US3096275A - Method for separating dirt from aqueous suspensions of pulp fibers - Google Patents

Description

e. H. TOMLINSON 1| 3,096,275 METHOD FOR SEPARATING DIRT FROM AQUEOUS SUSPENSIONS 0F PULP FIBERS 3 Sheets-Sheet 1 July 2, 1963 Filed Sept. 26, 1961 II I! J7 15 N I I I 7 I 9 I j 88 A INVENTOR GEO/F85 75/74/N60M [I Y 74 //I\\ BY zwfv l ATTORNEYS July 2, 1963 e. H. TOMLINSON n 3,

METHOD FOR SEPARATING DIRT FROM AQUEOUS SUSPENSIONS OF PULP FIBERS Filed Sept. 26, 1961 5 Sheets-Sheet 2 H5SUI l 1 l & 4 I 30 I f V l I l t I k I' l l l l mazwrzr I I I 8/ I z* I l 10- -r I l INVENTOR 65am: fa.

BY v fiirroamssgs J y 1963 G. H. TOMLINSON n METHOD FOR SEPARATING DIRT FROM AQUEOUS SUSPENSIONS OF PULP FIBERS 5 Sheets-Sheet 3 Filed Sept. 26, 1961 E N w M a m H wr "W 25 g M r N 5 8, m 4 m 4 5 I I II I I J CENT/7f 0F INVENTOR I w/ H.

ATTORNEYfi United States Patent 3,13%,275 METHOD FOR SEPARATING DIRT FROM AQUEOUS SUSPENSIONS 0F PULP FIBERS George H. Tomlinson 1!, Long Sauit, Ontario, Canada Filed Sept. 26, 1961, Ser. No. 140,702 3 Claims. (Cl. 2092) This invention relates to a new and elficient method for cleaning an aqueous suspension of pulp or papermaking fibres whereby objectionable foreign material, including so-called dirt, is removed by means of combined centrifugal force and shear action in a device which can be classed as a vortex type separator, and this application is a continuation-in-part of my earlier filed application, Serial No. 363,277, filed June 22, 1953, now Patent No. 3,037,628 issued June 5, 1962, which was a continuation-in-part of my earlier filed application, Serial No. 205,655, filed January 12, 1951, now abandoned.

In the conversion of wood or other fibrous material to pulp and paper three main classes of foreign matter or dirt are encountered.

(1) Portions of the wood which have not completely broken down into individual fibres as a result of the chemical and/ or mechanical action. A particle of such material may be relatively large such as an undefibred chip or knot, but when broken down to size such as might be imbedded in a sheet of paper it is referred to as a shive.

(2) Pieces of bark which do not disintegrate in cooking and remain as black or dark coloured specks in the fin al pulp.

(3) Foreign matter of non-fibrous origin that enters the system with the wood or at any subsequent stage, this including sand, pipe-scale, fly-ash, cinders, etc.

On account of the extremely divergent physical character and size of the dirt particles that may be present many varied methods of cleaning cellulosic pulp have been devised and have been conventionally employed in combination in order to obtain a final product which is reasonably uncontaminated with foreign materials that would mar the appearance and serviceability of the final sheet of pulp or paper.

Prior to the present invention the most frequently used method of cleaning cellulosic pulp was screening a dilute aqueous suspension of the pulp. However, in order to prevent plugging and matting it is necessary to use slots or holes many times greater than the fibre width. Inasmuch as screens are not completely selective in their action, multiple screening has been often used, but in spite of this, some shive normally carries through. By cooking the pulp using relatively drastic conditions the quantity of initial shive is diminished, and in bleaching, further chemical cleaning of shive is obtained due to removal of li-gnin which binds the fibres together. In addition to chemical cleaning, mechanical attrition results from the action in pumps through the pulp mill, as well as in the beater, jordan or other refining equipment in the paper mill whereby many of the individual shives are torn apart. By a balance of multiple screening with chemical and mechanical defibring it has been possible to control shive content in the final fine paper to a reasonably low level, with the chemical cleaning normally acting as the control point-that is, the amount of chemical or temperature or time at the cooking or bleaching stages being increased when the shive count appears to Patented July 2, 1963 be unreasonably high at any point along the system, this normally resulting in lower pulp yields and sometimes in excessive degradation of the pulp.

The control of bark dirt has been an even more serious problem. Whereas the shive is normally somewhat longer than it is wide this enhancing its chance of being rejected by the screen, bark dirt tends to be random in shape, approaching a sphere. Moreover, it is less susceptible to bleaching out. Thus conventionally, to obtain a fine paper free of bark dirt, it has been necessary to take extra precautions with the wood used. This has meant recycling of logs through the barking drums to remove all traces of adhering bark. However, logs contain ingrown bark around knots, etc., and many mills have restorted to treatment with rossers, Woodpeckers, etc., to remove both wood and bark around the most seriously affected areas.

The minimization of high specific gravity extraneous dirt has been normally accomplished by taking precautions to avoid its entry to the system by providing for thorough washing of the logs and filtration of process water, etc., by riffling in the pulp mill, and by the use of stationary cylindrical centrifugal cleaning tubes in the paper mill, all three methods usually being used in combination.

It is thus apparent that through direct and indirect expense the cost involved in producing a clean sheet of pulp or paper is extremely high. The individual conventional means of cleaning are normally specific for one type and/ or size of dirt, and even for that type and/or size the etficiency is often low, while for types and sizes of dirt other than that for which the equipment was designed, the efficiency may be essentially zero.

By the methods of the present invention shive, bark dirt and high specific gravity dirt can all be removed simultaneously from the fibre, and, within certain wide limitations, this is accomplished regardless of the size of the particles. Furthermore, by the choice of certain dimensions of the individual cleaning units, which will modify the flow of the separated dirt within the unit, it has been found possible to extend the range in .the size of particles removed in one direction or the other as may be desired.

This combination of results, which has been obtained in a vortex type separator, has never been described as having been accomplished by any previously available means or device.

Prior to the present invention it had not been considered possible to efifectively separate shives from pulp by means of centrifugal cleaners. Sampson and Group in U.S. Patent 2,377,524, who developed the hydrocyclone, a conical centrifugal unit somewhat similar to a gas cyclone in general shape, for continuously removing dirt from aqueous cellulosic pulp suspensions, stated that the shive components of the dirt have a reaction to centrifugal force which is so nearly the same as that of the individual fibers that they are hardly separable by centrifugal action from the regular fibres. In one of their most efiicient cleaning units, which had a diameter of 3.08 inches and a capacity of 20 gallons per minute, they noted that 44.7% of the shives which were not disinte grated due to mechanical action were rejected with 23.5% of the good fibres. Thus 76.5% acceptable pulp contained 55.3% of the shives. From this it follows that parts accepted fiber would contain of the non-disintegratable shive carried from the initial pulp indicating that such equipment is obviously unsuitable for producing a shive-free pulp.

According to the teachings of said U.S. patent a dirt particle is subjected to centrifugal force which Carries it to the wall where it advances to the apex of the cone and is rejected. However, opposing this effect is the inward radial flow so that material not readily separated by centrifugal force will be carried in to the center and out the top outlet With the accepted stock. Thus they concluded that the only way to improve the efiiciency of the unit was to use a unit of decreased diameter which increases the angular velocity and therefore the centrifugal force and to increase cone length which decreases the inward radial flow. They concluded that a useful unit for pulp cleaning would be in the size range of 2.0 to 4.2 inches diameter and would have a length of to 15 times this diameter.

I have now found that excellent removal of shive, to a degree considered impossible in the prior art, is achieved in a hydrocyclone unit having a markedly greater diameter, and hence considerably lesser centrifugal force, than those obtained in said U.S. patent.

This surprising finding results from the previously unrecognized fact that shive and other large light particles, particularly of elongated shape, are drawn to the center of small diameter units as a result of the high velocity gradient or shear which prevails, the particles passing to zones of higher velocity gradient which are towards the center of the unit, in spite of the high opposing centrifugal force.

I have found, more generally, that when an aqueous suspension of fibres is introduced tangentially into a conical chamber, such as a hydrocyclone, to produce a whirling motion about a central axis, the suspension as it whirls inwardly and downwardly forms a conical vortex wherein two zones are clearly distinguishable; an outer zone characterized by angular velocities increasing inwardly along the radius and an inner zone wherein the angular velocity of the liquid is substantially constant. The shear forces arising from the angular velocity gradient are instrumental in separating the wood dirt or shive from the fibres. The shive thus freed from the fibres are carried by the centrifugal forces, simultaneously developed, toward the wall of the conical chamber whence it is carried to the apex of the cone and eventually rejected. I have found that effective separation of shive from the fibre 'and subsequent ejection of shive through the cone apex will occur only where the centrifugal force and angular velocity gradient generated in the vortex bear a ratio to one another above a certain minimum value.

I have further found that the magnitude of the centrifugal force and the angular velocity gradient thus generated, and consequently the ratio of such forces to one another, are determined by the geometrical variables of the conical chamber, viz, diameter of chamber, cone angle, inlet diameter, top outlet diameter and that there is a unique relationship between any particular combination of values for such variables and the ratio of centrifugal force to angular velocity gradient, and hence the separation efliciency of the chamber.

Accordingly, the present invention provides a continuous method of separating, from an aqueous suspension of cellulose fibres, dirt particles of Widely varying sizes including those having essentially the same specific gravity as the cellulose fibres, wherein the suspension is introduced under pressure tangentially into the wider end of a conical separation chamber thereby to produce a vortex of conical form, said vortex having an outer zone of radially inwardly increasing angular velocities and an inner zone wherein the angular velocity is constant, the centrifugal forces and the shear forces being at maximum in the outer zone of increasing angular velocities at a point close to the radius of transition between the outer zone and the zone of constant angular velocity, the ratio of the centrifugal forces to the angular velocity gradient at said maximum having a value of 16 1O or greater, and the angular velocity gradient having a value of 600 'sec.- ft. or greater.

Proceeding now to a more detailed description of the invention reference will be had to the accompanying drawings, in which:

FIGURE 1 is a View, partly in vertical section and partly in side elevation, of one type of a vortex separator embodying my invention;

FIGURES 2 and 3 are sectional views showing the applications of conical fittings to vary the effective diameter of the dirt discharge aperture of the conical separating chamber forming part of the apparatus shown in FIG- URE 1.

FIGURES 4, 5, 6 and 7 show the variation with radius within the unit of pressure, velocity, centrifugal force and angular velocity gradient respectively in three selected units.

The separator shown in FIGURE 1 comprises a relatively large diameter hollow truncated cone section 5 of a length considerably greater than its maximum diameter. The upper or larger end of cone section 5 is joined to a cylindrical head section 7 provided with a tangential stock inlet 8, preferably of rectangular cross section, elongated in the direction of the axis of the unit. A cleaned stock outlet pipe 9 extends below the level of the tangential stock inlet 8 in axial alignment with rejects discharge aperture 6, situated at the truncated apex of the cone, to provide a stock outlet of greater diameter than said rejects discharge aperture.

The cleaned stock may be delivered through outlet pipe 9, pipe reducer 15 and elbow 16 to a throttling valve 17 and thence, through suitable pipe connections 19, to any suitable point of delivery such as to the next process stage. A pressure gauge 20 may be provided immediately ahead of throttling valve 17.

The included cone angle of cone section 5 is an important factor and may be of the order of 10 to 18. When the area of stock inlet 8 is approximately 12.5 square inches and the inside diameter of cylindrical head section 7 is 12 inches, the inside diameter of cone section 5 at aperture 6 should be approximately 2 inches. For certain purposes, a smaller diameter at aperture 6 may be desirable. In this event, separately formed flanged conical fittings 12 and 13 may be provided and selectively secured to flange 14 of cone section 5, as shown in FIG- URES 2 and 3. The minimum internal diameter of the cones of conical fittings 12 and 13 are 1.75 inches and 1.5 inches respectively so that an appropriate rejection outlet size may be provided, while maintaining the angle of the cone down to the smaller diameter. In like man ner, a rejects outlet diameter of 1.25 inches or any other desired size may be obtained by the use of additional fittings similar to the fittings 12 and 13. The stock outlet pipe 9 extends to the lower end of cylindrical head section 7. The internal cross section of outlet pipe 9 may be approximately equal to the cross section of inlet 8, this corresponding in the present instance to a diameter ofapproximately 4 inches. When stock is pumped through this unit a liquid-free column 21 forms at the axis extending between apex outlet 6 and outlet pipe 9, this column being surrounded by liquid stock suspension 22 which is in rapid rotation about this column.

When unrifiied and unscreened sulphite pulp at a pressure of 45 pounds per square inch was supplied to the unit it was found to have a capacity of 705 U.S. gallons per minute. When connected with a 1.25 inch apex outlet, and operating without back pressure, it was found to repect 4.0 U.S. gallons per minute at a consistency of 1.2%, this discharge being in the form of a hollow cone spray having an included angle of and amounting to 0.57% of the inlet volume and containing 1.31% of the total entry solids. This remarkable small flow through a 1.25 inch outlet results from the fact that the region surrounding the central axis is completely void of liquid, this resulting from the high centrifugal force within the unit. The rejection orifice is normally set to a diameter just larger than that of the liquid-free column, thus minimizing the quantity of rejects which must be rehandled, yet being of sufficient size to prevent pluggage. The accepted stock, in the foregoing example, was found to be somewhat cleaner than that obtained with the use of the conventional riffling and fine screen system which was running in parallel, using identical feed stock.

It was found that when the 15 conical section was replaced with one of it was necessary to increase the apex orifice diameter to 1.5 inches in order to obtain a rejection discharge. With this modification cleaning efficiency was improved, and the capacity of the unit increased from 705 to 800 U.S. gallons per minute.

An installation of vortex cleaners of the above dimensions, using the 10 cone, was made in an alkaline process pulp mill located at Cornwall, Ontario, producing 150 to ZOO-tons pulp per day from mixed hardwood. In this installation the stock from the brown stock washers was passed through a coarse screening system and then through three vortex cleaners connected in parallel. The rejects from the coarse screening system and the vortex cleaners were combined, diluted, and passed through another vortex cleaner with the accepted stock being recycled through the system and the rejects sewered. The stock was then thickened, bleached and next passed through a second bank of four cleaners in parallel, with the rejects being diluted and pumped to a secondary cleaner, the accepted stock being returned to the system and the rejected stock being sewered. The total sewered rejects at the unbleached system amounted to 0.15% by weight of the accepted stock and consisted of shive, bark dirt and miscellaneous gritty material together with a small amount of fiber. The total rejection at the bleached stage amounted to 0.067% of the accepted stock.

After the above installation was made it was found that the previously used conventional system for handling the unbleached screen rejects consisting of a rotary screen and a flat screen was no longer required. This was also the case with the bleached stock rifliers and fine screens (primary and secondary systems) and all of this equipment was removed from the plant. Hourly dirt count tests in this mill, taken over an eight month period following the installation, showed an average count of 1 square millimeter equivalent black area per 1000 square inches of pulp sheet surface with only 2.9% of the counts being above 4. In the corresponding period prior to this installation none of the counts were as low as 1 and 98% were above 4. Because of the high cleanliness with the new system it was found possible to use 14% less chemical at the digester than previously inasmuch as it was no longer necessary to periodically raise the charge for dirt control.

With conventional operation, it had been necessary to use only wood from which the bark had been removed and which had been carefully cleaned. It was found that with the vortex separators in operation, totally unbarked wood could be successfully used and that essentially all of the bark particles were removed. With unbarked wood, the total rejection rate at the unbleached stages increased from approximately 0.22% of the accepted stock to approximately 1.66% of the accepted stock due to the large number of bark particles.

It has been found that the type of rejection of the vortex cleaners is dependent, in an inter-related way, on a number of the dimensions of the unit which include both the absolute and relative size of the inlet and outlet connections, the diameter and the cone angle. It has been possible to study the nature of the motion within the unit and its effect on entrained solid material, all as related to the dimensions involved so that it thus becomes possible to predict the type of action which may be obtained with any proposed unit which may be built.

When a single large object, such as a woody knot, say /2 inch x /2 inch x inch enters the unit described above it is not rejected but orbits against the cone wall until it is worn by attrition to a size that will reject. By using a transparent cone it was possible to measure the velocity of the orbiting knot. The orbit established was specific for its size and as the knot was worn away its position advanced towards the apex of the cone. A rubber Number 2 laboratory stopper, which showed similar behavior but without changing orbit due to attrition, was introduced with the feed to a 12 inch diameter unit having a 12.5 square'inch inlet and a 4 inch diameter top outlet and a 10 cone, this being operated at 45 pounds per square inch inlet pressure and without pressure at the outlet. The throughput was 800 U.S. gallons per minute, from which its calculated velocity, at point of entry, was 20.5 feet per second. Thet stopper was found to orbit at the point where the cone was 1.8 inches in radius.

The velocity of the orbiting stopper, as measured with a strobortac, was 4,300 r.p.rn. corresponding to a velocity of 67.6 feet per second at the cone wall.

When -a body of a fluid rotates about the central axis of a vessel the extreme cases of either a free vortex or a forced vortex may develop depending on the conditions that initiate and maintain the whirl. In the free vortex both the angular and tangential velocities increase as the axis of rotation is approached, according to the relationship.

where V and V are the tangential velocities at radii R and R respectively.

In the forced vortex the angular velocity is constant, that is, the fluid rotates as though it were a solid wheel, with the tangential velocity decreasing as the axis is approached. The corresponding relationship is From the above equations, it was possible to calculate the theoretical velocity at 1.8 inches radius for the two types of flow assuming that the tangential velocity at the outside of the cylindrical section (having a radius of 6 inches) is the same as the entry velocity, namely 20.5 feet per second. Also since the orbiting stopper occupied a wide radius band it was not certain as to the actual radius at which the liquid velocity coincided with that of the velocity of the stopper, but for the purpose of this calculation it was taken at the cone wall. A comparison of the experimental value with those calculated for free and forced vortices are shown in Table I.

Comparison of the data of Table I shows that the remarkable acceleration of the fluid can be explained by the formation of a free vortex in the unit. In Table I the centrifugal force values were calculated according to the equation,

where F=the centrifugal force in number of gravities V=velocity in feet per second R=radius in feet g=acceleration due to gravity=32.2 feet per second At a radius of 1.8 inches the calculated value of centrifugal force is 968 with free vortex flow compared with only 7.9 for forced vortex flow illustrating the magnitude of the separating forces that can be developed in this type of equipment.

Due to the increasing velocity obtained with free vortex flow considerable shear forces develop as the fluid spirals toward the axis. The shear forces arise as a consequence of the angular velocity gradient S defined as:

dw 21rdR (4) where w is the angular velocity in radians per second and R is as defined in Equation 3.

It is to be understood that the more common definition of the angular velocity gradient is given by the formula dw S d R simply by multiplying the given values by 6.28.

The desirability of shear can best be appreciated from a consideration of the results obtained when centrifugal force only is applied. When a suspension of pulp fibres with randomly distributed dirt of the character normally present is placed in a laboratory centrifuge tube and subjected to a high centrifugal field, high specific gravity material such as sand will settle at the lower end of the tube, and while both the fibre and wood dirt concentrate towards the end of the tube, the wood dirt remains randomly distributed amongst the fibre. That is, the migra tion of wood dirt through the fibre is largely prevented in the absence of shear. By contrast, in the apparatus of the present invention a high concentration of such wood dirt results from the shear at right angles to the centrifugal field which is characteristic of free-vortex flow, and which breaks up the fibre flocks, allowing such material to work through the fibre.

The shear forces also play an important role in keeping the pulp fibre from centrifuging out from the suspension. The fibres contained in pulp stocks are of appreciable size, approximately 20 to 30 microns in diameter and 1000 to 3000 microns in length, with their mass approximating that of a solid sphere of 65 microns diameter. It might be expected that due to the high centrifugal forces developed in the equipment of the present invention they would centrifuge out, and in fact there is considerable separation of fibre as indicated by the increased consistency of the rejects discharge. Fortunately, however, it is possible to control the apparatus so that the major portion of the fibre is carried in the upward helical flow. It is believed that this results from the flexibility and shape of the fibres, with their high ratio of length to diameter. Thus their tendency to settle towards the wall is partially counter-balanced by the tendency of the ends of the fibres to be caught in the more rapidly rotating inward spirals which carry them toward the centre of the unit. The shear elfect is most marked and concentrated in regions of small diameter.

The basic flow pattern Within the unit is as follows:

The stock, entering under pressure initiates a swirling motion in the whole unit, the liquid simultaneously spiralling inwardly and downwardly at increasing velocity. An upflowing helical stream, having a maximum diameter approximately equal to that of the accepted stock outlet pipe, develops about an air column which forms at the central axis, this upfiowing stream being maintained by a liquid flow transferring from the downward spiral flow. The apex outlet size, when set to a diameter just larger than the air column, allows for the fractionation or classification to take place. This fractionation depends on the fact that the material being removed follows a different pattern than the main liquid stream. Under normal operating conditions most of the fibre follows the main liquid stream. Particles such as fly ash or bark, under the influence of centrifugal force, work their way through the pulp, which is in a non-flocculated state because of shear, to the cone wall, where they are carried by the downward component of flow to the apex outlet where they are rejected. Heavy objects such as knots are also carried to the cone wall and are carried down by the downward liquid flow. However, as they approach the apex the high centrifugal force developed tends to push them back to zones of greater radius so that an equilibrium orbit is established.

With shive, a difierent pattern occurs. Whereas the presence of a shear field is essential to the primary separation of the dirt from the fibre, it can also give rise to re-entrainment towards the apex of the cone. Shive normally also has a somewhat greater length than thickness or breadth and in a unit such as that described it tends to be drawn into the up-flowing stream from regions of high shear but because of its greater mass as compared with that of the individual fibres its tendency to be thrown out from the upfiowing column towards the wall is very much greater. The tangential momentum developed by the shive in the high centrifugal field of the up-flowing column is generally sufiicient to carry it out towards the cone wall against the in-flowing stock and through regions of high shear to regions of lesser shear at the wall whence it carries down and is eventually rejected. This recycling of the shive within the unit results in an increased concentration as compared with the feed to the unit in spite of the fact that its mass is too small to allow the establishment of an orbit of the type described for knots. Thus a counterbalance is set up between the shear effect drawing the concentrated shive into the centre column, from the region of the apex of the cone, and the centrifugal action which throws it back to the wall. If the shear force is relatively great, in relation to the centrifugal force, a high concentration of shive develops, and some carries through in the accepted stock. By modification of the dimensions of the unit the relation between the centrifugal and shear forces can be varied thus increasing or decreasing the efliciency of separation of such material which for certain units may be zero.

The movement of the shive into zones of high shear appears to result from the Magnus effect as referred to by D. L. Streeter in Fluid Dynamics p. 143 (McGraw-Hill, first edition 1948). By this elfect a rotating cylinder will move at right angles to the direction of fluid flow as a result of the velocity gradient obtained by this motion. The Flettner rotor ship, with circular rotating cylinders with axis vertical to the ship, operates on this principle. The air flowing past the cylinder on the side rotating in the direction of the Wind moves at a faster rate than that on the opposite side, which is turning against the wind, this resulting in a thrust into the direction of high air velocity, i.e., at right angles to the direction of the wind. In the case of pulp suspensions in the vortex cleaners the shive is rotated as a result of the shear not only in its axial direction, but also end over end, etc. and in zones with very high shear rates in relation to shive size the thrust obtained can be greater than that obtained by the centrifugal force.

With a 12 inch diameter unit with 2 inch diameter inlet and top outlet connections and a 10 cone small shive of A inch length and inch thickness, and smaller was rejected with very high efficiency. However, unlike the unit previously described which rejected large shive with high efficiency, shive having a length of A1 to /2 inch and a thickness of /s to 7 inch, as it approached the apex, was carried back into the up-flowing column from which it recycled to the cone wall as described above with about 10% being accepted and the balance being rejected. Still larger sizes of shive were all carried in the accepted stock.

However, a unit of similar diameter, and cone angle, and having the same 2 inch diameter inlet, but with a four inch diameter top outlet gave excellent rejection of large shive with relatively little, if any, recycling of the shive to the up-flowing column. When the velocity was measured at 1.8 inches radius, using the method previously described, it was found that the velocity was materially less than for the similar unit but having a 12.76 square inch inlet and that the free vortex showed a transition to forced vortex in the general region in which the measurement was being made. Unfortunately it was diflicult to accurately establish the true transition radius with this technique because of the relatively wide radius band occupied by the orbiting rubber stopper.

It was found possible, by means of a different technique, to establish the flow pattern with somewhat greater accuracy. Prandtl and Tietjens, Fundamentals of Hydro and Aeromechanics, p. 214 (1st edition, McGraw-Hill, 1934) show that the change in pressure as a function of radius for a free vortex surrounding a forced vortex is R is the corresponding radius, V is the corresponding tangential velocity, and p is the density.

Test cleaners were fitted with taps for pressure measurement placed down the full length of the cone and the values of pressure as obtained by means of a mercury column were plotted against the radius. Typical curves as obtained with 40 pounds per square inch on the header are shown in FIGURE 4. Each of these curves is convex at large radii, such curve shape being typical of the pressure profile for free vortex motion; at smaller radii the curves show an inflection point followed by a concave section, a concave shape being typical of the pressure profile for forced vortex motion. The three curves differ appreciably in shape and in the value of the radius at which the transition from free vortex to forced vortex motion occurs, the value of the transition radius being largest with the unit having the two inch diameter inlet and four inch diameter outlet, and smallest with the unit having the two inch diameter inlet and two inch diameter outlet. At still smaller radii, in each unit, the pressure decreases to zero and thereafter the central axis of the unit is void of liquid.

When the pressure-radius curves of FIGURE 4 are analyzed in accordance with Equation 4 the velocityradius curves shown in FIGURE 5 are obtained. In each case, the velocity increases exponentially as the radius decreases reaching a maximum value at the transition radius and then decreases linearly as predicted by the forced vortex equation.

It was found with this improved technique that the velocity-radius function does not follow the free vortex equation exactly, i.e.,

1 1= a 2 the departure being due to axial and radial motion, fricl 9 tion and liquid viscosity, all of which reduce the velocity below that predicted by the free vortex equation. The free vortex equation can be rewritten as:

where 11" is the velocity exponent which has a value of 1 for an ideal free vortex and 'l for a forced vortex. The velocity exponents as obtained by fitting the data to Equation 6 for the three units are given in Table II, line 11. I

Similarly, the peripheral velocity, as calculated from the pressure-radius relationship was found to be somewhat lower than the entry velocity, the latter being calculated from the size of the entry connection and the volumetric throughput of the liquid. The relationship is given in the following equation:

V,=CV (7) where V, and V are inlet and peripheral velocities respectively and C is the inlet velocity coefficient. Values for the inlet velocity coefficient are given on line 5 of Table II.

'From the tangential velocity-radius relationship the centrifugal force and angular velocity gradient can be calculated using Equations 3 and 4 respectively. The relationship between centrifugal force and radius and between angular velocity gradient and radius are illustrated in FIGURES 6 and 7 respectively.

The principal data of FIGURES 4 to 7 inclusive are summarized in Table II. Comparing two units with the same ratio of inlet to outlet, unit A having a 4 inch diameter inlet and outlet and unit B having a 2 inch diameter inlet and outlet it will be noted, that unit B (with the smaller inlet-outlet connections) has markedly smaller throughput, much lower values of velocity, centrifugal force and angular velocity gradient at the periphery. However, as a consequence of the low peripheral velocity, the value of the radius at which transition between the free vortex and forced vortex zones occurs is only half that of unit A i.e., 0.6 inch compared to 1.2 inches and since the tangential velocities at the transition radius in this case are nearly the same the maximum value of the centrifugal force in unit B is approximately twice that of unit A and the maximum value of the angular velocity gradient is approximately four times as great. Thus the ratio of centrifugal force to shear in unit B is approximately one-half that of unit A, i.e. 29X 10 compared to 61 x 10- As previously noted, unit B rejected bark particles and small shive with extremely high efficiency, gave intermediate efliciency on medium size shive 'which recycles between the inner and outer stream, and zero efficiency for large shive, which, being drawn into the upgoing stream by the shear forces, cannot penetrate through the zones of high shear and therefore carries with the accepted stock from the top of the unit.

In contrast to this, unit A having lower centrifugal force but also having lower shear and higher ratio of centrifugal force to angular velocity gradient, rejects all sizes of shive with good efliciency though its efficiency on very small sizes of dirt is somewhat less than that of unit B.

The effect of increasing the top outlet diameter in relationship to the inlet diameter can be seen by comparing unit C, with the 2 inch diameter inlet and 4 inch diameter outlet to, unit B. The peripheral velocity is increased, which has the effect of moving the transition radius away from the centre of the unit i.e., from 0.6 inch to 1.5 inches and thereby reducing the centrifugal force to about onethird, and the angular velocity gradient to one-eighth the value for unit B. The ratio of centrifugal force to angular velocity gradient is also increased by a factor of 2.5. These changes have the effect of allowing all sizes of shive to be rejected with good efficiency and with a minimum of recycle between the upward and downward streams. The efliciency on small dirt is reduced because of the lower maximum centrifugal force.

A number of units having other combinations of inlet and outlet connection diameters, maximum cone diameters and cone angles were constructed and the pressure radius relationship established when operating at 40 pounds per square inch pressure. The various relationships, as shown in Table II, were calculated for these units. The important values obtained are summarized in Table III, the three units already described being shown again as Examples 3, 1 and 6.

The data are arranged in groups of decreasing diameter, and, in each group the units are arranged in order of decreasing peripheral velocity. It can be seen that for each group and in the range examined, increasing the top outlet diameter has the effect of increasing the peripheral velocity. In each of these sub groups, decreasing the inlet size also has the effect of increasing the peripheral velocity. High peripheral velocity results in a large transition radius, a low centrifugal force, a low velocity gradient and a high ratio of centrifugal force to angular velocity gradient, with the actual magnitude of each being dependent on the diameter of the unit. For instance the ratio of centrifugal force to angular velocity gradient is greater for units of 12 inch diameter than for geometrically similar units of smaller diameter.

Dealing specifically with the 12 inch diameter unit the highest ratio of centrifugal force to shear is obtained with the unit having a 3 inch inlet and a 4 inch outlet, Example 2, Table III. This unit will give complete elimination of all sizes of shive from pulp suspensions of low fibre concentration and has excellent efliciency at fibre concentrations of about 0.75% which is a normal concentration prevailing in the manufacture of pulp and paper. It is also effective on bark and other types of dirt and has a high capacity. The value of 725 for centrifugal force and 925 sec.- ft. for angular velocity gradient are adequate. This unit is preferred to the unit with a 2 inch inlet and 4 inch outlet (Example 1) which has excellent efii ciency on all sizes of shive at low fibre concentration but because of the lower values of centrifugal force and angular velocity gradient tends to lose in efficiency rapidly as the fibre concentration is increased. It is also more effective than the unit with the 4 inch inlet and 4 inch outlet (Example 3) because of the higher ratio of centrifugal force to angular velocity gradient, i.e. 79 x10- compared to 61 *10- The unit with the 3 inch inlet and 3 inch outlet (Example 4) has a centrifugal force to angular velocity gradient ratio of 49 10- Its high centrifugal force and velocity gradient give it improved cleaning performance on types of dirt other than large shive. It is particularly adapted for use in the production of newsprint which otherwise contains considerable small shive and other dirt but does not contain very large shive.

The unit with the 2 inch diameter inlet and 2.5 inch outlet (Example 5) is a lower capacity unit and has a centrifugal force to velocity gradient ratio of 3 8 10 This unit is useful on bleached pulps and fine papers where extreme cleanliness in regard to fine particles is required and where only small shive may be present in the initial stock. This unit is preferred to the similar unit but with a 2 inch diameter outlet (Example 6) which has a rather similar value for centrifugal force and a lower ratio of centrifugal force to velocity gradient.

This last named unit, with 2 inch diameter inlet and outlet was initially used to remove conducting particles from a paper machine furnish for the production of tabulating card stock with outstanding results compared to any other cleaning device used previously for this grade of paper.

The dimensions of the 9 inch diameter unit (Example 7) were so selected as to give a similar result to that obtained with the 12 inch diameter unit with the 2 inch inlet and outlet connections referred to above (Example 6 It is also possible to achieve approximately the same results using a 6 inch diameter unit but using outlet diameters considerably larger than the inlet diameter. In

tively light in construction, are particularly useful in re-' moving chop and birdsced which are small, short chunks of undefibered wood occurring in chip groundwood as produced by refining chips, high yield semichemical pulps, etc.

Similarly it is also possible to construct units larger than 12 inch diameter having essentially the characteristics of say the unit with 3 inch inlet and 4 inch outlet. However, when this is carried to its extreme, say units of 36 inches diameter, the unit does not develop sufficient centrifugal force to allow efficient removal of shive and fine dirt. However, such units may be used for removing knots and other heavy dirt.

The effect of change in cone angle, other dimensions equal, can be seen in the case of the 6 inch diameter unit by comparison of data obtained with a 7 /2" cone (Example 11) and a 15 cone (Example 12). With the smaller cone angle the throughput is somewhat greater which has the effect of producing a high entry velocity, a larger transition radius, and therefore lower centrifugal force and shear. In these examples the ratio of centrifugal force to shear is little different in the two units.

The 3 inch diameter unit described in Example 13 of Table III having a 0.5 inch inlet, a 0.63 inch outlet and 5 cone angle was the unit selected by Sampson and Group for actual use in cleaning pulp suspensions. It was found to have a very small transition radius, 0.07 inch, high centrifugal force and extremely high velocity gradient. The ratio of centrifugal force to velocity gradient of 3.5 l0 is too low to allow removal of any shive more than inch in length.

By enlarging the outlet size to 1.06 inches (Example 14) it was found possible to improve the rejection of shive so that it is possible to reject shive having a length of A to /2 inch and a thickness of A; to inches similar to that obtained with the unit having a 12 inch diameter and fitted with a 2 inch inlet and outlet connections (Example '6). However this unit is effective only at very low fibre concentrations.

The ratio of centrifugal force to velocity gradient is greater for geometrically similar units being approximately twice as great for units of twice the diameter therefore giving better shive removal for that unit. However, with any unit size the ratio of centrifugal force to velocity gradient can be increased by increasing the outlet pipe in relationship to the inlet pipe. However, when this is carried to extremes the centrifugal force and shear values become low and the unit is ineffective for direct removal except at extreme low fibre concentrations.

When the pressure to a unit is increased the throughput, tangential velocity, centrifugal force and velocity gradient are all increased. However the location of the transition radius is not altered and within the range of 10 to 60 pounds per square inch there is no major change in the classification of shive that is obtained, although the cleaning efficiency is increased.

Generally, I have found that a centrifugal force to velocity gradient of 16 10 or higher is necessary to give adequate removal of shive from fibre in commercial pulps. Units working at a centrifugal force to velocity gradient of 60 l0 and higher will reject dirt of all type and have particularly high efficiency on large shive of the type found in unscreened unbleached pulps. Units working in the range of 25 to 60 '10 are suitable for pulps which have already been subjected to some precleaning and do not contain the largest shive sizes, being particularly effective on bark specks, on shive having a size of /4 inch length and inch thickness and even up to /2 inch length and A5 inch thickness which are difficult to bleach.

13 TABLE II Characteristics of 12 Inch Diameter Units Fitted With 10 Cones Operating at 40 Pounds per Square Inch Pressure 14 angular velocities and a second zone wherein the angular velocity is substantially constant, the first and second zones meeting at a radial distance from the central axis at which the centrifugal force and angular velocity gradient are at a maximum, the ratio of centrifugal force to angular l a Unitimot size'outlet diameter velocity gradient at said radial distance bein at a value greater than 16x10 and an angular velocity gradient n at said radial distance being less than 29,000 sec. ft? 128 2 inches 2 inches while operating at 40 p.s.i. feed pressure, and continuously inichelzlsfl 2inchcs 4i11ches withdrawing from the central portion of the vortex at no es the apex end said dirt and a minor proportion of the (1) Pressure to unit Minn? 40 4O 40 cellulose fibers and withdrawing at the end remote from r nr ou i ip t o s ais/mmn 392 3 28 the apex end a stream of suspension which contains a e V8 061 y, 5G0 (4) Peripheral eiocit ft./ec 1 98 mayor proportion of the cellulose fibers. 5) Inlet veloci y coe cien .98 .4 E Peripheralpressqre,Msecm 33.5 360 2&0 2. A method according to claim 1 wherein the ratio (7) Peripheral centrifu al force, 24 8 7 1 34 6 of centrifugal force to velocity gradient has a value withaVl 16S i a re r i i iem i angu lar velocity u 3 21 6 1n the range of 16 1()- Ito 79 1O gra ien seer (9) Centrifugal mm to Shear 20 3. The method of separating from an aqueous pulp ratio at periphery x10 124 63 16 suspension of cellulose fibers dirt particles of widely v-ary- (10) Transition radius, inche 1.20 0.60 1.56 I (11) velocity exponent 66 46 mg sizes characterized by the use of an apparatus having (12) I 498 4M 445 a diameter at the larger end thereof between 6 in. and

13 entrifu a1 orce a k c o mm menus, gramme? n 770 1490 490 12 1n., a cone angle between 7.5 and 15 an inlet diam- 14) Angular velocity gradient at eter between -1 1n. and 4 1n., an axial accepted stock outlet transition radius, sec. mi 1240 5150 too be ween 1 1n. and 4 1n., and a re ected stock outlet of (15) i iiiiigi t zi ciigiii g ir an smallersize than said accepted stock outlet, whirling the sition rad X 0 61 29 74 suspension in the separating chamber to produce a vortex 0 of conical form, said vortex having a zone of high shear 1 Equivalent to a 4-inch diameter radius. 3

TABLE III Behaviour of Vortex Cleaners 0 Various Dimensions OPERATING AT POUNDS PER SQUARE INCH PRESSURE Diameter, inches Transition radius-inches Cone Capacity Velocity Periph- Inlet Veloc- Transi- Ex, angle, at 40 p.s.i., at inlet erial velocity ity extion Ratio cen- Top deg. U.S. gals/ ft./secs. velocity, coefliponent radius Veloe- Centrifugal Velocity trifugal C one Inlet outlet nun. it./secs cient in inches ity, force gradient, force to ft./sec. gravities seeftr velocity gradient 12 2 4 10 380 38. 9 23. 6 61 46 1. 5 44. 5 490 660 74 10- 12 3 4 10 540 24. 5 21.0 86 68 1. 5 54. 0 725 925 79Xl0- l2 4 4 10 790 20.6 19. 8 98 57 1.2 49. 8 770 1, 240 61 X10- 12 3 3 10 420 19.1 15.0 79 70 90 56. 5 1, 340 2, 730 49x10- 12 2 2. 5 10 266 27. 2 14.0 51 63 72 53. 6 1, 490 3, 900 321x10 12 2 2.0 10 230 23. 6 10. 7 46 66 60 49.0 1, 490 5, 150 29x10" 9 1. 66 2.0 10 165 24. 6 15.0 61 56 51. 6 1, 990 7, 410 27x10- 8 1.5 1. 5 10 112 20. 4 11.3 .55 .66 36 49. 8 2, 560 14, 700 17Xl0' 6 1.0 1. 63 15 83 34 21. 6 63 53 54 53. 3 1, 630 6, 420 25Xl0' 6 1. 0 1. 36 15 73 29. 9 16.8 .56 45 52. 5 2, 280 9, 580 24 10- 6 1.0 1.0 7. 5 28. 7 13. 6 .48 54 33 44. 8 2, 260 14, 500 16x10- 6 1.0 1.0 15 56 23 12. 5 .54 63 27 56.8 4, 680 29, 000 17x10" 3 0. 5 .63 5 21.2 34. 6 15. 5 .45 .23 07 31.0 5, 180, 000 2. 9X10- 3 0. 5 1. 06 5 28. 0 45. 7 33.0 69 19 .35 25.0 670 3, 820 17. 5X10- I claim: forces as measured by the angular velocity gradients act- 'l. The method of separating from an aqueous pulp suspension of cellulose fibers dirt particles of Widely varying sizes characterized by the use of an apparatus comprising a separating chamber having at least a portion thereof of conical form, a tangential inlet at the larger end thereof, an axial accepted stock outlet at said larger end and an axial dirt discharge outlet at the apex of the chamber, the diameter at the larger end of said chamber being from 3 in. up to less than 36 in., the cone angle being between 5 and 18, said inlet diameter being greater than 0.5 in., and said axial accepted stock out-let being not less than 1 1n., the method comprising introducing an aqueous suspension of cellulose fibers into the tangential inlet, the suspension containing dirt particles of widely varying sizes including shive, whirling the suspension in the separating chamber to produce a vortex of conical forms having a first zone of radially increasing ing simultaneously with centrifugal force, an angular velocity gradient created in said zone not greater than 29,006 sec. ftf while operating at 40 p.s.i. feed pressure, a ratio of centrifugal force to angular velocity gradient at the point substantially of maximum value of said forces being greater than 16x10" whereby dirt particles are concentrated in the outer centrifugal field of the vortex together with a minor proportion of the cellulose fibers, and continuously withdrawing from the rejected sto'ck outlet said dirt particles including shive and a minor proportion of fiber and withdrawing from the accepted stock outlet a stream of suspension containing a major proportion of the cellulose fiber.

References Cited in the file of this patent UNITED STATES PATENTS 2,878,934 Tomlinson Mar. 24, 1959

Claims (1)

1. THE METHOD OF SEPARATING FROM AN AQUEOUS PULP SUSPENSION OF CELLULOSE FIBERS DIRT PARTICLES OF WIDELY VARYING SIZES CHARACTERIZED BY THE USE OF AN APPARATUS COMPRISING A SEPARATING CHAMBER HAVING AT LEAST A PORTION THEREOF OF CONICAL FORM, A TANGENTIAL INLET AT THE LARGER END THEREOF, AN AXIAL ACCEPTED STOCK OUTLET AT SAID LARGER END AND AN AXIAL DIRT DISCHARGE OUTLET AT THE APEX OF THE CHAMBER, THE DIAMETER AT THE LARGER END OF SAID CHAMBER BEING FROM 3 IN. UP TO LESS THAN 36 IN., THE CORE ANGLE BEING BETWEEN 5* AND 18*, SAID INLET DIAMETER BEING GREATER THAN 0.5 IN., AND SAID AXIAL ACCEPTED STOCK OUTLET BEING NOT LESS THAN 1 IN., THE METHOD COMPRISING INTRODUCING AN AQUEOUS SUSPENSION OF CELLULOSE FIBERS INTO THE TANGENTIAL INLET, THE SUSPENSION CONTAINING DIRT PARTICLES OF WIDELY VARYING SIZES INCLUDING SHIVE, WHIRLING THE SUSPENSION IN THE SEPARATING CHAMBER TO PRODUCE A VORTEX OD CONCAL FORMS HAVING A FIRST ZONE OF RADIALLY INCREASING ANGULAR VELOCITIES AND A SECOND ZONE WHEREIN THE ANGULAR VELOCITY IS SUBSTANTIALLY CONSTANT, THE FIRST AND SECOND ZONES MEETING AT A RADIAL DISTANCE FROM THE CENTRAL AXIS AT WHICH THE CENTRIFUGAL FORCE AND ANGULAR VELOCITY GRADIENT ARE AT A MAXIMUM, THE RATIO OF CENTRIFUGAL FORCE TO ANGULAR VELOCITY GRADIENT AT SAID RADIAL DISTANCE BEING AT A VALUE GREATER THAN 16X10**-2 AND AN ANGULAR VELOCITY GRADIENT AT SAID RADIAL DISTANCE BEING LESS THAN 29,000 SEC.-1FT. -1 WHILE OPERATING AT 40 P.S.I. FEED PRESSURE, AND CONTINUOUSLY WITHDRAWING FROM THE CENTRAL PORTION OF THE VORTEX AT THE APEX END SAID DIRT AND A MINOR PROPORTION OF THE CELLULOSE FIBERS AND WITHDRAWING AT THE END REMOTE FROM THE APEX END A STREAM OF SUSPENSION WHICH CONTAINS A MAJOR PROPORTION OF THE CELLULOSE FIBERS.

Images