Process and apparatus for controlling the density of the apex discharge of a cyclone

^Aug· ^I9fiU and 4Pa«^^HcJ„^NG THE DENSI 2,648,433 Filed Dec. 5, 1950 °^F ™ “ ^DISC™^{GE 4} °™® ,

Sheets-Sheet 1

<img file="US2648433A_D0001.tif" />

ATTORNEY

Aug. 11, 1953 q u WRIGHT γτδι 9RAAA11

PROCESS AND APPARATUS FOR CONTROLLING THE DENSITY ’

Filed Dec. 5, 1950 °^F ™^{E APEX D}?^SCHARGE °^{F A CYCL0NE} . ₊

Sheets-Sheet 2

<img file="US2648433A_D0002.tif" />

^Aug· Access «, the dehsit?’⁶⁴⁸’⁴³³

Plied Dec. 5. 19S0 °^{Γ APBX DISCHAEGE} °^{E Λ} “«Μ·: , ,_{h 4}

Sheets-Sheet 3

<img file="US2648433A_D0003.tif" />

ATTORNEY

Patented Aug. 11, 1953

2,648,433

UNITED STATES PATENT OFFICE

2,648,433

APPARATUS FOR CONTROLLING THE DENSITY OF THE APEX DISCHARGE OF A CYCLONE

Orrin Hughitt Wright and Joseph Leonard Weaver, ^Brewster&gt; Fla·, and Elliot Bryant Fitch, Westport, Conn.; said Wright and said Weaver assignors, by mesne assignments, to Naamloze Vennootschap Maatschappij voor Kolenberwerking Stamicarbon, Heerlen, Netherlands, and said Fitch assignor to The Dorr Company, Stamford, Conn., a corporation of Delaware

Application December 5, 1950, Serial No. 199,309

Claims.

This invention relates to cyclones. More specifically this invention relates to a process of controlling the consistency of the discharge from the ^apex of a hydraulic cyclone and to apparatus for effecting this control automatically.

Cyclones have long been widely employed for the dewatering, desliming, and classification of fluid slurries. They have also been used with finely divided media for the performance of separations based on specific gravity differences utilising centrifugal force to effect the separation Generally, they are constructed in the form of a hollow cylinder or cone wherein the separatory process takes place. For simplicity their operaion will be described with particular reference to the conical type as each type of cyclone operates according to the same general principles, and the phrase “apex orifice” will be used hereinafter to describe the opening from which the heavy fraction is emitted not only from the conical type cyclone but from the cylindrical type cyclone as well.

The feed, a pumpable slurry, enters the cone tangentially through a feed inlet port near the base at a pressure which may be as low as 5 lbs./in.² or lower, and as high as 200 Ibs./in²or higher, and spirals towards the apex of the cone in circles of decreasing diameter with increasing angular velocity. As a result, two concentric vortices form, creating a central air column. Both vortices rotate in the same direction but their flows are opposed, the inner vortex moving axially to the base of the cone and the outer vortex moving axially to the apex The heavier constituent of the feed is thrown by centrifugal force into the outer vortex, and the lighter constituents, usually water and slimes in the case of desliming operations, and water and the lighter fraction of the feed in the case of specific gravity separations, are forced towards the inner vortex. The resultant of the forces within the cone causes these lighter constituents to travel along the inner vortex towards the base of the cone, exiting through a base orifice or passageway axially disposed therethrough, while the heavier constituents travel in the outer’vortex towards the apex and are discharged through an orifice of definite size.

The operation of the cyclone is influenced by many factors—including feed pressure, area of (Cl. 209—211) feed entrance, percent solids in feed, size distribution of solids in feed, size of base outlet, and size of apex orifice. Since it is practically impossible in ordinary operations to keep the feed variables 5 constant or to change the basic cyclone structure, there is, as a result, a tendency with changes in feed for the characteristics of the two discharges from the cyclone to vary in their composition. Consequently, in attempting to control this, the 10 most critical of the factors above mentioned which are capable of practical control is the size of the apex orifice.

When the cyclone is used as a thickener or classifier, it is of course desirable to produce as dense an apex discharge as possible. In a similar manner, some cyclone separatory processes have to operate with a maximum density of apex discharge in order to effect separation at a sufficiently high gravity with lighter and cheaper solids such as magnetite.

Unfortunately when a cyclone operates under conditions closely approaching maximum consistency of the apex discharge, small changes in any of the factors referred to above may increase the consistency to the point where the apex discharge becomes sausage-like or, as it is usually described “ropy.” When this condition prevails there is no longer a hollow central air column or core extending through the apex. Even when this condition sets in, the ropy type of discharge frequently extends up into the cyclone resulting in plugging. It has therefore been necessary in the past to operate cyclone thickeners with an apex discharge having a consistency sufficiently more fluid than that of a ropy discharge so that there is a sufficient margin of safety. Even the most critical manual supervision can not be relied on to prevent plugging if the cyclone is to be operated very close to maximum apex discharge consistency. Plugging soon results in serious contamination of the overflow product discharge, and when the plugging is complete, no thickening or classification takes place, all the feed being discharged through the base of the cyclone. In the case of cyclones operating as separators, plugging changes the gravity at which separation takes Place and results in contaminated products due to incomplete separation.

According to the present invention it has been found that the operation of a cyclone can be con15

2,648,433 trolled manually and preferably automatically, near maximum apex discharge consistency without any danger of overloading or plugging. The invention is carried out by applying a partial vacuum to the central air column in the vortex finder of the cyclone. Preferably this is done by applying a suction to the central air column or core which persists in the middle of the vortex of the overflow after it has been discharged from the cyclone. It has been found that the vacuum in the central air column does not vary with the consistency of the apex discharge in a uniform manner. At very low apex discharge consistencies, when the consistency of the discharge is extremely fluid, the discharge is emitted as a spray and. the increase in vacuum with increase in apex discharge consistency approaches linearity in this zone, the slope of the curve in this zone varying with different feeds and with different designs of cyclone. When the apex discharge begins to approach the consistency of a cohesive discharge, the curve changes its character and becomes much steeper, the steepness in this second zone increasing as the consistency of the apex discharge becomes ropy. At a point where the discharge becomes sausage-like and there is no significant remaining central air column in the apex the vacuum reaches a maximum. Thereafter, in this third zone there is no increase in vacuum with further decrease in apex diameter as the cyclone begins to plug up.

The sharp change in characteristics of the curve as the danger point is approached makes it possible by means of the present invention to effect reliable control at a maximum apex discharge consistency approaching maximum discharge density for the particular operation. The steep part of the curve, that is to say the second zone where the rate of increase of vacuum with , increase of discharge consistency is markedly greater than that of the lower portion of the curve, permits accurate control because there is a relatively much greater change in vacuum with each change of density. As pointed out above, the exact shape of the curve varies with the nature of the feed and design of the cyclone. It is possible to use the present invention to control the apex discharge density on the steep zone of the curve. There is, therefore, a choice of the particular density at which the apex discharge is to be maintained. Controlling it at the lower part of the steep zone of the curve where the discharge is usually of the spray type gives maximum safety but does not give as high an apex discharge density as is possible by controlling within a higher range of the curve where the discharge is usually of the cohesive, non-spraying type. The exact range depends on the relative economic factors involved and is further discussed below.

As a result of these discoveries, it has been found that when a pressure responsive means is made to communicate with the partial vacuum in the central air column and when said means is made to actuate a valve controlling the flow of discharge through the apex orifice so that the valve is opened somewhat when the vacuum in the central air column increases above a predetermined value, and so that the valve is closed somewhat when the reverse takes place, and the vacuum is thereby maintained substantially constant, a discharge of substantially uniform consistency is obtained despite changes in the density of the feed and the danger of plugging is com10 pletely and automatically obviated. In brief, the flow of apex discharge is controlled so as to oppose changes in the partial vacuum.

As a further result, it has been found possible to operate a cyclone so that irrespective of other variations it uniformly yields an apex discharge which has a consistency which closely approximates the highest which can be discharged through the apex.

Hereinafter the degree of the vacuum referred to above may be referred to as “pressure differential,” that is, as the difference in pressure between atmospheric pressure and the pressure in the central air column, and the scale of inches of water will be uniformly used in connection therewith.

It is a further advantage that conventional cyclones can be readily altered to permit the practice of the present invention and that otherwise no particular form of apparatus is required when the instant invention is practiced automatically.

The invention will be further described with reference to the drawings in which:

Fig. 1 is a graph showing a plot of the air pres&gt; sure differentials existing in the central air column of a cyclone during a typical run illustrating the present invention against the respective apex discharge densities obtained;

Fig. 2 is a vertical section of one form of ap) paratus according to the present invention; and

Fig. 3 is a vertical section of a second form of apparatus according to the present invention.

Fig. 1 is a plot of data accumulated during a typical run made in a cyclone modified according to the present invention, with manual control of the apex valve. This run was made in a twelveinch cyclone fed with a slurry of about 43.7% crushed limestone and water having a specific gravity of about 1.4 which was pumped into the cyclone at a pressure of about 25 lbs./in.² through a feed entrance about 5.72 in.² in area. The cyclone was provided with a suction leg which was capable of producing a maximum vacuum equal to about 70 inches of water. In this run the diameter of the apex discharge valve was at a maximum at the beginning of the run, and was continuously constricted manually to a minimum at which plugging of the cycle occurred. The specific gravity of the apex discharge and the corresponding pressure differentials in the central air column were noted at intervals during the run. At the beginning, the discharge was whirled out in so fluid a spray that the rush of air entering the apex could be detected by ear, the specific gravity of the apex discharge approximated that of the feed, the pressure differential in the central air column was low and in its rate of increase was essentially linear. In that zone, with gradual constriction of the apex valve both the specific । gravity of the discharge and the pressure differential in the central air column slowly increased. When, however, the density of the apex discharge reached about 1.89 at a pressure differential of about 2 inches of water, formation of a slightly cohesive discharge began and departure from the essentially linear increase in the central air column pressure differential began. With further constriction of the valve the consistency of the discharge and rate of increase of the pressure differential increased more rapidly, while the density increased relatively slightly. Plugging occurred when the pressure differential was about 70 inches of water corresponding to an apex discharge specific gravity of about 2.08. At that point the central air column was blocked at the

2,648,433 apex, feed began to pass into the overflow conduit, and the curve abruptly leveled off. Continuing to decrease the diameter of the apex valve caused no increase in vacuum above this maximum, although minor fluctuations occurred thereafter.

The points observed during this run were as follows:

70...

20.4-.

6.8...

5.1—

2.752.50..

1.875.

Apex Discharge

Vacuum, Inches HiO

Specfiic Gravity

Percent Solids

2.08 1.985 1.940

1.925 1.918 1.900 1.892

79.0

76.0

74.1

73.5

73.2

72.5

72.0 ⁶ forming a liquid seal to the cone. The slurry, now substantially freed of its rotational movement but containing the entrained air, passes through suction leg 9, which is sufficiently long to produce the desired degree of vacuum in the central air column. Conduit 6 is perforated to admit hollow vacuum probe 12 through air-tight gland 25 which terminates within the base outlet passageway on cyclone axis 27 and which is connected to a conventional vacuum relay 15 through trap 13 and conduits 14 and 28. The vacuum relay is provided with compressed air supply 16, compressed air exhaust 17, and supplies rubber pneumatic sphincter valve 18 in valve housing 29 with compressed air through line 19. As a result, the sphincter valve inflates and partially closes when the pressure differential at orifice 11 is below a predetermined value, and partially opens when the pressure differential is above that predetermined value. The discharge from apex (0 passes first through apex conduit 20 and then through spray shield 21.

The above type of apparatus permits completely automatic control of the consistency of the discharge. When manual operation is desired, valve 24 may be closed to cut off the relay, and the pressure differentials are read directly on the manometer 23 by opening valve 22. Manual control of the air supply to sphincter valve 18 is effected by operating the supply valve in relay 15 which being conventional is not shown.

Fig. * ’

Fig. 2 ciples. passes casing ___ _ vortices in cone 33, described above. Tl._ vortex comprising the lighter constituents trav25

In the second zone, that is, when the pressure differential in the central air column was more than about 2 inches of water, an increase of two units of density in the apex discharge caused more than twice as great an increase in the pressure differential in the central air column as was caused by an increase of one unit of density in the density of the apex discharge.

As will be discussed more in detail hereinafter there is nothing critical about any of the magnitudes set forth in Fig. 1. The specific gravity of the slurry being treated may differ widely from that employed. The maximum pressure differential may be greater or less than that shown with a resulting accentuation of diminution of the slope of the curve, and the point at which the curve begins to rise steeply may vary as the result of a combination of factors. In every instance, there will be found a zone where control according to the process of the present invention is feasible.

Fig. 2 is a section of one form of apparatus suitable for the practice of the instant invention, and will be described according to its mode of operation. A conventional cyclone I is shown having feed conduit 2 with tangential inlet port 3 in head casing 4 and a base outlet passageway or tubular exit from the base. In one form, this tubular exit may have a section 5 axially located within the cyclone and extending toward the apex beyond port 3, but it essentially comprises curved overflow discharge conduit 6, horizontal conduit 7, curved conduit 8, and vertical suction leg 9. Feed entering under pressure rotates ------₇----------within head casing 4 and is impelled into cone 55 *^rol_^{tile air ram} in ram housing 41 by unbalI where the speed of rotation increases as the ^r ’ feed proceeds in the direction of apex 10. As a result, two vortices are created rotating in the same direction but having opposed axial flow - _____________.......

components. In normal operation, the central 60 unbalancing by return motion lever 40 which reair column described above extends through the cone, beginning at the apex and continues through the base outlet passageway. The lighter components of the feed are forced inwardly, be- „ _______________________ come part of the inner vortex, travel upwardly, 65 ^c°ntrols the loading on the conventional measurand leave the cyclone at orifice 11. The heavier ' ’ components are thrown outwardly, become part of the outer vortex, and are discharged with a rotational movement which varies with their consistency through apex 10.

is a different form of the apparatus of utilizing, however, the same general prinFeed enters through conduit 30 and tangentially through port 31 into upper 32 wherein it rotates and forms the two . ----------------- The upper els through base outlet section 34 into overflow sipated by an abrupt change in its direction, the overflow leaving through pipe 36 which connects with a suction leg of the same design as that shown under Fig. 2, with similar entrainment of air. In ordinary operation the discharge through the base outlet section 34 does not completely fill chamber 35. A partial vacuum varying with the vacuum in base outlet section 34 therefore normally exists in the central portion 50 of this chamber and is taken off by vacuum probe 37 through vacuum probe line 37α to conventional vacuum relay 52 of the mechanical type. Compressed air is supplied to this relay at inlet 38. Fluctuations in the vacuum in line 37α conancing the balanced proportions of compressed air or oil which is supplied through lines 39 and 48 to opposed ends of the ram. Provision is made for limiting travel of the ram during this balances the air or oil in lines 39 and 48 when the ram has travelled a predetermined distance. Maintenance of the desired vacuum in line 37α isjrffected by adjusting control knob 50 which ing diaphragm of the apparatus. The air is exhausted from the relay in a pulsating stream at orifice 49 corresponding to pulsations in vacuum line 37α. The ram housing 41 is mounted on ad....Mu,?.. α,,υΛ i«, -θ justable jack 42, so that it may be adjusted to

The discharge through orifice 11 is a slurry ⁴^{* i-}' containing entrained air rapidly rotating about its central air column. It passes into conduits 6, 7 and 8, provided with air-tight couplings 26 chamber 35 where the central air column is dis45 line 37α. The ram housing 41 is mounted on adthe optimum operating position of apex orifice plate valve 43. This valve is mounted on shaft 44 which passes through apex spray shield conduit 45 within gland 46 and opens and closes wherein the central air column is dissipated, 75 apex orifice 47, IZm^shTg SaXly Ξ

8,648,433 same result in substantially the same manner as the sphincter valve of Fig. 2. The discharge from orifice 47 and spray shield 45 flows through transfer conduit 51 to waste or to storage as may be desired.

Any type of vacuum probe may be used so long as it will permit determination of the pressure differential in the central air column without undue disturbance of the vortex. As shown in the figures this may be an open ended pipe of j relatively small diameter, but is not limited thereto, and includes vacuum probes of other types including the electric diaphragm type. Preferably it enters the cyclone centrally, that is, on an axis of the base orifice.

The position of the sensitive portion of the probe is of greater interest. It need not terminate in the same positions shown in Figs. 2 and 3. All that is necessary is that it terminate or have its vacuum sensitive part located at such a point that it will indicate the pressure differential in the central air column of the cyclone. In Fig. 3, the probe is shown as terminating in the air space formed in the center of chamber 35, as it has been found that the air pressure here varies very closely with variations in the pressure in the central air column in the vortex finder. Accordingly, this position satisfies the requirement for most operations. Unusually turbulent flow of liquid through the overflow chamber 35 occasionally results in the development of back pressure in the probe, and it has been discovered that the ideal location for the probe, suitable for virtually all types of operations, is within the vortex finder itself. Ordinarily it is not advantageous to extend the probe so that it passes completely through the vortex finder and into the body of the cone as the same back pressure noted above occasionally develops with the probe in this position. A preferred location is shown in Fig. 2 where the probe terminates about one-half a diameter of the vortex finder below the outlet end of the vortex finder. In liquid cyclones without vortex finders the sensitive portion of the probe is preferably terminated on the axis of the base passageway within the central air core, and slightly above the opening of that passageway into the hydraulic cyclone, and in no event, of course, should the pressure responsive portion of the probe terminate in a current of slurry.

The vacuum relays are not critical in the instant invention. They are conventional and may be replaced by any device which will cause the apex valve to open and close somewhat with changes of vacuum in the central air column. In fact in small cyclones the vacuum relay may be omitted entirely and the apex valve, which may be of the iris type, may be actuated by a mechanical linkage operated by the vacuum probe. In larger installations, the relay becomes necessary.

One preferred form of relay, specially suitable for operating the sphincter valve shown in Fig. 2, is constructed by bringing the vacuum line to one end of an open-ended mercury manometer connected by pneumatic and mechanical linkages which operates a controlling and recording controller, such as a Foxboro Model 40 Stabilog Controller provided with manual control sub-panel and followed by a suitable air booster to give the required range of output pressure. This controller, when operating on an air supply of about 80 p. s. i. pressure, rapidly maintains the sphincter valve at such aperatures that the vortex

G5 vacuum is maintained very constant with an almost unnoticeable degree of hunting. Another alternative controlling device is that shown in Fig. 3. In devices of the latter type, any change in the air pressure on the measuring diaphragm is communicated to a delicately controlled multiplying valve. When the forces on the measuring diaphragm change, full air or oil pressure is thrown on the power piston moving it to a new position until the measuring diaphragm is once more in a state of equilibrium. One such device is the widely-known commercial Republic Pneumatic Type Regulator Series “65” which appears to have the advantage of reliability but which also appears to be incapable of producing as rapid control as the Foxboro pneumatic relay system described above. The Republic flowmeter is described at pages 1-3 of Section E of “Data Book No. S-13” copyrighted in 1938 by Republic Flow Meters Co.

While the automatic apparatus discussed above gives the best results and is therefore preferred, not all of this equipment is required to carry out the essentials of the instant invention. As stated, it is possible to practice the instant invention, manually, and this may become necessary during an emergency. To permit manual operation, the apex valve is made manually operable, the relay system is shut off and the vacuum probe is connected directly to an open ended manometer or equivalent vacuum gauge. The apex valve is then manually controlled to maintain the pressure differential in the central air column substantially constant at a predetermined value. With sufficient practice a surprisingly uniform discharge can be thus obtained, and it is a particular advantage that by use of the vacuum gauge it is not necessary for the operator to watch the apex product at all. Manual operation, while feasible, requires close supervision and ordinarily is not preferred as fully automatic operation gives more uniform results with elimination of one factor of expense.

The maximum degree of vacuum developed in the central air column by the suction leg or ' other vacuum producing device when the apex valve is closed is not critical as all that is necessary is that there be a measurable difference between the degree of partial vacuum existing when the cyclone is being operated normally and when ' the central air column is blocked or plugged. It is difficult to maintain an adequate degree of control when this difference amounts to only a few inches of water. Ordinarily, the maximum vacuum should be at least about six inches of water as this gives a sufficiently large measurable difference when the normal operating pressure differential is about 2.5 inches of water. Much better results are obtained when the maximum vacuum is increased to about 70 inches of water since the slope of the curve in this range is steeper. When discharges of a high degree of consistency are desired, a maximum degree of vacuum of about 125 to 150 inches of water is usually found preferable because with these higher vacua, the curve rises much more abruptly, a large difference in the vacuum corresponding to only minor differences in the density of the apex discharge and much more precise control becomes possible. No limit, can be set on the maximum degree of vacuum which should be provided for any one installation as this depends upon such variables as the specific gravity of the apex discharge, size of the particles in the feed, the boiling point of the liquid component of the slurry, and the tem75

2,848,483 io lialf-opened position, if it at brie extreme of ^the other of closure, it should of Course be replaced with a valVe of more suitable size. The presstire differential in the behtfai air coltuiiri is theil noted and the automatic regulator ciit in to mairitairi this differential cChstaiit thereafter . During this procedure of adjustment the apex and the overflow discharges may be combined and be returned to the feed by subsidiary circuits which are no part of the present invention.

hi the draftings the cyclone has been shown as mounted vertically. Since the separatory efficiency of the cyclone does hot depend on gravity but results from the application of centrifugal forces, the body of the cyclone iriay be mounted m any plane. The stiction leg, however, should preferably extend in a generally downward direction.

^In addition to its use in densifying arid thickening cyclones, aS described above, the procedure ^{oi the} Present invention is also of value in the control of Cyclone separators,- for exaihple Such W ^{described in} Driessen et al. Dutch Patent No. 58,482 expiring October 16, 1964, particularly Where the density of the desired discharge lies oh a sufficiently steep portion of the curve of the type of Fig. 1. The applicability of the instant procedure to these separatory Cyclones where a medium is used, is influenced by such factors as the specific gravity, particle Size and concentration of the medium solids as well as- of couise, by the character of the coal, ore or other matter in the medium. For example,’ when a ferrosilicon medium of low gravity and thin consistency is employed, the density of the apex discharge may be on the straight portion of the curve, but when magnetite is added to such a feed to give a medium of a greater consistency containing a greater proportion of particles, fully automatic Control on the steep portion of the slope usually becomes possible.

In the drawings, only conical cyclones have been shown, but it will be understood that the invention may be practised in cyclones of the cylindrical type as well. Accordingly, in the nnrrinn ^thf “^apex” ^{is used to} designate the portion of the cone or cylinder through which the heayier fraction of the feed is discharged Alsq, m the drawings and in the foregoing specification, a. conventional hydraulic cyclone has lieen ^escribed with a conventional tubular exit from the base. In the drawings, such an exit and element S in Fig. 1 and element 34 in Fig 3 sometimes are referred to in the industry'as ^flnders”; but the description of the base orifice passageway or tubular exit from the base that embodies the passageway, is not intended to restrict the invention to the use of vortex finders since the invention is operable when embodied ^{a hydraullc} cyclone Without a vortex finder.

. w.^{lth or} without a vortex- finder,· the essential· partial vacuum will appear iri the air core at cdridltiofik ^{the 0386} °^{UtlGt Undei}’^{the disclosed}