A New Discrete Formulation of the Rock’n’Roll Resuspension Model for Time-Dependent Eulerian Flow Solvers

Particle resuspension significantly contributes to particle exposure, yet its integration into air quality models remains limited. Existing couplings between resuspension models and CFD codes predominantly rely on threshold velocity approaches that do not fully represent the progressive depletion of the surface particle load or the role of turbulent fluctuations in driving detachment. This study presents a discrete formulation of the Rock’n’Roll resuspension model that overcomes these limitations by treating the adhesion force distribution as a prognostic variable. This formulation is compatible with time-stepping Eulerian flow solvers and is demonstrated through its implementation in the PALM Large-Eddy Simulation code. The discretized formulation is validated against experimental data for 10 and 20 μm particles, showing good convergence and agreement across a range of friction velocities. Remaining discrepancies are attributed to the log-normal adhesion force distribution hypothesis rather than to the coupling itself. Analysis of the simulated resuspension dynamics reveals that the flow acceleration phase dominates total particle removal, with up to 87% (respectively 94%) of 10 μm (resp. 20 μm) particles resuspended before the steady state is reached. Flow heterogeneity alone generates significant spatial variability in remaining fraction with implications for the interpretation of experimental measurements. The proposed framework is extensible to other kinetic resuspension models, paving the way for improved resuspension modelling across a broader range of particle sizes, surface conditions, and complex flow environments.