Instead, possible solutions to this problem is presented by using and developing suitable key indicators, which explain how the driver behaves and the limitations in the CarMaker driver behavior model. The research also showed that a valid vehicle model is crucial for the results. All in all, the results from simulations, rig tests and vehicle tests correlate in a good way. The key indicators span approximately the same span when comparing the different tests. This is valid for both key indicators on transmission shaft fatigue load and energy consumption at drive shafts, which shows on a good potential for future development and use.