Particle resuspension modelling in turbulent flows

European Aerosol Conference 2011 / 4-9 septembre 2011

The work presented is concerned with the way small particles attached to a surface are resuspended when exposed to a turbulent flow. The process is important in a number of environmental and industrial processes. Of particular concern to this work is the release of radioactive particles from the primary circuit of an LWR as a consequence of a loss of coolant accident (LOCA). In this particular case the focus is on small particles < 5 microns in size, where the principal force holding the particle onto a surface is derived from van der Waals inter molecular forces. In general even for nominally very smooth adhering surfaces there exists a very broad spread in the adhesive forces. Here wedevelop and assess an improved version of the Rock’n’Roll model [Reeks & Hall, 2001] based on a statistical approach to resuspension which gives rise to a resuspension rate constant for the release of particles from a potential well due the action of the fluctuating aerodynamic force du to the turbulence. The analogy is with desorption of molecules from a surface in which the variance of the fluctuating aerodynamic force is the analogue of temperature. In this work we improve on the model by using measurements of the statistical fluctuations of both the stream wise fluid velocity and acceleration close to the wall from both LES and DNS of turbulent channel flow, translating those measurements into the statistical moment of the drag force acting on the particle attached to the surface. The original model assumes that aerodynamic forces and their time derivative are uncorrelated and have a Gaussian distribution. Here we examine the influence of non-Gaussian forces on the resuspension rate.

The ultimate model is a hybrid development of R’n’R model adapted for application to multilayer deposits based on the Friess and Yadigaroglu multilayer model [Friess & Yadigaroglu, 2001]. The particle size distribution is considered in the model in order to study the coveragent effect of the deposit layers. The model results are compared to the STORM SR11 test [Castelo et al., 1999] and BISE experiment [Alloul-Marmor, 2002]. As a general conclusion, the results showed that the models with small adhesive spreads give better comparison with the resuspension measured in the two
experiments.