Locomotion on soft soil is important for many navigation and exploration operations, including explorations for coastal, petroleum, and fresh water resources, extraterrestrial terrain navigation, mapping operations for mine and radiation detection, and search and rescue operations. In general, motions on soft or loose soil require high energy expenditures compared to motions on packed surfaces. Through support of this award, fundamental research will be pursued to gain further understanding of locomotive interactions with different granular media. This research will draw upon knowledge from various areas, including soil modeling, numerical simulations, vehicle dynamics, and data mining. A salient impact is expected to be the advancement of our understanding of vehicle maneuverability in granular media and related strategies for energy efficiency. The cross-disciplinary research will provide exceptional learning opportunities for the researchers involved and also usher in new generation of researchers trained in computational modeling and sciences.
Guided by the team's prior efforts in computational and experimental dynamics, nonlinear phenomena, and fluid-structure interactions, this research effort will be pursued with a focus on understanding granular media interactions associated with wheeled and legged locomotion on soft soils. Original experiments will be conducted by using photo-elastic granules to examine inter-granular force chain development in response to local surface disturbances generated by external loading. These experiments are expected to generate data for simulations of granular interactions based on discrete element models. These simulations will leverage GPU computing. Continuum, geomechanical-hydrodynamic simulations will be used to associate observed force chains with specific soil types. The research effort will help answer basic questions such as the following: i) Can force chains be used to predict slipping in soft-soil interactions? ii) Does an optimum gait speed exist at which soil slipping is minimized? iii) Can spheres, combined with complex friction models, accurately simulate interactions of sand grains and other non-spherically shaped bodies in a statistically averaged sense? and iv) Can one apply statistical distributions such as the Tracy-Widom distribution to interactions with granular media? The findings are expected to provide a unique means for interrogating granular reactions to locomotive interactions in a way not tractable through experimentation alone. The outcomes are expected to be useful for developing efficient propulsion strategies for robot platforms that span a range of different sizes, weights, speeds, and energy consumption targets.
