Overview of mechanistic particle resuspension models: comparison with compilation of experimental data

Abstract

In this study, several mechanistic models for particle removal from substrates were introduced and their predictive ability was tested. Empirically optimized parameters for a selected set of particle-substrate combinations were obtained and the ability of these models to predict resuspension for a large number of particle-substrate combinations and flow conditions using the published experimental data was explored. Our analysis showed that accurate predictions for a broad spectrum of particle-substrate combinations require accounting for substrate roughness and particle non-sphericity. From analysis of our compiled experimental data set, it was determined that the shear velocities required for particle resuspension follow a log-normal distribution, with geometric standard deviations in the range of 1.4 to 1.7. From comparison of the calculated and measured critical shear velocities, we found that a critical roughness parameter value of Δc = 0.774 is a reasonable estimate for most real substrates. We show that the combination of parameters determined from our study and models that account for roughness and particle non-sphericity, it is possible to predict particle resuspension from most substrates with reasonable accuracy.