The cleaning of fine coal, particularly for metallurgical use, is receiving greater attention. An increase in the proportion of fine coal in the run-of-mine product has been created by changes in mining practice, and a concurrent deterioration in the quality of the fines has occurred because of more widespread full-seam mining. These two factors, augmented by the depletion of premium beds of metallurgical coal in some areas and the resulting shift to “dirtier” beds, are providing the impetus for adoption of Improved cleaning techniques. For example, froth flotation, which has been used for years in a few preparation plants, now is gaining more rapidly in acceptance. Similarly, the dense-medium cyclone, which is used widely in Europe for making difficult separations, recently has been introduced to the United States.

Regardless of the cleaning process employed, the finer sizes of coal can be treated with maximum effectiveness only when they are removed from the run- of-mine product and treated separately. No present process can treat the full range of sizes together with complete effectiveness. Flotation is applicable only to the coal finer than a top size ranging between 14- and 48-mesh, and cyclones require removal of all material finer than 0.5 to 0.3 mm. from the feed. Thus, the trend toward more-effective cleaning of the fine coal imposes the attendant problem of fine screening. Whereas in the past there was little need to screen finer than about ¼-in., present developments indicate that fine screening will become increasingly prevalent.

Fine screening is troublesome because screen capacity is low, efficiency is relatively poor, and blinding often is serious. Thus, with an eye to growing American requirements, the Federal Bureau of Mines arranged to investigate a radically different type of screen known as the sieve bend, developed by the Dutch State Mines and first described in 1954. The sieve bend is simply a stationary section of wedge-bar screen bent to an arc. Its attributed are freedom from moving parts, high capacity, ability to operate below the size range of conventional screens, and relative freedom from blinding.

The smallest commercial-size sieve bend, 1 ft. wide, was used in the investigation. A variety of screen surfaces having openings from 0.3 to 3.5 mm. were employed in Bureau of Mines experiments herein reported, and the testing program embraced natural and synthetic feeds exhibiting a wide range in size composition. The capacity of the screen when operating at peak efficiency varied with the size of the openings, and ranged from 100 gallons per minute at 0.3 mm. to 500 gallons per minute at 3.5 mm. These flow rates corresponded to 5 to 26 tons of coal per hour. Unlike conventional screens, which exhibit a steady decrease in efficiency with increase in load, the sieve bend has an optimum capacity for each size of opening. Efficiency decreases with feed rates that are either higher or lower than the optimum range.

The sieve bend effects a separation at a size substantially smaller than the openings. With 0.3 mm. openings a separation at about 0.2 mm. or 65-mesh was obtained. With 3.5 mm. openings the separation was at 1.3 mm. or about 14-mesh. However, the exact size of separation depends to some extent upon the size composition of the feed. For example, the presence of a high proportion of particles having about the same size as the screen openings tends to coarsen the separation.

As with conventional screens, the efficiency of the sieve bend depends largely on the size composition of the feed. Higher efficiencies are obtained when the feed contains a large proportion of undersize material than when the percentage is small. Similarly, the presence of an abnormally high proportion of particles having about the same size as the screen openings militates against high efficiency. With the sieve bend, these factors appear to outweigh the size of openings in determining screen efficiency. Efficiencies in the range of 70 to 80 pct. characterized operation on feeds exhibiting an ordinary degree of screening difficulty, and a peak efficiency of 84 pct. was achieved under especially favorable conditions.

As a dewatering device the sieve bend is inferior to conventional screens. The oversize product is distinctly wetter than that obtainable with vibrating screens or high-speed shakers.

Acknowledgments

The experimental work described in this report was conducted at the Seattle Coal Research Laboratory, Bureau of Mines, which functions in cooperation with the School of Mineral Engineering of the University of Washington.

Description of Equipment and Test Procedures

The sieve bend employed was of American manufacture. It consists of the screen surface, which is a 1-ft. wide section of wedge-bar screen bent to a 60° arc on a 30-in. radius, and a housing that comprises a feed box, undersize hopper, and support for the screen. Slurry flows from the feed box onto the upper end of the screen surface in the form of a vertical sheet, which is tangent to the curvature of the screen. The feed box provides a head of about 3 ft., and is vented to prevent any build-up of additional pressure. The diagram in figure 1 illustrates what happens as the feed slurry progresses down the screen. According to Fontein, each successive wedge bar slices a layer off the bottom of the stream. The thickness of these layers is equal to about one-fourth of the net bar spacing, and the layers carry particles of a size up to approximately half the net bar spacing. Coarser particles extent so far into the stream traveling along the surface of the deck that they are carried over the openings and discharged at the end of the screen as the oversize product.

Coal was fed from a bin with a variable-speed feeder, which discharged into a launder leading to the screen. A metered water supply discharging into this launder provided means for varying the percentage of solids in the feed. Owing to the high capacity of the screen, only a 4-inch width was employed; the remainder was blanked off. However, all capacities are expressed per foot of screen width.

In making a test, the individual coal and water flow rates were adjusted to give the desired throughput and percentage of solids, and this slurry allowed to pass over the screen for a few minutes to establish equilibrium conditions. Then the oversize and undersize products from the screen were diverted simultaneously into tared containers for a sampling period of 12 to 15 sec. Even with such short sampling periods and the use of a 4-in. screen width, the samples often weighed over 300 lb. The samples collected in this manner were weighed, dried, reweighed, and the coal subjected to screen analyses. Screens in the Tyler square-root-of-two series were used.

Most of the experimental work was done with a washed coal from Carbon County, Utah having a top size of about 1/8-in. The screen analysis of this coal, and those of the other coals used, are shown as part of the detailed test data in the appendix.

Method of Evaluating Efficiency

After examining a number of formulas for assessing the efficiency of screening, one developed by Rabling and used by Holbrook and Fraser was adopted. This formula defines efficiency in terms of percentage elimination of undersize material, and is written as follows:

Efficiency = A x B x 100/C

where:
A = Percentage of undersize material in the undersize product
B = Percentage of undersize product
C = Percentage of undersize material in feed.

With a conventional screen having square openings, the term undersize material would refer to material finer than the screen aperature, for this is an exact limiting size. The sieve bend, however, does not separate at the size of its openings; therefore, undersize material was defined as that finer than the size of separation. Size of separation, in turn, was defined as the size at which there was 5 pct. of oversize material in the undersize product.

The size of separation and percentage of undersize in the composite feed were obtained by plotting the screen analyses of the undersize product and feed on log-probability paper. Unusual care in plotting and reading the curves was required to insure consistent efficiency values. Relatively small deviations in the assigned separation size were found to affect the efficiency values disproportionately.

Influence of Operating Variables

The only operating variables known to influence the performance of the sieve bend on a given material are the rate of feed and the proportion of solids in the feed. Therefore, the first step in the investigation was to evaluate these two factors separately. This was done by making two series of tests; first the proportions of coal and water in the feed were held substantially constant while the feed rate was varied, then the proportion of coal in the feed was varied while the flow rate was held constant.

Feed Rate

Preliminary operation indicated that the sieve bend performed best on feeds containing 15- to 25-pct. solids, and therefore the tests at variable feed rate were made within this range; most of them were at substantially 20-pct. solids. The detailed results of tests made at various feed rates with 0.3-, 0.5-, 1.8-, and 3.5-mm. openings are shown in tables 1 through 4 (all tables are in the appendix).

Figure 2 illustrates the relation between feed rate and efficiency developed in this series of tests. With each size of screen opening, a peak efficiency occurred at an intermediate feet rate. Feed rates either higher or lower than the optimum flow caused a substantial reduction in the efficiency of screening. In this respect the sieve bend differs from conventional screens, which exhibit a steady decline in efficiency with increase in feed rate. When the sieve bend is underfed, dewatering occurs too rapidly, the

material moving down the screen surface loses mobility, and screening is impaired. The slopes of the curves in figure 2 are flatter for the coarser screen openings, indicating greater latitude in the rate at which the sieve bend can be fed without impairing efficiency. With the smaller openings feed rate is more critical.

Figure 3 shows that the relationship between screen opening and optimum feed rate is essentially linear in the range investigated, and varied from 100 gallons per minute at 0.3 mm. to nearly 500 gallons per minute at 3.5 mm. These flow rates correspond to 5 and 26 tons of coal per hour, respectively.

Figure 3 shows also that the relationship between the size of opening and the size at which the sieve bend separates is essentially linear. With the smallest openings the size of separation was 63 pct. of the net bar spacing, whereas with the largest openings it was 37 pct. Thus, the generalization that the sieve bend separates at half the size of the openings is only an approximation.

The peak efficiencies obtained in this series of tests were 74, 72, 67, and 78 pct., respectively, for the 0.3-, 0.5-, 1.8-, and 3.5-mm. openings. The fact that the second coarsest screen gave the lowest efficiency is ample evidence that factors other than size of opening have a major bearing on efficiency, even when the feed rate is optimum.

Pulp Ratio

The tests made on feeds containing various ratios of coal to water were conducted with the feed rates found to be optimum for each size of screen opening-150 gallons per minute for the 0.5-mm. deck and 500 gallons per minute for the 3.5-mm. deck. Tables 5 and 6 present the results of these tests, and figure 4 illustrates the relation between feed-pulp ratio and efficiency. With the 3.5-mm. deck, varying the percentage of solids in the feed from 13.5 to 26.4 had comparatively little effect on efficiency, which ranged only from 76 to 78 pct. The maximum efficiency occurred at about 22 pct. feed solids. With the 0.5-mm. screen the influence of feed solids was more pronounced. At 18 to 20 pct. solids the screen was distinctly more efficient than at either lower or higher concentrations of solids in the feed.

Additional information on the influence of feed-solids concentration was obtained from a second series of tests on the 0.5-mm. screen, in which the feed rate was increased to 250 gallons per minute. The results of these tests are not included in the report because of space limitations, but they also indicated that maximum efficiency was obtained at about 20 pct. feed solids.

The data in table 6 for the 3.5-mm. screen show a progressive and significant increase in size of separation with increase in feed solids, amounting to about 50 pct. over the range investigated. As the concentration of solids in the feed increased, more coarse material was drawn into the undersize product. No similar trend is evident in the data for the 0.5-mm. screen.

Influence of Feed Composition

The operation of conventional screens is influenced by the size composition of the feed; therefore, this factor was investigated with the sieve bend. Two aspects of feed composition that affect screen performance are the amount of undersize and the proportion of particles having about the same size as the screen openings. These two factors were investigated separately with feeds of synthetic size composition. A large lot of coal was screened into 3- to 28-, 28- to 48-, and 48-mesh to 0 size groups. For the 0.5-mm. screen used in this part of the investigation the 28- to 48-mesh fraction represented “near-aperture size” material, and the 48-mesh to 0 fraction represented undersize. In one series of tests the amount of undersize material was varied while the amount of near-aperture size material was held constant; in the second series the amount of undersize was held constant while the proportion of near-aperture size material was varied. All tests were made at a feed rate of about 150 gallons per minute and 20-pct. solids, the conditions previously found to be optimum for the 0.5-mm. screen.

Influence of Undersize Material

Table 7 shows the results obtained in four tests made with the proportion of undersize decreasing progressively from 65 to 6 pct. The separation size remained substantially constant at 48 mesh in this series, but efficiency fell markedly as the proportion of undersize in the feed decreased. As shown by the upper curve in figure 5, in the range from 65- to 20-pct. undersize, the drop in efficiency was moderate, but further decrease in the amount of undersize caused a sharp reduction in efficiency. Whereas at 65 pct. undersize the efficiency was 84 pct., at 6 pct. undersize it was only 50 pct.

These results indicate that the sieve bend would be highly efficient as a scalping screen, where the proportion of oversize was small, but relatively inefficient in removing a small amount of fine material. In this respect the sieve bend resembles conventional screens.

Influence of Near-Aperture Size Material

Table 8 shows the results obtained in a series of four tests in which the percentage of near-aperture size material (28- to 48-mesh) was increased progressively from 0.2 to 26.9 pct. As shown by the lower curve in figure 5, increasing the proportion of near-aperture size material through the range of 0.2 up to about 12 pct. had negligible effect on

the efficiency of the screen, but continued increase above this range caused a steady reduction in efficiency. In the range from 12 to 27 pct. of near-aperture size material, efficiency dropped from 75 to 59 pct.

The proportion of near-aperture size material in the feed affected not only the efficiency of the screen but the size of separation as well. As shown in figure 6, the separation size increased progressively from 190 to 352 µ as the percentage of near-aperture size material was increased from 0.2 to 26.9 pct. Thus, at an unusually-low percentage of near-aperture size material the separation was at 65-mesh, whereas with an unusually high proportion of such material the separation was at 35-mesh; with the natural feeds tested previously, the separation was at 48-mesh.

Tests on Natural Feeds

The use of feeds having synthetic size compositions to provide individual control over the amount of near-aperture and undersize materials necessarily resulted in some screen analyses that were greatly different from those ordinarily encountered with coal. Therefore, to insure that these unusual compositions had not resulted in anomalous test results, a final series of tests was made on coals exhibiting natural variations in size composition. An anthracite, a firm-structure bituminous coal, and a very friable bituminous coal were prepared for screening by crushing in a hammer mill or by stage crushing in rolls. The variation in the strength of the coals and in the crushing methods provided a wide range in size composition, even though all but one of the samples had a top size of about ¼-inch.

Table 9 shows the results of trials made with four of these samples on the 0.5-mm. screen. The relation between efficiency and the size composition of the feed is shown in the following tabulation.

Except for test 29, in which the amount of near-aperture size material was unusually high, efficiency increased steadily with increase in amount of undersize material. Comparison of tests 30 and 5 shows that the proportion of undersize material in the feed outweighed the amount of near-aperture size material in determining efficiency; a 2-percentage-point increase in the undersize material more than compensated for an increase of 4 percentage points in the near-aperture size material.

Table 10 shows the results of trials made with three of these samples on the 3.5-mm. screen, and the following tabulation shows how efficiency varied with the composition of the feed.

Tests 14 and 32 illustrate the high efficiency attainable with feeds containing a large amount of undersize coal. Of particular interest in this connection is the fact that the efficiency of 83 pct. for test 32 is about the same as that obtained with a similar synthetic feed on the 0.5-mm. screen. Thus, the size of the screen openings is less important than the composition of the feed in determining efficiency.