Attrition Scrubbers
If clays are not liberated, the final frac sand product could fall outside required turbidity
specifications. The general use of attrition scrubbers is to break down these clays, whether found
in sand size particles or as coatings on the surface of sand grains, to allow for their rejection.
Attrition scrubbing cannot change or alter the grain shape. If the particles arc not round, no
economically justifiable amount of scrubbing will change that fact.
Attrition scrubber‘s drives can either be gearbox driven, or via belts and sheaves. Gearbox driven
units tend to be higher in capital costs, but offer long-term operation without any sacrifice in
performance. Belt drives tend to be less expensive, but as the belts and sheaves wear, they need
to be replaced, adding to maintenance cost and down time Often, operators w ill add water to
prevent plugging of the attrition cells rather than replacing the worn parts. The lower percent solids
created by this added water prevents attrition scrubbers from operating as designed so product
quality is ultimately sacrificed.
Attrition scrubbing is the process where a mineral is scrubbed primarily by the action of the slurry
particles impacting one another. This type of particle scrubbing achieves the desired “cleaned”
particles while minimizing wear on the Attrition Cells.
Description
Description
Attrition Scrubbers Provide Intense Scrubbing Action at Densities of 70%-
80% Solids
The Attrition Scrubber provides a simple, economical solution to many of today’s
beneficiation problems. Many products can be made marketable by the removal of
surface films, coatings or slimes. An important example is the glass sand industry
where removal of iron stains and slimes results in a premium product. Certain
applications require the disintegration of clay balls or bituminous matter. In many
cases important mineral values occur as slime coatings on sand grains or as
cementing materials. Attrition Scrubbers remove these coatings, thus upgrading
the mineral values. Other applications include high density reagent conditioning for
better flotation such as is required in the treatment of phosphate and certain other
non-metallic ores.
The scrubbing action must be carried out at high density (between 70% and 80%
solids by weight) which results in grain-to-grain attrition within the mass. This
intense scrubbing action requires high horsepower consumption.
The Attrition Scrubber was developed to impart this horsepower with the highest
degree of efficiency. The high efficiency is achieved with the rubber – covered
Turbine Type Propellers which are designed to move a large volume of dense
material through the propeller zone with a minimum horsepower input. And, of
equal importance, is the fact that the design of the turbines minimizes abrasive
wear and cavitation.
Each cell assembly has two turbines of opposite pitch. The particles impinge on
each other in the zone between the turbines. The turbines, in addition to being of
opposite pitch also have a different degree of pitch which imparts a shearing
action, thus keeping the entire cell volume action.
Sizes range from the 11″ x 11″ tank to the 56″ x 56″ tank. Attrition Scrubbers are
available in 2-cell, 4-cell and 6-cell units. On cells of 48″ x 48″ and 56″ x 56″
enclosed reducer drives are supplied. Turbine assemblies and tanks are available
with rubber, neoprene or abrasion-resistant steel liners as required for the
application.
Attrition Scrubbing
An attrition scrubber, patterned after the U. S. Bureau of Mines model, was built at
the M.I.T. Mineral Engineering Laboratory and tested in the scrubbing of western
carnotite and roscoelite ores. Tests indicated that the sand product increased in
purity as the amount of work put into the scrubbing increased. The time of
scrubbing was of maximum importance. Speed, pulp density and size of charge
had progressively less effect on the purity of the sand product.
The choice of a scrubbing procedure will depend chiefly on economic
considerations, that is, the value of the product less the cost of power required and
the cost of subsequent treatment per ton.
The Fall Creek roscoelite ore and the Gypsum Valley carnotite ore responded best
to scrubbing, while the Rico roscoelite ore and the Mexican Hat carnotite ore gave
the poorest response.
The object of attrition scrubbing is to produce clean grain surfaces by the removal
of surface stains or cements without appreciably reducing the size of the discrete
grains. In some cases the cleaned grains are the valuable product, as in glass
sands, or where a selective flotation or electrostatic separation is to be made, as in
the case of spodumene and chromite, but in others the stain or cement is the
valuable constituent, as exemplified by carnotite and roscoelite ores.
A number of machines may be used to accomplish the scrubbing, among which
are grinding mills, both with and without a grinding medium, flotation machines,
paddle, screw or rake washers, and especially constructed agitators or attrition
scrubbers.
The attrition scrubber, a relatively new tool in the field of mineral dressing, is a
laboratory machine that was invented by Mr. James Norman and developed at the
Eastern Experiment Station of the Bureau of Mines. Its function is properly
described by the word scrub, which means “to rub something hard, or to cleanse
by rubbing”, and the redundant name was conferred on it to emphasize the
intensity of its rubbing action. The motor that drives the attrition scrubber is larger
than the container for its charge. The attrition scrubber has had limited application
up to now, but some installations of scrubbers of this type have been made on a
commercial scale. The history of the development of the attrition scrubber,
including a bibliography, was given in a letter from Mr. John Dasher to Mr. J. K.
Gustafson dated April 5, 1948, a copy of which forms appendix G of this report.
An attrition scrubber, patterned after the Bureau of Mines model was constructed
at the M.I.T. Mineral Engineering Laboratory and tested in the scrubbing of four
western carnotite and two western roscoelite ores. The origins of these ores,
together with comparative analyses, are given in Topical Report MITG-207 dated
September 30, 1918, but for purposes of ready reference Table 1 below duplicates
some of this information.
SUMMARY
An attrition scrubber was built and tested in the scrubbing of four Western carnotite
and two Western roscoelite ores. The tests showed that the sand product
increased in purity as the amount of work put into the scrubbing increased, the time
of scrubbing being of maximum importance. Speed, pulp density and size of
charge had progressively less effect on the purity of the sand product.
Metallurgically, there is small choice between the quality of work done by the
attrition scrubber and some other methods of scrubbing, such as treatment in a
flotation machine or attrition with steel punchings in a mill, so the choice of
scrubbing procedure for any particular ore will depend chiefly on economic
considerations, that is the value of the product less the cost of the power required
and the cost of subsequent treatment per ton. Since work with the carnotite and
roscoelite ores was terminated before power requirements for the different
scrubbing procedures could be determined, this question is still unresolved.
Comparison of the results obtained with the attrition scrubber under standard
conditions with the best results by other scrubbing procedures for each of the ores
investigated, is shown in Table 2.
The Fall Creek roscoelite ore and the Gypsum Valley carnotite ore gave the best
response to scrubbing, while the Rico roscoelite ore and the Mexican Hat carriotite
ore were the most difficult to scrub.
More extensive testing of all the ores was outlined. It was planned to determine the
power requirements for each of the different scrubbing procedures, to make
additional tests at longer intervals of self-attrition, and at shorter intervals of
scrubbing with steel punchings, so that more comparative results might be
obtained. Also an investigation of other scrubbing media, such as rubber balls or
rubber covered steel rods should still be made.
DESCRIPTION OF APPARATUS USED
Attrition Scrubber
The attrition scrubber constructed at the M.I.T. Engineering Laboratory is an
agitator, with a rotor and a stator which almost fill the space where the pulp is
confined. The space between the rotor and stator is small, but it is several times
the diameter of the largest grain to be scrubbed, so that no grinding takes place,
but a severity of agitation sufficient to rub off the coating on mineral grains without
breaking the grains is attained.
The stator is made from a section of five inch steel pipe, nine inches long, welded
to a square steel plate. The pipe is lined with 0.25 inch of Haveg, a laminated
phenol-formaldehyde plastic, and three sets of blades, four in each set, 0.5 inch x
1.5 inches x 0.125 inch are attached to the inside of the pipe with threaded pins.
Each blade in a set is placed at an angle of 90 or 180 degrees with the other
blades in that set and the blades are covered with rubber.
The rotor consists of a shaft, 0.5 inch in diameter, to which three sets of blades,
four in each set, similar to those in the stator but 0.25 inch longer, are welded
cruciformly. The blades are two inches apart and are so placed that when the shaft
is rotated, and the stator is in place, they pass between the stator blades. The rotor
blades are also covered with rubber.
The rotor is mounted on a steel framework, which carries the Doall Speedmaster
(Model 4B) used to change or control the rotor speed, and the electric motor to
drive the rotor. A 0.5 hp, 1725 rpm, single phase electric motor is used.
A direct-reading tachometer is connected to the top of the rotor shaft to give
continuous speed readings and an ammeter and voltmeter are connected to the
motor for power measurements.
A tight-fitting, grooved cover, that is held in place by four springs anchored to the
stator base, fits over the top of the stator and prevents loss of charge during
operation.
Plate 1 shows the attrition scrubber closed and ready for operating.
Plate 2 shows the stator dropped and the rotor ready for cleaning.
Plates 3 and 4 show the details of construction and method of mounting the
tachometer and ammeter.
Rotating Mills
For the self-attrition tests an assay-size porcelain mill running at 72 rpm was used,
and for the tests where steel punchings provided the scrubbing medium, a slowly
rotating (6 rpm) cast iron mill, 8″ in diameter x 7″ long, charged with 4 kilos of 0.5
inch x 0.125 inch steel punchings, was employed. The cast iron mill was turned by
placing it on two rotating steel rolls.
Flotation Machine
A standard laboratory model Fagergren flotation machine having an impeller speed
of 1775 rpm was used for the scrubbing tests at 20 per cent solids, and this same
impeller and housing, stripped of its “wings”, was used for some of the tests at 50
per cent solids by lowering the mechanism into a 1500 ml beaker containing the
pulp. The beaker rested or a rubber mat to prevent rotation.
EXPERIMENTAL METHODS
Standard Conditions for the Attrition Scrubber
For the purpose of comparing different methods of operating the attrition scrubber,
certain arbitrary conditions were selected and designated as “standard, operating
conditions.” These conditions were based on published data obtained in the
operation of a similar attrition scrubber and appeared to be a satisfactory starting
point for testing the MIT machine.
Standard speed was selected as the speed obtained in operating the scrubber
when one-quarter horsepower of energy was being expended on 500 grams of
charge at 50 per cent solids. It was determined by running the scrubber empty, to
obtain a measure of the power consumed by the motor, bearings, pulleys and
belts, then adding the pulp and adjusting the scrubber speed until the ammeter
showed that one-quarter horsepower was being impressed, in addition to the
power required by the empty scrubber – a power factor of 85 per cent was
assumed for the small fractional horsepower motor used.
Standard time of scrubbing was selected as the time necessary to give a work
input of 20 hp-hrs/T. It was calculated from the following formula.
hp-hrs/T = input (hp) x treatment time (min.) x 15100/sample weight (grams)
and was determined as 2.7 minutes, for a 500 gram charge at 50 per cent solids,
1500 rpm. and 0.25 net hp input.
Standard pulp charge for the scrubber was 500 grams of ore at a pulp density of 50
per cent solids.
Standard Conditions for Other Scrubbing Methods
In making all the self-attrition tests, as well as those wherein steel-punchings were
the scrubbing medium, the standard ore charge was 500 grams and the standard
pulp density 67 per cent solids.
Five hundred grams of ore was also the standard charge for the Fagergren
flotation mechanism tests, with 20 per cent solids standard for the regular flotation
cell tests and 50 per cent solids for the beaker tests. Standard time of scrubbing for
all of these tests was 20 minutes.
Sedimentation Procedure
Fractionation of the scrubbed pulps from all tests was carried out in the following
manner. Four pounds of Daxad 23 was added per ton of ore, usually to the pulp
during scrubbing, to insure thorough dispersion. The scrubbed pulp was
transferred to an eight liter jar, 23.5 cm in diameter and 23.5 cm deep, and distilled
water added until the pulp depth became 13 cm. This gave a pulp dilution of 10 per
cent solids for the tests with 500 gram ore charges, 14.3 per cent solids for the
tests with 750 gram ore charges and 18.2 per cent solids for the tests with 1000
gram ore charges. The pulp was then vigorously stirred and allowed to settle for 16
minutes, at the end of which time the suspended solids were removed by a special
glass siphon with a turned-up tip on the short end. The depth of pulp removed was
11.5 cm out of the total depth of 13 cm. The volume was next made up to the
original point with distilled water, the pulp agitated, allowed to settle for 16 minutes
and a second fraction of suspended solids removed as before. Both 16 minute
fractions were combined and treated as a single fraction.
The volume was made up to the original point again and two fractions removed, as
described above, after 4 minutes of settling and also after 1 minute of settling.
The solids remaining after the removal of the second one-minute fraction were
considered to be sands.
All products were dried in an electric oven, without the removal of any liquid except
by evaporation, and the dried material was weighed, sampled and assayed.
Four fractions were made in each test rather than the two required, because, at the
time the tests were run, it was not possible to get rapid assays of the test products,
and consequently it could not be determined whether the split should be made
after sedementation for one minute, two minutes, four minutes or some other time.
RESULTS OF ATTRITION SCRUBBER TESTS
The influence of four variables on the quality of work done by the attrition scrubber
was investigated by tests on two different ores. One a carnotite, that was relatively
easy to scrub, as shown by self-attrition tests in a porcelain mill without scrubbing
medium, and the other a roscoelite, that similar tests showed was difficult to scrub.
Tests were run on each of these ores wherein the time of scrubbing, the pulp
density, the speed of scrubbing and the size of the charge were varied one at a
time.
Effect of Time of Scrubbing
The effect of varying the time of scrubbing is shown in Table 3. The results indicate
that as the time of scrubbing increases, and more work is done on the ores, the
weight per cent of the sands fraction decreases, as well as the content of both
uranium and vanadium in the sands fraction.
The carnotite ore showed the best response to scrubbing. Doubling the time of
scrubbing this ore resulted in a lowering of both the uranium and vanadium content
of the sands by about 10 per cent each time it was doubled.
The roscoelite ore was more refractory. Doubling the time of scrubbing resulted in
only about 4 to 6 per cent reduction in the uranium and vanadium content of the
sands of this ore each time, and a twenty-fold increase in scrubbing time still left
better than 38 per cent of the total uranium in the sands.
Effect of Per Cent Solids
The effect of varying the pulp density of the material being scrubbed is shown in
Table 4.
With the exception of the power consumed, the results follow different patterns for
each of the two ores tested.
Progressive increases in the pulp density of the carnotite ore resulted in a
corresponding increase in the uranium content of the sands fraction, although
almost no change in the weight per cent of the sands was shown. The vanadium
acted in about the same way as the uranium.
The roscoelite ore, however, showed a decrease of about 10 per cent in both the
uranium and vanadium content of the sands fraction as the pulp density was
increased from 50 per cent solids to 70 per cent solids. A small decrease, about 3
per cent, was also obtained in the weight per cent of the sands.
Increase in the pulp density, though, resulted in a sharp decrease in the
horsepower-hours per ton consumed by both the carnotite and roscoelite ores, and
this decrease is about the same for each ore.
It should be pointed out that the calculated horsepower-hours per ton shown for
some of the tests is probably high because of a mistaken belief that the attrition
scrubber had been properly worked-in at the time the blank measurement of the
amount of power consumed by the scrubber itself was made, and that there would
be no appreciable change in this amount during the short time the tests were being
made. After completion of the scrubbing tests some anomalous results prompted a
check of the blank, and it was found that a considerable reduction in power
consumption had taken place. The scrubber drew 9 amperes at first, but only 7.5
amperes at the end of the tests, so the horsepower calculations were made on the
basis of this last figure and are, without doubt, unduly high in some instances. The
effect of the blank reading is large, for the standard ampere increase due to the
pulp is only 2 amperes, and the 1.5 amperes difference in the blanks means that
the 20 hp-hrs/T intended as work input may actually have been 35 hp-hrs/T.
Effect of Speed
The effect of running the attrition scrubber at 80 per cent and 120 per cent of the
standard speed is shown in Table 5.
Operation of the scrubber at 80 per cent of the standard speed resulted in lowering
the work input by about 50 per cent, in scrubbing both the carnotite and roscoelite
ores, but about 10 per cent more uranium was left in the sands fraction of the
carnotite ore, and 6 per cent more in the sands fraction of the roscoelite ore by this
action.
Increasing the speed to 120 per cent of the standard resulted in an increase of
work input with both ores, but the increase was not uniform, being greater with the
roscoelite ore than with the carnotite ore.
An optimum speed at which the scrubber should be operated is indicated by the
tests, especially those with the softer carnotite ore, where higher speed not only
resulted in an increase in work input but also in the amount of uranium contained in
the sands fraction. The optimum scrubber speed for the roscoelite ore was
evidently not attained, because progressively higher speeds showed less and less
uranium in the sands fraction.
Effect of Size of Charge
Table 6 shows the effect of varying the size of the charge to the attrition scrubber
while operating at 1200 rpm, or 80 per cent of standard speed.
The results show that, aside from efficiency of operation, the size of the charge to
the attrition scrubber has little effect on the quality of work done by the scrubber.
For maximum efficiency, therefore, the scrubber should be operated at its full
capacity.
OTHER METHODS OF SCRUBBING AS COMPARED TO ATTRITION
SCRUBBING
Both the carnotite and roscoelite ores were tested by methods of scrubbing other
than with the attrition scrubber. These methods comprised self-attrition in a rotating
porcelain jar for 2, 4 and 6 hours, scrubbing in an iron mill with steel punchings for
2, 4 and 6 hours, agitation in a Fagergren flotation machine at 20 per cent solids
for 20 minutes and agitation in a beaker at 50 per cent solids with a Fagergren
machine mechanism, minus its wings, for 20 minutes.
The six graphs, Figures 1-6, which follow show the results obtained with these
different scrubbing procedures as compared with scrubbing in the attrition scrubber
under standard conditions, each individual ore being represented by a single
graph. The data from which these graphs were drawn may be found in Appendixes
A to F, inclusive.
The graphs were constructed by plotting the per cent of U3O8 in the sands against
the weight per cent of the sands, with 100 weight per cent and zero per cent U3O8
at the extreme right. It follows, then, that the lower down on the graph and the
further to the right that a point appears, the better the result it represents.
Keeping this fact in mind it appears that the work of the attrition scrubber was
outstanding only with the Gypsum Valley carnotite ore (sample D3). With the other
ores there is little choice between the results obtained with the attrition scrubber
and with those obtained by other methods of scrubbing, excepting that the self-
attrition tests gave consistently poorer results than those of any other scrubbing
method.
The problem of choosing the best method of scrubbing resolves itself therefore,
into one of economics. Is it more economical to scrub three minutes in the attrition
scrubber than twenty minutes in the Fagergren cell or 2 to 6 hours in a mill with
steel punchings? Tests were contemplated where the power consumed in these
different scrubbing procedures would be measured, but the work was discontinued
before they could be made, so the question is still unresolved. Other tests were
also contemplated to get more comparable results, specifically, tests of self-attrition
at longer time intervals and of scrubbing with steel punchings at shorter time
intervals. Also tests of other scrubbing media, such as rubber balls or rubber
covered rods, were planned.
