Investigation of load behaviour of an industrial grinding mill
No Thumbnail Available
Date
2009-03-24T07:22:06Z
Authors
Couvas, Constantinos
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The dynamic behaviour of the ball load (specifically toe and shoulder positions) within a
dry Φ4.74m by 7.4m long, 2.15MW powered, ball mill was investigated as a function of
worn and new liners. The mill operating conditions and charge composition where kept
constant. Three types of experimental probes were designed, manufactured and mounted
strategically on the mill to determine the orientation of the load inside the mill. These
included a combination of electrical conductivity, movement and photo-detector probes.
Data recorded from each probe was processed, analysed and compared with each other
and to Millsoft’s 2D theoretical numerical simulation model. Achieving good electrical
contact proved to be a very difficult task, hence conductivity did not compare well to
Millsoft simulations. It was deduced that the conductivity probe design and electronics
data capturing system limited accurate recording of the ball trajectory positions and tended
to represent more of the load ‘locked-in-state’ region (or the period where maximum
pressure is applied to the mill walls) instead of the extreme load toe and shoulder
positions. The movement probe data compared best with Millsoft predicted load toe and
shoulder results at 133 and 282 degrees, respectively. The probe design was however
susceptible to forces from different directions thus decreasing its accuracy. Only two tests
from only one photo detector probe (measuring ball reflection) were successfully recorded.
The one test showed high variances while the other was affected by an inaccurate datum
signal; the low number of effective detector data resulted in defining the detector data as
unreliable. Measured experimental power was compared to three published power
prediction models and Millsoft. The relative error percentages concluded that only Millsoft
best predicted mill input power for both the worn and new liners, at 1871kW and 1887kW,
respectively. These results have provided valuable information on probe design for dry
industrial type mills and mill internal load dynamics.