Experimental validation of a differential variational inequality-based approach for handling friction and contact in vehicle/granular-terrain interaction

Abstract

The observation motivating this contribution was a perceived lack of expeditious deformable terrain models that can match in mobility analysis studies the level of fidelity delivered by today's vehicle models. Typically, the deformable terrain-tire interaction has been modeled using Finite Element Method (FEM), which continues to require prohibitively long analysis times owing to the complexity of soil behavior. Recent attempts to model deformable terrain have resorted to the use of the Discrete Element Method (DEM) to capture the soil's complex interaction with a wheeled vehicle. We assess herein a DEM approach that employs a complementarity condition to enforce non-penetration between colliding rigid bodies that make up the deformable terrain. To this end, we consider three standard terramechanics experiments: direct shear, pressure-sinkage, and single-wheel tests. We report on the validation of the complementarity form of contact dynamics with friction, assess the potential of the DEM-based exploration of fundamental phenomena in terramechanics, and identify numerical solution challenges associated with solving large-scale, quadratic optimization problems with conic constraints. 1. DEM for Terramechanics: Modeling and Numerical Solution Strategies Adopted This contribution is motivated by an ongoing effort to identify predictive modeling approaches that can characterize dynamics of off-road vehicles. The salient feature of off-road maneuvers is the presence of deformable terrain. Owing to the complex soil behavior, deformable terrain continues to pose significant hurdles that limit the spectrum of scenarios that can be analyzed through computer modeling. The task undertaken is timely, given that it is difficult and expensive to evaluate a vehicle's performance during a majority of off-road maneuvers using physical experiments. Indeed, the range of scenarios that can be considered for physical testing is limited due to time and cost constraints. It is thus desirable to employ computer modeling in a virtual prototyping exercise that, when drawing on physics-based, predictive models, can improve designs, compress the release cycle, and reduce costs.

Tensor Train Accelerated Solvers for Nonsmooth Rigid Body Dynamics

“…The complementarity method for frictional contact, as well as the solution methods for its CCP relaxation have been thoroughly validated and contrasted with experimental data [MJN16,MFJN17]. We 4.1.…”

Truck tyre-terrain interaction modelling and testing: literature survey

“…Later in 2009, Slade (2009) developed an off-road rigid ring model for truck tire using FEA. On the other side, few studies have employed mesh-less smoothed particle hydrodynamics (SPH) (Lescoe, 2010;Goodin and Priddy, 2016) and discrete element (DE) (Nakashima et al, 2010;Melanz et al, 2016) methods for modelling tyre-soil interactions. One year later, Lescoe (2010) developed a soil model in a tire-soil interaction using FEA and smoothed-particle hydrodynamics (SPH).…”

Deformable soil with adaptive level of detail for tracked and wheeled vehicles

“…Among the various approaches presented in literature, we mention the use of FEA soil with elasto-plastic continuum coupled with FEA tyres (Xia, 2011) and the use of the discrete element method (DEM) for the soil, both with rigid wheels (Zhang and Li, 2006;Nakashima et al, 2007) or with deformable FEA tyres (Nakashima and Oida, 2004;Wasfy et al, 2014). We remark that an alternative to using DEM is the non-smooth dynamics approach, based on differential variational inequalities (DVI), discussed in Smith et al (2014) and Melanz et al (2016).…”