Contamination Study for Heavy Duty Vehicles using Smoothed Particle Hydrodynamics

Abstract

Advancement in automobile technology has lead to increasingly complex driver assistance technology. Many of these involve the placement of sensors on the external surface(s) of vehicles. In the case of heavy duty vehicles, these sensors could be exposed to harsh driving conditions which could lead to contamination or soiling of their surfaces. Sensors must perform reliably and safely in such extreme conditions and must be kept clean for optimal performance. Additionally, contamination can also adversely affect the visibility of the driver as well as other road users. The durability and lifetime of components are also affected by contamination. Thus, studying contamination is vital. Contamination study for heavy duty vehicle development includes a combination of physical testing and numerical simulations. Traditional numerical simulations use a computational mesh, which makes the inclusion of multiphysics and replication of vehicle kinematics a formidable task. An alternative to the current method, broadly known as purely particle based methods, removes the need for a computational mesh and allows various topological changes to be captured more easily. The purpose of this thesis was to evaluate particle based methods for studying contamination, in particular Smoothed Particle Hydrodynamics (SPH). Three software developed on SPH were used to simulate splashing phenomena of a simplified as well as a detailed truck model driving through a pool of water. These simulations were used to study the capabilities of the software in terms of accuracy of the modelled physics, computational demand and ease of usage. The three SPH-based software studied were LS-Dyna, Preonlab and Simcenter SPH. SPH being of Lagrangian framework allowed inclusion of kinematics needed for replicating the motion of a truck moving through a pool. The methodology included an initial parameter study to understand the behaviour of certain tuning parameters in each software needed to setup realistic simulations. The comparisons of the results from the parameter study was achieved by capturing the splash height and wetting area on the surface(s) of the vehicle. Further, a detailed truck model was used to replicate a splashing scenario in the three software whose results were then compared against a similar physical test. The study has shown that SPH is a reliable method to capture the large scale phenomena such as waves seen while driving heavy duty vehicles through water pools. Splash and spray phenomena in some of the tools evaluated might require a smaller SPH particle sizing to capture the required droplet density seen in the same phenomena in reality. It was also noted that the resolution of mesh used in importing the geometry has an influence when evaluating the surface area wetted after the simulations due to the interpolation techniques used to display the results. Further, the cost of computation of the three software were compared as part of the parameter study.