Physicochemical Problems

Physicochem. Probl. Miner. Process. 50(2), 2014, 783−794

of Mineral Processing
ISSN 1643-1049 (print)
www.minproc.pwr.wroc.pl/journal/
ISSN 2084-4735 (online)

Received November 11, 2013; reviewed; accepted, March 8, 2014

THE ROLE OF ORE PROPERTIES

IN THICKENING PROCESS

Majid UNESI*, Mohammad NOAPARAST**,

Seiyd Ziaedin SHAFAEI**, Esmaeil JORJANI*

*
Department of Mining Engineering, Science and Research branch, Islamic Azad University, Tehran, Iran
**
University of Tehran, Tehran, Iran
*
Corresponding author. E-mail address: noparast@ut.ac.ir

Abstract: The role of ore properties (density, particle size, and mineralogy) in thickening process was
studied in this research. The shaking table was used to prepare the sample for the tests. The tailings were
continuously fed on the table by gravity to separate the tailings in three products as slime, middling and
coarse particles. The solid density and particle size of the samples were different. To study the effect of
mineralogical properties, the sedimentation behavior of the feed and middling samples were tested. The
results showed that the free settling velocity of the feed (2–6 mm/s) was less than that of the middling
sample (18–23 mm/s), and the compressibility of middling (density: 0.63–0.86 Mg/m3) was more than
that of feed (density: 0.33–0.47 Mg/m3). This was due to the amount of clay reduction in the middling
sample. The sedimentation behavior of the slime and the coarse samples were also compared in order to
study the effect of particle size and density. The settling velocity of the slime and the coarse particles was
obtained as 0.1-0.4 and 26 mm/s, respectively, and the maximum underflow density were obtained as
0.35 and 1.57 Mg/m3, respectively. Therefore, the particle size and density reduction reduced the
thickener performance. In order to study the effect of particle size, the sedimentation behavior of the
slime and coarse samples were compared, and it was obtained that the settling velocity and underflow
density increased with the increasing in the particle size.

Keywords: density, particle size, mineralogy, settling flux, underflow density, thickening

Introduction
Mineral processing techniques are performed in environments which contain a
significant amount of water, and in many cases, the most of water is removed with
tailings. Therefore, it is necessary to use dewatering equipments such as thickeners in
order to avoid environment contamination and water recycling (Burger et al. 1999).
There are many models and experiments which have been presented by researchers
to understand the thickening process and to predict the relationship between process

http://dx.doi.org/10.5277/ppmp140229
784 M. Unesi, M. Noaparast, S.Z. Shafaei, E. Jorjani

parameters. A sedimentation relationship was first proposed by Coe and Clevenger in

1916 which was not practical for flocculated suspensions. Kynch provided a
breakthrough in suspension dewatering theory in 1952 which shows that the
relationship between concentration and height can be used when the slurry is treated
with a flocculating agent. Although the Kynch theory is not suitable for all mineral
slurries, it is still used for thickener design calculations. Talmage and Fitch (1955)
applied the mathematical approach of Kynch for thickener design and thereafter
Yoshioka et al. (1957) obtained the thickener area by using a settling flux. Moreover,
Oltmann (1975) suggested a simple empirical approach to the critical flux
determination, and hence thickener area. The approximate depth of the thickened
sludge layer was readily determined by the method outlined by Osborne in 1977, and
the Wilhelm-Naide model (1981) used the underflow concentration to define the unit
area which is still used for thickeners design. Also Dahlstrom and Fitch (1985) and
Yalcin (1988) proposed a relationship between settling flux and unit area. A graphical
method was presented by Kelly and Spottiswood (1989) to determine the settling flux
(Buscall et al. 1987). Furthermore, several models have been developed to simulate
dewatering behavior by chemical engineers (Bustos et al. 1999, Concha et al. 1996,
Deventer et al. 2011). A 1-D model of dewatering was developed by Buscall and
White (1987) which quantify the solids volume fraction, the compressive yield stress,
and the hindered settling function (Garrido et al. 2003a). In addition, Landman and
White (1994) used a phenomenon model which was a combination of empirical and
theoretical models, and they described the solid flux function and effective solid
stress. They also determined the forces acting on particles in the sedimentation and
consolidation process (Garrido et al. 2003b). The BW theory, which was developed by
Green (1997), properly modeled the comprehensive yield stress Py(φ) and hindered
settling function R(φ) in the suspension bed, but not sedimentation above the bed
(Gladman et al. 2010). The Kynch theory, developed by Garrido et al. (2003),
modeled the sedimentation-consolidation process (Green 1997). Garrido et al. (2003)
developed a software for design and simulation of batch and continuous thickening.
Thereafter, further studies were conducted by other researchers. An algorithm has
been developed by Usher et al. (2005) to account for the underflow concentration and
sediment bed from fundamental suspension properties (Landman et al. 1988). The
suspension dewatering equations were proposed by Usher et al. (2009) based on the
aggregate densification where by aggregate compact and become smaller when
subjected to forces in thickening process. They presented the liquid flow velocity
around and through aggregates (Lester et al. 2010). The validation of the Usher
algorithm was studied by Gladman et al. (2010). They stated that this model was most
accurate at the shortest residence times and lowest bed heights, and became poorer for
longer residence time and higher beds. This was due to changes in dewatering
properties of flocculated aggregates over time which is not adequately considered in
the Usher algorithm. Lester et al. (2011) developed the 2-dimensional model of BW.
They used the continuity, separation, and transport equations in the modeling. Since
 The role of ore properties in thickening process 785

solving this model by computational methods was difficult, they used the CFD method
for modeling (Mcfarlane et al. 2005). Van Deventer et al. (2011) developed the Kynch
theory which based on aggregate densification behavior. Additionally, the aggregate
densification theory was used to predict the final equilibrium bed height by
densification rate and bed compression. Also, the relationship between aggregate size
and thickening time was obtained (Usher and Scales 2005).
As mentioned above, the role of ore properties has not been studied by researchers,
and most of them have focused on the effect of suspension properties in dewatering
process. Therefore, the aim of this study was to study the influence of ore properties
such as solid density, particle size, and mineralogy on the thickening process in order
to better understand the dewatering behavior in these thickeners.

Materials and methods

The tailings obtained from the Sarcheshmeh copper flotation plants in Iran were used
for all experiments. As the shaking table could separate the different particles based on
their density and size, the tailings samples were initially continuously fed on the table
by gravity to separate the tailings in three products as a slime “c”, middling “b”, and
coarse “a”. The particle size analysis, XRD and mineralogical studies were carried out
to understand the sample properties. The laboratory screens and cylosizer were used
for particle size analysis, and the results are shown in Table 1 and Fig. 1. The five
Warman cyclosizers were used to analysis the particle size less than 44 μm (i.e. +44,
+33, +23, +15, +11, and –11 μm). This cyclosizer worked at 2130 N/m2 pressure for
20 min. The solid density was determined by a pycnometer (100 cm3) and the results
are presented in Table 2 along with the XRD analysis shown in Fig. 2. The
microscopic mineralogical analysis was carried out to find out the metallic minerals in
the samples (Table 3). In order to study the effect of the ore properties on the
thickening process, the experiments were carried out on the plant conditions. The
samples were taken as a suspension from the shaking table and prepared by tailings
waste water, hence the thickening behavior can only be attributed to the ore properties.
The sample suspension density was measured by Marcy scale and prepared to required
density based on experimental conditions. The pH of the sample was kept as 11 which
was the pH in the plant.
A high molecular weight of anionic polyacrylamide (NF43U from SNF) was used
to flocculate the suspension and prepared at 0.25 g/dm3 which were the industrial
dosage. A 500 cm3 of cylinder was used to study the thickening behavior and a
chronometer was used to record the mud lines at different times.
786 M. Unesi, M. Noaparast, S.Z. Shafaei, E. Jorjani

100

80

Cumulative Undersize (%) 60

40

Table Feed
20 Table Coarser Size
Table Slime
Table Middlings
0
1 10 100 1000
Particle Size (μm)

Fig. 1. Particle size distribution of feed, slime, middling, and coarse samples

Table 1. Particle size analysis of the samples

Feed Slime Middling Coarse

d80 (μm) 120 32 120 160
d75 (μm) 105 28 105 140
d50 (μm) 40 8.5 35 70
d25 (μm) 8 2 6 20

Table 2. XRD analysis and solid density of the samples

Minerals Feed Slime Middling Coarse

aluminum silicate
75.6 84.2 43 –
(clinochlore, illite, chamosite, anorthite) (%)
quartz: SiO2 (%) 22 15.6 53.6 15.6
pyrite: FeS2 (%) 2.3 - 2.4 84.2
solid density (g/cm3) 2.8 2.7 2.8 4.7

Table 3. Mineralogical analysis of the samples

Sample Chalcopyrite Pyrite Sphalerite Hematite Non-metallic Minerals

Table Feed (%) 0.26 11.621 0 0.302 87.8
Table Slime (%) 0.087 12.244 0 0.392 87.243
Table Middling (%) 0.166 12.352 0.088 0.276 87.066
Table Coarser (%) 1.646 55.256 0.118 0.736 42.243
 The role of ore properties in thickening process 787

(a) Feed

(b) Slime

(c) Middling

(d) Coarse

Fig. 2. XRD patterns of (a) feed (b) slime (c) middling (d) coarse samples

Results and discussions

Since the amount of flocculant consumption for the thickening process is usually
between 20 and 40 g/Mg, the tests were carried out with a wider range of dosage
between 15 and 60 g/Mg. The sedimentation curves for the samples are shown in Figs.
788 M. Unesi, M. Noaparast, S.Z. Shafaei, E. Jorjani

3 to 6. As presented in Tables 2 and 3, the most of the feed is composed of clay

minerals and quartz, and the most amount of the slime is composed of clay minerals,
but the content of clay minerals in the middle part of the shaking table is significantly
lower than the content of them in the feed. In addition, clay minerals were not
observed in the coarse sample, and much of them (about 55%) consist of pyrite. The
sedimentation tests were performed on the feed and middling in order to investigate
the tailings sedimentation behavior in mineralogical view because these two samples
have the same solid density as 2.8 g/cm3 and their d80 particle size are similar. In this
study, the term "clay" is used as a mineralogical term, i.e. any of a diverse group of
fine-grained minerals, not a size fraction.
The negative surface charge (OH- in alkali solution), swelling characteristics, and
small particle size of clay minerals negatively affect the settling behavior of the slimes
fraction within the thickening. These factors also affect the rheology of the slurry as it
influences the solids concentration and the manner in which the particles stack during
settling. As can be seen in Figs. 3 and 5, at different amounts of flocculant, the free
settling velocity for the feed sample (2 to 6 mm/s) is much less than that for the
middle part sample (18 to 23 mm/s) and in the compaction zone, the compressibility
for the middling sample (in concentration range of 0.63 to 0.86 Mg/m3) is more than
that for the feed sample (in concentration range of 0.33 to 0.47 Mg/m3), which this
stems from the reduction of clay content in the middling sample. Also, in Fig. 3, it is
observed that by the increasing the flocculant dosage to 30 g/Mg, a good
compressibility was obtained in the compaction zone. However, an excessive amount
of the flocculant caused the reduction of bed compressibility for the feed sample, and
accordingly the final height value in the settling curve increased to 64 mm (for 20
g/Mg) and 68 mm (for 60 g/Mg). The cotton-mode was formed in this zone due to the
increasing in the size of the flocks, the networked water was remained inside of the
flocks, and therefore the underflow concentration of paste thickeners was reduced.
This case was not observed in the middling sample (Fig. 5), and with the increasing of
the flocculant dosage, the compressibility increased as well as the free settling velocity
and the value of the final height in the settling curve decreased from 43 mm (for 20
g/Mg) to 33 mm (for 60 g/Mg), which could be caused by the interactions between
coarse and fine particles in the feed sample, and interaction between the coarse
particles in the middling sample.
The packing of the particles on top of each other during settling influences
rheology and density of the thickened underflow. The most familiar packing types are
referred to as “edge-face” also referred to as “houses of cards” and “face-face” also
known as a “band structure” packing relationship (McFarlane et al. 2005). Clays
typically stacks into honeycomb structure. The honeycomb packing retains large
amounts of water which fills the voids formed by this type of arrangement.
Great effort and highly effective dewatering systems are required to remove the
interstitial water therefore the thickener performance will decrease by increasing the
clay content in the ore. Since the coarse and high density particles have higher settling
 The role of ore properties in thickening process 789

velocity than ones that were fine and lower density, thus they could be divided into
two parts in view of settling behavior. Solid density of coarse particles (a) and fine
particles (c) were 2.7 and 4.7 g/cm3, respectively, and this difference at the density
values was caused by the existence of clay minerals (aluminosilicates) in part (c) and
metallic minerals (such as chalcopyrite and pyrite) in part (a). Since there was much
difference between them in terms of size and density, there was also much difference
between their settling velocities (Fig. 7).
There is a direct relation between the solid density and amount of metallic
minerals. Therefore, the effect of metallic minerals in the settling behavior is defined
as higher solid density. In addition, the surface charge on the particles (as
mineralogical effects) plays an important role in the flocculation performance. This
effect is not significant for the metallic minerals because the higher solid density is
much more effective than the surface charge. Therefore, the metallic minerals (as ore
properties) are more significant in the settling behavior than the flocculant dosage
(operating parameter) because the changes in the flocculant dosage showed no
significant effect on the thickening behavior of the sample part (a). In the case of this
sample, the little amount of flocculant is needed to improve the clarity of the thickener
overflow. It was postulated that the dominant minerals (such as clay and metallic
minerals) in the samples would cause affecting the thickener performance.
Additionally, part (c) and feed can be compared in a view of the particle size.
Although their density was close together, they had different particle size distribution.
The effect of this particle size differences was visible in their settling velocities. The
smaller size would cause the lower settling velocity, hence that settling velocity for the
slime sample was obtained between 0.1 and 0.4 mm/s but that was achieved between 2
and 6 mm/s for the feed (Fig. 7).
Meanwhile, the bed compressibility, which was one of the basic parameters in
thickeners, was studied for these two samples. As can be seen in Figs. 4 and 6, the
final height for sample (a) was about 19 mm, and for sample (c) was about 85 mm. In
other words, the compressibility for sample (a) was very high and for sample (c) was
very low, which is due to the parameters such as particle size, density, and
mineralogy. With the increasing in the flocculant dosage, the compressibility in part
(c) increased (Fig.4) and the final height increased from 92 mm (for flocculant dosage
of 15 g/Mg) to 85 mm (for flocculant dosage of 60 g/Mg). The significant increase in
the compressibility was observed with the increasing in the flocculant concentration
from 15 g/Mg to 35 g/Mg, and from 40 g/Mg to 60 g/Mg. This change was also visible
in the free settling zone, and there was no significant effect on the settling velocity,
and the compressibility by the increasing in the amount of 5 g/Mg flocculant. The
behavior of cotton accumulation of flocks (which was shown in Fig. 3 for feed) was
considered minimal in this sample. The increasing in the flocculant dosage showed no
effect on its compressibility for the sample part (a) because a high density and large
particle sizes of this sample led to little need for the flocculant (Fig. 6).
790 M. Unesi, M. Noaparast, S.Z. Shafaei, E. Jorjani

300 300
15 g/Mg 20 g/Mg 25 g/Mg 15 g/Mg 20 g/Mg 25 g/Mg

250 30 g/Mg 35 g/Mg 40 g/Mg 250 30 g/Mg 35 g/Mg 40 g/Mg

45 g/Mg 50 g/Mg 55 g/Mg 45 g/Mg 50 g/Mg 55 g/Mg

200 200

Height (mm)
Height (mm)

60 g/Mg 60 g/Mg

150 150

100 100

50 50
(a) Overall
(a) Overall
0 0
0 1000 2000 3000 4000 0 2000 4000 6000 8000 10000
Time (s) Time (s)
300 300
(b) Lineer part (b) Lineer part

270
250 Height (mm)
Height (mm)

240
200
210

150 180

100 150
0 50 100 150 200 0 300 600 900 1200 1500
Time (s) Time (s)
76 120
(c) End part (c) End part
74
110
72
Height (mm)
Height (mm)

70 100

68
90
66

64 80
0 1000 2000 3000 4000 500 3000 5500 8000 10500
Time (s) Time (s)

Fig. 3. Feed sedimentation curves Fig. 4. Slime sedimentation curves

(a) overall (b) linear part (c) end part (a) overall (b) linear part (c) end part

Comparing Figs. 3 to 6 indicated that the cotton-mode in the compaction zone,

observed only in the feed, can be removed by using this classification because this
causes water trapping among the flocks and reduction of underflow concentration in
the thickeners and reduces water recovery. Furthermore, with the separation of coarse
and high density particles, the consumption of the flocculant for this fraction can be
reduced or eliminated.
 The role of ore properties in thickening process 791

300 300
15 g/Mg 20 g/Mg 25 g/Mg 15gpt 20 g/Mg
15 g/Mg 20gpt
25 g/Mg
250 15gpt 20gpt 25gpt 250 25gpt 30gpt
30 g/Mg
30gpt
35 g/Mg
35gpt
40 g/Mg
40gpt
30 g/Mg
35gpt 35 g/Mg 40 g/Mg
40gpt
45gpt 50gpt

Height (mm)

Height (mm)
200 45 g/Mg
45gpt 50 g/Mg
50gpt 55 g/Mg
55gpt 200 45 g/Mg
55gpt 50 g/Mg 55 g/Mg
60gpt
60gpt
60 g/Mg 60 g/Mg
150 150

100 (a) Overall 100

(a) Overall
50 50

0 0
0 1000 2000 3000 4000 0 200 400 600 800 1000
Time (s) Time (s)
320 340
320
(b) Lineer part (b) Lineer part
290
270
270

(mm)
Height (mm)
Height (mm)

240
220
220

Height
190
170
170
140
120

120
90
70

70 40
20
0 5 10 15 0 20
20 40
40 60
60
Time (s) Time (s)
Time
46 19.5
(c) End part (c) End part
44
19.4
Height (mm)

Height (mm)

42
19.3
40
19.2
38

36 19.1

34 19
0 1000 2000 3000 4000 0 200 400 600 800
Time (s) Time (sec)

Fig. 5.Middling sedimentation curves Fig. 6. Coarse sample sedimentation curves

(a) overall (b) linear part (c) end part (a) overall (b) linear part (c) end part

30

25
Settling Rate (mm/s)

20 Feed

Slime
15
Middlings

Coarser
10

0
0 20 40 60 80
Flocculant Dosage (g/Mg)

Fig. 7. Free settling curves of the samples

The Wilhelm-Naide model was used in order to compare the settling flux, and the
results were shown in Fig. 8. As seen from Fig. 8, the maximum settling flux for the
coarse, middling, feed and fine particles were obtained as 120, 55, 9, and 0.25
792 M. Unesi, M. Noaparast, S.Z. Shafaei, E. Jorjani

Mg/m2/h, respectively. Also, the maximum concentrations in the underflow were

obtained as 1.57, 0.87, 0.45, and 0.35 Mg/m3, respectively. In other words, the
maximum settling flux and concentration in underflow zone for the coarse and the
middling parts were more than these amounts in the feed, and for the slime part were
less than these amounts in the feed. This is due to changes in the ore properties which
were illustrated before. The increasing in the flocculant consumption showed no effect
on the settling flux for the coarse particle; it was due to the high density and coarse
10 0.3
(a) Feed (b) Slime
15gpt
15 g/Mg
8 15 g/Mg
15gpt 20gpt
20 g/Mg
20 g/Mg
20gpt 25gpt
25 g/Mg
25 g/Mg
25gpt 0.2 30gpt
30 g/Mg
Flux (t/m 2/h)

Flux (t/m 2/h)

6 30 g/Mg
30gpt 35gpt
35 g/Mg
35 g/Mg
35gpt 40gpt
40 g/Mg
40 g/Mg
40gpt 45gpt
45 g/Mg
45 g/Mg
45gpt 50gpt
50 g/Mg
4 50 g/Mg
50gpt 55gpt
55 g/Mg
55 g/Mg
55gpt 0.1 60gpt
60 g/Mg
60 g/Mg
60gpt
2

0 0
0.1 0.2 0.3 0.4 0.5 0.1 0.2 0.3 0.4 0.5
C (t/m 3) C (t/m 3)

60 150
(c) Middling (d) Coarse
15gpt
15 g/Mg 15gpt
15 g/Mg
20gpt
20 g/Mg
20gpt
20 g/Mg
25 g/Mg
25gpt
25 g/Mg
25gpt
40 30 g/Mg
30gpt 100 30gpt
30 g/Mg
Flux (t/m 2/h)
Flux (t/m 2/h)

35 g/Mg
35gpt
35 g/Mg
35gpt
40 g/Mg
40gpt
40 g/Mg
40gpt
45 g/Mg
45gpt
45 g/Mg
45gpt
50 g/Mg
50gpt 50 g/Mg
50gpt
55 g/Mg
55gpt
20 50 55 g/Mg
55gpt
60 g/Mg
60gpt 60 g/Mg
60gpt

0 0
0.2 0.4 0.6 0.8 1 1.25 1.35 1.45 1.55 1.65 1.75
C (t/m 3) C (t/m 3)

Fig. 8. The settling flux in different underflow density

(a) feed (b) slime (c) middling (d) coarse samples

size of metallic minerals in this part. As can be shown from Fig. 8, the initial flux
values of the slime sample changed due to the changes in the settling rate. It means
that the flocculant cannot prevail on the initial turbulence which was created because
of the low settling velocity of the slimes. Also, this behavior was observed in the feed
sample with a low concentration of flocculant (15 g/Mg). Therefore, the minimum
amount of required flocculant can be obtained by studying the competition between
initial turbulence and settling velocity which can be achieved in the settling flux curve.
According to the measurements and calculations, the weight of solid in the slime,
middling, and coarse particle size parts were 46%, 50%, and 4% of the feed,
respectively. Thus, half of the feed was transferred to the middling part but it could be
understood from Figs. 7 and 8 that a major and effective factor in particles settling was
the amount of slime in the feed.
 The role of ore properties in thickening process 793

Conclusions
The role of ore properties (density, particle size, mineralogy) in thickening process
(settling velocity, bed compressibility, settling flux) were studied in this work using a
Sarcheshmeh tailing sample. Initially a shaking table was employed and sample was
applied to achieve three different slime, middling, and coarse products. As a result, the
free settling velocity of feed (2–6 mm/s) was less than middling sample (18–23 mm/s),
and the compressibility of middling (density: 0.63–0.86 Mg/m3) was more than feed
sample (density: 0.33–0.47 Mg/m3). This was due to the amount of clay reduction in
middling sample. Also, it was observed that an excessive amount of flocculant caused
reduction of bed compressibility for the feed sample, and the final height value in the
settling curve increased to 64 mm (for 20 g/Mg) and 68mm (for 60 g/Mg). The cotton-
mode was formed in this zone due to the increasing in the size of the flocks, hence the
networked water was remained inside of the flocks, and hence the underflow
concentration of the paste thickeners reduced. Therefore, it caused remaining the
networked water inside of the flocks and the underflow concentration of the paste
thickeners would decrease. This case was not observed for the middling sample, and
the compressibility increased with the increasing of flocculant dosage as well as free
settling velocity. The value of the final height in the settling curve decreased from 43
mm (for 20 g/Mg) to 33mm (for 60 g/Mg), which could be caused by the interactions
between coarse and fine particles in the feed sample, and interaction between the
coarse particles in the middling sample.
Also, it was concluded that the clay minerals (aluminosilicates) and metallic
minerals (pyrite, chalcopyrite) caused to reduce and to increase the settling velocity,
settling flux, and underflow density, respectively, hence the settling velocity of the
slime and the coarse particles were obtained as 0.1–0.4 and 26 mm/s, respectively. The
maximum flux of the slime and the coarse particles were obtained as 0.25 and 120
Mg/m2/h, respectively, and the maximum underflow density of slime and coarse
particles were obtained 0.35 and 1.57 Mg/m3. Therefore, the lower particle size and
density would reduce the thickener performance. The cotton-mode in the compaction
zone, which was observed only for the feed, could be removed by using this
classification. Furthermore, with the separation of coarse and high density particles,
the consumption of the flocculant for this fraction would be reduced or eliminated. In
addition, the effective and major parameter in the sedimentation was the clay content
of the feed.

References
BURGER R., BUSTOS M.C., CONCHA F., 1999, Settling velocities of particulate systems: 9.
Phenomenological theory of sedimentation processes: numerical simulation of the transient behavior
of flocculated suspensions in an ideal batch or continuous thickener, International Journal of Mineral
Processing, 55, 267–282.
794 M. Unesi, M. Noaparast, S.Z. Shafaei, E. Jorjani

BUSCALL R., MCGOWAN I.J., MILLS P.D.A., STEWART R.F., SUTTON D., WHITE L.R., YATES
G.E., 1987, The Rheology of strongly flocculated suspensions, Non-Newtonian Fluid Mechanics, 24,
183–202.
BUSTOS M.C., CONCHA F., BURGER R., TORY E.M., 1999, Sedimentation and Thickening, Kluwer
Academic Publishers, Dordrecht, Netherlands, 285.
COE H. S., CLEVENGER G. H., 1916, Methods for determining the capacities of slime thickening
tanks, Trans AIME, 55, 356–384.
CONCHA F., BUSTOS M.C., BARRIENTOS A., 1996, Phenomenological Theory of Sedimentation,
Chapter 3 in Sedimentation of small particles in a viscous Fluid, Computational Mechanics
Publications, Southampton, UK, 51-96.
DAHLSTORM D.A., FITCH E.B., 1985, Mineral Processing Handbook, N.L. Wiess (ed), SME/AIME,
Chapter 9, 2-14.
DEVENTER B.G., USHER S.P., KUMAR A., RUDMAN M., SCALES P.J., 2011, Aggregate
densification and batch settling, Chemical Engineering Journal, 171, 141–151.
GARRIDO P., BURGOS R., CONCHA F., BURGER R., 2003, Software for the design and simulation of
gravity thickeners, Minerals Engineering, 16, 85–92.
GARRIDO P., CONCHA F., BURGER R., 2003, Settling velocities of particulate systems: 14. Unified
model of sedimentation, centrifugation and filtration of flocculated suspensions, International Journal
of Mineral Processing, 72, 57–74.
GLADMAN B.R., RUDMAN M., SCALES P.J., 2010, Experimental validation of a 1-D continuous
thickening model using a pilot column, Chemical Engineering Science, 65, 3937–3946.
GREEN M.D., 1997, Characterization of suspensions in settling and compression, PhD thesis,
Department of Chemical Engineering, University of Melbourne Parkville Victoria 3052, Australia.
KELLY E.G., SPOTISSWOOD D.G., 1989, Introduction to Mineral Processing, Mineral Engineering
Services, Denver.
KYNCH G. J., 1952, a theory of sedimentation, Trans Faraday Society, 48, 166–176.
LANDMAN K.A., WHITE L.R., BUSCALL R., 1988, The Continuous-flow gravity thickener: Steady
State Behavior, AIChE Journal, 34, 239–252.
LESTER D.R., RUDMAN M., SCALES P.J., 2010, Macroscopic dynamics of flocculated colloidal
suspensions, Chemical Engineering Science, 65, 6362–6378.
MCFARLANE A., BREMMELL K., ADDAIMENSAH J., 2005, Microstructure, rheology and
dewatering behavior of smectite dispersions during orthokinetic flocculation, Minerals Engineering,
18, 1173.
OLTMANN H.H, 1975, Filtration and Separation, 12, 6, 636.
OSBORNE D.G., 1977, Solid-Liquid Separation, Chapter 5, L. Svarovsky (ed), Butterworth, London,
Boston, 75–99.
TALMAGE W. P., FITCH E.B., 1955, Industrial and Engineering Chemistry, 47, 38.
USHER S.P., SCALES P.J., 2005, Steady state thickener modeling from the compressive yield stress and
hindered settling function, Chemical Engineering Journal, 111, 253–261.
USHER S.P., SPEHAR R., SCALES P.J., 2009, Theoretical analysis of aggregate densification: Impact
on thickener performance, Chemical Engineering Journal, 151, 202–208.
WILHELM J. H., NAIDE Y., 1981, Sizing and operating continuous thickeners, Mining Engineering,
1710–1718.
YALCIN T., 1988, Bulletin of the Canadian Institute of Metallurgy, 81, 910, 69.
YOSHIOKA N., HOTTA Y, TANAKA S., NAITO S., TONGAMI S. 1957, Continuous thickening of
homogeneous slurries, Chemical Engineering, Tokyo, 21, 66-74.