Driver-vehicle interaction under influence of crosswind gusts

The ease with which a vehicle can be driven on motorways has become a major safety and refinement issue for the motor industry. A vehicle that requires minimal correction, instructs the driver how much to apply, and then proceeds to execute the driver's command accurately, is preferable. When driving an automobile at high speed, the feeling of directional stability is depending on the vehicle reaction to external disturbances and to driver input. To be able to evaluate that, good models both of vehicle behaviour, external disturbances and driver behaviour is needed. This work has focused on vehicle and driver behaviour due to external disturbances caused by natural crosswind gusts. A literature study on on-centre handling and aerodynamic disturbance is presented in the opening chapters of this work. To address these questions, computer simulation and vehicle measurement are normally used. Here, the two are complemented by a driving simulator including a driver, to bridge the gap between computers and real world. A vehicle computer model is used for simulations of directional stability due to crosswind gusts. The actual model describes a Volvo V40 with a large number of chassis details included. A method to identify basic tyre properties using modal analysis is presented. Also a method to ease the development of vehicle models, especially considering the implementation of alternative tyre models is developed. To validate computer models, an experimental vehicle equipped with measurement systems was developed. The aim was to create a vehicle that, with short preparation time, delivers reliable measurements. The computer model was tested through all developmental steps and displays reasonable results for the applicable level of detail. A model for external disturbance generated by a natural side wind gust, represented by the lateral force and yaw torque is proposed: The Generalized Crosswind Model. The method is intended for simulation purposes, being generated by a combination of field measurements and inverse vehicle simulation. To understand how a driver reacts to, and compensates for vehicle movement, a study of driver response was performed using the moving base driving simulator at the Swedish National Road and Transport Research Institute, VTI. Driver response to the generalised crosswind model was studied and compared to a natural gust and to the step and the ramp models. Furthermore, different gust amplitudes were tested to give an indication of the overall sensitivity of drivers. Driver behaviour was studied and the drivers categorized. Drivers were found to be characterized by different behaviour patterns concerning reaction time, compensation speed, compensation precision, tracking and normal driving strategy. Since it is clear that there are several possible combinations of driver behaviour, a vehicle should be designed to be robust for all these driver patterns. Thus, the identified patterns of driver characteristics should be considered when deriving evaluation methods of vehicle-driver response. (A)