Semiempirical model of a wheel-soil system

Abstract

Developing new concepts for the prediction of off-road traction is becoming essential as an increasing number of sport utility vehicles (SUVs), sport activity vehicles (SAVs), and other off-road vehicles arrive on the market. The prediction of off-road traction is a difficult problem because of the number of factors affecting wheel performance on soft soils or snow. Most of the existing methods for wheel-soil thrust prediction are based on Coulomb’s equation, in which stresses acting on the wheel-soil contact patch are analyzed. The objective of this work was to develop a new method to infer mathematical models of a wheel-soil system in which soil-stress states are correlated with the forces acting on a road wheel. A set of model equations is obtained by means of the system identification method with the use of experimental data gathered in the field with real vehicles. This paper includes a description of the experimental setup, procedures, and methods used in the tests. The experimental method is based on simultaneous measurements of the soil-stress state and wheel forces during test runs of an instrumented vehicle over soft soil surfaces. Soil-stress states consist of three principal stresses and their direction cosines, which are determined with the use of a stress state transducer (SST) placed at a depth of 15–30 cm under the wheels. Longitudinal, vertical, and transverse wheel forces are measured with a six-element wheel force transducer. The wheel forces obtained in the test runs are correlated with the respective soil stresses. Based on the data, three families of models were obtained by means of the system identification method.