XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3889
Clark, D., Rodabough, D., Kane, A., Kone, B., Rajala,
J., Avenido, N., Bartholomew, K. and McIvor, R.E.
(2023). Improvements in grinding performance at the
Fekola gold mill, Mali West Africa. SAG Conference
Proceedings, Vancouver, 25 p.
Conger, W., DuPont, J-F, McIvor, R.E. and Weldum, T.P.
2018. Ball mill media optimization through functional
performance modelling. Mining Engineering, Nov.,
28–37.
Davis, E.W. 1925. Ball mill crushing in closed circuit with
screens. Bulletin of the University of Minnesota School
of Mines Experimental Station, v. 28, no. 42, p. 14–18.
Davis, E.W. 1946. Pulp densities (and size distributions*)
within operating ball mills. AIME Trans., v. 169,
p. 155–159. (*important content not mentioned in
title of article.).
Epstein, B. 1948. Logarithmico-normal distribution of
breakage of solids. Industrial and Engineering Chemistry,
v. 40, no. 12, p.2289–2291.
Finch, J.A. and Ramirez-Castro, J. 1981. Modeling min-
eral size reduction in the closed-circuit ball mill at Pine
Point concentrator. International Journal of Mineral
Processing, v. 8, p. 61–78.
Hinde, A.L. and Kalala, J.T. 2009. The application of a
simplified approach to modelling tumbling mills,
stirred media mills and HPGR’s. Minerals Engineering,
v. 22, p.633–641. (ag sag Mills +HPGRs uses.).
Laplante, A.R., Finch, J.A. and del Villar, R. 1987.
Simplification of the grinding equation for plant simu-
lation. Trans. IMM, Sec. C, v.96, C108–112.
McIvor, R.E., Bartholomew, K.M., Arafat, O.A., Daavettila,
S. and Savola, E. 2021. Coarse particle retention test-
ing on a ball mill at Lundin’s Eagle Mine Humboldt
concentrator, Proceedings of Australian Institute of
Mining and Metallurgy Mill Operators Conference,
Brisbane.
McIvor, R.E. 1988. Technoeconomic analysis of plant
grinding operations, Ph.D. thesis, McGill University,
Montreal.
McIvor, R.E., Lavallee, M.L., Wood, K.R., Blythe, P.M.
and Finch, J.A. 1990. Functional performance char-
acteristics of ball milling. Mining Engineering, March,
p.269–276.
McIvor, R.E. 1988. Classification effects in wet ball milling
circuits. Mining Engineering, Aug., p. 815–820.
McIvor, R.E. 1984. A material balance calculation proce-
dure for grinding circuit hydrocyclone selection. CIM
Bulletin, Dec., p. 50–53.
McIvor, R.E., Duval, L. and Leclercq, L. 1991. Use of func-
tional performance analysis for plant grinding media
testing and selection. Cobre ’91, v.2, p.145–155.
McIvor, R.E., Roux, Y. and Roy, M. 1994. Comparative
efficiency determinations of open circuit ball milling at
la Compagnie Miniere Quebec Cartier. CIM Bulletin,
July-Aug., p. 69–70.
McIvor, R.E., Weldum, T.P., Mahoski, B.J. and Rasmussen,
R.S. 2000. Systems approach to grinding improve-
ments at the Tilden concentrator. Mining Engineering,
Feb., p.41–47.
McIvor, R.E. 2006. Industrial validation of the functional
performance equation for ball and pebble milling cir-
cuits. Mining Engineering, Nov., p. 47–51.
McIvor, R.E. and Finch, J.A. 2007. The Finch-McIvor
functional performance-based grinding circuit model-
ing system, Proceedings of the Annual Meeting of the
Canadian Mineral Processors, Ottawa.
McIvor, R.E. 2014. Plant performance improvements
using grinding circuit “classification system efficiency.”
Mining Engineering, Sept., p.72–76.
McIvor, R.E., Bartholomew, K.M., Arafat, O.A. and Finch,
J.A. 2017. Ball mill classification system optimization
through functional performance modelling. Mining
Engineering, Nov., 18–31.
McIvor, R.E. and Makni, S. 2023. The effects of discharge
design on wet ball milling performance. SME Annual
Meeting, Denver, preprint 23-057, 5 p.
McIvor, R.E., Finch, J.A., and Makni, S. 2023. The truth
about population balance modelling. CIM Journal,
DOI: 10.1080/19236026.2023.2204708.
Metcom Technologies, Inc. 2023. On-line training program
to measure and improve the processing performance of
plant grinding operations. www.metcomtech.com.
Plitt, L.R. (1976). A mathematical model of the hydrocy-
clone classifier. CIM Bulletin, Dec., p.114–123.
Roberts, E.J. 1950. The probability theory of wet ball mill-
ing and its application. AIME Trans., v. 187, Mining
Engineering, Dec.: p. 1267–72.

Previous Page Next Page

Extracted Text (may have errors)

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3889
Clark, D., Rodabough, D., Kane, A., Kone, B., Rajala,
J., Avenido, N., Bartholomew, K. and McIvor, R.E.
(2023). Improvements in grinding performance at the
Fekola gold mill, Mali West Africa. SAG Conference
Proceedings, Vancouver, 25 p.
Conger, W., DuPont, J-F, McIvor, R.E. and Weldum, T.P.
2018. Ball mill media optimization through functional
performance modelling. Mining Engineering, Nov.,
28–37.
Davis, E.W. 1925. Ball mill crushing in closed circuit with
screens. Bulletin of the University of Minnesota School
of Mines Experimental Station, v. 28, no. 42, p. 14–18.
Davis, E.W. 1946. Pulp densities (and size distributions*)
within operating ball mills. AIME Trans., v. 169,
p. 155–159. (*important content not mentioned in
title of article.).
Epstein, B. 1948. Logarithmico-normal distribution of
breakage of solids. Industrial and Engineering Chemistry,
v. 40, no. 12, p.2289–2291.
Finch, J.A. and Ramirez-Castro, J. 1981. Modeling min-
eral size reduction in the closed-circuit ball mill at Pine
Point concentrator. International Journal of Mineral
Processing, v. 8, p. 61–78.
Hinde, A.L. and Kalala, J.T. 2009. The application of a
simplified approach to modelling tumbling mills,
stirred media mills and HPGR’s. Minerals Engineering,
v. 22, p.633–641. (ag sag Mills +HPGRs uses.).
Laplante, A.R., Finch, J.A. and del Villar, R. 1987.
Simplification of the grinding equation for plant simu-
lation. Trans. IMM, Sec. C, v.96, C108–112.
McIvor, R.E., Bartholomew, K.M., Arafat, O.A., Daavettila,
S. and Savola, E. 2021. Coarse particle retention test-
ing on a ball mill at Lundin’s Eagle Mine Humboldt
concentrator, Proceedings of Australian Institute of
Mining and Metallurgy Mill Operators Conference,
Brisbane.
McIvor, R.E. 1988. Technoeconomic analysis of plant
grinding operations, Ph.D. thesis, McGill University,
Montreal.
McIvor, R.E., Lavallee, M.L., Wood, K.R., Blythe, P.M.
and Finch, J.A. 1990. Functional performance char-
acteristics of ball milling. Mining Engineering, March,
p.269–276.
McIvor, R.E. 1988. Classification effects in wet ball milling
circuits. Mining Engineering, Aug., p. 815–820.
McIvor, R.E. 1984. A material balance calculation proce-
dure for grinding circuit hydrocyclone selection. CIM
Bulletin, Dec., p. 50–53.
McIvor, R.E., Duval, L. and Leclercq, L. 1991. Use of func-
tional performance analysis for plant grinding media
testing and selection. Cobre ’91, v.2, p.145–155.
McIvor, R.E., Roux, Y. and Roy, M. 1994. Comparative
efficiency determinations of open circuit ball milling at
la Compagnie Miniere Quebec Cartier. CIM Bulletin,
July-Aug., p. 69–70.
McIvor, R.E., Weldum, T.P., Mahoski, B.J. and Rasmussen,
R.S. 2000. Systems approach to grinding improve-
ments at the Tilden concentrator. Mining Engineering,
Feb., p.41–47.
McIvor, R.E. 2006. Industrial validation of the functional
performance equation for ball and pebble milling cir-
cuits. Mining Engineering, Nov., p. 47–51.
McIvor, R.E. and Finch, J.A. 2007. The Finch-McIvor
functional performance-based grinding circuit model-
ing system, Proceedings of the Annual Meeting of the
Canadian Mineral Processors, Ottawa.
McIvor, R.E. 2014. Plant performance improvements
using grinding circuit “classification system efficiency.”
Mining Engineering, Sept., p.72–76.
McIvor, R.E., Bartholomew, K.M., Arafat, O.A. and Finch,
J.A. 2017. Ball mill classification system optimization
through functional performance modelling. Mining
Engineering, Nov., 18–31.
McIvor, R.E. and Makni, S. 2023. The effects of discharge
design on wet ball milling performance. SME Annual
Meeting, Denver, preprint 23-057, 5 p.
McIvor, R.E., Finch, J.A., and Makni, S. 2023. The truth
about population balance modelling. CIM Journal,
DOI: 10.1080/19236026.2023.2204708.
Metcom Technologies, Inc. 2023. On-line training program
to measure and improve the processing performance of
plant grinding operations. www.metcomtech.com.
Plitt, L.R. (1976). A mathematical model of the hydrocy-
clone classifier. CIM Bulletin, Dec., p.114–123.
Roberts, E.J. 1950. The probability theory of wet ball mill-
ing and its application. AIME Trans., v. 187, Mining
Engineering, Dec.: p. 1267–72.

Help

Destination page number Search scope Search Text

loading

3890
Vertimill Wear Sensing
Sudarshan Martins, Xiangjun Qiu
Metso
ABSTRACT: The Vertimill is a highly effective and reliable grinding solution consequently, it has been widely
adopted. However, like all other grinding technologies, it has faced challenges in managing wear life and
measuring charge levels. To address these issues, a new detection system has been developed. It offers a novel
way to determine and manage wear life. This solution is described, highlighting its capabilities and benefits.
Furthermore, the results of DEM simulations of the Vertimill, including wear processes, are presented. The
simulations confirm that the measurements provided by the new solution can estimate the state of wear of the
liners, producing an important tool for effective maintenance and management of the Vertimill. Overall, this
new sensor solution represents a positive advancement in the field of grinding technology and has the potential
to improve the efficiency and effectiveness of the Vertimill.
INTRODUCTION
The Vertimill is a vertical stirred mill used in the mining and
minerals processing industry for grinding materials to a fine
particle size. The motor turns the Vertimill screw, stirring
the grinding media and the slurry, as shown in Figure 1.
This stirring motion promotes a charge motion that
creates an efficient grinding environment for the ore.
However, this effect is not confined to the ore. The grinding
media and the screw liners are also subject to the effects of
the charge motion—they wear. The monitoring of the state
of the liner wear is addressed here. First, a description of
the major components of the technology is given. Next, the
results of one of the field deployments is presented. Finally,
the features and capabilities of the system are discussed.
THE VERTIMILL DEM
DEM, a computational mechanics method addressing
granular flow, has been widely used in the comminution
industry (Qiu 2016). In 2001, a DEM model, enhanced
with Archard’s wear mechanism, was developed to predict
liner wear in horizontal mills such as AG/SAG and Ball
Figure 1. A Vertimill

Previous Page Next Page

Extracted Text (may have errors)

3890
Vertimill Wear Sensing
Sudarshan Martins, Xiangjun Qiu
Metso
ABSTRACT: The Vertimill is a highly effective and reliable grinding solution consequently, it has been widely
adopted. However, like all other grinding technologies, it has faced challenges in managing wear life and
measuring charge levels. To address these issues, a new detection system has been developed. It offers a novel
way to determine and manage wear life. This solution is described, highlighting its capabilities and benefits.
Furthermore, the results of DEM simulations of the Vertimill, including wear processes, are presented. The
simulations confirm that the measurements provided by the new solution can estimate the state of wear of the
liners, producing an important tool for effective maintenance and management of the Vertimill. Overall, this
new sensor solution represents a positive advancement in the field of grinding technology and has the potential
to improve the efficiency and effectiveness of the Vertimill.
INTRODUCTION
The Vertimill is a vertical stirred mill used in the mining and
minerals processing industry for grinding materials to a fine
particle size. The motor turns the Vertimill screw, stirring
the grinding media and the slurry, as shown in Figure 1.
This stirring motion promotes a charge motion that
creates an efficient grinding environment for the ore.
However, this effect is not confined to the ore. The grinding
media and the screw liners are also subject to the effects of
the charge motion—they wear. The monitoring of the state
of the liner wear is addressed here. First, a description of
the major components of the technology is given. Next, the
results of one of the field deployments is presented. Finally,
the features and capabilities of the system are discussed.
THE VERTIMILL DEM
DEM, a computational mechanics method addressing
granular flow, has been widely used in the comminution
industry (Qiu 2016). In 2001, a DEM model, enhanced
with Archard’s wear mechanism, was developed to predict
liner wear in horizontal mills such as AG/SAG and Ball
Figure 1. A Vertimill

3888 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
prevalent. These programs can also be used to simulate the
effect of feed ore changes on circuit production capabil-
ity and link multiple stages of plant comminution circuits,
thereby proving useful for production planning.
The limitations of the breakage and selection mill models
used in population balance modeling have recently been
documented (McIvor et al, 2023). Paramount among these
is the false assumption of a (given, arbitrary) self-similar
breakage function (a fixed, relative to only to the particle
size, breakage event product sizing), which is needed to
allow calculation of individual particle breakage rates. One
can change the breakage function that was assumed to
construct the mill model, and the selection function (indi-
vidual particle size by size grinding rates) will simply adapt
to create a fit to the data. While the ball mill model can
correctly calculate trends in the mill product sizing as mill
feed throughput rates, size distributions, and ore grindabili-
ties change, the assumed breakage function and subsequent
selection function used are fictitious. As a result, the model
cannot deal with changes to the mill grinding environment,
like media sizing, the slurry percent solids, or slurry hold-
up, without even more egregious assumption-making.
Population balance models do not require accurate
data, only balanced data, which, unfortunately, can be
derived from surveys conducted with poor sampling.
Circulating load ratio estimates made from survey data are
sensitive to the size distributions used in the calculation.
However, since pumping is not considered in the popu-
lation balance-based circuit modeling methodology, inac-
curate circulating load ratios resulting from poor survey
data are tolerated. In contrast, the ACGR mill and cyclone
separation models use discrete values calculated only from
accurate survey data, and cyclone feed pump mass flow rate
and total dynamic head are integral to the circuit model.
Exploiting pump (and its motor) capacity are critical CSE
improvements, and as such, poor quality data is identified
and stricken from further consideration. Accordingly, when
applying ACGR and Functional Performance methods,
great efforts are made to ensure sampling practices result
in high-quality data needing little adjustment in the mass
balancing step.
CONCLUSIONS
Functional Performance identifies two ball mill circuit effi-
ciencies, each related to specific circuit design and operating
variables. These two circuit efficiencies provide quantitative
metrics for ball mill circuit performance, allowing bench-
marking and optimization. For example, CSE is maxi-
mized through intentional, step-by-step adjustments to
the pump and cyclones or potentially the use of alternative
classification equipment, and MGrEff is maximized through
alterations to the mill grinding environment.
The Apparent Cumulative Grinding Rates (ACGR)
model is well-proven and provides practitioners a way of
evaluating ball mill circuit classification changes without
unnecessary and unverifiable assumptions built in into
breakage-selection-based models. Additionally, including
the cyclone feed pump in the circuit model is critical to
practical optimization exercises.
Functional Performance Analysis coupled with the
ACGR mill model and discrete, point-by-point (non-curve
fit) classifier performance models (including pumping)
provides a powerful, practical, and proven alternative to
classical circuit computer modelling and simulation.
Continued development of the understanding of the
ACGR model in relation to mill environment parameters,
such as percent solids and media size, is ongoing and pro-
vides a ripe field of study unclouded by the complexity of
population balance methods. By focusing on the grinding
rate at the size of interest for a given ball mill circuit, which
corresponds to the Functional Performance mill grinding
rate, practitioners now have a clear optimization objective
for ball mill studies.
ACKNOWLEDGMENTS
The authors acknowledge the Tilden Mine and the Fekola
Mine for providing the opportunity to collect the data
used in the examples in this paper. The authors also thank
Professor Jim Finch for his ongoing guidance, especially in
development of the ACGR mill model.
REFERENCES
Amestica, R., Gonzales, G.D., Barria, J., Magne, L.,
Menacho, J. and Castro, O. 1993. A SAG mill circuits
dynamic simulator based on a simplified mechanis-
tic model. Proceedings of XVlll IMPC, Sydney, v. 1,
p.117–129.
Bartholomew, K.M., Arafat, O.A., McIvor, R.E. 2018.
A guide to maximizing ball mill circuit classification
system efficiency (CSE) for operators and equip-
ment designers. Proceedings of Australian Institute of
Mining and Metallurgy Mill Operators Conference.
Bowdish, F.W. 1960. Theoretical and experimental studies
of the kinetics of grinding in a ball mill. AIME Trans.,
v. 217, p. 194–202.
Ciribeni, W., Bertero, R., Tello, A., Puerta, M., Avellq,
E., Paez, M. and Menendez Aguado, J.M. (2020).
Application of the cumulative kinetic model in the
comminution of critical ores. Metals, v. 10, no. 925,
11p., DOI: 10.3390/met10070925.

Previous Page Next Page

Extracted Text (may have errors)

3888 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
prevalent. These programs can also be used to simulate the
effect of feed ore changes on circuit production capabil-
ity and link multiple stages of plant comminution circuits,
thereby proving useful for production planning.
The limitations of the breakage and selection mill models
used in population balance modeling have recently been
documented (McIvor et al, 2023). Paramount among these
is the false assumption of a (given, arbitrary) self-similar
breakage function (a fixed, relative to only to the particle
size, breakage event product sizing), which is needed to
allow calculation of individual particle breakage rates. One
can change the breakage function that was assumed to
construct the mill model, and the selection function (indi-
vidual particle size by size grinding rates) will simply adapt
to create a fit to the data. While the ball mill model can
correctly calculate trends in the mill product sizing as mill
feed throughput rates, size distributions, and ore grindabili-
ties change, the assumed breakage function and subsequent
selection function used are fictitious. As a result, the model
cannot deal with changes to the mill grinding environment,
like media sizing, the slurry percent solids, or slurry hold-
up, without even more egregious assumption-making.
Population balance models do not require accurate
data, only balanced data, which, unfortunately, can be
derived from surveys conducted with poor sampling.
Circulating load ratio estimates made from survey data are
sensitive to the size distributions used in the calculation.
However, since pumping is not considered in the popu-
lation balance-based circuit modeling methodology, inac-
curate circulating load ratios resulting from poor survey
data are tolerated. In contrast, the ACGR mill and cyclone
separation models use discrete values calculated only from
accurate survey data, and cyclone feed pump mass flow rate
and total dynamic head are integral to the circuit model.
Exploiting pump (and its motor) capacity are critical CSE
improvements, and as such, poor quality data is identified
and stricken from further consideration. Accordingly, when
applying ACGR and Functional Performance methods,
great efforts are made to ensure sampling practices result
in high-quality data needing little adjustment in the mass
balancing step.
CONCLUSIONS
Functional Performance identifies two ball mill circuit effi-
ciencies, each related to specific circuit design and operating
variables. These two circuit efficiencies provide quantitative
metrics for ball mill circuit performance, allowing bench-
marking and optimization. For example, CSE is maxi-
mized through intentional, step-by-step adjustments to
the pump and cyclones or potentially the use of alternative
classification equipment, and MGrEff is maximized through
alterations to the mill grinding environment.
The Apparent Cumulative Grinding Rates (ACGR)
model is well-proven and provides practitioners a way of
evaluating ball mill circuit classification changes without
unnecessary and unverifiable assumptions built in into
breakage-selection-based models. Additionally, including
the cyclone feed pump in the circuit model is critical to
practical optimization exercises.
Functional Performance Analysis coupled with the
ACGR mill model and discrete, point-by-point (non-curve
fit) classifier performance models (including pumping)
provides a powerful, practical, and proven alternative to
classical circuit computer modelling and simulation.
Continued development of the understanding of the
ACGR model in relation to mill environment parameters,
such as percent solids and media size, is ongoing and pro-
vides a ripe field of study unclouded by the complexity of
population balance methods. By focusing on the grinding
rate at the size of interest for a given ball mill circuit, which
corresponds to the Functional Performance mill grinding
rate, practitioners now have a clear optimization objective
for ball mill studies.
ACKNOWLEDGMENTS
The authors acknowledge the Tilden Mine and the Fekola
Mine for providing the opportunity to collect the data
used in the examples in this paper. The authors also thank
Professor Jim Finch for his ongoing guidance, especially in
development of the ACGR mill model.
REFERENCES
Amestica, R., Gonzales, G.D., Barria, J., Magne, L.,
Menacho, J. and Castro, O. 1993. A SAG mill circuits
dynamic simulator based on a simplified mechanis-
tic model. Proceedings of XVlll IMPC, Sydney, v. 1,
p.117–129.
Bartholomew, K.M., Arafat, O.A., McIvor, R.E. 2018.
A guide to maximizing ball mill circuit classification
system efficiency (CSE) for operators and equip-
ment designers. Proceedings of Australian Institute of
Mining and Metallurgy Mill Operators Conference.
Bowdish, F.W. 1960. Theoretical and experimental studies
of the kinetics of grinding in a ball mill. AIME Trans.,
v. 217, p. 194–202.
Ciribeni, W., Bertero, R., Tello, A., Puerta, M., Avellq,
E., Paez, M. and Menendez Aguado, J.M. (2020).
Application of the cumulative kinetic model in the
comminution of critical ores. Metals, v. 10, no. 925,
11p., DOI: 10.3390/met10070925.