US2702630A

US2702630A - Classification of particles

Info

Publication number: US2702630A
Application number: US100024A
Authority: US
Inventors: Roderic M Scott
Original assignee: Sharples Corp
Current assignee: Sharples Corp
Priority date: 1949-06-18
Filing date: 1949-06-18
Publication date: 1955-02-22
Anticipated expiration: 1972-02-22

Description

Feb. 22, 1955 Filed June 18, 1949 R. M. SCOTT CLASSIFICATION OF PARTICLES- Fine Frau/:0

Flue

3 Shegts-Sheet 1 Coarse Fraction Classifier l/VW/VTOR Radar/c M. Scott Attorney Feb. 22, 1955 R. M. scoTT CLASSIFICATION OF PARTICLES 3 Sheets-Sheet 2 Filed June 18, 1949 INVENTOR. Roderic M. 8001/ BY v Attorney Feb. 22, 1955 R. M. SCOTT CLASSIFICATION OF PARTICLES Filed June 18. 1949 3 Sheets-Sheet 3 Coarse F'eed Fine Fraction Y 7 Caarsc I Fine Fracfian y Coarse Hne Fracfion I Coarse Eng 7 Coarse Pr (y Feed /70 F /g.7. Fme

7 PCoarse ee I Fme h 7,

// Fracfion Coarse Fme Coarse Coarse Fracfion F/ne m+l F e Coarse m //V VE N 7' 0R F/ne Rodric 'M. Scarf Coarse Fraction A ffarney United States atent O CLASSIFICATION OF PARTICLES Roderic M. Scott, Wynnewood, Pa., assignor to The Sharples Corporation, a corporation of Delaware Application June 18, 1949, Serial No. 100,024

13 Claims. (Cl. 209133) This invention relates to a method and apparatus for the separation or classification of solid particles, and more particularly solid particles below sieve size.

The precise classification of powders consisting of particles ranging widely in size, into fractions entirely free of particles larger or smaller than a particular size, or into fractions containing only particles of a particular narrow size range, is an exceptionally difiicult problem of great importance, which is frequently encountered in the preparation of a wide variety of products such as catalysts, adsorbents, pigments, abrasives, and metal powders, for example. In many cases the best quality of product obtainable for a specific application is directly related to the limiting degree of sharpness with which the particle classification can be made by the use of existing methods and available apparatus.

I have discovered a method for making very sharp separations or precise classifications of solid particles which comprises passing a feed mixture of particles through a primary particle classifying zone in a manner to produce an efiicient separation of the mixture into two fractions containing a preponderance of particles larger and smaller, respectively, than a selected critical or cut point size, passing the fine fraction separately through a second particle classifying zone in a manner to produce a sharp separation into preferably a larger amount of fine fraction containing a preponderance of particles finer than the selected critical or cut point size and preferably a smaller amount of a reject coarse fraction containing a preponderance of particles larger than the selected critical or cut point size; passing the coarse fraction from the primary classifying zone through a third particle classifying zone in a manner to produce a separation into preferably a larger amount of a coarse fraction containing a preponderance of particles larger than the selected critical or cut point size and preferably a smaller amount of a reject fine fraction containing a preponderance of particles smaller than the selected critical or cut point size, and recirculating the smaller amounts of coarse and fine reject fractions back through the primary classifying zone together with the feed mixture. In certain cases in which it is important that only the fine fraction or only the coarse fraction, respectively, be sharply separated from the original mixture of particles, I prefer to pass only the fine fraction or only the coarse fraction, respectively, through a second particle classifying zone in a manner described above for the succeeding classification in each particular case, and obtain preferably a sharper fine fraction as final product and recirculate the preferably smaller amount of coarse reject fraction, or obtain preferably a sharper coarse fraction as final product and recirculate the preferably smaller amount of fine reject fraction, respectively, back through the primary classifier together with the feed mixture.

Referrring now to the drawings:

Figure 1 is a schematic drawing showing a preferred embodiment of my invention employing three classifiers;

Figure 2 is a set of graphs showing particle size distribution curves of the type obtainable by the use of three classifiers as shown in Figure 1;

Figure 3 is a set of graphs showing particle size distribution curves of the type obtained by the use of three classifiers when the reject fractions are combined instead of being recirculated through the primary classifier;

Figure 4 is a flow sheet showing the flow of materials in the three-classifier system of Figure 1;

Figure 5 is a flow sheet showing the flow of materials in a classifier system employing a large number of classifiers;

Figure 6 is a flow sheet showing the flow of materials "ice in a two-classifier system in which the coarse reject fraction is recirculated; and

Figure 7 is a flow sheet showing the flow of materials in at two-classifier system in which the fine reject fraction is recirculated.

In a preferred embodiment of the invention, as shown in Figure 1, the classifier is an improved device having some of the characteristics of the apparatus disclosed and claimed in my co-pending application, Serial No. 100,023; filed June 18, 1949.

It will be understood however, that for the purpose of the present invention, any one of a number of particle classifiers may be used. The classifier 10 shown as partly sectioned in Figure 1, comprises an outer cylindrical casing 11 which terminates in a conical collector 12, connected to a delivery conduit 13. The classifying region 14 comprises a pair of oppositely disposed radially tapering surfaces 15 and 16 into which the material to be classified is fed through an annular passage 17. Upper tapering surface 15 is provided with an annular boss or extension 18 which fits into the axial bore 19 of surface 16 and communicates with conduit 20.

The material to be classified is fed to the machine from a hopper 21 and vibratory feeder 22, the latter being connected to an injector 23. Air manifold 24 is connected to a convenient source of air pressure (not shown) and is also connected to injector 23 by means of air line 25. The particles to be classified are carried to classifier 10 through conduit 26 which communicates tangentially with annular passage 17 to provide an initial vortex or circulating motion to the particles. The air introduced into air line 25 through valve 27 is preferably sufficient only to carry the solids from the injector to the first classifier. The main air supply for the classifier is controlled by means of valve 28 and air line 29 which are in turn connected to injector 30. Exhaust air from the classifier is carried through conduit 31 and valve 32, the latter of which also functions to control the pressure within the classifier 10.

The main air supply is introduced tangentially into casing 11 through conduit 30a and rotates in the classifier 10 as it movesdownward and into classifying zone 14 between inwardly converging surfaces 15 and 16. The rotating air moves inward in vortex fashion from the outer periphery of the'classifying zone 14. As a result of the vortex action and the convergence of surfaces 15 and 16, the radial velocity of the air continuously increases as it moves inward from the periphery to the axis of the classifying zone 14. The solid feed particles introduced into the vortex through annular passage 17, as described above, smoothly slide into the rotating vortex. By virtue of the combined action of the outwardly directed centrifugal force and the inwardly directed drag force of the air on the individual particles, a classifying process takes place. By carefully regulating the rate of air flow in the classifying zone 14 particles larger than a particular size are forced outward to the periphery of the zone by the centrifugal force, while particles smaller than that size are carried inward toward the axis of the zone by the drag force.

Thus the solid particles which are introduced into the classifying region 14 are separated into a coarse fraction which is collected in funnel 12 and fed through conduit 13 and valve 35 to a second classifier 36 which may be identical with classifier 10. The fine fraction of solid particles, which is classified by means of region 14, passes out of the classifying zone in suspension in the air by making a sharp nearly right angle turn into annular passage 19. Here the drag force acting on the fine particles in suspension is vertically downward while the centrifugal force is still radially outward. Thus the fine suspended particles are forced out of the air against the walls of the passage 19 in a downward and outward direction. The air stream also makes about a turn in leaving passage 19 through conduit 31. In thus re-- versing the direction of flow of the air, the fine particles are very efficiently separated in passage 19. These par ticles are fed to classifier 38 (which may be identical with classifier 10) through conduit 20 and valve 37.

The fine fraction from classifier 36 is returned to the original feed by means of conduit 39, valve 40 and conduit 41, whereas the coarse fraction from classifier 36 is retained in receptacle 42. Exhaust air from this classifier is expelled through valve 43 into the common exhaust line 44. In like manner, the fine fraction from classifier 38 is delivered to receptacle 45 through conduit 46, whereas the coarse fraction from classifier 38 is returned to the initial feed through valve 47,

conduits

48 and 41. Exhaust air from classifier 38 is expelled through conduit 49 and connects to air line 44 through valve 50. Air for

classifiers

36 and 38 is provided from manifold 24 through

air lines

55, 56 and

valves

57, 58, respectively.

By closing valve 32 sutficiently to cause a slight pres sure buildup in classifier 10, a slight excess of pressure is provided at its particle outlets over the pressure in classifier 36 and classifier 38, respectively. These respective differentials in pressure provide the respective pressure heads necessary to cause the flow of the particle fractions into

classifiers

36 and 38, respectively.

Valves

50 and 43 provide means for the regulation of relative pressure drop between

classifiers

36 and 38, as well as the pressure drop across each said classifier.

It will thus be understood that the basic cycle contemplates an initial separation of the aggregate into a fine fraction and a coarse fraction respectively containing a preponderance of particles smaller and larger than the selected cut point size. Each of these original fractions is then run through a separate classifier to be separated preferably operating with substantially the same selected cut point size as used in the initial sep aration. The coarse fraction from the second-stage classifier 38, together with the fine fraction from the second-stage classifier 36, is mixed with the feed and recirculated through the first-stage classifier. This system enables all of the particles to be subjected to at least two classifying actions so that any particle that would be misclassified would of necessity have to get into the wrong fraction at least twice by mistake. Any particle which has been improperly classified in the first stage, still has an opportunity to be correctly reclassified in the second stage, whereupon it is either removed from the system as correctly classified or returned to the system where it will eventually reach the correct fraction.

For the purpose of illustration, I have shown a system in which the pressure differential between the first and second stages is controlled by means of valves. Any other suitable manner of accomplishing this result may be used. Furthermore, the exact manner of feeding the air and solid aggregate to the classifiers is intended to be illustrative only.

Referring now to Figures 2 and 3, the advantages of the above described system of particle classification can be clearly illustrated by means of distribution curves. The solid curve A in Figures 2 and 3 presents the distribution curve of a powder in which the percentage by weight of the particles is plotted against the logarithm of the diameter of the particles. The performance of a classifier may be judged by the steepness of such a powder distribution curve in the vicinity of the selected critical or cut point size.

Figure 3 shows the distribution curves of a classification system in which the aggregate is run through the first classifier, whereupon the fine fraction is run through a second classifier and the coarse fraction is run through a third classifier, without any recycling. In such a system, three fractions are obtainable.

After the first run through the classifier, the particles are divided into two fractions having a distribution as shown by curve 60 for the fine fraction, and curve 61 for the coarse fraction. These curves are not symmetrical in shape because the performance of the classifier may be different with respect to particles smaller than the critical size, than with respect to particles larger than the critical size. These two fractions are now run through additional classifiers adjusted to make the cut at substantially the same diameter as in the first classifier and this additional run divides each fraction into two parts, the distribution curves for which are shown as curve 62 for the fine, and curve 63 for the coarse. Curve 64 shows the intermediate cut which comprises the fine fraction of the second coarse classifier and the coarse fraction of the second fine classifier.

Careful examination of the distribution curves shown in Figure 3 will reveal that:

(a) The average size as indicated by the maximum of each curve has been made smaller for the fine fraction since the peak of curve 62 is farther to the left than the peak of curve 60; likewise, the average size of the particles in the coarse fraction has been increased since the peak of curve 63 is farther to the right in Figure 3 than is the peak of curve 61;

(b) The quality of the classification as determined from the slope of the curves at the diameter at which the cut is made has not been noticeably improved by the second treatment; and

(c) The third fraction, comprising nearly a quarter of the aggregate, has a distribution curve similar to that of the original feed as shown by a comparison of curves A and 64 (Figure 3), except that it covers a more narrow particle size range.

Referring now to Figure 2, the distribution curves of an embodiment of the invention using three classifiers as shown in Figure 1, may be compared with the curves of Figure 3. The output from the first machine, as shown in curves 60' and 61', is substantially the same as curves 6@ and 61 in Figure 3. The distribution curves for the final fractions however, show a very striking difierence to the curves for the final fractions in Figure 3. Curve 64 shows the distribution in the final fine fraction and curve 65' shows the distribution in the final coarse fraction. When compared to the first run curves, it will be noted that the average size of the fine cut is larger and the average size of the coarse cut is smaller, as noted by comparing curve 64 with curve 60' and curve 65' with curve 61'. At the cutting point, the slope of the distribution curves 64' and 65 is much greater than the slope of curves 60 and 61. Furthermore, all the material in the feed has been classified into two fractions having relatively good characteristics.

It will thus be apparent that the basic unit of the system, comprising three classifiers, may be advantageously employed to provide three important advantages when accuracy of classification is essential. First, the average particle size in each fraction is closer to the size at which the cut is made; second, the distribution in each fraction is greatly improved; and third, the above increased efiiciency is achieved without additional apparatus.

Figure 4 illustrates diagrammatically the flow of materials in the fundamental unit of my improved system and is applicable to the system shown in Figure 1.

The same principle is readily suited for use with any desired number of individual units, and Figure 5 illustrates a flow diagram for such an arrangement. There are for the purposes of illustration, m-l-n-l-l classifiers connected in accordance with the invention. The two end units n and m from which the final fine and coarse fractions are removed respectively, and the first-stage classifier 70 into which the aggregate is initially fed, together with any number of intermediate classifiers designed as 1 to 11-1, and 1 to m+l, comprise the system.

Figure 6 illustrates the flow of materials in a system used when the primary object is to prepare a precisely classified fine fraction of particles and a well classified coarse fraction of particles which does not need to be as precisely classified as the fine fraction. Only two classifiers, 71 and 72, are required for this system and only the coarse reject fraction from the second-stage classifier 72 is recycled through the first-stage classifier 71.

Figure 7 illustrates the corresponding situation when the primary object is to prepare a precisely classified coarse fraction of particles. In this case only the fine reject fraction from the second-stage classifier 72 is recycled through the first-stage classifier 71.

The system has been illustrated herein in connection with a classifier of the type having a stationary classifying zone. However, it will be clearly understood that any other classifier may be employed in the system, for example, a whizzer-type or other machine may be used with improved results.

While preferred embodiments of the invention have been described, it will now be obvious to those skilled in the art that modifications may be made and the features of one modification may be utilized with or in substitution for other features of other modifications, all Within the scope of the appended claims.

I claim:

1. A method of classifying particles of solid material which comprises pneumatically classifying an aggregate of finely divided solid particles in a pneumatic classifying zone to provide a coarse fraction and a fine fraction, pneumatically classifying the said fine fraction in a second pneumatic classifying zone to provide a coarse fraction and a fine fraction, pneumatically classifying said firstmentioned coarse fraction in a third pneumatic classifying zone to provide a coarse fraction and a fine fraction, segregating the fine fraction from said second classifying zone, recycling the coarse fraction from said second classifying zone through said first classifying zone, segregating the coarse fraction from said third classifying zone, and recycling the fine fraction from said third classifying zone through said first classifying zone.

2. A method of classifying solid particles which comprises passing a feed mixture of finely divided solid particles through a primary pneumatic classifying zone to produce a coarse fraction containing a preponderance of particles larger than a selected cut point size and a fine fraction containing a preponderance of particles smaller than said selected cut point size, passing said coarse fraction through a second pneumatic classifying zone to produce a coarse fraction containing a preponderance of particles larger than said selected cut point size and a reject fine fraction containing a preponderance of particles smaller than said selected cut point size, passing said first-mentioned fine fraction through a third pneumatic classifying zone to produce a fine fraction containing a preponderance of particles smaller than said selected cut point size and a reject coarse fraction containing a preponderance of particles larger than said selected cut point size, separately collecting the coarse fraction produced in said second classifying zone, separately collecting the fine fraction produced insaid third classifying zone, and recycling through said primary classifying zone said reject fine fraction and said reject coarse fraction for classification therein together with said feed mixture.

3. A method of classifying particles of solid material which comprises pneumatically classifying an aggregate of finely divided solid particles in a pneumatic classifying zone to provide a coarse fraction and a fine fraction, pneumatically classifying said fine fraction in a second pneumatic classifying zone to provide a coarse fraction and a fine fraction, segregating said last-mentioned fine fraction, pneumatically classifying said first-mentioned coarse fraction in a third pneumatic classifying zone to produce a coarse fraction and a fine fraction, segregating said last-mentioned coarse fraction, recycling the coarse fraction from said second classifying zone and the fine fraction from said third classifying zone through said first-mentioned classifying zone, introducing air tangentially into each of said classifying zone to induce desired particle separation, and controlling the flow of air between said classifying zones to provide pressure differentials in the second and third classifying zones.

4. A finely divided solid particle classifying system comprising a first pneumatic classifier having an intake, at least one coarse fraction discharge, and at least one fine fraction discharge; a second pneumatic classifier having an intake, at least one coarse fraction discharge, and at least one fine fraction discharge; means, including a particle conduit, for connecting the fine fraction discharge of the first classifier to the intake of the second classifier; means, including a particle conduit, for connecting the coarse fraction discharge of the second classifier to the intake of the first classifier; a third pneumatic classifier having an intake, at least one coarse fraction discharge, and at least one fine fraction discharge; means, including a particle conduit, for connecting the coarse fraction discharge of the first classifier to the intake of the third classifier; means, including a particle conduit, for connecting the fine fraction discharge of the third classifier to the intake of the first classifier; means for collecting particles from the fine fraction discharge of the second classifier, and means for collecting particles from the coarse fraction discharge of the third classifier.

5. A particle classifying system in accordance with claim 4, in which all the classifiers have substantially the same particle cut point characteristics.

6. A particle classifying system in accordance with claim 4, having a source of compressed gas, and means for directing the compressed gas into the classifying zone of each classifier.

7. A particle classifying system in accordance with claim 4, in which the particle conduits are provided with valve means for controlling the relative fiow of material between the classifiers.

8. A particle classifying system in accordance with claim 4, in which each classifier has an exhaust air conduit, and valve means in said exhaust air conduits for controlling the pressure in the associated classifiers.

9. A finely divided solid particle classifying system comprising a first pneumatic classifier, a particle intake for said first classifier, a gas intake for said first classifier, a coarse fraction discharge for said first classifier, a fine fraction discharge for said first classifier, a second pneumatic classifier, a particle intake for said second classifier, a gas intake for said second classifier, a coarse fraction discharge for said second classifier, a fine fraction discharge for said second classifier, a connection between the fine fraction discharge of the first classifier and the particle intake of the second classifier, a valve in said connection, a connection between the coarse fraction discharge of the second classifier and the particle intake of the first classifier, a valve in said connection, a third pneumatic classifier, a particle intake for said third classifier, a gas intake for said third classifier, a coarse fraction discharge for said third classifier, a fine fraction discharge for said third classifier, a connection between the coarse fraction discharge 6f the first classifier and the particle intake of the third classifier, a connection between the fine fraction discharge of the third classifier and the particle intakes of the first classifier, means for injecting particles into the particle intake of the first classifier, and means including a source of compressed gas for moving the particles in the connections and intake.

10. A method for classifying an aggregate of finelydivided solid particles into end fractions respectively of fine particles and of coarse particles which comprises initially pneumatically classifying said aggregate into a fine fraction and a coarse fraction in an intermediate classifier of a series of at least three pneumatic classifiers in each of which a fine fraction and a coarse fraction is pneumatically produced, feeding the fine fraction produced in each said classifier into the next classifier of said series in one direction for pneumatic classification therein, feeding the coarse fraction produced in each said classifier into the next classifier of said series in the opposite direction for pneumatic classification therein, and separately collecting the final fine fraction and the final coarse fraction at the opposite ends of said series.

11. A method as defined in claim 10 wherein all classifications are made with substantially the same selected cut point size.

12. A multi-classifier system for the classification of an aggregate of finely-divided solid particles into end fractions respectively of fine particles and of coarse particles Which comprises at least three pneumatic particles-size classifiers arranged in series, each said classifier adapted to produce a fine fraction and a coarse fraction, means for feeding an aggregate of finely-divided solid particles to be classified to a classifier positioned intermediately of said series, means for feeding the fine fraction produced in each said classifier to the next classifier in said series in one direction and for separately collecting the fine fraction produced by the end classifier in said direction, and means for feeding the coarse fraction produced in each said classifier to the next classifier in said series in the opposite direction and for separately collecting the coarse fraction produced by the end classifier in said opposite direction.

13. A multi-classifier system as defined in claim 12 wherein each of the classifiers have substantially the same cut point characteristics.

References Cited in the file of this patent UNITED STATES PATENTS 1,145,903 Lehrack et al. July 13, 1915 1,629,593 Stebbins May 24, 1927 1,660,683 Stebbins Feb. 28, 1928 1,928,702 OMara Oct. 3, 1933 1,985,947 OMara Jan. 1, 1935 2,087,645 Hermann July 20, 1937 2,125,086 Rourke July 26, 1938 2,329,299 Rourke Sept. 14, 1943 2,377,524 Samson et al. June 5, 1945