Multiphase CFD simulation of water separation in automotive air intake systems

Abstract

Clean and dry air is of the essence for the automotive combustion process as well for the cab climate control systems. Thus, it is important to achieve e cient water separation in the air intake system. Depending on the size and velocity of the water droplets, the water will be separated from the air and exit through the drainage, enter the air intake system or create a thin lm on the inner walls of the air intake. This thin lm of water might later break up into droplets and once again mix with the air. With a reliable and e cient numerical methodology for prediction of water droplets and water lm behavior in the air intake system, it will be possible to optimize the air intakes before any prototypes are built. To predict the behavior of air and water ow in the air intake systems, two multiphase models in the CFD software STAR-CCM+ were investigated and compared on a simpli ed geometry. The models used were the Lagrangian Multiphase model (LMP) and the Dispersed Multiphase model (DMP). The investigation was done by conducting several small studies of the LMP and DMP models together with their submodels. Due to its advantages, such as particle tracking and more developed submodels, more time was spent on investigating the LMP model than the DMP model. The Lagrangian Multiphase model was also tested on a complete air intake system. To model the lm inside the air intake Fluid Film modelling and Volume of Fluid (VOF) were tested. Using the results from these studies, together with the ndings of the literature review, conclusions were made that the LMP model is more suitable to use for simulate water separation in the automotive air intakes. Due to limitations with the stripping model, VOF could not be used to model the lm inside the air intake. Therefore is the Fluid Film model recommended. The volume fraction of water has a great impact on the percentage of injected water that accumulates in the lter for some droplets diameter. This needs to be taken into consideration when performing simulations and experiments.