Impact of the adhesion-force lever-arm “a” on the Rock ‘n’ Roll resuspension model and how to compute it from contact mechanics

Abstract

The paper proposes a methodology to fill a gap in the context of mechanistic particle resuspension models related to the adhesion and lift force lever arms (a): to date, only one experimental observation exists from a centrifuge experiment with a specific particle/surface pair which found rP/a = 100 and assumed that the contact points were symmetric with respect to the particle center (Reeks, M.W., & D. Hall (2001) Kinetic Models for Particle Resuspension in Turbulent Flows: Theory and Measurement. Aerosol Science, 32, 1–31). However, the Reeks and Hall’s (2001) model is highly sensitive to this parameter, which therefore cannot be used as-is for different particle diameters or materials. This paper proposes a methodology to compute the adhesion and lift force lever arms (aA and aL, respectively) based on Atomic Force Microcopy (AFM) surface roughness measurements and contact mechanics arguments. In practice, a spherical, perfectly smooth particle is brought in contact (computationally) with the measured surface roughness and is then rotated by the point of contact to determine the two or more asperities over which the particle would sit on the surface. A method to compute the asperity radius of curvature is devised to compute the asperity deformation and the correct partitioning of the total deformation between the particle and the asperity is presented. For the glass surface under consideration, the asperity deformation was negligible, i.e., it was below the van der Waals radius, which is the limit of applicability of contact mechanics. The Reeks and Hall’s (2001) model is then extended to account for the joint probability of aA and aL. Significant differences in particle resuspended after a day in common atmospheric conditions are found between the approach that uses the lever-arm distributions or their mean values. Also, important differences are found when comparing the lever-arm distribution and rP/a = 100. Hence, one should be cautious of using rP/a = 100 for surface/particle pairs differing from those in the original study. Limits of applicability of the theories used are presented together with considerations regarding how to extend the method to non-spherical particles and different materials.