The number most often quoted in cyclone selection is the cut size (d₅₀): the particle diameter at which efficiency is 50%. On its own it is misleading, because the real question is: how much of the dust distribution you actually have will be captured?
Two cyclones can share a d₅₀ and still differ markedly on the fine side if the slopes of their efficiency curves differ. And it is usually the fine side that sets the emission limit.
Loading is frequently overlooked
At low dust loading, separation occurs through the motion of individual particles in the centrifugal field. Above a certain loading a different mechanism takes over: particles are swept to the wall en masse, and efficiency comes out higher than single-particle theory predicts.
The Muschelknautz method treats these two regimes separately and therefore gives more realistic results at industrial loadings than classical approaches; the book by Hoffmann and Stein covers the method and the limits of its application in detail [1].
The practical consequence: carrying an efficiency curve measured at low loading in the laboratory over to a heavily loaded process oversizes the design unnecessarily.
Modelling: two difficulties at once
The swirl is anisotropic. In the rotating flow inside a cyclone the tangential velocity follows a profile made up of forced- and free-vortex regions. Two-equation models that assume isotropic turbulence get this profile systematically wrong; a Reynolds stress model or large eddy simulation is required. See the turbulence model note.
The vortex core precesses. The low-pressure core on the axis does not stand still; it swings around (precessing vortex core). That is a transient phenomenon, invisible to a steady solution, which also tends to underpredict the pressure drop.
On the particle side. In discrete phase tracking the drag coefficient must be computed as a function of Reynolds number; the piecewise fit of Morsi and Alexander is widely used for this [2]. Running with a single mean diameter erases the most critical quantity — the escape fraction of the fines. See the multiphase note.
Geometric details matter disproportionately
- Vortex finder immersion depth. Too short and the incoming flow short-circuits straight to the outlet.
- Inlet geometry. Tangential, scroll or vaned inlets produce different tangential velocity profiles at the same flow rate.
- Dust hopper. Reverse flow in the hopper can re-entrain separated dust; this is the cause of many field efficiency shortfalls, and it lies in the hopper rather than in the cyclone itself.
References
- Hoffmann, A. C. & Stein, L. E. (2008). Gas Cyclones and Swirl Tubes: Principles, Design, and Operation, 2nd ed. Springer. DOI: 10.1007/978-3-540-74696-6
- Morsi, S. A. & Alexander, A. J. (1972). An investigation of particle trajectories in two-phase flow systems. Journal of Fluid Mechanics, 55(2), 193–208. DOI: 10.1017/S0022112072001806