Model Based HandlingAnalyses

Abstract: The thesis consists of building vehicle models for lateral dynamics that provide accurate results, while being simple enough to highlight clear relationships between vehicle properties and handling behaviours. The approach that characterizes the present work is built on the foundations provided by the bicycle, or single-track, model (STM). Nonetheless, its developments and improvements substantially differ from the usual sequence found in literature. Most studies based on this analytical vehicle model satisfy the quest for increased robustness by involving nonlinear description of the contact forces generated between tyres and ground. The presented analysis, instead, embraces the challenge of using completely linear models to describe vehicle lateral dynamics in a variety of driving situations. Hence, the aim of this thesis is to provide the reader with a deeper understanding of the effects of various physical properties on the actual dynamic behaviour, by complementing numerical correlations obtained from simulation outputs with analytical descriptions of their causes. The first step of the presented work deals with the selection and design of driving scenarios aiming at comprehensively evaluating vehicle handling, as well as defining the boundaries of the investigation. As in most engineering problems, a compromise has to be sought, since the set of maneuvers has to be broad enough to point out various dynamic characteristics, while limiting the time needed to perform evaluation of a real vehicle. Furthermore, the space occupation of each maneuver must allow execution in an actual testing facility. The design of suitable driving scenarios triggers an iterative process, in which the initial formulation of the single-track model is complemented with additional states and sources of complexity, based on the limitations that simulation outcomes point out. A complex model, previously built with the aid of Multi-Body Dynamics software, is initially accepted as a \black box", and adopted as a reference to estimate qualities and deficiencies of the results. The next step of the work deals with physical vehicle testing and correlation. After post-processing the data and comparing it to the analytical vehicle model, it is seen that the vehicle model represents the physical vehicle behaviour very well. The area where the model lacks is in transients, which in-turn can be explained by the lack of compliance in the system. The last part of the thesis brie y touches upon the vehicle behaviour in a dynamic driving simulator along with effects of motion cueing gains.