Microscopic observation of aerosol particle deposition in turbulent channel flows

The transport behaviour of carbonaceous dust in the primary circuit of a High Temperature Reactor (HTR) plays an important role in the safety assessment during a Design Basis Accident (DBA). Carbonaceous dust is formed mainly due to friction between graphite fuel elements (Kissane, 2009). In a pebble bed reactor the dust forms due to abrasion of graphite material between the pebbles. During reactor operation it is a safety issue to precisely predict the dust deposition and the corresponding resuspension rate of particles released into the containment during a DBA.
Deposition of aerosol particles has been investigated e.g. by Sippola & Nazaroff (2004). These experiments have been performed in steel ducts at Reynolds numbers and with particle relaxation times observed in a HTR. Their results follow the “v-shaped” curve of the non-dimensional deposition velocity against particle relaxation time and show a dependency on flow speed, particle size and orientation of the duct.
More recently, an Euler-Lagrange CFD simulation of particle deposition in a turbulent square duct flow has been performed by Lecrivain (2012). It was found that the friction velocity significantly influences the particle deposition velocities.
Deposition experiments of liquid and solid particles in a fully developed horizontal turbulent square duct flow are performed to further study the deposition behaviour of aerosol particles. The oil liquid particles (DEHS) are generated by a condensational aerosol generator (TOPAS, SLG 270). The aerodynamic particle size distribution and the particle number concentration of the suspended particles are determined by isokinetic sampling using an Aerodynamic Particle Sizer Spectrometer (TSI. APS 3321). The particle size distributions are fairly monodisperse for particles with daero = [1.5, 2.5, 3.5, 4.5] µm. The solid aerosol particles are microspheres (AkzoNobel, Expancel DU, d50 = 6.5 µm) and are injected into the flow field by means of a solid aerosol generator (TOPAS, SAG 410). The particle mass flow rate and the particle concentration are precisely adjusted by the feed rate of the SAG. The particle size distribution of the airborne particles is measured by means of Scanning Electron Microscopy (SEM). The size distributions gained by the APS and the SEM analysis are used to calibrate the size distributions obtained by an optical microscope.
A commercial light microscope equipped with a CMOS camera is mounted underneath the test section of the channel. It is focused on the inside surface of the channel floor which consists of a glass plate coated with indium tin oxide to remove electrostatic charges. This allows a time-resolved in situ observation of the particle deposition processes. A standard LED light source illuminates the microscope’s 6 x 4 mm² field of view and the CMOS camera records the scatter light of the wall deposited particles. The measurement uncertainty for particles larger than 2 µm is assumed to be 24%.
Figure 1 illustrates the time averaged deposition velocities for varying friction velocities. The CFD results of Lecrivain (2012) and Sipolla & Nazaroff (2004) are also plotted for comparison purposes. In the particle relaxation time range τ+ = 0.001..10 the deposition velocity increases with decreasing friction velocity. which is also observed elsewhere. It is assumed that the effect of gravitational settling leads to an increase of deposition velocity for decreasing flow speed. The results of this study also capture this tendency. Nevertheless, the scatter in the data has to be further investigated.