Modelling the effect of spray breakup, coalescence and evaporation on vehicle surface contamination dynamics

Abstract

Vehicle surface contamination is an important design consideration as it affects drivers' vision and the performance of on board camera and sensor systems. Previous work has shown that eddy resolving methods are able to accurately capture the flow field and particle transport, leading to good agreement for vehicle soiling with experiments. What is less clear is whether the secondary break-up, coalescence and evaporation of liquid particles play an important role in spray dynamics. The work reported here attempts to answer this and also give an idea of the computational cost associated with these extra physics models. A quarter scale generic SUV model is used as a test case in which the continuous phase is solved using the Spalart-Allmaras IDDES model. The dispersed phase is computed concurrently with the continuous phase using the Lagrangian approach. The TAB secondary break-up and the stochastic O'Rourke coalescence models are used. The spray's rate of evaporation is calculated based on the relative humidity encountered on a typical October day in Britain. The secondary break-up model is found to be redundant, possibly due to the properties of spray. The coalescence model predicts high coalescence of particles close to the source and improves agreement with experiment, although at a high computational cost. Including evaporation removes small particles from the simulation and reduces overall contamination. When used along the coalescence model, evaporation is found to be negligible as it does not influence large particles to the same extent as it affects small particles. This suggests that droplet physics models need to be considered together as they can have a strong effect on each other as well as vehicle soiling. Here, we show that coalescence can be accounted for by using the time-averaged spray, obtained outside the region of high coalescence. This gives a very good agreement with experiment.