Chert, which is very hard, and dolomite, which is moderately hard, were selected for use in a study of the effect of different speeds and ore charges in batch tests. Mill, ball load, and pulp consistency always were the same. Results for chert are shown in the upper section of table 13, and for dolomite in the lower section. Each section is divided into six series and each series includes six or seven tests.
Usually the ore charges were 200 to 35 pounds, 72 pounds being the amount, as determined by test, required to fill the interstices of the ball load at rest. (In similar rod-mill tests, analyzed in tables 8 and 9, 39 pounds were required to fill the interstices.). When the great variety of speeds and sizes of charges is taken into account, the similarity of the sizing analyses is remarkable. The highest capacity in each series in the upper section was with an ore charge of from 50 to 75 pounds, but the highest efficiency was with an ore charge of from 150 to 200 pounds.
The lower section, which is for dolomite, shows these same characteristics in a general way, but there is less spread between the charges required for best capacity and best efficiency.
The broad spread of “time” in the different grinding periods should be noted with the similarity of the products. The time periods for grinding ranged from 66.6 minutes for the heavy chert charges at slow speed to 1.6 minutes for the light dolomite charges at high speed.
The tests indicate that for high capacity with the hard ore in continuous work a large discharge should be emphasized, and that for efficiency a smaller discharge would be desirable. When the ore is softer, like dolomite, this distinction is not so important. An earlier paper shows by diagrams “ that the balls in the tests analyzed in table 13 would be somewhat too large for dolomite, but it is believed that this does not alter the above-mentioned effect.
The capacity increased with the speed, and when the ore charges were large the greatest efficiency was achieved at about 50 percent of critical. Although this speed is the one found to be best in some of the first tests, in which slipping was reduced by using a very short mill, it should be interpreted broadly as indicating a range rather than a specific setting.
When the chert charges were small, efficiencies were about the same at all speeds, showing a small increase at high speed; but when dolomite charges were similarly, small, better efficiencies occurred at lower speeds. This difference illustrates how disputes can arise when experimenters draw conclusions from evidence that is too limited.
A common characteristic of the 12 series of tests is that the selective grinding of the coarse particles was better with a heavy ore charge. However, the grinding rate was low with these same heavy ore charges. It may be difficult to get the idea, but the tests indicate that except for time spent in dumping and reloading, a batch wet mill would give the greatest output with a small ore charge.
In each series in table 13 the highest point on the power curve usually coincided with the highest point on the capacity curve. This will be seen to be different in the dry grinding, as shown in table 14 and figure 1.
Under analogous conditions, the power for dolomite was higher than for chert; at the respective speeds 30 to 80 percent of the critical and with the same ore charges, respectively, the increase in power was about 4 percent. One factor that surely contributed to this increase was the smaller volume of dolomite; the ore charge was determined by weight, not by volume. The volume of a unit weight of dolomite was about 8 percent less than the same weight of chert. Another fact to be recorded is that at the end of the runs there was more coarse dolomite than flint. By referring to table 5, it may be seen that the mean particle size affects power, and there also the maximum power was attained with the coarsest particle size.
Another factor to be considered is the type of grind. For a given charge of dolomite, the type of grinding is almost identical at all speeds. However, from a similar comparison of equal weights of the chert products it might be said that the high speed did give a little more selective grinding. These variables are mentioned primarily to call attention to the close accord.
Table 13 is valuable in connection with the old contention that high speed and impact are important. Certainly the record of efficiency does not support advocates of impact. At 30 percent speed the balls had only a slow, sluggish, rolling action, yet the grinding, when correlated with the power, was satisfactory. It seems, therefore, that the literature accumulated during the last 20 years, which attempts to elucidate the principles of ball milling by discoursing on “balls thrown in the air”, is not comprehensive.
In a given time, the batch mill, with about 75 pounds of ore charge, did more grinding than the mill with a lighter or heavier ore charge. This is proved by taking a given series and dividing weight by time, and also it may be seen in the table in the item “surface tons per hour.”
Experimenters have been inclined to substitute number of turns of the ball mill for power readings and to speak of it as torque. They can find in table 13 evidence to dissuade themselves of that practice. If the products of speed and time shown for the first four series of tests for dolomite are considered (in which the tests with 50 pounds are given), they will be found to be identical—120; that is, the number of turns was the same in each test. However, if power is similarly multiplied by time, the energy input will be seen to vary to a considerable degree. The number of turns is not always an exact measure of torque. The determination of power is the only accurate means of measuring torque.
Dry-Batch Ball Milling with Various Ore Charges
The tests shown in table 14 were conducted on the two ores in the same manner as shown in table 13, except that the grinding was dry and at only 50 percent speed. The percentage weight of products on 14-mesh shows only moderate changes in the type of grind; in the wet grinding reported in table 13, the heavy ore charge received selective grinding.
Capacity and efficiency were both maximum when the ore charge was 75 to 50 pounds. This capacity characteristic is like the wet grinding, but the high efficiency, also with the same light ore charge is unlike the results of wet grinding, where a heavy ore charge was required for efficiency. That is, in dry grinding, the light ore charge was the best both as to capacity, and efficiency. In this way dry grinding was like wet grinding in the rod mill, as shown in tables 8 and 9.
The peak of the power curve shown in table 14 at 150 pounds of ore is contrary to what would be predicted from the findings in wet grinding, in which the peak was reached with the small charge. In dry grinding it was a new experience to see the capacity curves and power curves of each series sloping in opposite directions instead of running nearly parallel, as they did in wet grinding. (See fig.1.)
The dissimilarity of the power curves in wet and dry grinding may be examined by considering the amount of ore and water in the mill. The amount or volume of material filling the interstices between the balls in action has been shown to influence power, and it may be noted that the volume of a small amount of ore plus water, which required maximum power in wet grinding, is about the same as the volume of a large amount of dry ore, which had the same effect.
Dry-Batch Ball Milling with Equal Ore Charges and at Various Speeds
Having found from table 14 that about 75 pounds gave the most advantageous dry charge, this amount was selected for runs at speeds from 40 to 90 percent of critical. The results are shown in table 15. In a general inspection of this table, the 90-percent speed should be ignored after noting that the speed was too high. With that exception, the products for the different speeds are almost identical. Unlike wet grinding of chert in 75-pound charges, in which the efficiency was about the same at all speeds, the maximum efficiency in dry grinding was attained at high speed. It is possible that the tendency of the dry charge to stick to the grinding medium at slow speeds reduces grinding efficiency. Although the best work was done at high speeds, nevertheless the 90-percent speed was too high and illustrates how the work of a high-speed mill could be improved by reducing the speed.