CEE April-June 2012

- diameter of 160 mm. The classifier drive can be regulated from 600 rpm to 4000 rpm, thus permitting cut-points from 280 IJm top-cut with limestone at a materials' density of 2.72 g/cm 3 . Modes of mill operation Three modes of operation can be selected with this machine: operation without recirculation, with recirculation (with return or with partial/total removal of the grit material) and as an edge mill with no classifying element. The requirement was the construction of the mill in such a way that all these modes of operation could be implemented without major conversion work to the mill. In the mode of operation without external grit recirculation, the mill operates identically to an industrial vertical mill. Only the longer up-flow zone resulting from grit recirculation should be emphasized. Mass flow and grit dispersity can be registered in the mode of operation with grit recirculation. The grit can be sampled via a bypass, which permits complete diversion of the flow of grit for the purpose of obtaining representative samples. For its edge mill mode of operation, the mill is designed without the use of system air. The air duct under the grinding table was designed and constructed in such a way that descent of the materi al through the nozzle ring is assured. In this mode of operation, material input is also accomplished via the central feed system. This facility was designed, constructed and commissioned within a twelve-month period; all engineering work was performed internally. Application of the design of experiments An initial estimation of the amount of testing necessary, aimed an answering various questions concerning the behaviour of the vertical mill, concerning possibilities for up-scaling and evaluation of the weighting of influencing parameters, indicated the necessity for more than two hundred individual tests. A study of the relevant technical literature confirmed this level of testing. Reference describes a procedure with standardized laboratory tests. A standard laboratory test using a reference mill is thus performed in order to generate information on the necessary mill power, the potential throughput, and other specific mill parameters. This approach can only be applied safely given knowledge of a large number of results obtained from tests performed with working mills. The use of a simulation based on a CFD software package was considered as an economic concept but, despite the enormous recent advances in this technology, was ultimately rejected as being too risky and too unreliable, although it has been used very successfully for the verification of existing systems. It is also difficult to find answers to many questions concerning the internal mass flow and the effectivity of comminution and of internal components. Application to restricted sub-sectors of the overall system, however, is very useful in terms of high informational value . Therefore a combination of classical test series and design of experiments was used. This is intended to provide information on the influence exerted by the various parameters and on any interactions that are present. Such a test plan can take the form of a fractional factorial screening schedule. In this test plan, the test parameters are tested in various combinations, and the parameter variations are fixed at two levels for each parameter. The levels are selected in such a way that the values for the parameters coincide with minimum and maximum process values (Table 2). Unlike classical test plans, in which only one single parameter is changed at one time, in order to study its influence, multiple parameters are always changed at once in a DoE test plant. The most important influencing parameters were isolated on the basis of the process knowledge already available. Criteria which must be fulfilled if the test is to be regarded as valid were defined as follows during this process: x Stable mass balance x Constant pressure drop across the mill x Constant product dispersity x Homogeneous system-air flow A test can be considered stable on the basis of these four criteria. An observation period of 60 minutes from the attainment of all four criteria applies here for the time factor. The constancy of the mass flow and of product dispersity is checked by means of sampling downstream of the filter. 27