MAVS VTI Model

The Mississippi State University Autonomous Vehicle Simulator (MAVS) incorporates multiple vehicle-terrain interface (VTI) equations to simulate tire-terrain interaction on various surface types including sand, clay, dry and wet pavement, snow, and ice. The MAVS vehicle simulation has several unique features which will be highlighted in this work. First, it is dynamic, meaning that the simulation updates at discrete time intervals, $dt$, during which aspects of the tire-soil interaction such as the tire deflection, angular velocity, slip, and load will change. Second, the VTI works on rough, uneven terrain, meaning the load and resulting VTI must be calculated individually for each tire at every time step. Finally, the MAVS vehicle simulation is modular, meaning the tire model and the vehicle-terrain interaction model are separable. This implementation is similar to the one in the CREATE-GV Mercury software [Goodin 2017].

Steps in the VTI Calculation

MAVS uses a rigid-body multibody dynamics (MBD) model to simulated the vehicle. In the MBD implementation, elements can be manipulated by applying forces and torques. Dynamic state variables like position, orientaiton, and velocity are then calculated by the MBD code. VTI forces are applied at the hub of each wheel, and each wheel is connected to the sprung mass (chassis) by a slider joint with a linear spring-damper system. The forces applied at the hub come from the tire and VTI model. The following subsections detail the sequence of the VTI calculation in MAVS. A basic overview of the MAVS tire/terrain update step is outlined below in the figure below.