Experimental and Numerical Investigations of Skim Milk Powder Stickiness and Deposition Mechanisms

Abstract

The particle gun method and Computational Fluid Dynamics (CFD) modelling was used to study stickiness and deposition mechanisms of skim milk powder in an impingement jet hitting a stainless steel plate. The particular focus was on the effect of jet velocity and particle size distribution on deposition. Low jet velocities of 10.3, 14.8 and 19.4 m/s were studied at fine particle size levels of 30, 51, and 61 mm, using a jet-plate height to jet diameter ratio of 4. For skim milk powder with a bulk particle size (d(0.5) = 61 μm), lowering the air velocity from 19.4 m/s to 10.3 m/s increased the level of deposition and decreased the point at which deposition first occur as measured by the temperature difference between the glass transition temperature (Tg) of amorphous lactose and the air jet temperature of the particle gun. This point is called (T-Tg)critical. The critical point decreased from 39.0 °C to 18.6 °C as the velocity decreased from 19.4 to 10.3 m/s and the (T-Tg)critical obtained at the lower velocity is in closer agreement with previously reported fluid bed rig results. The (T-Tg)critical point and level of deposition was also found to be highly dependent on particle size. Increasing the average particle size from 30 μm to 61μm increased the (T-Tg)critical from 8.2 °C to 18.6 °C and 14.8 °C to 39.0 °C for jet velocities of 10.3 m/s and 19.4 m/s respectively. Levels of particle deposition also dramatically decreased for both velocity ranges. Ring shaped deposit morphologies were observed with increasing particle stickiness. Beyond (T-Tg)critical powder deposits formed at the periphery of the plate creating a large round clear zone which decreased until a striped deposit ring formed and finally deposits formed only at the centre. Particles were observed to bounce radially from the centre of the plate before sticking. Milk powder deposits are therefore governed by the kinetic energy of the impinging particle in addition to particle surface stickiness. The particle gun method was modelled using Fluent CFD software as an adhesion phenomenon arising from the particle-surface contact dynamics of a particle laden impingement jet contacting a vertical collection plate. The development of a wall boundary condition for specifying the particle-surface interaction has been the focus. A particle is captured by the wall if the impinging kinetic energy is below the prescribed critical normal kinetic energy; otherwise the particle rebounds with reduced kinetic energy. The model was developed through the User Define Function option of Fluent. The CFD model confirmed that particles rebound radially from the collection plate several times before sticking. Circular deposit morphology results from such modelled contact dynamics which are similar to the observed experimental deposits. The level of deposition predicted by CFD increased with increasing levels of critical normal kinetic energy, in the same way experimental deposits increased with increasing particle stickiness. The current model did not considered the contribution from the tangential velocity component to particle stick/rebound behaviour, but it is expected the tangential velocity may also play a significant role and should be included in future CFD models. It is recommended that the particle-surface interaction needs to be studied in more detail, preferably with imaging systems such as Particle Image Velocimetry (PIV), so that individual particle trajectory and deposition behaviour can be followed and analysed.