This paper was produced for the 2019 NAFEMS World Congress in Quebec Canada

Resource Abstract

In this study it will be shown how uncertainty quantification (UQ) is carried out together with terramechanics simulations to generate stochastic mobility maps. The goal of creating the mobility map of the Speed Made Good for a given vehicle is to have a map that provides the maximum obtainable speed at any location in the map. In the traditional NATO Reference Mobility Model (NRMM), only the nominal deterministic values of the variables involved in the terrain properties and terramechanics simulation models are considered in generation of off-road mobility maps (Speed Made Good and GO/NOGO). Thus, the deterministic mobility maps would not be reliable and cannot be used effectively in mission planning of NATO forces under different terrain scenarios and for selection of capable vehicles. To support the Next Generation NRMM, the generation of stochastic off-road mobility maps are developed in this work using full stochastic knowledge of terrain properties and modern terramechanics modelling and simulation capabilities. This work will show how UQ is used for reliability assessment for Speed Made Good and GO/NOGO decision based on the input variability models of the terrain elevation and soil property parameters in order to generate stochastic mobility maps. For the demonstration of this work, the region of interest selected is Michigan Tech’s Keweenaw Research Center (KRC) in Michigan. For the simple terramechanics simulation model the FED Alpha (Federal Efficiency Demonstrator) vehicle is used. The UQ is carried out to create the stochastic mobility maps of the KRC area. The stochastic mobility maps generated have different probability levels, e.g., the 90% Speed Made Good map means that there is 90% probability that the maximum obtainable speed is greater than or equal to the value shown on the map. It is shown how the deterministic map has probability levels ranging between 0% to 90%, meaning the deterministic map does not provide reliable information of what the maximum obtainable speed is. This demonstrates the need for considering the variability so that accurate Speed Made Good maps can be generated and have a given probability level associated with them, to provide reliable information to the decision maker.