Prediction of nanoparticle deposition in channel and duct flows: a simple method to implement new turbulent dispersion models in CFD tools

In most current industrial applications, Reynolds averaged turbulence models (RANS) are used due to their relative simplicity and computational efficiency. Such models are generally combined to a Lagrangian approach to predict particle transport and deposition in turbulent flows. The prediction requires accurate modeling of the airflow turbulence and the interactions between particles and eddies (turbulent particle dispersion model). The default dispersion model used in most CFD codes is the Eddy Interaction Model (EIM), which frequently overestimates the deposition rates. In this work, a simple method is proposed to implement a stochastic one-step dispersion model, based on the Langevin equation, in the Fluent (R) code. The performance of the prediction of nano- and micro-particles transport and deposition in turbulent channel flows with this model is compared to available numerical data obtained with Fluent (R) simulations using the EIM. The use of an anisotropic Langevin model is shown to improve the accuracy of deposition prediction.