Evaluation of off-road uninhabited ground vehicle mobility using discrete element method and scalability investigation

Abstract

Full-scale military vehicles are teaming up with uninhabited ground vehicles (UGVs) to improve the success rate of tactical operations on off-road terrains. UGV can perform initial mobility testing on soft soils during missions to evaluate the performance (e.g., go/ no-go) of full-scale military vehicles. Therefore, the concept of scale model testing is proposed. The scale model testing can be divided into two parts, i.e. the scalability of soil and the scalability of the tyre-soil interaction. The scalability of soil is defined as a relationship between the mechanical properties of an in-situ terrain (heterogeneous) system and a laboratory (homogeneous) soil system while accounting for the differences in sand, silt and clay particle shapes and size distributions. Physical properties such as moisture content, bulk density, compaction, and interparticle forces are kept the same for laboratory and in-situ terrain conditions. The 2NS and fine-grained sands were modelled using the discrete element method with Edinburgh elastic-plastic adhesion contact model. It was found that the scalability depends on the testing conditions and the soil’s nature. The heterogeneity of soil affects the cohesive and adhesive forces present in the soil system. The pressure-sinkage and shear stress vs shear displacement relationships are found scalable. The cone index vs depth relation is not scalable. The scale model testing can be divided into two parts, i.e. the scalability of soil and the scalability of the tyre-soil interaction. The scalability of soil is defined as a relationship between the mechanical properties of an in-situ terrain (heterogeneous) system and a laboratory (homogeneous) soil system while accounting for the differences in sand, silt and clay particle shapes and size distributions. Physical properties such as moisture content, bulk density, compaction, and interparticle forces are kept the same for laboratory and in-situ terrain conditions. The 2NS and fine-grained sands were modelled using the discrete element method with Edinburgh elastic-plastic adhesion contact model. It was found that the scalability depends on the testing conditions and the soil’s nature. The heterogeneity of soil affects the cohesive and adhesive forces present in the soil system. The pressure-sinkage and shear stress vs shear displacement relationships are found scalable. The cone index vs depth relation is not scalable.Further, the scalability of tyre-soil interaction is established using the dimensional analysis method to establish similarity in the full-scale and scaled systems. The developed non dimensional parameters are kept the same in both systems. In this research, the lightweight Armoured Personnel Carriers such as the FED Alpha and Land Rover are considered as the full-scale systems (upper boundary) and UGVs for example, the Husky or Warthog as a scaled system (lower boundary). Consequently, the tyre-soil interaction behaviour is similar in this specific tyre size and loading range. The full-scale tyre modelled is FED Alpha tyre 335/65R22.5 and is scaled down by size to scales 0.7, 0.5 and 0.25. Six different terrain simulation models of both sands were prepared with cone indexes ranging from 14.79 kPa to 149 kPa. It was found that the drawbar pull and tractive efficiency vs slip relations are scalable. The mean error in drawbar pull prediction w.r.t. NATO experiments is 12% and 9% for 2NS and fine-grained sand, respectively. The drawbar pull varies from square to cubic power w.r.t. the scale of the system. The gross traction varies with square power w.r.t. the scale of the system. The tractive efficiency is constant w.r.t. the scale of the system. It is concluded that a 0.5 scale system can predict the full-scale system’s mobility performance on sands. This key finding can be used to design and develop cost-effective and lighter UGVs to support full-scale military vehicles on the battlefield. The limitation of the DEM technique is that it is computationally expensive as the number of particles increases.