In table 3 of a recent publication a report was made on the use of pipes (instead of rods) stuffed with different materials to give grinding media of different densities when the dimensions were always the same. The mill had a large discharge opening. There it was shown that through a range of densities from 2.89 to 7.30 and a range of speeds from 40 to 70 percent critical, the sizing analyses of the products were about the same (differences in sizing analyses will be discussed later) when the ratio of tons of feed to horsepower was constant. These results seemed to be a step toward the answer to an old question about the effect of the density of the grinding medium. The considerations of that work will now be carried farther by reference to table 3 of the present report, which has additional data on the same tests.
The tests were made at four speeds, and at each speed the pipes had five different loadings to give the specific gravities indicated. The power is recorded under A and the densities of the stuffed pipes are under B. The column B/A is notable because of the moderately regular variation of the values therein, but when taken by itself does not permit a satisfactory interpretation. However, when the weight of ore and water are taken into account in C, and the ratio C/A is given as in the last column, ratios of remarkable accord are obtained. From these ratios it is concluded that under the conditions specified, the power varied as the weight of medium plus ore plus water.
Doubtless the weight of the ore in a mill is a factor in power consumption. This subject was touched upon when the statement was made that heavy minerals used more power than do flint or dolomite.
The difference in sizing analyses mentioned above are shown in table 4. This table is made up of selected parts of old table 3, and is intended to show how much the light media failed. The results of grinding at 50 percent critical are all brought together. The selection of any other speed would serve the purpose as well.
Although the calculated efficiencies are about the same at the various densities of media, it is plain that the light media gave nonselective grinding; that is, the products from the light pipes had too much in the coarse sizes. This failure is not brought out by the surface calculations because the change in amount of calculated surface is so small when coarse material is crushed to an intermediate size. Non selective grinding by the light media inspires the idea that whereas much attention of operators has been given to the size of media for certain particle sizes of ore, the weight of the media also should be taken into account. That a light medium fails by nonselective grinding will be shown later in table 21, in which pebbles are considered. There the pebbles failed when the ore was very hard