CA1296673C - Density classification of particulate materials by elutriation methods(improved) - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

This specification relates to a counter-flow sedimentation separator and process which is designed to separate mixed particulate materials on either side of a preset cutoff density and to a method of separating such materials. The separator has been designed to highgrade gold, zinc, tin, lead, barite and gold tailings, but can be used to perform a similar function with other materials given an adequate density difference between the materials to be separated. It employs a series ofcounterflow separation units used in conjunction with screening operations in such a way as to effect density separations. This process is specially provided with new and inventive embodiments which increase systems efficiency and density selectivity, while decreasing water and energy requirements.

Description

a~73 TITLE QF~rl~ INVEN rlON

Density Classification of Particulate Material by Elutriation Methods This invention relates to a process and apparatus particularly applicable to the mining indushy by which particles are separated according to density.

BACKGRO~D C)F TE~E~ IN~NTION

In the mining industry many minerals such as alluv;al gold, lead, zinc, tin and barite are often benificiated at least in part by density separation systems.
15 Present techniques such as heavy media separators, jigs, or tables can be employed to accomplish these tasks, however the cost of performing such separations on a large scale can be quite high. Other more dynarnic methods such as sluice boxes must be operated on a batch process basis, and are often inefficient with respect to recovery of particles at the fine end of the particle size 20 spectrum.
Forms of elutriation techniques have been in use in laboratories for decades where particles are classified according to diameter, as in a "Cyclo-sizer" (Trademark of Warman International). This system utilizes the differential sedimentation rates of particles with different diarneters to effect 25 separations according to size. In this application particle density must be constant in order for consistent size classifications to be made. I
A patent of background interest to the present invention is United States Patent number 2,429,436, issuedOct. 21, 1947 to G.B. Walk.er. Walker teaches 30 that the "dynamic sedimentation rate" of a particle, th~t is the rate at which a particle under the influence of gravity settles against a vertical flow of "separation medium", is determines by the partical diameter and density, and by the density and flow rate of the separation medium. (The separation medium is described as a liquid with solids in suspension.) Thus, where particles are 3S settling at an equal rate against a given flow, the smaller particles of this group will be of relatively higher density with respect to the larger particles settling at the same rate. Walker then outlines how a screening step following the dynamic sedimentation or flow step is employed to separate the larger particles from the ,, .,~,............................ .

'7~

smaller higher density particles. Thus by applying this screening step to the underflow of a given vertical flow rate, a density separation a~fect can be achived, where the undersize underflow will comprise the high density material, and the oversized underflow will be of relatively lower density Walker further teaches how this idea can be applied to effect density separations over a wider size spectrum by starting with a relatively high separation medium flow rate in the ~irst separation chamber such that the "sink" or underflow mateAal which settles out is essentially comprised of only the largest of the denser particulate constituents, and thus does not require screening In Walker's prefered embodiment, the overflow, or the "float" fraction from the first separator flowsfreely into a second separator in which the upward flow of the separation mediumis adjusted to permit the settling of somewhat fimer densç particles and the largest of the lighter particles The sink product of the second separator is now subjected to a screening process wherein the screen mesh size is selected to allow the smaller, more dense, particles to pass while the larger less dense particles are retainedCorrespondingly the float from the second separator is introduced into a third which has an even lower vertical flow velocity of separation medium such that the finer fractions in the particle spectrum can be addressed This trend continues through the system in order to separate fimer and finer fractions at each step until the system is no longer econornical to operate as the mass of the remainder decreases.
However, as the discharge or float from each step is approximately equal in volume to the swm of the inflow plus the vertically flowing separation medium, each step in the series must be physically larger than the previous step in order to accommodate the increasing lnflow. This problem affects the economic utility of the system, and becomes especially critical when multi-separator systems are required to perform density separations over a wicle size spectrum Walker further teaches how the density of the separation medium can be increased by suspending high density colloidal or semi-colioidal solids in the liquid being employed By increasing the density of the separation medium towards that of the less dense mateAal, the ratio of the diameter of low and high density particles which will settle at the same rate through such a liquid increases This allows for the flow or elutriation based operations to be affected by partical density to a grester extent On the other hand both the high and low density particles will settle at a somewhat lower rate, thus the system cannot be operated at quite as high a rate.Walker specifically describes how a high density magnetizablepowder can be carefully diluted to the desired density, used as the separation medium, reclaimed by a rnagnetic separator, demagnetized by passing it through coils in order ~3~ 3 toprevent magnetic agglomeration, then thickened prior toredlluticsnfor rewse in the separation medium. Unfortunately, the processes for demagnetizing such a material, or thickening a colloidal or semi-colloidal suspension both pose majortechnical problems. Further, a magnetic separation medium could cause problems during screening operations by sticking to itself, to other particles or to the screens Walker stated thatif such a magnetizable material cannotbe found or employed ~or some other reason the same liquid weighting principles can be used, however a gravity separation system will be required to recover the weighting agent This again presents a forrnidable technical problem, this being performance of gravity separation on particles which are colloidal or semi-colloidal in sizeThus, utilizing a medium of this nature in order to perforrn a density separation simply results in having to perform another more difficult density separation In practice, the shortGommings of this design become more evident.
Many have already been pointed out regarding the shortcom~ngs of the process, asdescribed by Walker, but the apparatus itself, as descr~bed by WaLker, is also fraught with inherent inefficiencies, and logistical problems which will now be outlined. In his preferred embodiment, Walker illustrates and describes the series of separators as inverted cones. The angle at the apex of these cones is apparently in the range of 45 to 90 degrees, and the cones are equipped with a spigot having an outlet valve, and a "bustle" for the admission of the separation medium. Firstly, the effective area through which the separation medium flows at the desired speed or "design velocity" is at the apex of the cone, and is quite small. This area represents the effective area across which the separation is made, thus limiting it will limit the rate at which the separation can be made. liurther, as the area across which the separation is being performed is so small, the design velocity will accordingly increase as the effective separation area is decreased by underflow particles passing through the area.
Anotherproblem relating tomaintaining the design velocity across the effective area is that Walker includes no provision to ensure "plug flow" is achieved in the effective area. That is to say that the ac,tual upward velocity of the separation medium will vary across the effective area of the separator Thesevariations in flow velocity will inevitably lead to variations in the sedimentation characteristics of the underflow and result in errors in the density classification.
Another shortcoming of note is that Walker makes no provision to prevent the separation medium from simply flowing through the discharge spigot without effecting the required vertical flow through the upper portions of the cone as required. Therefore a considerable portion of the separation medium is not performing its designated task. The remainder of the separation medium which does flow upward quickly loses speed as the cross sectional area of the cone rapidly increases. Thls geometry results in a large fraction oP khe material to be separated remaining trapped in the upper portion of the cone since it is too small or of insufficient density to escape as under~low, but too large or dense to be floated off as overflow. Given time this situation will result in bedding of the material in the cone with channels running through the beds to allow the separation medium and the inflow to flow directly into the next cone. The obvious solution to this problem is to reduce the size of the cone such that the overflow takeoff occurs before the velocity is reduced to any appreciable extent. However, due to the variations in the design flow profile, as outlined above, early overflow of the float product would xesult in some of the particles, which should have been underflow, being erroneously contained in the overflow.
It is the object of this invention to efPect density separations by means of combining screening operations with flow related separation steps in such ways as to overcome the problems inherent in the apparatus such as that of Walker.

~U~MARY OF HE INVEN~ION

According to the present invention, there is provided a process Por separating mixtures of particulate materials according to particle density. The process comprises pass:iny the particulate materials through a series oP aounter~low separation units. Each unit uses the classifying effect of a flow of~ liquid to produce underflow and overflow ~ractions, with a similar one of said fractions being introduced to the next unit in the series. These counterflow separation units will be any suitable apparatus which performs a Plow separation and will consist for example oP
elutriation columns or hydrocyclones~ In elutriation columns, the mixed particulate materials settle against an upwardly flow column of separation medium. Those particles with sedimentation _ 4 _ ~,~C~ 3 velocities great enough to overcome the vertical flow become the underflow fraction whereas the lighter particles~ khat is those particles with a lesser sedimentation velocity~are carried up as the overflow fraction. Hydrocyclones can also be employed to produce two flow fractions. The fraction discharging throuyh the apex of khe hydrocyclone can be considered the underflow, while the vortex discharge will constitute the overflow fraction. Several parameters relating to the specific dimensions and operation of the hydrocyclone will determine the relative size/density characteristics of the two flow fractions produced by the separator. These various parameters, foremost among these being the "tangential flow rate", are analogous to the vertical flow rate of the elutriation ~

- 4a -L~

jÇj'A7~

column. Therefore, the concept of the "vertical ~low rate or velocity" shall be employed throughout this specification ;n reference to any parameters which can be manipulated in order to change the sizeldensity related separation criteria of any counterflow separation unit. The vertical or upward flow rate of the separation medium in each unitdiffers progressively through the series. The vertical flow velocity for each unit is selected such that rnaterials with progressively different sedimentation velocities are contained in the unde~flow of each successive unit. The otherof said flow ~ractions from each unit is subjected to a screening operation utilizing screen mesh sizes selected according to the verticalflowrate in order to pass particulate materials above somepreselected density, but retain the particulate materials below that density. Where elutriation columns serve as the counterflow separation units, one or more units in the series wi~l employ a primary extraction zone where liquid bearing solids in suspension is extracted and employed in whole or in part as the separation medium within the system.
In addition the invention provides ~an apparatus for separating particulate high density materials from associated particulate materials having a lower density.
The apparatus comprises a pluralitv of counterflow separation units in a series.The units are provided with means to pass a similar fraction from each unit into the next unit of the series and means to control the vertical flow rate in each successive unit. Fach unit i,, provided with an outlet for discharging the -fraction to be screened, and a means to collect the discharge from each Imit. Screens are associated with individual units in the series where re~quired. The screen mesh si~es being selected according to the vertical flow rate in order to pass particles above a selected density, but to retain particles of lesser density. The counterflow separation units can be elutriation columns, hydrocyclones, or a combination of the two. When elutriation columns are used as the counterflo~v separation units, one or more unit in the series will employ a primary extraction tank from which l;qu;d bearing solids in suspension will be extracted and employed in whole or in part as the separation medium within the system.
The process typically begins by passing the materials to be separated through a screen into a feed hopper with a means to introduce the particulate material to be separatedinto an upward flowing column of separation medium contalned in an "elutriation column". Where the higher density material is the desired ~raction the initial flow velocity will be selected such that it does not allow even the largest particles of the waste material to settle. The rate of flow of the separation medium will depend on the size density characteristics of the particles to be separated, but the initial flow rate could be as high as 5 meterg per second. The high density sink or underflow is introduced into the next classifying ~low which allows smaller high density particles along with the largest waste particles to settle out by virtue of its lower flow veloci~. The underflow from the second separator is S screened such that the material which is underflow due to its relatively large size (low density) becomes the discarded oversize, while the undersize, being of higher density is retained. The overflow from this second elutriation column is introduced into the next column which again has a lower flow rate, and similarlythe underflow is subjected to a screening process with a smaller mesh size as isrequired to effect the proper separation. Again it is the undersi7ed underflow which is retained and theoverflow continues to be introduced to successively lower velocity flows and finer screening processes until the overflow fraction is so fine or small that it is believed to contain so little of the desired material that the separation becomes non-economic. Generally this system is designes to separate particles which are less than 3/4 inches in diameter down to particles which canpass through 400 mesh screens. Dependent on the size/density spectrum of $he materials being separated the flow rates could vary between 5 meters per second for the initial flow for very large particles to 0.1 cm/sec. for the finest particles in the group. The number of steps required will depend both on the width of the size spectrum as well as the difference in the density of the two materials.
It should be noted that at the very fine end of the size spectrum the flow related separations will ideally be performed with hydrocyclones rather than theelutriation columns which cannot be operated at a high rate with very fine particles.
As with the columns of the system, the hy~rocyclones would be connected in series whereby a similar fraction from each step forms the inflow for the next cyclone in the series.
A "recirculation flow" in the process of the invention eliminates the need to artifically weight the sseparation medium by the addition of solids. Tn practice, the recirculation flow embodies a large diarneter overhead chamber or "primary extraction tank" from the top of which liq.uid containing sn~all particles is extracted, and reintroduced into the lower portion as the separation medium. This ~echniqueuses the material being separated as the weighting agent, and thereby meets the need for a weighting agent to maximize the efficiency of the separation. Further, using this recirculation method, the overflow from one column to the next does not grow substantially as it does when this technique is not used. This precludes the necessity of having to increase the physical size of each successive separation column, which is a serious problem when dealing with a wide particle spectrum ~P~ 73 requiring many stcps. This also allows standardization of the columns of the apparatus across successive steps. This standardization will also permit mass production, simplify replacement of worn sections in the ~eild, and simplify rnodifications required for the system to separate materials with different size/density distributions. Therefore, by avoiding the use of artificial weighting agents, no secondary separation process is required to recover the weighting agents.
An alternative weighting technique which can be used is to dissolve some compound in the liquid. For example, where the chemistry of swbsequent benificiation steps pennits, calcium chloride (a relatively inexpensive chemical) could be dissloved in water, br~nging the solution's density above 1.35 grams per cubic centimeter before adding suspended solids to further weight the medium.
This action wouid result in a two fold weighting effect, firstly by increasing the effective liquid density, but also by allowing larger size solids to remain in suspension. Finally, if this technique is employed, the system could be operated at a much lower ambient temperature due to freezing point depression, a property which may be of some benifit in certain climates.
A typical example of the apparatus might consist of an interconnecting series of elutriation columns with decreasing flow rates such that the overflow from each column is introduced into the next column of the series. Valves and meters used to control the recirculation pumps are one way to control the recirculation flow in the columns and thus ensure that each column has a successively lower flow thanthelast. Whenhydrocyclones replace elutriationcolurnns ;n per~ormingthe flow related separations they will also be connected in series. Control over the characteristicsof the hydrocyclonesepclration canbe achievedbyan adjustable solidsextractionmechanismwhich cancontrol thedischarge flowrate. As the size of hydrocyclones is effectively limited, ;n some cases it may become necessary to use more than one series of hydrocyclones connected ;n a parallel maner to one another in order to process large volumes of matenal.
Ideally an apparatus utilizing columns would em~loy a constant ~ed, and be designed to effect separations across a wide size spectrum, at a high rate with good density selectivi~ at a low expense. Some of the embodiments employed to achieve these goals will be summarized below by way of example. It will however be understood that it is not intended to limit the invention to such embodiments.
On the contrary, it is intended to cover all alternatives, modifications and equivalents as may be included within the spirit and scope of the invention as defined by the appended claims.

The specific geometry of the columns allows for a narrowing of the cross sectional diameter just below the "primary separation section", where the separation medium is to effect the separation. This configuration aids in ~maxirniæing the plug flow characteristics through the primary separation section, thereby giving the system better density selectivity. The lower portion of the columns consists of the sedimentation chambers. The recirculation flow enters the column at the primary sedimentation chamber, then flows up into the primary separation section. The secondary sedimentation chamber being located below the primary sedimentation chamber does not have a large vertical flow rate within, thus the underflow can settle to the bottom of the secorlda~y sed~nentation chamber where the solids are removed. A progressive cavitation pump is ideal ~or this purpose as it can remove the underflow with a minimum of associated liqu;d loss,while at ~e same time operating at a known volume per time, thereby allowing theliquidloss to be compensated for. Immediately above the primary separation 1~ section is the "secondary separation section". It is in this section that the material being separated is introduced into the vertical flow of separation medium. The diameter of the secondary separation section is selected to be sufficiently larger than that of the primary separation section so that the vertical flow in the secondary portion is slightly lower than the design velocity for that column. This slight reduction in velocity is designed to prevent any underflow particles ~rom erroneously reporting to the overflow, while at the same time minimizing the residence time for the overflow material within the line which takes the overflow from a given column and introduces it into the next column or separator in the series.
Above this point is located the "primary extraction tank" which is comprised of a rapid increase into a large diameter chamber. The primary extraction tank is designedtoreduce the vertical velocity enough so that most solids will remain near the bottom, and eventually be introduced into the next column. At the top of this tank, liquid with solids in suspension is extracted andpumpedbackintothe system via the primary sedimentation chambers. Thus, this liquid with solids in suspension serves as the separation medium. The rel,ative size of the solids remaining in suspension will depend on the vertical flow rate of the liquid in the primary extraction tank, which will in turn depend upon the internal diameter ofthe tank, and the rate of extraction from the tank.
In order to prevent any of the separation medium (liquid containing suspended solids) from contaminating the under~low from which the final product will be screened, some method of "scrubbing" can be employed. One method is to add some "clean" liquid to the secondary sedimentation chamber. A

restriction in diameter at the top of this chamber ensures a vertical flow of sufficient magnitude to prevent any of the undesired solids from settling out. ( In this instance "clean" liquid refers to liquid which has an acceptally small amount of solids contained in the liquid.) One possible method of obtaining this "clean"
liquid is through the use of "secondary extraction tanks". In another embodimenthydrocyclones could be ernployed to extract liquid with a minimum of suspended solids. By locating each hydrocyclone such that the recirculation pump feeds thetangential inlet, and by usingvalves to control the two discharge apertures, themajority of the volume will discharge through the apex and be used as separation medium, while the volume extracted through thevortex fimder will be used as "clean" scrubbing liquid. Another possible method of scrubbing entails introducing the separation medium via an upwardly directed nozzle such that the medium will plume out to the full diameter of the column just inside the lower section of the primary separation section. Thus, only those weighting particles involved in eddy currents near the edges of the column will settle out as all others will be carriedup by the vertical ilow. These solids which do settle outcan then be separated from the true undel~low via a double screening technique; for example, in a case where the higher density material is the desired product, thecontaminated underflow is subjected to the appropriate screening operation which will result in an oversize discard product, and a contaminated undersi~e concentrate.
This concentrate can then be rescreened using a finer mesh size selected to retain the density concentrate, and pass the contaminant which had been suspended in the recirculation flow. It should be noted that in this second method some of the material may be lost, and the velocity profile across the critical portion of the column may be adversely affected. ID various circumstances, economics rnay favour one method of scrubbing over the other, or perhaps parts of both methods might beutilized within the same system.
The system used to introduce the initial slurry feed to the first separation column will consist of a hopper with a constant outflow rate as e~fected by an auger or progressive cavitation pump. The inflow solids may then be diluted withwater from the "main holding tank" to form a slurry with a lcnown solids to liquids ratio. The "~nain holding tank" mentioned above embodies a large tank into whichall waste liquid such as the final overflow is returned. In most cases, some form of thickenerwill be employed to remove most of the suspended solids in the water entering the main holding tank.
Often the initial feed rate for the f~rst column may prove too high for the columns down the series to accommodate the high inflow. That is, the velocity in the secondcuy active portion will become too high in compa~ison with the primaryactive portion. To remedy ~his situation it will be necessary to reduce the over~low volume at certain spots throughout the system. Again this can be easily accomplished through the use of a secondary extraction tank, where the required volume reduction can be achieved by diverting a portion of the cyclone overflow back to a main holding tank. In the example embodiment sumarized above all recoverable liquids are sent back to the main holding tank, thus the absolute minimum liquid requirements for the system will be that amount of liquid which cannot be economically recovered from the solids which constitute the concentrate, the overflow fines, and the coarser discardmaterials.

BRIEF DESCRIPIION ~)F THE DRAWINC;S

In drawings which illustrate example embodiments of the invention:
FIGIJRE 1 is a graph illustrating typical theoretical design considerations used to set flow rat~s, to decide how many columns should be employed, and to decide where screening procedures should be employed in the apparatus and process of the present invention:
FI(:;URE 2 is a cross sectional schematic view of a typical elutriation column of an example apparatus according to the present invention FIGURE 3 is a schematic representation of an example five-stage elutriation process and apparatus according to the present invention; and FIGURES 4 and S are schematic representations of possible apparatus configurations employing hydrocyclones as the counterflow separation units ~n ~eprocesæ and apparatus according to the present invention While the invention will be described in conjunction with example embodiments, it will be understood that it is not intended to limit the invention to such embodimel ts. On the contrary, it is intended to cover all alternatives, rnodifications and equivalents as may be included within the spirit and scope ofthe invention as defined by the appended claims.

ETAILED DES~R~EION OF THE INVENTION
The elutriation process described here refers to a type of dynamic counter-flow sedimentation in which the solid particulate material settles a~ainst a vertical flow of liquid separation medium. Pa~ticulate materials of di~fering densities t7~3 andrestricted sizeranges, generally less than 1/2 inch inparticle dianleter, areideally suited for density classification by elutriation methods. The purpose of the design is to effect such density classifications on the described materials via an economically desirable application of the elutriation method.
The theoretical principles on which this system operates will initially be discussed in terms of separating two m~xed particulate materials. It will further be assumed for the purpose of the following explanation of the principles affecting the operation of this system that the desired particles have a density of 5 grams per cubic centimeter~ and the waste particles have a density of only 3 grams per cubic centimeter. However, it shall be lmderstood that the speci~ic values included m this explanation are included only ~or the purpose of explaining the operating principles which affect this process, and that the process is not limited to this specific method of application or to materials as described above. Alternative methods of employing the principles of the process will be described later in this speci~lcation.
Sir Isaac Newton described the settling velocity of a spherical particle through a liquid with the equation:
v2= 4 D g (ds-dl) 3 Q dl where "v" is the settling velocity, "D" is the particle diamet~r "g" is the acceleration due to gravity, "ds" is the density of the solid particle, "dl" is the density of the liquid, and "Q" represents an experimentally determined constant related to the characteristics of the separation medium. (The value of "Q" ~or pure water is in the order of 0.40) As applied to the elutriation process, the value of "v" is used to select the vertical flow rate of the separation metlium such that particles of a given density "ds", and a desired diameter "D" will be at terminal velocity within the rising liquid, and therefore, effectively suspended. Particles of greater diameter or density will drop to the bottom as underflow, and particles of a lesser size or density will be washed away by the separation medium, as over~low. Thus the elutriation method can be used to perform particulate separations with respect to size or density as long as the other variable (ie densi~ or slze) is a constant within that group of particles.
In practice, Newton's equation holds roughly true only in predicting the terminal velocity of particles generally greater than 14 mesh in size. Settling characteristics of particles finer than about 65 mesh are best described by Stoke's Law:

, 11 ~, , .
, D g (ds-dl ) Y=
l8 Jl where"ll"represents theviscosity ofthe separation mediumin unitsconsistent ~ with the other variables. All other variables in Stoke's Law represent the corresponding physicalcharacteristics as employed in Newton's description of ~ the phenomenon. Settling characteristics of particles which are too small to fall in the Newtonian range, but too large to be in the Stokesian range, can be determined experimentally. These particles are said to be in the transition settling range. With respect to separations of very fine particles ~generally less than 150 mesh) the sedimentation rates are so slow that separating large volumes becvmes problematic as large cross-sectional areas are required. Therefore hydrocyclonesare used to perform flow separations in this size range. As with Stokes Law, theparticle diameter to density relationship relavent to hydrocyclone separations is also an inverse square function, and thus hydrocyclones can serve the same function as elutriation columns in per~orming density separations. Thus, for particles with different size/density characteristics, different methods may be required to calculate the terrninal velocity. Nevertheless, it is true across the entire size spectrum that as the density of the separation medium approaches that of the lower denstiy material, the relative diameter of low densi~ particles at terminal velocity increases with relationto the diameter of the higher density particles also at terminal velocity in a given flow. This effect is due to the "(ds-dl)" term of the equations, which approaches zero sooner when applied to the lower density solids than when "ds" has the higher value associated with the more dense particles. Applying this pAnc;ple, under Stokesian conditions, to particles with high versus low density, where all particles are subjected to a vertical flow of separation medium ( withdi~fering density values ), the following table can be generated:

Separation High: Low Medium Densit~ Ratio of Density P~rticle D~ameters ~:

2 ~ :

3~ 2.75 , 3~ 7~3 As the values indicate, by increasing the density of the separation medium, the density difference between the two types of materials has a more pronounced affect on particle diameter. Since it is the function of the flow related steps in the system to effect separations where the underflow is composed of relatively smaller high density particles mixed with relatively larger low density particles, the differences in size are best accentuated by using a higher density separation medium. This increase in the ratio of high versus low density particle diametersfacilitates the screening steps which follow each elutriation step in order to effect thedensity classifilcation.
The principles outlined above can be employed to make density classifications by combining flow and screening steps as described. However there will inevitably be a certain amount of error associated with the classifications as the specific sedimentation characteristics for some particles will deviate from predictions due to things such as variation in the velocity profile, or the variations in shape of the specific particles involved. In general these variations lead to a normally distributed error curve being superimposed on the desired separation. In terms of FIGURE 1 this would be represented as a certain amount of uncertainty or fuzziness associated with the curves. This becomes a serious problem when a highdegree of system selectivity is required to separate two materials with a small density difference. The selectivity of the apparatus will be inversely proportional to the variance of this errordistribution,howeveradditionaltreatmenteffcctscan yield vast improvements in total system efficiency Additional treatment effects, as alluded to above, may vastly improve the total system sensitivity to particle density as the additional treatments reduce what is known in statistical terms as the Type II error. One such additional treatment which might be employed would be to take advantage of the shorter time it takes ~or high density particles to approach terminal velocity. This principle, on which jig separations rely, could also cause a density related settling ratio increase thereby increasing the density selectivity of this system The actual modification to thepresent system which would be required to make use of this additional treatment effect could be as simple as using a pulsating pump on the recirculation flow line.
Of course this is just one possible method of application of this idea, and many other methods of adding this effect are possible. Also mentioned earlier was the effect variations in particle shape could have on increasing the e~ror distribution. ~ fact, in certain cases the differences in friability between the desired and the wasteproduct can lead to a consistant difference in particle shapes. In some cases, the 3~

screens employed can be selected to have opening shapes which will favour passing one product more than the other. Again this fact can be used to increase system selectivity.
~lternative methods of application of the principles of the process include such things as startin~ with the low flow rates, and screening the overflow fromeach stage while the underflow is introduced to the next higher flow. Again at the next stage the overflow is screened such that the overflow/oversize is the lowerdensity material, and the undersize is the high density product. Another vari2tion of the application of the pr~nciples set forth is to e~fect the screening procedures first, then introduce the various size fractions to the classifying flows of separation medium. It should also be pointed out that the system can be employed in cases where the lower density material is the one which is desired, and the high density material is the waste. In these cases simple modifications of the apparatus and system, obvious to anyone skilled in the art can be made to save the low density product and discard the high density product. It should also be evident to anyone skilled in the art that this system is not limited to separations involving materials of only two densities, in factby using double ormultiple screening decks for each underflow product, the system can separate as many density fractions as desired. ~or example, one screen yields twodensity fractions; two screens ~the larger mesh on top of the smaller) will result in the lowest density as the oversize,the middle density fraction onthe second screenl andthehighest density material passing both screens. Finally, in instances where the density separations must be made only at the very fine end of the size spectrum, elutriation columns could be abandoned completely in favor of a series of hydrocyclones. When one series of hydrocyclones could not process the required volume, an additional series of hydrocyclones could be employed to operate in a configuration parallel to the first series.
These variations on the inventive concept, as outlined above, can be comb;ned with one another resulting in compounded variations of applications of the inventive idea. However, all such variations are considered within ~e scope of th;s patent.
Turning to FIGURE 1, ideally line "X" represents the desired density separation line. In practice, however, separations as along line "X" must be 3~ approximated as indicated by the shaded areas both above l;nes A through E, where these areas represent the s;ze/density eharacteristics of theunde~lowproducts from successive elutriation columns, and to the right of the vertical dashed lines, "F" through "I" where these lines represent the screening steps appropriate to each ~, t3~~7~3 underflow fraction. In this example application of the system~ it is the undersize/underflow which is the high density material which is being benefici;3ted, thus the particles left of the screen lines would be discarded as waste, The curved lines, A through E, are approximations to the solutions of the settling equations S given earlier where the vertical flow rate "v" of a separation medium with a density of 2 grams per cubic centimeter is set and the sizeldensitS/ line represents characteristics of particles with terrninal velocity "v". All the curves asymptotically approach the density of the separation medium itself, and decreasing the flow rate has the effect of broadening the cuIve to the right, As the flovl/ rate is decreased to zero, the parabola is flattened into a straight line along the density of the separation medium. It should be noted that although increasing the density of the separation medium does increase the settling ratio, the particles will not settle as fast, therefore the system cannot be operated at as high a rate. Shaded areas below the density value of 3 would indicate that particles with those size/densi~ characteristics would erroneously be classified with the high density product, Similarly, non-shaded areas above the value of 5 would be indicative of the desired high density material being discarded, In any given application, the related operating parameters of rate, contamination and efficiency must be weighed against eachother in order to use the system to its best advantage under the circumstances at hand, It will also be noted that the underflow of the first elutriation step is not indicated as being subjected to a screening step, In many cases this may be the case as the feed into the system may already have been reduced to some limited size, In these cases the screening step has effectively been performed prior to the first elutriation step, In other cases, where only a Yery small portion of the particle population will report to this underflow, and the majority of these particles are OI the higher density material, screening may be deemed unnecessary, FIGURE 2 is a representation of a typical elutriation column as might be employed in conjunction with other sirnilar columns~ and additional components as required, to comprise the total apparatus, The column in PIGURE 2 is presented by way of example, and should not be considered to embody the only possible method of application of the inventive idea, Changes in such features as the scrubbing technique employed may necessitate changes in the structure of the columns employed. Such changes will be evident to one skilled in the art, and thus will not be described in this specification. Further, it is not necessary that each colurnn in an apparatus consist of all the parts illustrated in FIGURE 2, and described below, Slight variations in the application of the inventive idea which may rendersome components of the column unnecessary in certain instances will '~' 1S

~",, be outlined later in this specification.
Each column can be both physically and functionally divided into ten important parts, namely: the inflow line ~ l), the overflow line ( 2), the secondary separation section also known as the secondary active portion ( 3), the primary extraction tank ( 4), the recirculation line ( 5)~1 the primary active or ~separation section ( 6), the primary sedimentation chamber ( 7), the secondary sedimentation chamber ( 8), the clean inflow line ( 9), the solids extraction mechanism (l0), and the screen (l l). ~ach part will be described in terms of its contemplated function, and the speci-fic embodiments used to achieve these goals.
The inflow line ( l) ca~ies the materials to be separated into the system, from the ~eed hopper (15) in the case of the first column, or from the previous column inall subsequent cases. The solids will generally be suspended in some fluid whichacts as the transport medium. In most cases water will serve this purpose. The overflow line ~ 2) forms the inflow line ~ l) forthe subsequent column in all cases, although in the case of the last column, the overflow line ( 2) will not turn into another inflow line (l) as the series is ending. Communicating with both the inflow line ( l) and the ove~flow line (2) is the secondary separation section ( 3).
This portion comprises the area in which the inflow material is introduced to the vertical flow of the separation medium. The cross-sectional area of this section is selected to be such that the vertical velocity contained within is reduced slightly below that within the prima~y separation section ( 6~ even when the inflow line ( l) is adding material at the normal rate.
Immediately above the secondary separation section ( 3) and communicating therewith is the primary extraction tank (4). It is the purpose ofthis segment of the apparatus to extract liquid containing small suspended solids which will be used as separation medium from the top of the column. The diameterof this section is selected with relation to the1desiredcharacteristicsof the separation medium. If the diameter is relatively small, the vertical velocity in this section will be relatively large, thus the separation medium will consist of liquid with relatively large particles suspended in it. On the other hand, if the diameter is increased, the vertical velocity will be decreased, and therefore only relatively smaller particles will be suspended intheseparationmedium. Fluids extracted from the primary extraction tank (4) are pumped back into the primary sedimentation chamber ( 7) via the recirculation line ~ 5). As can be seen in FIGURE 2, the primary sedimentation chamber (7) is in direct communication with the primary separation section ( 6) immediately above it, and the secondary sedimentation chamber ( 8) directly below it. The primary separation section ( 6) is . 16 the section in which the critical flow sepa~ation is ef~ected. 'l'hus it is critical that the exact design velocity is maintained in this area. The secondary sedimentation chamber ( 8) is directly connected to the solids extraction mechanism (10), the primary sedimentation chamber ( 7), and the clean inflow line ( 9) which performs the scrubbingof the underflowpnor to itsremovalfromthe system.
In a prefered embodiment, the secondary sedimentation chamber ( ~) is designed such that the cross-sectional diameter OI the column entering this section is relatively sm~ll.
FIGURE 2 illustrates a narrowing towards the top of the secondary sedimentation chamber (8) which will cause a relatively flat velocity profile at the narrowest point. This reduction m diameter permits the vertical flow rate, necessary to the scrubbing action to be achieved with a minimum of "clean" liquid. Thus the separation medium is not diluted tc an excessivedegree as the recirculation flowwill constitute its majorcomponent. A simple alternative to this configuration would be to make the entire length of the secondary sedimentation chamber ( 8) of relatively small diameter. Although the velocity pro~lle may be adversely affected, the addition of "clean" liquid can stiil be minirnized. Below the solids extraction mechanism (10) is the screen (11) fvr such columns where a post flow screening step is required (generally for all but the first in the series, as the material is usually screened before being introduced to the flow steps). It should be noted that liquid entering the primary sedimentation chamber ( 7) will, by design of the column, be forced to rise and act as the separation medium. This is because scrubbing liquid is being added to the secondary sedimentation chamber ( 8) at ahigher rate than the solids extraction mechanism (1~) removes the underflow volume. This rising separation medium encounters a smooth but rapid nanowing of the column diameter just as it enters the primary separation section(6). I'his preferred embodiment provides for a flattening ~f the velocity profile in order to effect a more precise classification in the prim~y separation section ( 6).
E7IGI~ 3 schematically represents an example application of the process and apparatus as it might be applied to beneficiate a matçrial with characteristics identical to those described in the previous example. It should be noted that the diamond symbol used in FIGUR~ 3 represents a controlable pump such as a variable speed positive displacment pump. A pump, meter, and valve in series might alternatively be employed to perform the required function.
The materials to be separated are f~rst introduced into the system through a feed hopper (15). 1~ general it would be considered expedient to pre-screen the material entering the hopper (15,~ to some maximum size in order that oversized foreign materials are removed. This can be accomplished through the use of an appropriate size screen (11) covering the entrance to the feed hopper (153.
Failure to incorporate this embodiment may result in plugging of the system by oversized foreign objects. The feed hopper (15) itself should be of a size related to the size of the sy~tem such that the desired feed rate can be maintained at all times regardless of the technique being used to fill the feed hopper (15). That is to say, the feed hopper (15) must never be empty while the system is in operation, and should therefore be of adequate si~e so as to accommodate fluctwations in the rate at which material is added to it. Furthermore, the geometry o~ the feed hopper (15)shouldprevent bridging of the material which could result ~nacessationof material inflow. In general this means that the smallest dimension of the feed aperture should be at least five times the diameter of the largest particle size, or in this specific case, at least five times the aperture diameter of the screen covering the feed hopper (15). Also the slope angle of the feed hopper ~15) should be great enough to ensure adequate wall slippage.
Material to be introduced into the separation system should be extracted from the feed hopper ~15~ at a constant volume per unit time. A progressive cavitation pump such as a Moyno (Registered Trademark of Robbins and Meyers Inc.) may be suitable for this purpose. In order to reduce wear in this feed pump (14), and ensure consistent inflow characteristics, the material in the feed hopper (15) should be wetted to the saturation point. By adding excess liquid to the feed hopper (15), such that the excess sits on top, the material at the point of extraction will be saturated. Liquid ~rom the main holding tank (12~ would be suitable for this purpose, and through this pre-wetting, the solids to liquids ratio of the 2~ material entering the system is known and constant. Krlowing that the inflow material is saturated removes any doubt regarding charàcteristics of the inflow material, and facilitates accurate dilution. The purnpin~ action of the ~eed pump (14) also should break down agglomerated particles into thcir smaller constituents.
This is important as agglomera~es will erronously settle out early in the separation system due to their lar~e size, when in fact the p~rticl}late ,constituents should settle out later in the system.
Material leaving the feedpump (14) must thenbedilutedin orderthat the particles can settle against the separation medium relatively independently of 3S one another once in the primary and secondary separation sections (6) and (3) respectively. Theflowof separation medium will alsocontribute to therequired dilution e~fect, thus the entire dilution need not be achieved in the inflow line (1) for the first column in the series. In general, material in the inflow line ( 1 ) need not be more than rou~hly S0 percent solids by volume. However this may vary somewhat dependent on the design velocity in thc initial column.
The particulate solicls entering any given elutriation column as inflow can be conceptually divided into those particles which will eventually sink out as underflow, and those which will leave that column as overflow. Hereina~ter, for the purpose of this example, particles which will eventually constitute the underflow for a given column will be referred to as underflow while in any portion of saidcolumn The term ovel~ow will be used with similar intent hereina-fter. In order to facilitate adequate comprehension of the embodiments of this process and apparatus, such embodiments shall be discussed with relation to their contemplated function, as the path of the underflow then the overflow is outlined throughout the system The particles which will eventually comprise the underflow fraction from the first column enter the first column via the inflow line ~1) as described above.
The in~low line (1) flows down into the lower portion of the secondary active portion (3), forming an acute angle between the two. As mentioned previously, the volume of the inflow is selected with relation to the cross-sectional area of the secondary separation section (3), and the desired size/density characteristics of the underflow for that specific column, such that the net vertical velocity in the secondary separatioll section (3) is slightly less than that in the primary separation section ( 6). The rationale inherent in this slight reduction in vertical velocity is to ensure that underflow particles caught up in the turbulence due to the infl!ow mixing with the separation medium in the secondary separation section ( 3) will eventually settle back down to the primary separation section ( 6).
The vertical flow velocity in the secondary separation section ( 3) should generally be no more than 9S percent of the design velocity for that column, where the design velocity for the column refers to the vertical velocity in the primary separation section ( 6). In cases where the over~ow line ( 2) is in close proximity to the inflow line (1), the velocity should be ~urther reduced as the turbulenceeffects may still be largely undiminished at the overflow line ( 2). S~Sviously the vertical separation between the inflow line (1) and the o~erflow line ( 2) can be reduced when dealing with the lowerflow rates. Aithough such reductions decrease the size of the apparatusl they serve no other functional purpose, and therefore are not generally employed as this inclusion would preclude 3~ standardization of the components of the system. Excessive reduction of the vertical velocity in the secondary separation section ( 33 with respect to the design velocity will result in a net increase ;n the percent solids in the secondary separation section (3). This effect is due to the larger fraction OI solids with .

~ ~Y~ 7"~

size/dens;ty characteristics such that they cannot settle out as unclerflow, but, due to the lower flow rate in the secondary separation section (3) some particles will be quite slow to be exhausted as overflow Thus, the net effect of a substantial velocity reduction in the secondary separat;on section (3) is an increase in liquid density within this area. As mentioned earlier, this çffect will result in better density selectivity in the secondary separation section (3) due to the decrease in the (ds-dl) term in the applicable sedimentation equation. On the other hand, if the density gets too high in the secondary separation section (3), plugging is more probable, and theassociated increase in viscosity can both complicate the settling characteristics of particles (due to accentuations of deviations in settling rate as caused by variations in particle surface area versus mass based on particle shape, etc.), and increased power consumption related to the recirculation pumps ~16) required in the system. For these reasons, the velocity in the secondary separation section ( 3) should generally not be less than 70 per~ent of the design velocity for that specific column.
Immediately below the secondary separation section (3) and cornmunicating directly therewith is the primary separation section ( 6). As indicated by the name, the function of this section is to effect the critical flow related separation. Thus it is of paramount importance that the precise design velocity is maintained throughout the entire cross-section of this segment of the apparatus. To this end, the ideal geometry embodied to accomplish this consists of a cylindrical shape (to minimize wall friction in relation to cross-sectional area ) and a smooth but rapid narrowing of the cross-sectional area at the lowerrnost end of the primary separation section ( 63. This second feature serves to flatten the velocity profile through the section immediately above it, ancl therefore aids in ef~ecting a precise separation at thiæ point. With respect to the vertical length of the primary separation section (6), the selected length must be adequate to ensure that the net downward velocity of the inflow is effectively nullified~ with enough additionallength in which to perform the required flow separation. Should this add;tional length substantially exceed the required amount, the unde~flo~w particles will have an inordinately large residency time within this section. This will manifest itself in a greaterproportion of the solids occupying the separation area, and thus by effectively reducing said area, the design velocity will be increased. The net effect of excessive length of this secti~n is to cause the underflow to precipitate in waves. The reason for this phenom non is that the area reduction mentioned abovetends to result in an increase in the vertical velocity which clears most of the solids out of the area; the veloci~ then decreases in response, and another wave begins to settle outuntil the area is again overf;lled with solids. This willresult in a cycle which repeats itself with a period proportional to the excess length of this component. Associated with this oscillation about the desired dynamic equlibrium within the system will be a slight decrease in the separation precision 5of the system. The two methods to reduce or eliminate this oscillatory phenomenon are to reduce the feed rate to the extent that the effective area reduction caused by sedimenting particles is not appreciable enough to cause theoscillation, or simply to shorten the length of the primary separation section ( 6) to the extent that the period of the oscillation becomes so short that it effectively lO` disappears. For economic reasons, the latter solution is favored over a reduction in the feed rate to the system.
Immediately below the primary separation section ( 6), and directly communicating therewith is the primary sedirnentation chamber ( 7). It is via the recirculation line (5), which connects with the primary sedimentation chamber ~7) 15that the separation medium enters the column. Due to the design of this system this recirculation flow results in a positive vertical velocity within the primary separation section ( 6). The primary sedimentation chamber ( 7) embodies a somewhat larger cross-sectional area than the primary separation section ( 6) and thus has a lower vertical velocity within. Therefore, once any underflow particles 20have reached this portion of the apparatus, they will settle relatively unhindered to the solids extraction mechanism (lO). Below the primary sedimentation chamber (7) and communicating therewith is the substantially narrower secondary sedimentation chamber ( 8). In order to prevent the small solids suspended within the separation medium from settling to the solids extraction mechanism (lO), a 25slight vertical flow of "clean" liquid (where, as before, "clean" is defined as liquid with an acceptably low percentage of solids suspended in it) is maintained by the "clean" scrubbing liquid entering this segment via the clean inflow line ( 9). The vertical velocity of the scrubbing liquid leaYing this ar~a of the apparatus must 30exceed that in the primary extraction tank (4) from which the recirculation flow is derived, however it can be substantially lower than the ,design velocity for that specific column. It should be remembered that while the small upper diameter of the secondary sedimentation chamber ( 8) does not require a large volume per unit time to generate this flow, the liquid which is being extracted by the solids 35extraction mechanism (lO) must be replaced in this zone. Given that the vertical velocity exiting the secondary sedimentation chamber ( 8) falls within the aboveguidelines, the underflow willeasily settle to the solids extraction mechanism (lO), but the suspended particles within the separation medium will not, hence the .. . .

6~ 3 scrubbing will be effective.
When dealing with very fine solids, ( ie. less than 100 mesh) the clean in~low may be proportionally increased such that it can generate the design velocity independent of the recirculation flow, thereby precluding the need for scrubbingIn such cases, the clean inflow line ( 9) must be transposed to enter the columnwhere the recirculation line ( S) normally enters, as introduction at the usual point wouldresultin thevertical velocity withinthe secondary sedimentationchamber (8) exceeding the design velocity by vir~ue of its small diameter. Obviously failure to embody this change, which is illustrated in the final column of PIGURE
3, could result in no solids being able to underflow from the column. It should be noted that in this situation the secondary sedimentation chamber ( 8) becomes a vestiginalembodimentwhich may beomitted.
Finally, the underflow for a given column reaches the solids extraction mechanism (10) where the particles are removed ~rom the column. Invariably a IS small amount of liquid will be removed with these solids; however it is best to minimize these liquid losses as they must be made up for by the clean inflow. The solids extraction mechanism (10) could be a positive displacement pump capable of removing the solids without plugging, or some form of discharge control valve could be employed thereby decreasing the risk of particle breakdown duringextraction.
The solids extraction mechanism (10) clischarges the fraction to be screened to the screen (11) for that column. As mentioned earlier, and as illustrated in FIGURE 3, in certain cases, the screenin~ operation will not be required as the ent~re wnderflow fraction will consist of the high density material. The aperture size for the screen employed will be selected relative to the design velocity for the specific column based on the gwidelines described earlier. In the case of this example where the higher density material is being sought, the undersized under-flow will constitute the concentrate, and the oversi~e will be the low density waste product. As the screening step embodies no new or exceptional technology it will not be described further in this disclosur~.
Havingdescribed thepath throughthe system takenby theunderflow of any given column, the overflow shall now be discussed in a similar ~shion.
The majority of the overflow entering the secondary separation section (3) via the inflow line ( 1) is carried up by the separation medium, and leaves the column via the overflow line ( 2). It will be remembered however that the verticalveloci~ within the secondaTy separation section ( 3) isslightly less than the design veloci~7 thus there will be a small number of particles with size/density characteristics such that they cannot be carried up as in~nediate overflow, nor can they settle out as underflow. These retained particles will accumulate in the secondary separation section ( 3) of the apparatus slowly increasing tbe effective liquid density therein until such time as the density increase is sufficient to bring about the overflow of the very particles caus;ng it. The prime function of this section ~that is to prevent the erroneous overflow of any underllow particles) will still be accomplished in spite of the density increase which will occur for a time after initial system startup. This is due to the fact that the (ds-dl) term in the sedimentation equations will have a greater effect on the lower density particles resulting in preferential overflow of the low density particles. The net result is that any underflow particles which are erronously overflowed will most likely be composed of the lower density waste material, therefore, the selectivity of the system will not be adversly affected by such mistakes.
Some overflow material will be carried up into the primary extraction tank ( 4) from which the recirculation flow originates. This portion of the overflow is temporarily held in the column to act as the weighting agent in the separation medium. Theoretically, any given particle could remain in this loop within the colurrm. For practical purposes, however, the system will equilibrate soon afterinitial startup such thatthe inflow of particles of this class is equalledby the overflow rate of such particles, hence this loop cannot be considered infinite.
FIGURE 3 further illustrates secondary extraction tanks (17) which are incorporated in orderthat the inflowvolume toeach columndoes notincrease, but can in fact be made to decrease. The importance, and need for this has been earlier outlined herein. The design of these tan'ks is such that only an accepta'bly small amount of solids are removed with the liquid extracted. Thus a large diameter, and a low extraction rate are recommended. The -fluid extracted from these secondary extraction tanks (17) c.m be used as "clean" scrubbing liquid, returned to the main holding tanlc (possibly by way of the thickener ('l3)) or dealt with in any other desirable manner. In general, the fine solids recovered by thethicl~ener (t3) constitute waste material, and would be discarded with the wastefractions from the screening steps.
The'final discharge of overflow ~mes which are not introduced into another separation unit may be returned to the main holding tank (12) by way of a thickener (13) as is illustrated in PIGURE 3. This will al10w for the recovery of the liquids used where this is desirable. Each column need not be equipped with its own recirculation circuit. In some cases the separation can be effected entirely with clean liquid (especially where the flow rate is small) as is illustrated in the final ' . 23 5 ~ q column of FIGUR~ 3. Another possibility wowld be to extract additional volumes from the primary extraction tank (4) of one column in order to supply separation medium -for the adjacent columns. Here again the primary extraction tank ( 4~ must be designed such that the liquid extracted has the desired characteristics to 5 be used as separation medium for all those columns in which it is employed.
Similarly, the secondary extraction tanks (17) need not be associated with each and every column ~n the system. In the case of the secondary extraction tanks (17) the parameter dictating the minimum frequency of their occurrence is the acceptability of the ratio of the vertical flow rates within the prirnary and secondary separation sections ( 6) and (3) respectively for all the columns involved.
FIGURES a, and S illustrate two possible methods for the serial connection of the counterflow separation units, which are represented as hydrocyclones (18) in these drawings. In both cases it should be noted that the feed may come directly from a column at the end of a series of columns, or in the case of very fine materials which can be separaeed according to this process without columns, the hydrocyclone series can be fed directly by some form of screen/hopper/feed pump assembly. In the latter case the screens (11) associatedwith hydrocyclones (18A) or (181H) would not be required as the aforementioned feed system would have a screen incorporated. In order to control the parcuneters of the separations performed by the hydrocyclones (18) a solids extraction mechanism(lO)whichcould vary the extractionrate could be employed. A
variable speed positive displacementpump, or sorne formof adjustable valve could be used to serve this purpose. Further control over the separation can be achieved by modifying the diameter of the "tangential inlet" (20). The diameter of the "vortex discharge aperture" (19) is not critical as in all cases the solids extraction mechanism will dictate the flow rate through the vortex'dischar~e apertwre (19).FIGUR~ 4 embodies a serial connection wherein the overflow is the flow ~raction which is introduced into the next separation unit, and the screens addressmg theunderflow grow progressively finer as the series pro~resses. The apparatus of l~IGURE S on the other hand is connected such that the underflow -fraction is introduced via the tangential inlet (20~ to the next unit in the series. In this case the screens used to address the overflow fractions produced will employ progressively coarser mesh as the series continues.
Thus it is apparent that there has been provided in accordance with the invention an apparatus for separating rnixtures of particulate materials that fully satisfies the objects, aims and advantages set forth above. While the invention has been described in conjunction with speci~ic embodiments thereof, it is evident that many alternatives, modifications and var~ations will be apparent to those skilled in the art in light of the foregoing description. Accordingly, it is intended to embrace all such alternatives, modifications and variations as fall within the spirit and broad scope of the invention.

~5 ~5

Claims (24)

1. A process for separating mixtures of particulate materials according to particle density comprising passing the particulate materials through a series of elutriation column units in association with screening operations, each unit producing underflow and overflow fractions, with the overflow fraction from each unit being introduced to the next unit in the series, the vertical upward flow rate of the separation medium in each unit having a lower average vertical velocity than that in the previous unit, said vertical velocity for each unit being selected such that materials of decreasing size are contained in the underflow of each successive unit, the underflow from each unit besides the first unit being subjected to a screening operation utilizing screen mesh sizes growing progressively smaller through the series and being selected according to the vertical flow rate in order to pass particles with a density above a selected cutoff density, but to retain those particles below that density, one or more units in the series employing a primary extraction zone where liquid bearing solids in suspension are extracted and employed in whole or in part as the separation medium within the system.

2. A process for separating mixtures of particulate materials according to particle density comprising passing the particulate materials through a series of elutriation column units in association with screening operations, each unit producing underflow and overflow fractions, with the underflow fraction from each unit being introduced to the next unit in the series, the vertical upward flow rate of the separation medium in each unit having a higher average vertical velocity than that in the previous unit, said vertical velocity for each unit being selected such that materials of increasing size are contained in the overflow of each successive unit, the overflow from each unit besides the first unit being subjected to a screening operation utilizing screen mesh sizes growing progressively larger through the series and being selected according to the vertical flow rate in order to pass particles with a density above a selected cutoff density, but to hold those particles below that density, one or more units in the series employing a primary extraction zone where liquid bearing solids in suspension are extracted and employed in whole or in part as the separation medium within the system.

3. A process for separating mixtures of particulate materials according to particle density comprising passing the particulate materials through a series of hydrocyclone units, each unit producing underflow and overflow fractions, with the overflow fraction from each unit being introduced to the next unit in the series, the tangential flow rate in each unit having a lower average tangential flow rate than that in the previous unit, said vertical flow rate for each unit being selected such that materials of decreasing size are contained in the underflow of each successive unit, the underflow from each unit besides the first unit being subjected to a screening operation utilizing screen mesh sizes growing progressively smaller through the series and being selected according to the tangential flow rate in order to pass particles with a density above a selected cutoff density, but to hold those particles below that density.

4. A process for separating mixtures of particulate materials according to particle density comprising passing the particulate materials through a series of hydrocyclone units, each unit producing underflow and overflow fractions, with the underflow fraction from each unit being introduced to the next unit in the series, the tangential flow rate in each unit having a higher average tangential flow rate than that in the previous unit, said vertical flow rate for each unit being selected such that materials of increasing size are contained in the underflow of each successive unit, the overflow from each unit besides the first unit being subjected to a screening operation utilizing screen mesh sizes growing progressively larger through the series and being selected according to the tangential flow rate in order to pass particles with a density above a selected cutoff density, but to hold those particles below that density.

5. A process according to claim 1 wherein each unit is provided with a primary separation zone and a secondary separation zone immediately thereabove, the overflow fraction from each unit besides the first being introduced into a lower portion of the secondary separation zone.

6. A process according to claim 5 wherein liquid from each primary extraction zone of a unit is recirculated to the unit below the primary separation zone.

7. A process according to claim 6 wherein the liquid extracted from the primary extraction zone is reintroduced to the unit into a primary sedimentation zone located below the primary separation zone.

8. A process according to claim 7 wherein a secondary sedimentation zone is located immediately below and directly communicating with the primary sedimentation zone and wherein a flow of liquid having an acceptably small amount of solids contained therein is introduced into the secondary sedimentation zone to effect a vertical upwards velocity of sufficient magnitude that it prevents small weighting solids in the separation medium from settling into the underflow from each unit, whereby the underflow from each unit is thereby scrubbed.

9. A process according to claim 5 wherein a primary sedimentation zone is provided located below the primary separation zone and wherein a secondary sedimentation zone is provided immediately below and directly communicating with the primary sedimentation zone and wherein a flow of liquid having an acceptably small amount of solids contained therein is introduced into the secondary sedimentation zone to effect a vertical upwards velocity of sufficient magnitude that it prevents small weighting solids in the separation medium from settling into the underflow from each unit, whereby the underflow from each unit is thereby scrubbed.

10. A process according to claim 1 wherein one or more units of the system are provided with a secondary extraction zone communicating with and located between the primary extraction zone and the column and employed to extract liquid from the unit having acceptably small amounts of solids contained therein, said liquid to be used for scrubbing purposes within the system or removed from the unit to decrease the volume of overflow within the system.

11. A process according to claim 1 wherein the underflow from each of one or more of the units is subjected to a series of screening operations using progressively smaller screen apertures such that multiple density fractions are produced.

12. A process according to claim 1 wherein the separation medium is partially weighted by the addition of a soluble material to the liquid being used as the separation medium.

13. A process according to claim 1 wherein the vertical flow of the separation medium is pulsed in order to increase the density sensitivity of the separation.

14. An apparatus for separating particulate high density materials from associated particulate lower density materials comprising a plurality of elutriation column units in a series, the units being provided with means to pass the overflow from each unit into the next unit of the series, and means to progressively vary the average vertical flow rate in each successive unit, each unit being provided with an outlet for underflow, means being provided to collect the underflow from each unit, a screen being associated with each unit in the series besides the first unit, the screen mesh size varying progressively through the series and being selected according to the average vertical flow rate in order to pass higher density particulate materials but retain the lower density particulate materials, one or more units in the series employing a primary extraction tank from which liquid bearing solids in suspension is extracted and employed in whole or in part as a separation medium within the system.

15. An apparatus according to claim 14 wherein each column is provided with a primary separation section and a secondary separation section immediately thereabove.

16. An apparatus according to claim 15 wherein the primary separation section is provided with a narrowed column diameter to provide a flattening of the velocity profile of the separation medium and effect a more precise classification in that primary separation section.

17. An apparatus according to claim 16 wherein the secondary separation section has a larger diameter than that of the primary separation section.

18. An apparatus according to claim 15 wherein each column is further provided with a primary sedimentation chamber and a secondary sedimentation chamber, the primary sedimentation chamber being in direct communication with the primary separation section and immediately below it, and the secondary sedimentation chamber being directly below the primary sedimentation chamber.

19. An apparatus according to claim 18 wherein the primary sedimentation chamber has a larger cross-section area than the primary separation section, thus producing a lower vertical velocity within.

20. An apparatus according to claim 18 wherein liquid from the primary extraction tank is recirculated through the column and reintroduced to the column at the level of the primary sedimentation chamber.

21. An apparatus according to claim 14 wherein one or more of the counterflow separation units are provided with a secondary extraction tank communicating with and located between the primary extraction tank and the column, the secondary extraction tank to extract liquid which has an acceptably small amount of solids contained therein from the column to be used for scrubbing purposes within the system or removed from that unit in order that a decrease in the overflow volume be achieved.

22. An apparatus according to claim 21 wherein the secondary extraction tanks are hydrocyclones.

23. An apparatus for separating particulate high density materials from associated particulate lower density materials comprising a plurality of hydrocyclone units in a series, the units being provided with means to pass the overflow from each unit into the next unit of the series, and means to progressively vary the average tangential flow rate in each successive unit, each unit being provided with an outlet for underflow, means being provided to collect the underflow from each unit, a screen being associated with each unit in the series besides the first units the screen mesh size varying progressively through the series and being selected according to the average tangential flow rate in order to pass higher density particulate materials but retain the lower density particulate materials.

24. An apparatus for separating particulate high density materials from associated particulate lower density materials comprising a plurality of hydrocyclone units in a series, the units being provided with means to pass the underflow from each unit into the next unit of the series, and means to progressively vary the average tangential flow rate in each successive unit, each unit being provided with an outlet for overflow, means being provided to collect the overflow from each unit, a screen being associated with each unit in the series besides the last unit, the screen mesh size varying progressively through the series and being selected according to the average tangential flow rate in order to pass higher density particulate materials but retain the lower density particulate materials.