On Euler-Lagrange URANS turbulence models for predicting the transient dispersion of aerosols indoors

• Accurate prediction of indoor particle deposition rate is crucial in CFD studies. • Among URANS models, SST k-ω is a robust choice for indoor aerosol spread. • k-ε models over-predict particle deposition rates. • k-ε models predict stronger turbulence effects on particles near walls than SST k-ω. • Enabling the TKE production limiter enhances the accuracy of k-ε models. Accurate prediction of aerosol dispersion in indoor environments is vital for understanding airborne disease transmission and optimizing ventilation strategies. This study investigates the transient dispersion of aerosols in a controlled ventilated environment through a combination of experimental and numerical methods. A fully transient Euler-Lagrange approach utilizing Unsteady Reynolds-Averaged Navier-Stokes (URANS) turbulence models coupled with unsteady tracking of discrete particles using a stochastic method, was employed to simulate particle dispersion, focusing on three widely used models: RNG k-ε, Realizable k-ε, and SST k-ω. A custom-built experimental chamber provided low-concentration in-house data for validating the numerical simulations under precisely replicated conditions. The results demonstrate that while all three turbulence models captured general aerosol dispersion trends, the SST k-ω model most closely matched the experimental data. The findings reveal that the deposition rate is a significant source of error, particularly with the k-ε models, which predicted a decay curve with lower concentration and a different slope than the experimental data. These models tend to overpredict turbulent kinetic energy near surfaces, leading to the calculation of stronger artificial eddies that last longer, resulting in inaccuracies and an overestimation of the particle deposition rate. The study underscores the importance of selecting appropriate turbulence models for reliable predictions of aerosol behavior in indoor environments. Furthermore, the experimental data reported serves as a valuable resource for validating numerical approaches, particularly the Euler-Lagrange method, in future research.