This 911MPELY1050 laboratory (gold) shaker table has its inner deck made of aluminium alloy, it concentrate/tailing launder manufactured in stainless steel, and its strong shaker table stand if middle steel. This shaker’s deck is 1.1 m long X 50 cm wide. It weighs in at 150 Kg (330 lbs).
Its head motion mechanism emulates Wilfley’s bumping motion. The shaker table’ stroke can be adjust between 9 and 17 mm (3/8″ and 5/8″) at a shaking frequency of 280 to 460 cycles/minutes.
Select between a SAND or SLIMES Deck for your table.
First 90% of all people using concentrating tables use the sand deck.
The sand deck is efficient in separating high density from low density material (with a difference of 1 SG unit), in the particle size range ranging from 10 mesh to 150 mesh.
Slimes decks are used for particles in the 150 to 325 mesh range. A common observation made by those using the slime deck; the surface tension of water interferes with recovery in this particle size range. This is probably why 90% of people using the tables use the sand deck, also.
The laboratory shaking table is widely used for the gravity separation of sands too fine to treat by jigging. The physical principles utilised in tabling must be understood if preparation of feed and application of control are to be efficient.
Effect all Streaming Particles of Lateral Displacement
Consider a number of spheres rolling down a slightly tilted plane under the urging influence of a flowing film of water. Some of the spheres (shaded) in Fig. 170 represent heavy mineral and others (white) light gangue. The largest sphere travels fastest and the smallest one slowest, under the combined influence of streaming action and gravitational pull. Of two spheres having the same density, the larger moves faster. Of two having the same diameter, if the slope is relatively gentle and the hydraulic urge relatively strong, the lighter sphere travels faster. If during the otherwise free downward travel of these spheres the whole plane is moved sideways, then the horizontal displacement of the spheres varies in accordance with the length of time they take to roll down. This is represented here on the right, which shows that the largest light sphere has undergone the least horizontal displacement because it travelled fastest, whilst the smallest heavy one has been carried furthest to one side. From this it is seen that if a suitable displacing movement can be applied to a plane, the feed can be spread into bands according to the size and density of its constituent particles. If these bands are collected into separate vessels as they leave this deck, the feed will have been segregated into three main products:
Fastest-moving. Coarse light mineral (gangue).
Medium-moving. Fine light mineral plus coarse heavy mineral plus partly unlocked particles (middling).
Slowest-moving. Fine heavy particles (concentrate).
A particle light enough to respond mainly to the hydraulic influence of the flowing film of water moves down-plane with little horizontal displacement. A typical particle, unlike a sphere. will either slide or “skip” downward, rather than roll, provided it is reasonably free to move. Apart from the limited use of the automatic strake in concentrating metallic gold, continuous lateral displacement across the sorting plane cannot handle an adequate tonnage and is not used in the mill.
With the Laboratory shaking table a reciprocating side motion is applied to the sloping surface or “deck” down which the pulp is streaming. If this shaking action was applied symmetrically in both directions across the stream, each particle would move an equal distance in each direction, and separation into bands would not occur. The displacing stroke must be applied gently, so as not to break the grip between particle and deck. The deck accelerates, and in doing so imparts kinetic energy to the material on it. Then the deck motion is abruptly reversed so that it is snatched away from under the particles resting immediately above it. These continue to skid sideways (across the flow) until their kinetic energy has been exhausted. It is therefore essential to provide a differential side-shake which builds up gently and then breaks contact between deck and load.
Progressive Stratification along Riffle
This is provided by the shaking mechanism or head motion of the shaker table. The slower the particle travels downstream, the further it slides sideways under the influence of the shaking motion. Thus far discussion has been limited to a series of individual particles fed to the deck from one starting-point. If, instead, a layer several particles deep is fed from a starting-line, it becomes possible to handle a greatly increased load on the deck. The operating conditions have now changed. In the cross-section through such a layer, as seen normal to the direction of shake, the mixed feed first stratifies itself under the disturbing influence of the shaking action. The smallest and heaviest particles reach the deck, the largest and lightest stay uppermost, with a mixture of large heavy and small light grains between. This arrangement exposes the large, light particles to the maximum sluicing force of the film of water as it streams down the laboratory table. a force that can be controlled in intensity by varying the volume of water used and the slope of the deck. It is thus possible to exert some degree of skimming action to accelerate the downward movement of the uppermost layer without disturbing those below. The particles next to the deck are pressed to it by the material above, and therefore can grip it with greater firmness than would be given by their own unaided weight. They thus are able to cling during fast sideways acceleration, and are only freed and set skidding by the sudden reverse action.
The overlying particles have only a precarious hold. This aids the discriminating action of each stroke. The bottom particle travels furthest, breaks free at stroke reversal and is the first to skid. Those above it sway backward and forward and consequently receive less lateral movement. This accentuates the separating action by giving the bottom (heavy mineral) particles the maximum horizontal displacement per stroke and the upper (light gangue) grains the least. This aids the sorting discrimination. If the feed has been properly prepared by hydraulic classification, ensuring that all the grains have similar settling characteristics through vertical currents, film sizing can now take advantage of the variation in cross-section between the heavy and light particles in each stratum, sweeping down the lighter and leaving the heavier untouched. The particles thus segregated are then removed in separately discharged fractions, called bands, at the far end of the table’s deck. It would not be possible to form and maintain an evenly distributed thick bed of the kind called for by the foregoing considerations if a smooth plane deck were used. Riffles are therefore employed to provide protected pockets in which stratification can take place. They are usually straight and parallel with the direction of shake, but may be curved or slanted. The deck, instead of being plane, may be formed to provide pools in which the feed can stratify. The riffles must:
Arrest and spread the entering feed.
Aid transmission of shaking action to their enclosed load.
Expose the top layer of sand, after it has stratified, to the cross flow of wash water down the shaker table deck.
By suitably gentle tapering, promote delivery to succeeding riffles, wash planes or discharge point’s.
Thus (a) rules out as bad practice the use of “stopping” riffles set high above the rest, sometimes used to arrest and spread entering feed. If all riffles are not of similar initial height the stratifying action and transfer between them is upset. Smooth delivery is best achieved with a feed box integral with the moving deck, and aligned with the vibrator. It should let the feed down gently to the head riffles. Items (b), (c), and (d) are arguments against the use of curved riffles, which increase wall friction and upset stratifying action. A badly maintained mechanical action and deck coupling may mislead the engineer into redesigning his riffle plan, just as an incorrect stance may cause the unwary golfer to modify his swing instead of standing correctly. In the standard Wilfley table the riffles run parallel with the long axis, and are tapered from a maximum height on the feed side (nearest the shaking mechanism) till they die out near the opposite side, part of which is left smooth. Where the riffles stand high, a certain amount of eddying movement occurs, aiding the stratification and jigging action in the riffle troughs.
As the load of material is jerked across the Laboratory Shaker Table, the uppermost layer ceases to be protected from the down-coursing film of water, owing to the taper of the riffle. It is therefore swept or rolled over into the next riffle below. In this way the uppermost layer of sand is repeatedly sluiced with the full force of the current of wash water, riffle after riffle, until it leaves the deck. This water-film is thinnest and swiftest while climbing over the solid riffle, and the slight check and down pull it receives while passing over the trough between two riffles helps to drop any suspended solids into that trough.
At the bottom of the riffle-trough, then, the particles in contact with the deck are moving crosswise as the result of the mechanical shaking movement. At the top they are exposed to the hydraulic pressure of a controllable film of water sweeping downwards. In the trough of the riffle the combined forces-stratification, eddy action, and jigging-are arranging them according to density and volume.
Provided the entering particles have been suitably sorted and liberated, good separation can be achieved on sands in any appropriate size range from an upper limit of about i” to a lower one of some 300 mesh. The difference in density and mass between particles of concentrate and gangue determines the efficient size range which must be maintained by hydraulic classification or free-fall sorting of the feed. A further separating influence is applied hydraulically along each riffle as the water in it gathers energy from the deck’s movement. As it gathers speed in the forward half of its cycle, the water flowing along the trough parallel to the axis of vibration is accelerated. When the deck’s direction is abruptly reversed this flow is only gently checked relatively to the more positive braking force exerted on the skidding particles in the riffle. There is thus a mildly pulsed sluicing action across the Laboratory Shaker Table, in addition to the steady stream at right angles to it, down-slope. This cross-stream helps the particles to travel along the riffles. Since separation depends to a large degree on the hydraulic displacement of the particle, its shape influences its reaction. Flakes of mica, though light, work down and cling to the deck, and may be seen moving nearly straight across, even at the unriffled end where they meet the full force of the stream. Where there is no marked influence in density between the constituent minerals of a pulp, the shape factor aids a flat particle to move along the deck to the concentrates zone, and under like conditions helps an equi-dimensional one to move down-slope toward the tailings discharge. Shape factor can therefore help tabling in some cases, and be disadvantageous in others, depending on whether it reinforces or opposes differences in size between the classified particles of value and tailing.
A Modified Laboratory Concentrating Table
Small scale table concentration tests have many critics. Many metallurgists consider that such tests are of problematical value because of the difficulties involved in conducting and interpreting them. Many kinds of small-scale ore dressing tests are difficult to conduct, and there is, perhaps, good reason for thinking that table concentration tests are amongst the most difficult. Interpretation of results from small-scale tests is the responsibility of the metallurgists and engineers in charge, and it is often held that small-scale table concentration tests are particularly difficult to interpret.
Testing for Table Concentration
On analysis it appears that the difficulties encountered in conducting small-scale table concentration tests are of three fairly distinct kinds.
Firstly, there are difficulties due inherently to the small-scale nature of the operations; for example the smaller width of all mineral bands on the table and the less complete separation due to the shorter length of travel between the feed and discharge points.
Secondly, there are the effects of batch operation owing to the fact that the mineral particles behave differently during the initial period when the sample is just beginning to spread over the table, the middle period when feed and discharge are even and continuous, and the final stage, when the last of the sample has been added and the table is beginning to empty itself.
In the third place, there are the difficulties duo to lack of refinement in testing, for example, slackness in supports and slope adjustment, irregular water distribution and erratic feeding.
If the test must be conducted as a small-scale batch test, difficulties due to the first two causes are inevitable, but by proper attention to the equipment and technique used for laboratory table concentration tests, difficulties due to inevitable causes may be minimized.
Unfortunately, it is common to find that insufficient attention has been given to the careful design of laboratory concentrating tables, and it is believed that difficulties arising from this cause, combined with crude testing techniques, are largely responsible for difficulties in interpreting results. If proper attention is given to the points mentioned, there seems no reason why the results obtained should not be a reliable guide to the optimum performance of a commercial plant.
It is interesting to know that rather similar viewpoints are held by some investigators of the United States Bureau of Mines.
The present paper describes the development of the concentrating table used in the laboratory operated jointly by the Mining Department of the University of Melbourne and the Ore Dressing Section of the Commonwealth Scientific and Industrial Research Organization. Although the paper contains some discussion of the technique of table concentration testing, the bulk of it is devoted to describing the steps taken to improve the mechanical rigidity of the table and the convenience of its adjustments and controls.
Special Difficulties in Batch Tests
In order to comprehend the reason for the modifications made, it is helpful to consider, first, how a mixed feed of dense and light particles, say galena and quartz, behaves in an ordinary batch table concentration test.
It is supposed that the feed rate is uniform throughout the test and that the side slope and cross water are adjusted so that when stable conditions have been established on the table, the line of demarcation between galena and quartz will be on the concentrate end of the table 2 in. from the corner.
As the feed starts to flow on to the table, scarcely any galena appears for an appreciable time and the uppermost band of quartz appears well up along the concentrate end of the table.
Galena is scarce because the quartz moves more quickly; quartz appears well up the slope of the table because the forces tending to wash it across the table are not fully operative. There is little galena on the riffled portion of the deck, so that more quartz particles remain in the riffles where they have little opportunity to be forced by the galena to the top of the bed in the riffles, from where they would be washed down by the cross water.
As the feed continues to flow, more galena appears on the table, and when stable conditions have been established, the line of demarcation between galena and quartz moves down to a point 2 in. from the corner. This condition continues until feeding ceases. Shortly it will be noted that there is scarcely any quartz on the table and that the line of demarcation between the galena and the remaining quartz moves down the concentrate end of the table towards the corner.
The first effect occurs because the quartz moves across the table more quickly than the galena. The second effect occurs because the cross water washes the galena further down the unriffled part of the deck since there is practically no quartz to stop it.
It will be found, then, that if in a batch test a table is fed- uniformly and neither the cross, water nor the side slope is altered, the line of demarcation between concentrate and tailing will start at a point well up the concentrate end of this table, move gradually to a stable point and, at the end of the test, move rather quickly to a point much closer to the corner of the table.
If a clean separation is to be obtained, it will be necessary to move a splitter to follow this line of demarcation. However, it is common to find the movement of the separation point so great that moving a splitter is not alone sufficient to cope with the large changes which occur. In this case it is necessary to alter the side slope of the table.
The desirability of being able to change smoothly, easily and rapidly both the splitter position and the side slope of the table is thus made apparent.
The necessity for changing the side slope of the table during, a test introduces two corollary conditions:—
The method of changing the side slope should be such that no twisting strain is put on the driving mechanism.
Changing the side slope should not affect the cross water distribution.
Modifications to the Table Construction
To facilitate the conduct of table concentration test every aspect of the table construction was examined and a number of modifications introduced.
Shaker Table Head Motion
The table was originally used with a Wilfley head motion, a type which has proved very popular and has been used in a wide range of applications.
However, the head motion used on the laboratory table had been in service for a number of years, and had become badly worn. As alternative plans for a replacement were being considered, Mount Isa Mines Ltd. offered to donate to the laboratory a commercial Deister Plat-O head motion in excellent mechanical condition. This offer was gratefully accepted. For compactness, a frame was built to accommodate the table deck directly above the case containing the head motion, the movement being transferred through a lever arm pinned to the frame. The arrangement is illustrated in Figs. 1 and 3.
Lever arm lengths can be adjusted readily to give a stroke length ranging from 5/16 in- to 1½ in. The sharpness of the “kick” can also be adjusted. To date no experiments on the effect of either of these variables have been conducted. The speed is constant at about 300 strokes per min. and adjustment can only be effected by changing the driving pulley.
Shaker Table Frame
The frame is of welded construction. The base is made of 5 in. channels, and the rest of the frame of 3 in. channels and 2 in. and in. angles. The ample sections combined with the cross-bracing give a rigid frame.
Table Deck and its Supports
The commonest type of deck used on commercial tables is a wooden frame covered with linoleum. The riffles are usually of wood attached to the deck with copper nails.
A deck of this kind has only one major defect for test work—the difficulty of avoiding contamination of successive runs owing to solids lodging between the riffles and the linoleum surface. This trouble has been minimized by using a waterproof adhesive as well as the nails to attach the riffles. Another source of contamination in the old model table was a flat-bottomed feed box which was difficult to clean. The feed box now used was made from a short length of 1½ in. dia. pipe and may be seen in Fig. 1. This type of feed box is very easy to clean.
The deck is supported on four slipper rods which slide in seats arranged in independent pairs at each end of the table. Each pair of seats can move freely about a pivot, the pivots being aligned accurately. This arrangement provides a very rigid support, which accommodates itself easily to change of slope. A clear view of the rods may be seen at A, in Fig. 2, while the seats may be seen at A in Fig. 3.
The deck is connected to the head motion through a shackle and pin, (A and B, Fig. 5), while a spring attached at an angle beneath the deck keeps the slipper rods seated. A crank operated by hand-lever (A in Fig. 4) applies tension to the spring. Either one of two decks with slightly different riffling may be used. To remove the deck, spring tension is released by turning the hand-lever, and detaching the spring. The pin A (Fig. 5) is removed from the shackle B and the deck lifted off. To fit the other deck, these operations are repeated in reverse order. The changing of decks can be effected in about two minutes.
The table is provided with two adjustable splitters, a concentrate-middling splitter on the concentrate end of the table, and a middling-tailing splitter on the tailing side of the table. An external view of the splitters is shown in Fig. 6.
The concentrate end of the table is faced with a 1½ in. wide strip of 16 gauge brass sheet, its edge being flush with the edge of the linoleum deck surface. The splitter itself is a vertical sheet of brass, the top edge of which is about 3/8 in. below the deck surface. The splitter and its small attached launder are mounted on a split block which slides along two brass rods mounted on brackets underneath the table. The halves of the block are held against the rods by crossed leaf springs tensioned by a small knurled nut. The method of attachment is shown in Fig. 2. The cutter moves readily when slight pressure is applied, and maintains any set position.
The middling-tailing splitter rides on the outside edge of the launder; it is provided with a small capstan-headed lock screw, but it has been found that, generally, this is unnecessary.
Cross Slope Adjustment
The cross slope of the table is adjusted by a lever arm attached to the pair of slipper rod seats at the concentrate end. A second lever operates a locking nut at the back of the pivot. The two lever arms are shown in Fig. 4. When using this simple two lever arrangement, it has been found that when the locknut is released the cross slope of the table may change suddenly and jerkily. To improve this feature, a vertical screw type of adjustment is being attached to the lever arm B.
When the cross slope of the table is changed, a couple is applied to the bridge bar (D, Fig. 2) connecting the two slipper rods at the head motion end. To avoid applying a twist to the shackle E, the nut F tightens onto a shoulder on the pin G and not onto the bridge bar. The clearance is so small (0.001 in.) that there is no perceptible slackness although the shackle can twist quite freely.
Water Distribution Launder
The top edge of the table deck is not parallel to the axis about which the deck is tilted. Consequently, if the launder distributing cross water were attached to the deck, the water distribution would change when the cross slope was changed. To avoid this, the launder has been attached to the main frame by two pieces of 1½ in. x 3/16 in. flat steel bent appropriately. The launder is attached by hinges and may be folded up out of the way to facilitate changing of decks. The method of attaching the water launder is made clear in Fig. 4.
A common method of feeding a table for batch test work is by scoop. The discussion given of the behaviour of dense and light minerals in a batch test in which the feed is quite regular enables conditions to be foreseen when the table is fed by scoop. Suppose a somewhat extreme example in which a scoopful is fed onto the table in five seconds, and successive scoopsful added every 30 seconds subsequently. In the period immediately after adding each scoopful, the quartz added will move more rapidly than the galena, and so will push the line of demarcation between concentrate and tailing up. Subsequently, the corresponding amount of galena will arrive at the table edge, and so will push the line of demarcation down. This cycle will be repeated for each scoopful added. The result will be that the line of demarcation between concentrate and tailing will fluctuate. The extent of the movement will depend on the irregularity of the feed, and although with care the fluctuation may be minimized, the operation will inevitably be tedious and time-consuming, and even the best result will leave much to be desired.
Experiments with a launder feeding method have shown that it has decided advantages. The V-bottom launder used is shown in Fig. 1. The feed is spread fairly uniformly along the bottom of the launder, and the rate of feed regulated by the rate of feeding water to the head of the launder. About 90% of the feed will flow without further alteration, but some additional wash water is necessary near the end of a run to clean down the sides of the launder.
More elaborate launder feeding methods with progressing water jets, etc., have been proposed, but although these would appear to have further advantages, the simple method described has proved satisfactory. It does not give absolutely regular feed, but the changes occur, gradually and are easy to cope with.
Experiments with continuous circulation have also been conducted. The arrangement is shown, in Fig. 7. Concentrate, middling and tailing separate on the table and are deflected into a common pump, which discharges the, mixed feed into the dewatering cone shown. The overflow runs to waste and the discharge returns to the table. This system gives far more regular feed than any other method tried. It works very well for demonstration purposes, but quantitative tests have not yet been undertaken. The method proposed is to establish equilibrium conditions, and then take timed samples.
Three product hoppers are used, two small hoppers which are fixed, to the table framework being provided for concentrate and middling, while the tailing is collected in a large hopper fitted into a framework mounted on wheels. The large mobile hoppers of 30 gal. capacity are extremely useful in the laboratory for many purposes, such as the collection and settlement of slime, collection of jig and table tailings, and in fact any large quantities of ore pulp.
Both the fixed and mobile hoppers are closed with rubber bungs from inside, the bungs being fixed to long brass rods with T-handles. The clearance below the hopper outlets is sufficient for a 3 gal. bucket.
A laboratory concentration table was modified by incorporating a sturdier head motion, main frame and supports, and altering the controls so as to make them positive, convenient and independent of each other.
The advantages from the modifications to the table construction cannot readily be expressed in quantitative results. The important effect is that every operation, such as feeding the table, adjusting the side slope or product splitters, and handling the products, is easier, and the table itself is much less prone to erratic disturbances due to lack of rigidity in the framework, supports and adjustments. It is felt that these substantial mechanical improvements are bound to express themselves in improved metallurgical performance.