Wet Mills
In the early 1930s, the economic depression was in full force, and the only interest in new grinding machines was for milling gold ores, because gold was the only metal that was not falling in price. In 1933, gold rose from $21 to $35/oz. Alvah Hadsel, a mechanical engineer in California with considerable experience in the aggregate industry, devised a machine that he thought would replace an entire crushing-grinding circuit by lifting rocks high enough to cause them to shatter on impact when they were dropped onto a hard surface. In 1932, he built his first mill using this principle and installed it at the Beebe gold mine near Georgetown, California. It took the form of two parallel wheels (7.7 m x 1 m) that rotated at 2.66 rpm. Each wheel had 24 internal buckets (1.0 m x 0.6 m) to lift the ore and discharge it onto breaker plates (Hall 1935). This duplex mill, shown in Figure 7.21, was driven by a 74.6-kW motor. Water was added to the mill, and a 90-degree arc of each wheel was immersed in a concrete classifier tank. Fine particles overflowed the top of the tank, and the wheel collected the coarse particles for further breakage. The mill was reported to break 308 tpd from 0.3 m to 65% passing 75 pm.
By 1934, Hardinge had bought into Hadsel’s enterprise, and mills were being sold as Hardinge-Hadsel mills. The largest of these was 8 m in diameter and 1.45 m long, and it ground 210 tpd of passing 203-mm tough silicified schist to 96% passing 150 pm using a 74.6-kW motor (Hall 1935). The breakage mechanisms in the Hardinge-Hadsel mill were shattering and, to a lesser extent, autogenous grinding. But many of the rocks or their fragments had to be dropped up to 70 times to be broken completely, and this caused excessive wear on the buckets. The industry verdict was that it was suitable for softer ores but not for hard, abrasive ores. In the mid-1930s, Hardinge redesigned it, building a high-diameter, short-length mill with a high peripheral speed that was called the Hardinge Cascade mill. A number of installations were successful, although critical — size particles started to become a problem.
In one sense the Hadsel mill may be viewed as a failure because the concept of lifting a rock to a height at which it would be disintegrated by its own weight when dropped onto a steel plate was unsuccessful. In another sense, it was a success, because it led to the concept that using large rocks as grinding media in tumbling mills was an extension of secondary autogenous grinding. Fifty years later, large-diameter SAG mills with length:diameter ratios in the general area of 0.5-1.0 were in use and are still being used today to grind huge volumes of ore at low cost.
By 1935, the partnership between Hadsel and Hardinge had ended, and Hadsel turned his attention to dry autogenous grinding. Once more he established a process that has become very important. Hardinge continued to work on the Cascade mill, and one of his more successful inventions was the “electric ear,” which detected the change in sound emitted by the mill as the degree of filling rose or fell in response to ore change. The signal was used to maintain optimum grinding conditions automatically by controlling the feed rate. This was the start of feedback control for grinding mills.
Success for the Cascade mill came in 1958 after Hardinge presented a paper on autogenous grinding at the American Mining Congress. The paper’s presentation led to 12 wet mills (5.8 m x 1.6 m) being installed in an iron ore concentrator built by the Pickands-Mather Company in Wabush, Quebec, in 1959 (see Figure 7.22).
 |
FIGURE 7.22 Bank of twelve 5.8 m by 1.6 m (18 ft by 5 ft) wet autogenous Hardinge Cascade mills installed in an iron ore concentrator in Quebec in 1959 (Robinson 1980; reprinted by permission from CIM Publications)
The installation was successful and must have done much toward overcoming the prevailing skepticism about the value of autogenous mills. Fortune favored Hardinge on that occasion if not on others, because the low-grade iron ores in the Quebec Labrador trough had well-defined iron mineral granular structures and were excellent feed for fully autogenous mills.
In 1956, primary autogenous grinding was investigated at the Grootlvei Proprietary Mine in South Africa. A pebble mill (3.9 m x 5.2 m) was converted to receive run-of-mine ore, and a Williamson controller was used to ensure that a constant load of ore was maintained in the mill. Jack Williamson was a Canadian engineer who was in charge of the Union Corporation research laboratory in Springs, South Africa, in 1948. He developed a pebble feed controller that became indispensable to the operation of primary autogenous grinding when this process started at Union Corporation in 1958. The basis of the controller was that the load of ore in an autogenous mill changed as the feed rate, size distribution, or hardness of the ore changed, and the power used by the mill rose and fell with the load. The Williamson controller measured the power and automatically adjusted the feed rate to maintain the mill load and power constant. After 55 days of plant operation, the controller was deemed to be a success and it was built into control systems for all Union Corporation autogenous mills. The principle used in this controller is still widely used for controlling autogenous mills although refinements have been added.