Search Results

In the early stage of vehicle development process, it is customary to establish a set of goals for each kinematic and compliance (K&C) characteristic and try to find out design variables such as the location of hard points and bushing stiffness which can achieve these goals. However, since it is very difficult to find out adequate set of design variables which satisfy all the goals, many engineers should rely on their own experiences and intuitions, or repeat trial and error to design a new suspension and improve old one. In this research, we develop a suspension design process by which suspension K&C characteristic targets can be achieved systemically and automatically. For this purpose, design optimization schemes such as design of experiments (DoE) and gradient-based local optimization algorithm are adopted.

This paper presents a methodology of correlation between subjective and objective measures of vehicle on-center handling performance. The subjective measure is a professional test driver's rating of vehicle handling, while the objective measure assesses the handling performance via vehicle dynamic responses. Vehicle test data obtained from field testing has been analyzed to investigate links between the objective and subjective measures. Fifty-six physical parameters have been derived from on-centering hysteresis curves. Statistical tools are employed to obtain good correlation between driver rating and physical parameters. Using an interaction formula, a statistical model which relates the driver rating and principal physical parameters has been obtained. The proposed methodology will be used to show the physical parameters influence on subjective assessment and even to predict the subjective assessment of a vehicle handling performance.

Torsional stiffness properties were developed for both a 53-foot box trailer and a 28-foot flatbed control trailer based on experimental measurements. In order to study the effect of torsional stiffness on the dynamics of a heavy truck vehicle dynamics computer model, static maneuvers were conducted comparing different torsional stiffness values to the original rigid vehicle model. Stiffness properties were first developed for a truck tractor model. It was found that the incorporation of a torsional stiffness model had only a minor effect on the overall tractor response for steady-state maneuvers up to 0.4 g lateral acceleration. The effect of torsional stiffness was also studied for the trailer portion of the existing model.

Validation was performed on an existing heavy truck vehicle dynamics computer model with roll stability control (RSC). The first stage in this validation was to compare the response of the simulated tractor to that of the experimental tractor. By looking at the steady-state gains of the tractor, adjustments were made to the model to more closely match the experimental results. These adjustments included suspension and steering compliances, as well as auxiliary roll moment modifications. Once the validation of the truck tractor was completed for the current configuration, the existing 53-foot box trailer model was added to the vehicle model. The next stage in experimental validation for the current tractor-trailer model was to incorporate suspension compliances and modify the auxiliary roll stiffness to more closely model the experimental response of the vehicle. The final validation stage was to implement some minor modifications to the existing RSC model.

A determination of the vehicle states and tire forces is critical to the stability of vehicle dynamic behavior and to designing automotive control systems. Researchers have studied estimation methods for the vehicle state vectors and tire forces. However, the accuracy of the estimation methods is closely related to the employed model. In this paper, tire lag dynamics is introduced in the model. Also application of estimation methods in order to improve the model accuracy is presented. The model is developed by using the global searching algorithm, a Genetic Algorithm, so that the model can be used in the nonlinear range. The extended Kalman filter and sliding mode observer theory are applied to estimate the vehicle state vectors and tire forces. The obtained results are compared with measurements and the outputs from the ADAMS full vehicle model. [15]