The accurate prediction of commercial-vehicle ride and handling performance with computer simulation tools is dependent on the level of correlation between the computer model and experimental data. Correlating vehicle attributes to physical test data is often challenging due to the large number of degrees of freedom - and, correspondingly, the large number of tunable parameters - typically required to accurately model vehicle behavior. A high level of interaction between input parameters and vehicle attributes further complicates the task. As a result, this type of correlation is a multi-objective optimization exercise in which the judicious planning of supporting test activity is critical to achieving the right level of model accuracy with an acceptable amount of resource investment.
This paper discusses the methodology implemented in the validation of a tractor-semitrailer ADAMS model for both ride and handling simulations and presents the results obtained. A select set of vehicle-, subsystem- and component-level tests was performed to provide a basis for subsystem correlation and subsystem-interaction studies at the vehicle level. These included suspension characterization, modal testing, moment of inertia measurements and other specific component tests. Lastly, key vehicle ride and handling performance metrics were measured using skid pad maneuvers and on-highway data acquisition and these data used to validate the predictive capabilities of the model.
This multilevel approach proved an effective tool in developing a well-correlated model by helping to expose model deficiencies and better isolate their source. It also provided valuable insight into subsystem dynamic behavior changes due to interactions with other subsystems in the full-vehicle environment.