Page 182 - Instrumentation Reference Book 3E

166 Particle sizing

Manometric methods have been used to assess particles care has to be exercised at the injection
the sedimentation. A manometer fused into the point.
side of the column near the base and filled with
the clear fluid will register the change in mean Ceiztiifiigal methods For small particles, gravi-
pressure of the column with time and this of tational systems are very slow-acting. There is
course depends on the mass of material still in also a lower limit to the size of particle that can
suspension. Pressure differences are very small. be measured because of the effects of Brownian
Beta particle back-scattering has been used to
measure the mass of material at the base of a motion and convection. While it is possible to use
gravitational systems for particles as small as
column. The intensity is proportional to the 1 pm in water and about 0.5 pm in air, in practice
atomic number and the thickness, Le., the weight the lower limit is usually taken to be about 5 pm.
of sediment, provided the thickness does not build These problems can be reduced by the use of
up to saturation level.
centrifugal enhancement of the settling velocity.
The theory for centrifugal settling from a
Decantiizg If a homogenous fluidlsample mix- homogeneous fluid mixture is complicated in par-
ture is allowed to settle for time tl seconds and ticular because the terminal velocity varies with
the fluid down to a depth iz is then removed using distance from the center of rotation. Approxima-
a pipette, particles removed would all have a tions have to be made which limit the usefulness
terminal velocity less than hlt,. In the decanting of the techniques.
method, this process is repeated several times, The theory for two-layer systems is exact and
replacing the removed fluid with clear fluid and therefore these are more attractive. Unfortu-
re-mixing until the supernatant fluid is clear. The nately, a phenomenon known as “streaming”
process is then repeated but with a shorter time, can occur, in which particles tend to agglomerate
t2. The removed fluids are analyzed for dust con- and accelerate in a bunch instead of individually,
tent, each containing the fraction of the total behaving as a large particle; this renders the
mass of material of terminal velocity between results useless. Streaming has been prevented by
hlt,, and Itlt+~. The accuracy depends on the pre- using extremely low concentrations of particles in
cision of 12, t, the rate of removal of fluid. and the the starting layer. A technique using a third,
number of repeated decantations at each value oft; intermediate layer of immiscible fluid has also
so it is not high. been used successfully.
Theories always depend on the applicability of
Stokes’s law to the particle motion and this
Two-lajier methods If the upper layer is signifi-
cantly thinner than the lower layer at time t = 0, imposes a restriction on the maximum particle
size, depending on the speed of rotation. Further
then after a time t = fl, the only material to have problems exist for larger particles with respect to
reached the base of the column will be those the acceleration and decleration of the centrifuge.
particles with terminal velocity greater than hltl, In spite of the many problems, the techniques
where h is the height of the column, and a meas- have advantages for very small particles, as small
urement of the weight of those particles gives the
cumulative distribution directly. Liquid two-layer as 0.01 pm, and systems are available paralleling
methods are not common because of the difficul- many of the gravitational techniques. The most
promising appear to be the three-layer methods
ties of arranging the starting condition. The and the use of optical detection devices.
Granulometer uses a shutter system to maintain
separation until t = 0 and a manometer to meas-
ure the pressure changes as particles sediment. 11.8.2 Elutriation
The Sharples Micromerograph uses an air col-
umn approximately 2m tall. The sample is A group of particles suspended in a fluid that is
injected using a puff of compressed nitrogen moving upwards at a velocity V will undergo
between two concentric cones to disperse the par- separation, those particles with a terminal velocity
ticles and break up the agglomerates. An elec- less than V traveling upwards, the others settling
tronic balance coupled to a pen recorder monitors downwards. This process is called elutriation. The
the arrival of particles at the base of the column. fluid is usually water or air depending on the
The column is jacketed to maintain thermal sta- particle sizes. Strictly, an elutriator is a classifier
bility. Errors are experienced due to the loss of rather than a particle sizer. It divides the group of
some fine material to the walls of the column. particles into those above and those below a given
These can be reduced to some extent by anti- cut size. In the perfect system, the fluid would
static treatment. The time scale of the recorder clarify except for the few particles with terminal
is adjustable, fast for the initial phase, slow for velocity exactly V, and the settled particles could
the later phase of the sedimentation. With friable be removed and would contain no undersized