3875
Prediction of Vertimill Liner Wear by
Using Discrete Element Method
Xiangjun Qiu, Dilek Alkac
Metso
ABSTRACT: Having successfully applied discrete element method (DEM) to the prediction of horizontal
mill liner wear since 2001, we have now extended the predictions of mill liner wear to vertical mills, specifically
Vertimill. In the new algorithm, a parametric model is used to cover the liner surface geometry from new to
worn state, a least square method is used to smooth shear work done on the entire Vertimill screw liner worn
surface, and the liner volume loss due to wear can be translated into the changes of geometric parameters
ensuring the quality of re-mesh. We show the practicality of the DEM model for Vertimill liner wear prediction
with a sample numerical case for the Metso Vertimill (VTM ®).
INTRODUCTION
Many comminution devices have wear components.
Accurate estimation of wear progression on the compo-
nents is of great economic value in comminution industry.
As early as 1998, researchers had explored the possibility of
using DEM to simulate mill liner wear (Cleary 1998). In
2001, it was discovered (Qiu et al., 2001) that dry DEM
model can be used to simulate the whole history of mill
liner wear process within a time period of a few mill revo-
lutions as long as the DEM model meets two criteria: 1.
Linear Proportionality Criterion, which guarantees the
wear life can be reliably predicted by DEM through linear
scale up in time domain, and 2. Energy Spectra Deviation
Creation, which guarantees the particle collisional energy
spectra are not distorted by numerical shortening of the
wear span. Since then, numerous researchers have pub-
lished their numerical results, which were compared with
experimental measurements, demonstrating that using
dry DEM model to simulate the whole history of mill
liner wear process within a few mill revolutions is a valid
approach (see review by Weerasekara et al., 2013, Qiu
2016 and Xu et al., 2020). Since 2001, the discrete ele-
ment method (DEM) has been widely applied to the pre-
diction of liner wear in horizontal mills such as AG/SAG
and Ball mills and achieved great success in comminution
engineering (Qiu 2016). Since 2013, commercial DEM
software (Sarkar et al., 2013) has implemented the func-
tion to simulate horizontal mill liner wear. However, the
progress in transferring it to Vertimill liner wear predictions
has been very slow due to some challenges, mainly three (3)
different obstacles in applying the DEM simulations to the
case of Vertimill. Firstly, a Vertimill has only two to three
geometrically periodic units (screw liner flights), which
make it harder to obtain statistically stable average shear
work over the liner surface, compared to a horizontal mill,
which has much more geometrically periodic units (lifters).
Secondly, Vertimill screw liners cannot be geometrically
represented in a simple 2D profile as horizontal mill liners
can be. Finally, there is a difficulty involved in re-meshing
of worn liner surfaces without deterioration in the quality
of mesh grids. Some researcher attempted to use commer-
cial DEM software to model Vertimill liner wear (Esteves
et al., 2021). However, the model was not practical at all,