MODELING PARTICLE DEPOSITION IN TURBULENT DUCT FLOWS USING LAGRANGIAN APPROACH

摘要

This paper studied the ability and accuracy of a commonly used Reynolds-averaged Navier-Stocks (RANS) turbulence model, i.e. RNG k-ε model combined with a Lagrangian discrete random walk (DRW) model to predict particle deposition in vertical and horizontal turbulent duct flows. The performance of RNG k-ε mode in its original form and with near-wall correction was tested. A new and simple correction measure to turbulence kinetic energy was proposed for RNG k-ε model. It was found that RNG k-ε model greatly over-predicts the deposition velocity in its original form. The near-wall modification proposed in this paper is able to improve the simulation results on deposition. With the near-wall correction measure, RNG k-ε model can reflect the V-shape deposition of micro- and nano-meter particles.