Minerals Engineering 69 (2014) 24–28

Contents lists available at ScienceDirect

Minerals Engineering
journal homepage: www.elsevier.com/locate/mineng

The effects of chamber diameter and stirrer design on dry horizontal

stirred mill performance
Okay Altun a,⇑, Hakan Benzer a, Udo Enderle b
a
Hacettepe University, Mining Engineering Department, 06800 Beytepe, Ankara, Turkey
b
Netzsch Feinmahltechnik, Sedanstraße 70, 95100 Selb, Germany

a r t i c l e i n f o a b s t r a c t

Article history: This paper focussed on investigating the effects of chamber diameter and stirrer design on cement grind-
Received 21 August 2013 ing performance of a horizontal type dry stirred mill. Within the scope, pilot scale test works were under-
Accepted 11 July 2014 taken with two different chamber diameters (20.4 cm and 26.4 cm) having the same length and three
different stirrer designs (wing, cross and disc) having the same diameter (16 cm). The chamber diameter
tests were performed at the same stirrer design, media size and media filling. The studies concluded that,
Keywords: the use of larger chamber improved the grinding efficiency since 31.8% and 35.8% less energy was con-
Comminution
sumed than the smaller mill at the RRd50 of 1.41 and 1.66 respectively. This behaviour of the larger mill
Stirred mill
Dry grinding
can be attributed to the increased gap distance between the chamber wall and stirrer edge. With regards
Fine grinding to stirrer design, the statistical evaluations, grinding results and temperature measurements all indicated
Cement grinding that the disc design of stirrer ground the particles more effectively at high energy levels (>40 kW h/t). The
use of the disc design reduced the energy consumption by 21% (at RRd50 of 3.5). This was attributed to
dissipation of energy as heat since the temperature measured for the wing and cross types were higher
than the disc type.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction tested. The performances were compared by considering power

The idea of developing dry horizontal milling came from
increasing demand on energy efficient grinding machines for
cement grinding operation, which is an energy intense process.
Altun (2013) in his PhD study conducted pilot scale test works with
this grinding machine where the effects of operating conditions
and design parameters (chamber diameter, stirrer design) were
investigated. Within the scope of the paper, the effects of chamber
diameter and stirrer design on the cement grinding performance of
the dry horizontal stirred mill are to be presented. The influences
of these parameters were investigated in wet stirred milling stud-
ies. Kwade and Stender (1998) and Stender et al. (2004) discussed
the effects of different chamber sizes and concluded that the use
larger mill reduced the energy utilization at the same product fine-
ness for a given media size. The effects of stirrer design were stud-
ied by Stehr (1988) and Kwade (1999) who pointed out that pinned
design stirrers had higher power densities than the disc design.
For the dry horizontal stirred mill, 2 chamber diameters
(20.4 cm and 26.4 cm) with the same length and 3 stirrer designs
(disc, cross and wing) having the same diameter (16 m) were
⇑ Corresponding author. Tel.: +90 3122977600; fax: +90 3122992155.
E-mail address: okyaltun@hacettepe.edu.tr (O. Altun).
draw measurements, energy consumptions, size distribution anal-
yses and mill chamber temperature measurements which are pre-
sented in the following sections. The results obtained in this study
were discussed with the literature.
2. Materials and methods
2.1. Description of the experimental apparatus
The structure of the mill is similar to IsaMill (Clark, 2007) and
the key components are control panel, feed unit (feed hopper
and rotary valve), grinding chamber and product discharge
(Fig. 1). The technical specifications of the mill are given in Table 1.
For the dry horizontal stirred mill, the feed rate to the mill body is
adjusted via changing the speed of rotary valve placed under the
feed hopper. The throughput of the grinding operation is measured
from the product outlet by cutting the material flow. The power
draw of the mill, which displays on the control panel instantly, is
measured by a torque sensor installed to the mill. In grinding oper-
ation of the mill, the material to be ground flows from feed hopper
into the grinding chamber by gravity and at the same time air is sup-
plied from the feed inlet in order to improve material transportation

http://dx.doi.org/10.1016/j.mineng.2014.07.008
0892-6875/Ó 2014 Elsevier Ltd. All rights reserved.
 O. Altun et al. / Minerals Engineering 69 (2014) 24–28 25

1
2

Fig. 3. Particle size distributions of the materials used in the grinding tests.

4
Table 2
Fineness parameters of the materials used.
Fig. 1. Horizontal stirred mill on the left (1 – feed hopper, 2 – control panel,
3 – grinding chamber, 4 – product outlet). Final product Separator reject
d50 (lm) 15 66
d80 (lm) 32 121
Table 1
Blaine (cm2/g) 3277 710
Technical specifications of the dry stirred mill.

Motor power (kW) 18

Maximum feed rate (kg/h) 500
Stirrer tip speed (m/s) 1.08–9.76 Table 3
Maximum air flow rate (L/h) 1000 Milling conditions of the grinding tests.

Media material Steel

towards the discharge end. The ground particles leave the mill
chamber from the product outlet where a product separator exists
with the aim of retaining the media inside. Product separator is a
cage having openings smaller than the bead size and attached to
the discharge end of the mill directly as illustrated in Fig. 2.
2.2. Material and milling conditions
The material used for the grinding tests was collected from final
product and separator reject streams of a cement grinding circuit
during CEM I 42.5R type cement production. The particle size dis-
tributions and the fineness parameters e.g., mean size and Blaine,
are given in Fig. 3 and Table 2 respectively. Within this study, final
product material was used in testing chamber diameters and the
separator reject material was used for testing different stirrer
designs.
Chamber diameter and stirrer design test works were per-
formed under the milling conditions given in Table 3. Grinding
chemical, which was in liquid form, was used throughout the stud-
ies with the aim of improving particles’ transportation towards the
product outlet.
2.3. Characterization studies
The collected samples from the test studies were subjected to
characterization studies. In terms of characterization, the size dis-
Stirrer material Steel
Chemical dosage (g/t) 700
Air flow rate (L/h) 1000
tribution and Blaine measurements were undertaken via Sympatec
laser sizer (Germany) and Atom Teknik (Turkey) devices. Both
measurements were used in calculating reduction ratio (RR) and
surface area development (Eq. (1)) of the grinding process and then
correlated with the energy consumption.
Surface area development ðDSÞ ¼ Product Blaine  Feed Blaine
ð1Þ
3. Results and discussion
3.1. The effects of chamber diameter
In this section, the comminution results obtained from 2 cham-
bers having different diameters are presented. The dimensions of
both chambers used in test studies are given in Table 4. Through-
out the test studies final product material (Table 2) was ground,
the media composition of 60% 4 mm and 40% 3 mm was charged
and the disc stirrer design (16 cm in diameter) was preferred. As
can be understood, only the mill chamber was changed to examine
the effects on grinding performance. Table 5 presents the test plan.

Fig. 2. Discharge end of the mill with the product separator.

26 O. Altun et al. / Minerals Engineering 69 (2014) 24–28

Table 4 In Fig. 4, the power draws of both chambers measured at the

Dimensions of the two chambers. same operating conditions are illustrated comparatively. As
Mill 1 Mill 2 expected, at the same media filling, feed rate and stirrer tip speed,
Effective diameter (cm) 20.4 26.4 Mill 2 utilizes more power due to higher media content.
Effective length (cm) 74.0 75.0 The particle size distributions of all the test products were mea-
sured and in Fig. 5 some of them, which were obtained at 100 kg/h
feed rate, are illustrated with the operating conditions. The mea-
Table 5 surements show that, the higher the tip speed the finer the prod-
The test plan for the chamber diameter tests. uct. Additionally, operation with higher media filling resulted in
generating more fine particles as expected. The particle size distri-
60% Media filling 70% Media filling
Stirrer tip speed (m/s) Stirrer tip speed (m/s) butions also imply that, Mill 2 produces finer product than Mill 1
that was operated at the same operating conditions. Because Mill
Feed rate (kg/h) 4.3 6.5 9.8 4.3
p p p p 2 has higher media amount inside, performance comparison
100
p p p p depending on energy consumption is more appropriate.
260
p p p The power draw and size distribution data were used in evalu-
400
ating the energy-size reduction as well as the energy-surface area
development performances of the both chambers. Fig. 6 shows the
trends obtained from the test studies. Individual trends imply that
the specific energy consumption is directly proportional to reduc-
tion ratio (RR) parameter. Similar observations have been made by
Herbst and Sepulveda (1978), Stehr (1988), Breitung-Faes and
Kwade (2008) who concluded that energy consumption and the
product mean size were inversely proportional to each other.
When the grinding results of the both chambers are compared, it
is understood that Mill 2 has improved performance since it con-
sumes 31.8% and 35.8% less energy at the RRd50 of 1.41 and 1.66
respectively. Improved performance of Mill 2 can also be followed
from the change in surface area. The contribution of Mill 2 to sur-
face area development is higher than Mill 1 at each energy level. At
the energy level of 40 kW h/t, Mill 2 increases the surface area by
2000 cm2/g where Mill increases by 1500 cm2/g.
Fig. 4. Power draws of both chambers measured during the grinding operation. For wet horizontal stirred mills, the grinding results are indepen-
dent of chamber size when the layouts of the mills and operating

Fig. 5. Some of the particle size distributions of chamber diameter tests.

Fig. 6. Comparison of the grinding performances of the two chambers.

 O. Altun et al. / Minerals Engineering 69 (2014) 24–28 27

Fig. 7. Cross sectional views of Mill 1 and Mill 2.

conditions e.g., media size, media filling are the same (Weller et al.,
2000). Therefore, scaling up from lab to industrial scale is performed Fig. 9. The effects of stirrer design on no load power draw.
accurately since the energy-size reduction curves are the same.
However in this study, the layouts of the mill were changed since
Table 6
larger diameter chamber was used at the same stirrer design (Mill
Results of the paired sample t-test at 95% confidence level.
2) therefore the gap between the stirrer edge and mill chamber wall
increased (Fig. 7). For this reason, improved performance was Std. Std. error t Level of significance
obtained with Mill 2, even if the mills were operated at the same deviation mean (2-tailed)

energy level. As a conclusion of the studies, improved performance Pair 1 (disc-cross 0.02563 0.00969 6.93 0
of Mill 2 is attributed to change in internal dimension of the grind- type)
Pair 2 (disc-wing 0.03436 0.1299 6.049 0.001
ing environment. type)
Pair 3 (cross- 0.01773 0.0067 1.706 0.139
3.2. The effects of stirrer design wing type)

The stirrers are the key components for stirred milling therefore
their design is important since it directly affects the grinding effi-
ciency. Within the study, the grinding performances of 3 stirrer
designs having the same diameter (16 cm) were tested. Photo-
graphs of the stirrer designs tested are illustrated in Fig. 8.
Throughout the test studies, Mill 2 was used at 60% media filling
with 4 mm media inside and separator reject material was ground.
In order to produce different grinding results, the mill was oper-
ated at 2 different stirrer speeds, which were 3.3 m/s, 4.3 m/s,
and 3 different feed rates, which were 100 kg/h, 260 kg/h and
400 kg/h. Totally 12 sets of data were obtained.
In Fig. 9 the no load power draws of each design at different tip
speeds are illustrated. From the measurements it is understood
that the disc stirrer design draws less power than the cross and
wing designs. The statistical evaluations presented in Table 6 also
indicate that the disc design is significantly different from the Fig. 10. Some of the particle size distributions obtained from stirrer design tests.
other designs and it draws less power.
In Fig. 10, some of the product particle size distributions, which
were obtained at 260 kg/h feed rate, are illustrated comparatively. in energy consumptions can reach up to 21% (at RRd50 of 3.5). Sim-
The measurements show that, the higher the tip speed the finer the ilar observations are valid for surface area development as well. For
product. Additionally, it is observed that, the disc design stirrer the energy level higher than 40 kW h/t, disc design contributes
grinds the particles more effectively therefore produces the fine more to surface area development.
size distribution among the given designs. The literature reports that, the pinned stirrers are able to draw
For each stirrer design, the variation of reduction ratio (RR) and more power owing to its energy transfer mechanism from stirrer to
surface area with the specific energy consumption is illustrated in media (Kwade, 1999). These designs of stirrers lift and throw more
Fig. 11. It is understood that, at higher energy levels (>40 kW h/t) media per revolution due to its shape hence agitate the media
wing and cross design stirrers utilize more energy than the disc charge more vigorously. Consequently, the number of media to
design to achieve the same degree of size reduction. The difference media and media to chamber wall contacts increase. For this

Fig. 8. From left to right, alignments of wing, disc and cross stirrer designs.
28 O. Altun et al. / Minerals Engineering 69 (2014) 24–28
Fig. 11. The variation of reduction ratio with specific energy consumption.
Fig. 12. Mill chamber temperature measurements of the stirrer designs.
reason, it is highly probable that the kinetic energy gained by a
media is dissipated as heat due to the number collisions inside
and ultimately the mill performance is deteriorated. In this respect,
temperature measurements from the mill chamber can indicate
whether this phenomena is valid for the dry horizontal stirred mill.
The measurements were undertaken at different grinding time
intervals via infrared or non-contact thermometer for the tests per-
formed at around 70 kW h/t (Fig. 12). As illustrated in Fig. 12, con-
stant temperature (90 °C) applies after 25 min of grinding for the
disc design whereas the others continue to rise (>100 °C). As a con-
clusion, decreased performances of the wing and cross designs can
be attributed to the dissipation of energy as heat.
4. Conclusions
This paper presents the cement grinding results of the pilot
scale test works performed with dry horizontal stirred mill where
the effects of chamber diameter and stirrer designs were investi-
gated. Within the context of the study, different chamber diame-
ters having the same length were tested at the same stirrer
design (disc) and their performances were compared based on
the energy-size reduction and the energy-surface area develop-
ment relationships. The studies concluded that the larger mill (Mill
2) consumed 31.8% and 35.8% less energy than Mill 1 at the RRd50
of 1.41 and 1.66. The improvement in the grinding performances
can also be followed from surface area developments as well. For
wet horizontal stirred mills, the energy-size reduction curves are
independent of chamber size when the layouts of the mills and
operating conditions are the same. For the chamber diameter tests,
the layouts of the mill were changed since larger diameter cham-
ber was used at the same stirrer design (Mill 2) therefore the gap
between the stirrer edge and mill chamber wall increased. As a
conclusion, Mill 2 performed more efficient grinding operation
than Mill 1.
Stirrer design tests were conducted with Mill 2 for 3 different
designs named as cross, wing and disc having the same diameter
(16 cm). The results showed that, at higher energy levels
(>40 kW h/t) wing and cross design stirrers utilize more energy
than the disc design to achieve the same degree of size reduction.
The difference in energy consumptions can reach up to 21% (at
RRd50 of 3.5). Similar observations are valid for surface area devel-
opment parameter as well. Reduced milling performance of the
wing and cross designs were attributed to increased mill chamber
temperature that indicated most of the energy was dissipated as
heat.
Acknowledgements
Authors appreciate contributions of Assoc. Prof. Abdullah Obut
(Hacettepe University), Assoc. Prof. Aubrey Mainza (University of
Cape Town), Dr. Namık Aydogan, Research Assistant Firat Atalay
(Hacettepe University) for their assistance in writing, Research
Assistant Alper Toprak (Hacettepe University), Konuray Demir
(Loesche GmbH) and SET Italcementi Ankara Cement Plant for their
assistance in laboratory studies and Hacettepe University Research
Foundation Unit (Project No: 013 T06 604 006) for the financial
support.
References
Altun, O., 2013. Investigation of Dry Horizontal Stirred Milling Applications for
Cement Grinding Circuits. Ph. D. Thesis. Hacettepe University, Ankara, Turkey.
Breitung-Faes, S., Kwade, A., 2008. Nano particle production in high power density
mills. Chem. Eng. Res. Des. 86 (4), 390–394.
Clark, L., 2007. Grinding comparison test of IsaMills with Tower Mills using
Magnetite. Xstrata Technology, Technical Notes.
Herbst, J.A., Sepulveda, J.L., 1978. Fundamentals of fine and ultrafine grinding in a
stirred ball mill. In: International Powder and Bulk Solids Handling and
Processing: Proceedings held Rosemount Illinois, pp. 452–470.
Kwade, A., 1999. Wet comminution in stirred media mills – research and it practical
application. Powder Technol. 105 (1–3), 14–20.
Kwade, A., Stender, H., 1998. Constant grinding results at scale-up of stirred media
mills. Aufbereitungs Tech. 39 (8), 373–382.
Stehr, N., 1988. Recent developments in stirred ball milling. Int. J. Miner. Process. 22
(1–4), 431–444.
Stender, H., Kwade, A., Schwedes, J., 2004. Stress energy distribution in different
stirred media mill geometries. Int. J. Miner. Process. 74, 103–117.
Weller, K.R., Gao, M., Bowen, P., 2000. Scaling-up horizontal Stirred mills from a 4-
litre test mill. In: Powder Technology Symposium. Pennsylvania State
University, USA.