Simulating vehicle river crossing or fording
Driving a vehicle in deep water can be dangerous and may damage the vehicle. Electronics, air intake and other areas may be sensitive to water. With simulations of various configurations of water depth and speed, as well as vehicle CAD alternatives, manufacturers can precisely set maximum wading depth and reduce the vehicle’s vulnerability to water.
High-speed fording of shallow water generates high-pressure splashes that may impact the underneath structure of the vehicle. For example, safety covers located under the vehicle protect the engine and other critical parts from impacts and stress. Their resistance is assessed by manufacturers using experimental testing of prototypes. By bringing forward these critical tests to an earlier stage of the design process, the use of CFD simulation can prevent costly design issues.
A typical fording experiment involves various complex physical phenomena. As well as considering complex car geometry and motion, one must also account for free-surface flow and multiple atomization and coalescence events. Such complexity represents a significant obstacle for most conventional solvers, and also explains the difficult use of CFD to simulate fording until now.
Based on the SPH method, the SPH-flow solver is particularly well-suited to address this problem. The essential scattering of the fluid into multiple droplets and their possible coalescence can be properly captured.
Among the studies Nextflow Software has conducted on river crossing, a relevant one involves a 10 cm-deep and 100 meter-long water ford that must be crossed several times by the vehicle at a constant speed up to 40 km/h.
For this study, the air flow does not need to be simulated: only the water phase is accounted for. Free surface condition is therefore imposed at its boundaries. Besides, no viscous nor surface tension models have been turned on: although available in the SPH-flow solver, their effects can be proven to be negligible from a brief dimensional analysis.
The wide-scale discrepancy between the length of the ford and the centimeter characteristic length of the car body details can be considered as one of the main numerical challenges. Adaptive Particle Refinement (APR), implemented within the SPH-flow solver, drastically reduces the restitution time while keeping the simulation accurate. Indeed, by focusing the computing effort on highly discretized particles in the areas of interest, the 10-second simulation can be achieved in reasonable computation time without compromising on accuracy.
Data of interest for the designers mainly concern the forces on the safety covers. The SPH-flow solver allows a complete set of data to be analyzed, including total forces and local pressure fields, average value and temporal evolution. The wettability of the car components can also be visualized and processed.
For test conditions in which the water stress is expected to cause significant deformation of structural parts, Fluid-Structure Interaction (FSI) may need to be accounted for. In such cases, the SPH-flow solver can be coupled with a Finite Element structural solver. Different coupling strategies are available to strike a balance between robustness, accuracy and computational efficiency.