An Improved Slip-Based Model for Tire-Snow Interaction

The performance of a vehicle traversing over natural snowy terrain depends on the geometric and material properties of snow; these properties can change spatially and temporally depending on environmental conditions. The author's research group in recent years has been developing physically-based tire-snow interaction models that are snow-depth and slip dependent; these efforts are in contrast with previous models that are mostly empirical. An important element in tire-terrain interaction models is the pressure-sinkage relationship typically modeled empirically using plate indentation tests. Recently, a new mechanics-based indentation model for snow has been developed which significantly improves and simplifies the prediction of pressure-sinkage relationship. In addition, new vehicle-snow interaction slip-dependent test data have been obtained by the author's research group for the validation of vehicle-snow interaction models. In this paper, the new indentation model for snow is incorporated into a tire-snow interaction model; the results of the model are compared with the new test data. Parametric studies of the model indicate that the motion resistance and traction when normalized against the normal force on the tire decrease with the increase of normal force; normalized motion resistance and traction decrease with the decrease of snow depth. Softer snow leads to larger motion resistance. The minima and maxima of normalized drawbar pull and torque as measured by an instrumented vehicle are compared with the results from the improved model. The model with a larger coefficient of rolling resistance has reasonably good comparison with the test data; the slips associated with the minima of drawbar pull and torque also show good comparison with test data; on the other hand, slips associated with the maxima predicted by the model compare poorly with test data indicating that a better shear stress-shear displacement sub-model is needed. The range of tire sinkage from the model compares well with the range of sinkage of test data obtained using a 3-D laser-based profilometer.