Promoting excellence in mobility engineering

  1. FISITA Store
  2. Technical Papers

Virtual Experimentation on Vehicle Dynamics on a Multi-Physics Desktop - Part I: Dynamics and Active Control
HELSINKI2002/F02I247

Authors

Velardocchia, Mauro - Department of Mechanics Politecnico di Torino
Sorniotti, Aldo - Department of Mechanics Politecnico di Torino
Krief, Philippe - FIAT Auto Technical Direction
Suraci, Enrico - FIAT Auto Technical Direction

Abstract

The design of innovative vehicle components, like active devices for vehicle dynamic and traction control, needs a preliminary virtual experimentation for the due target setting and deployment operations, based on several approaches. Some mathematical models as well as several software environments are considered to support the steps of the product development. A preliminary computation of the dynamic response of the vehicle in time and frequency domains can be performed by simplified models of the vehicle on elastic suspensions with a reduced number of degree of freedom (d.o.f.), based on the equations of motion and developed in MATLAB and SIMULINK environments, while detailed analyses are usually provided through multibody codes including control and flexible bodies effects.

In a sort of a bottom up approach the vehicle can be studied in a pure MATLAB/SIMULINK environment, including chassis and powertrain, then by adding active control devices. In case of interaction with complex subsystems like tires it looks useful to run simultaneously different codes dealing with the control system, the tire, the vehicle, etc. in the same environment, i.e. SIMULINK, by connecting available and specialized blocks in a sort of a virtual desktop for the design including multiphysics numerical simulators. A similar concept can be developed in a multibody dynamics code by including several toolboxes in the same code or by co-simulating different blocks.

The aim of the paper is to show some examples concerning the methodology adopted during the co-operation performed by Fiat Auto with the Politecnico di Torino Vehicle Dynamics Team on the development of suitable methods for a realistic prediction of the vehicle behavior both from the point of view of the functionality and the safety and structural vehicle dynamics. In a first part of the paper shall be presented the modeling of the vehicle subsystems equipped with active control systems, like VDC, ABS/ASR/MSR, CDC, skyhook dampers and the suitable tuning of the relevant parameters performed on the experimental results acquired by Fiat Auto. A qualitative evaluation of the performances looks possible, within certain margins, also on simplified models running in light simulators, for instance in MATLAB / SIMULINK environments. They allow avoiding to run on workstations, under the control of specialized engineers, available only at the design units and offices. Furthermore new concepts and components can be tested and analyzed through simplified models, with a reasonable number of details and d.o.f. User is allowed to identify the main steps of the analysis and the blocks of the code performing numerical computations, dealing with the different subsystems of the vehicle (tires, suspension, chassis, powertrain). This way is possible to select the level of complexity of the model applied for the above calculations. Few examples show how the simulation of the motor vehicle dynamics in MATLAB and SIMULINK environments, that actually is definitely not new, can be provided by inputting data and formats available and tunable during real experimentation on road.

A detailed experimental validation has been performed comparing the numerical results of the implemented models and the experimental ones, acquired during road test maneuvers. The proposed nonlinear models have been implemented meeting the main targets of the road testing procedures. Inputs can be introduced in the format utilized by suppliers, particularly for what concerns suspensions and tires data. Outputs are presented through the conventional diagrams used for performance evaluation. Simple interfaces provide easy changes in parameters. Responses in time and frequency domains are considered and computed. Although frequency response is a delicate task of the analysis of nonlinear systems, the computation of power spectral density of the monitored signals is enclosed in the code in order to evaluate the main excitations applied to the vehicle, according to the standards.

Add to basket