Stirred mills are used in the minerals industry for fine and ultra fine grinding applications, where ore must be reduced to micron and sub micron size ranges for processing. They are becoming increasingly more common as easily processed orebodies are depleted and the need to process fine grained and complex orebodies increases. Traditional tumbling mills have been proved uneconomical for grinding below 20ìm.
One important issue in stirred milling is the selection of a grinding medium. Grinding media can be a source of high cost and loss of milling efficiency if not considered carefully. There are many factors that need to be considered, and compromises between them. A variety of media types are currently being used in industrial mills, but not always with the test work to prove their advantage.
Past research into stirred mills has mainly focused on the stirrer arrangement, stress intensity within the mills and scale up from lab-scale to full-scale operations. Research into media wear rate has mainly focused on steel ball media for SAG, semi-autogenous-grinding; and ball mills. Some work has been carried out for IsaMill media types, however a methodology for comparing and characterizing media wear has not been developed. Work carried out by Davis and Dawson (1989) showed that two wear rates were present (in grams per hour), especially for natural material such as sand. This makes comparison between manufactured and natural media types difficult. Smith (2001) noted that a lower milling efficiency was experienced during the first period of milling. Based on this the following hypotheses were formed:
Given media wear at different rates.
A methodology can be developed for comparison of wear rate for one media type against another.
The hypotheses were tested using four different types of grinding media:
.. River sand (from Nambucca Heads)
.. Lead slag (from MIM lead smelter)
.. Sample C
.. Ceramic balls
The media were tested in a laboratory scale vertical mill with simulated slurry of silica at 60% solids. The each test ran for two hours, after which the mill contents were screened and dried to determine the weight of media remaining in the original size fraction. The wear rate for each medium was analysed in grams of loss per kilowatt-hour of power draw (g/kWh).
Further testing, using the same procedure but varying the testing time, was carried out to investigate the change in wear rate over time. This test work was conducted for one media type, but it showed that analysing wear rate in terms of g/kWh produces a constant number for a given material, thus allowing for comparison between different media types.