Results (
Thai) 2:
[Copy]Copied!
1. Introduction
The separating action of a hydrocyclone treating
particulate slurry is a consequence of the swirling flow
that produces a centrifugal force on the fluid and suspended
particles. The feed slurry is injected tangentially
into the hydrocyclone at high velocity, to produce a
large centrifugal force field. The feed moves down the
wall rapidly and generates a helical vortex, which extends
beyond the lower end of the vortex finder. This
swirling flow is highly turbulent and three-dimensional.
In the centrifugal field, the particles move relative to the
fluid with respect to the balance of centrifugal and drag
forces acting upon particles in the radial direction, such
that classification occurs. The coarser or heavier particles
move toward the wall and are swept downward to
the apex of the cone. The fluid phase which carries the
smaller or lighter particles, approaches the apex and
reverses in the axial direction spiralling upward and
leaving the hydrocyclone through the vortex finder.
Along the axis, an area of low pressure is created by the
very high angular momentum. This may cause the formation
of a rotating free liquid surface at the centre. If
the hydrocyclone is open to the atmosphere, air is inhaled
through the apex and forms an air core. In that
case, the pressure at the air-liquid interface is equivalent
to atmospheric pressure (neglecting both the surface
tension and viscous forces).
The hydrocyclone, despite its ubiquitous applications
in the chemical, metallurgical and other industries, still
requires specific investigation since the flow-field is not
completely understood. Fisher and Flack (2002) recently
published experimental studies of the flow in hydrocyclones.
This has been informative with regards to the
internal flow-field dynamics, but important aspects are
still not understood. For example:
The frequently reported but anomalous ‘fish-hook’
effect which results in an excess of fines reporting to the
underflow, has not been categorically explained. This is
also the subject of other papers to be presented in this
international congress volume.
Hydrocyclone modelers have largely ignored features
such as the nature of air-core development, with
simplified air-core assumptions being made.
Detailed knowledge of the flow structure is required
if one is to consider such issues as energy saving, costeconomy or product quality. Several benefits could arise
from this knowledge, for instance: areas of high erosion
may be identified and potentially minimised or accounted
for in design; design modifications for improved
separation or reverse-design of cyclone geometry could
be obtained.
The drivers behind the application of simplified
physical models of hydrocyclone behaviour are principally
the issues of: complex flow behaviour arising from
the three-dimensional flow entry; multi-phase interactions;
and the mechanisms governing the formation of
an air-core (when the hydrocyclone is open to the
atmosphere). Consequently, computational studies have
been in general, limited to low particle-concentration
flows and to simplified geometries of the hydrocyclone
entry region. Advanced theoretical and experimental
techniques are still needed to obtain a better understanding
of the very complex physical phenomena
affecting the performance of hydrocyclones. The
knowledge of phenomena such as particle–particle,
particle–fluid, and particle–wall interactions would open
the way to the description of particle effects for suppression
or generation of turbulence and for non-
Newtonian slurry flows.
This paper reviews the existing models and addresses
how the computational algorithms could be extended to
include complex interactions between continuous and
dispersed phases. Specifically, the paper considers
methodologies for velocity field prediction and modelling
of the particle distribution. A strategy for future
developments is outlined. Also, essential measurements
for the validation of hydrocyclone predictions and
appropriatemethods for achievement of these are identified.
Being translated, please wait..
