Particle dispersion model for RANS simulations of particle-laden jet flows, incorporating Stokes number effects

An improved particle dispersion model for the Reynolds-averaged Navier–Stokes (RANS) simulation of particle-laden round, turbulent jets is proposed, which incorporates the effects of Stokes numbers on the flow and particle characteristics within a discrete random walk (DRW) framework. Modifications of the DRW dispersion model were implemented by correcting the particle–eddy interaction time and the local fluid velocity fluctuation of particles. Additionally, the dependence on the Stokes number of the key coefficients in the improved model was considered from a phenomenological perspective. The modified model was developed and validated by comparing the particle statistics in jets with well-designed direct numerical simulations (DNS) from the present study and with the DNS data from the previous study, respectively. It was found that the new model reduces the relative error in predicting the particle mass flux distribution (with respect to DNS data) by half over the conventional particle dispersion model, whilst further slightly improving the already good reproduction of the velocity phase. Also, the applicability of the new modelling approach to other dimensionless parameters, as well as to more complex flows, the confined swirling jet, is discussed.