Sie sind auf Seite 1von 11

Powder Technology 294 (2016) 5565

Contents lists available at ScienceDirect

Powder Technology
journal homepage: www.elsevier.com/locate/powtec

Effect of otation froth properties on froth rheology


Chao Li , Kym Runge, Fengnian Shi, Saeed Farrokhpay
Julius Kruttschnitt Mineral Research Centre, The University of Queensland, 40 Isles Road, Indooroopilly, Queensland 4068, Australia

a r t i c l e

i n f o

Article history:
Received 14 December 2015
Received in revised form 8 February 2016
Accepted 10 February 2016
Available online 13 February 2016
Keywords:
Froth rheology
Bubble size
Fraction of lamellae covered by solids
Flotation

a b s t r a c t
Froth otation is a widely used process of particle separation exploiting differences in surface properties. The
froth performance in a otation cell is expected to be affected by the froth rheology, as it affects the froth residence time that determines the probability of recovery of valuable minerals in the froth phase. Flotation froths
have a similar structure to the gasliquid foams whose rheology has been widely studied. However, to date,
very little work has been done in the rheology of otation froths owing to their instability and the presence of
solid particles (on bubble surfaces and in the plateau borders) that are believed to inuence froth rheology
and complicate any investigation. In this paper, the effects of froth properties on froth rheology were studied
by examining the results of 33 otation tests performed under various conditions that resulted in changes in
the froth properties and, consequently, the froth rheology. The experiments were performed in a 20 L continuous
otation cell. It was found that the bubble size and the fraction of lamellae covered by solids dened the froth
rheology, while the presence of particles in the plateau borders contributed very little to the froth rheology. A
model structure was developed by taking into account froth properties to predict froth viscosity.
2016 Elsevier B.V. All rights reserved.

1. Introduction
In otation, the froth phase plays the role of transporting the hydrophobic minerals from the collection zone to the concentrate launder.
Froth transportation consists of both vertical and horizontal motion:
the vertical motion is dened by the ow of bubbles that carry particles
moving from the pulpfroth interface to the launder lip level, while the
horizontal motion describes the motion towards the overow weir [1].
Drainage of valuable minerals occurs during froth transportation owing
to bubble bursting and bubble coalescence. Froth residence time, which
determines the probability of recovery of valuable minerals, is a function of the time bubbles take to move both vertically and horizontally.
The vertical motion is driven by the supercial gas velocity. The horizontal ow is a consequence of three factors that include the force of
gravity, the froth stability and the resistance to froth ow. The inuence
of the froth stability on the efciency of froths in recovering valuable
minerals has been extensively studied [25]. The resistance to froth
ow is supposed to be directly associated with froth rheology, and,
therefore, the importance of froth rheology on otation performance
has also been recognised. Shi and Zheng [6] and Farrokhpay [7] have
clearly shown that froth rheology can affect the froth mobility as
well as the froth stability, and ultimately inuence the otation performance. However, to date, it is not clear what froth properties affect froth rheology.
Rheology is a measure of the ow characteristics of a substance. It is
usually represented by a rheogram which plots the shear stress of a uid
Corresponding author.
E-mail address: c.li7@uq.edu.au (C. Li).

http://dx.doi.org/10.1016/j.powtec.2016.02.018
0032-5910/ 2016 Elsevier B.V. All rights reserved.

when subject to different shear rates. In general, a substance can either


exhibit Newtonian or non-Newtonian behaviour, with the latter including dilatant, plastic, pseudo-plastic and Bingham behaviours [8,9].
Various types of rheograms are illustrated in Fig. 1. Viscosity as a key
rheological term is a measure of the resistance of a material to deformation. It is a constant in Newtonian ow but shear rate dependent in nonNewtonian ow.
In order to fully understand what froth properties determine the
froth rheology, it is rstly necessary to gain insight into the froth characteristics. From the study of aqueous foams, when the gas volume fraction is less than 0.73, bubbles disperse in the liquid phase without
becoming attached to one another; at a gas volume fraction greater
than 0.73, the bubbles start to pack and are separated by thin-planeparallel lms forming polyhedral cells (lamellae) [10]. The thin lamellae
meet in lines (plateau borders) and the lines meet at vertices [11,12].
Flotation froth has a similar structure to dry foam. The air volume fraction in otation froth usually exceeds 0.90 especially in deep froth due
to rapid drainage. Furthermore, otation froth is a gasliquidsolid regime; there are solid particles present in the froth phase. Hydrophobic
particles are mainly attached to the lamellae while both hydrophilic
and hydrophobic particles (detached owing to bubble coalescence and
bubble bursting) are present in the plateau borders and vertices. A typical froth structure with particles is shown in Fig. 2.
Flotation froths have a similar structure to foams, making it possible
to begin the study of froth rheology by considering the rheology of
foams (i.e. leaving aside the presence of solid particles). Practically,
foam rheology is associated with bubble size and foam quality (the volume fraction of air in the foam) [1316]. When foam is dry, its rheology
is dominated by the bubble size [15,17]. However, a otation froth

56

C. Li et al. / Powder Technology 294 (2016) 5565

Fig. 1. Schematic diagram of shear rate as a function of shear stress for different types of uid.
After Mewis and Wagner [8].

in the plateau borders and vertices form the local solidliquid suspension. According to the study of suspension rheology, the solids volume
fraction in the suspension inuences the rheology [19]. Hence, whether
the solids volume fraction in the plateau borders and vertices is also a
crucial factor affecting froth rheology needs to be investigated.
Previously, the authors have investigated the effects of various otation conditions (viz. froth height, gas rate, impeller speed, feed particle
size and feed grade) on froth rheology. A Central Composite Rotatable
Design (CCRD) study was carried out, in which 33 otation tests were
performed in a 20 L continuously-operated otation cell [20]. The froth
rheology was measured using a method developed recently by the authors [21]. Ultimately, the otation conditions inuence the froth rheology through their effect on the froth properties. In this current work, the
results of the previous experiments (from the CCRD study) are examined
to fundamentally investigate the direct effect of the froth properties on
froth rheology. The froth properties of interest were identied above as
bubble size, fraction of lamellae covered by solids and the solids volume
fraction in the plateau borders and vertices. The method of evaluating
these froth properties is developed below.
2. Experimental

differs from a two phase foam in that there are also solid particles
present both as attached particles on the lamella of the bubble surface
and in the plateau borders which form between the bubbles. It is not yet
clear how the presence of solid particles affects the rheology of otation
froths. The hydrophobic particles attached on the lamellae in otation
froth function as the surfactant adsorbed on the interface lm in aqueous foam. The surfactant adsorbed on the interface lm lowers the interface tension and resists bubble coalescence, stabilizing the foam. The
presence of surfactant on the lm is characterized as its adsorbing thickness [18]. In this study, the presence of solid particles attached to the lamellae can be represented by the fraction of lamellae covered by solid
particles which takes into account the mass of solid particles attached
to the lamella per unit area and the size distribution of the solid particles. It is expected that the particles on the lamellae change the bubble
rigidity and smoothness. In addition, the particles trapped unselectively

Fig. 2. Schematic of otation froth ( hydrophilic particle;


After Ventura-Medina and Cilliers [10].

hydrophobic particle).

2.1. Flotation tests


The 33 otation tests were performed in a bottom driven 20 L otation
cell with cross sectional dimensions of 30 by 30 cm. Table 1 shows the details of the 33 tests in which there are seven repeat tests (i.e. Tests 1, 9, 10,
12, 18, 19, and 26). The aim of this work is to create otation froths with
different froth properties and evaluate the effects of these froth properties
on the froth rheology.
The otation feed was a mixture of pure chalcopyrite and silica. The
chalcopyrite was purchased from Geo Discoveries as bulk rock. The
Table 1
The conditions in the CCRD otation tests [19].
Test

Froth
height
(cm)

Supercial
gas velocity
(cm/s)

Impeller
speed
(rpm)

Chalcopyrite
particle size
P80 (m)

Copper
grade (%)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

7
6
6
8
6
6
6
6
7
7
8
7
8
8
8
6
8
7
7
8
8
6
7
5
7
7
7
7
7
7
9
7
7

1.4
1.0
1.8
1.8
1.0
1.8
1.0
1.8
1.4
1.4
1.8
1.4
1.0
1.8
1.0
1.8
1.8
1.4
1.4
1.0
1.0
1.0
1.4
1.4
0.6
1.4
1.4
1.4
1.4
2.2
1.4
1.4
1.4

900
750
1050
750
750
750
1050
1050
900
900
750
900
1050
1050
750
750
1050
900
900
750
1050
1050
900
900
900
900
1200
900
600
900
900
900
900

80
50
50
50
110
50
50
110
80
80
110
80
110
110
110
110
50
80
80
50
50
110
140
80
80
80
80
80
80
80
80
80
20

1.0
1.4
1.4
1.4
0.6
0.6
0.6
0.6
1.0
1.0
0.6
1.0
0.6
1.4
1.4
1.4
0.6
1.0
1.0
0.6
1.4
1.4
1.0
1.0
1.0
1.0
1.0
1.8
1.0
1.0
1.0
0.2
1.0

C. Li et al. / Powder Technology 294 (2016) 5565

57

silica was purchased from Sibelco Australia as ne particles (P80 =


73 m). Before each otation test, a measured quantity of the chalcopyrite was ground to the targeted particle size distribution, mixed with silica to achieve the desired feed grade, and diluted with Brisbane tap
water in a conditioning tank. The solids concentration was maintained
at 40 wt.% in all the tests. Sodium ethyl xanthate (2.0 g/t) and Dowfroth
250 (14.7 ppm) were used as the collector and the frother, respectively.
The otation tests were operated continuously in a closed circuit by
recycling the concentrate and tailing. Samples of the feed, concentrate
and tailings were collected, and weighed before and after drying.
Sub-samples were assayed for copper to determine the otation recovery. The concentrate ow rates in terms of mass and volume were also
measured.
A pictorial diagram of the experimental set-up is shown in Fig. 3.
More details of the experiments may be found in previous work [20].

size and the froth velocity prole towards the cell launder lip. A single
light source was mounted above the froth surface as this results in a
single bright light on each bubble a requirement of the froth analysis
algorithm. An Anglo-Platinum bubble sizer was used to measure the
bubble size in the pulp at the end of each test.

2.2. Froth rheology measurements

The estimation of the fraction of lamellae covered by solids and the


solids volume fraction in the plateau borders needs the value of the
gas hold-up in the froth. Researchers [1,23] have found that water and
particles drop back into the pulp phase mainly just above the froth
pulp interface, above which the froth phase is relatively constant in
terms of its properties. Hence, the mean gas holdup in the froth above
the lip level remains relatively constant.
In a continuous and stable otation process, the froth phase is in dynamic equilibrium. Air bubbles, particles and water continuously enter
the froth phase from the pulpfroth interface and ow over the launder
lip. Some air is lost from the froth surface during its transportation to the
launder lip (Fig. 4). By assuming that the drainage in the upper froth
phase is negligible, the gas hold-up in the upper froth zone can be
regarded as the ratio of the air ow rate to the total froth ow rate (including air, water and solids). The froth ow rate consists of the air ow
rate (VA) and the concentrate slurry ow rate (Vcs). As the gas entrained
into the tailings is negligible [24], the air ow rate is equal to the air aerated into the otation cell. The air ow rate and the concentrate slurry
ow rate were both measured; hence, the mean gas holdup (f) in the
upper froth zone can be calculated by Eq. (1).

The froth rheology measurements were conducted using a 6-bladed


vane (22 mm diameter and 16 mm height) attached to an air-bearing
rheometer (Anton Paar DSR301). A tube (74 mm diameter and 150 mm
height) was used to encircle the vane to eliminate the effect of the horizontal froth ow [21]. The vane was positioned in the middle of the cell
with its upper edge immersed 2 cm into the froth. During the froth rheology measurement, the torque values were measured by evenly increasing
the vane speed from 1 rpm to 15 rpm in equal increments (i.e. 1.0 rpm,
4.5 rpm, 8.0 rpm, 11.5 rpm and 15.0 rpm), with a 5 s interval between
each measuring point. A total of ve torque values were measured in
each test. Each series of torque measurements was replicated ve times
to determine the average value. The variation of the torque measurement
at each speed was determined and found to be reasonably low [20].
The vane was only immersed in the froth for the period of the rheology
measurements and then moved away not to impede the otation froth
movement.
2.3. Froth and pulp bubble characterization
A digital video camera (Sony ACC-FV50B) was mounted above the
otation cell to record the froth movement. The video images were
analysed by a contracted software [22] to determine the froth bubble

3. Estimation of froth properties


This section introduces the methods of estimating a number of froth
properties, viz. gas hold-up, fraction of lamellae covered by solids and
the solids volume fraction in the plateau borders and vertices. To perform the estimations, some assumptions were made based on reasoning
provided in the literature.
3.1. Gas hold-up (gas volume fraction)

VA
V A V cs

Fig. 3. A pictorial diagram of the experimental set up [19].

58

C. Li et al. / Powder Technology 294 (2016) 5565

the concentrate is roughly equal to the mass ow rate of chalcopyrite from the pulp to the froth.

Rc  R f
 100
Rc  R f 1Rc

The bubble surface area ow rate entering the froth phase can be determined by using the bubble surface area ux (Sb) which is a measure
of the fresh bubble surface area per cross-sectional area being supplied
to the cell. It is determined by the supercial gas velocity (Jg) and Sauter
mean bubble diameter (D32) as shown in Eq. (3) [27]. The cell crosssectional area (A) is known. Hence the bubble surface area ow rate in
the pulp (p) can be calculated by Eq. (4).

Fig. 4. The schematic of froth ow.

Sb

6J g
D32

p Sb  A

3.2. Fraction of lamellae covered by solids


To calculate the fraction of lamellae covered by solids, the mass of
solid particles attached to the lamellae per unit area (i.e. bubble
loading) and the size distribution of particles needs to be known.
The bubble loading in the froth phase can be derived from the bubble loading in the pulp phase. The estimation of bubble loading in
this work is based on the assumption that particles that detached
through bubble bursting and bubble coalescence do not reattach
on the lamellae in the froth phase. This assumption can be supported by some previous studies. Vera, Franzidis and Manlapig
[25] reported that froth recovery was unselective; thus, detached
chalcopyrite particles stay in the plateau borders and vertices with
no reattachment occurring. This is consistent with the nding reported by Ventura-Medina and Cilliers [11], who concluded that
the recovery of hydrophobic minerals in otation concentrate mainly came from the plateau borders and vertices because of bubble
bursting and bubble coalescence, indicating that reattachment was
not signicant.
Silica, a hydrophilic mineral, does not attach on bubble surfaces
and is recovered into the froth phase by entrainment, which contributes negligibly to bubble loading. Hence, bubble loading is associated with hydrophobic chalcopyrite particles only. In the pulp
phase, the chalcopyrite particles attached onto dispersed bubble
surfaces are transported to the froth phase. As a result of the quick
drainage of water in the froth, bubbles start to attach to one another.
Given that bubbles are densely packed in the froth phase, each lamella is shared by two neighbouring bubbles. The particles originally attached on dispersed bubble surfaces redistribute on lamellae.
Consequently, the bubble loading in the froth phase is double that
in the pulp phase.
Bubble loading in the pulp phase can be calculated as the ratio of
the mass ow rate of chalcopyrite entering the froth phase to the
bubble surface area ow rate entering the froth phase. The mass
ow rate of chalcopyrite entering the froth phase is equal to the
mass ow rate of chalcopyrite in the nal otation concentrate
when the drainage of chalcopyrite in the froth phase is negligible.
This is a reasonable assumption when the nal otation recovery
is close to 100%. As shown in Eq. (2), otation recovery (R) is determined by both pulp recovery (Rc) and froth recovery (Rf) [26]. Froth
recovery must be close to 100% when otation recovery is close to
100%, and therefore drainage from the froth will be minimal. Once
chalcopyrite particles are recovered into the froth phase, they will
survive to be in the nal concentrate. As shown in Table 2, recoveries in the otation tests were high (mostly N 85%). It is therefore a
reasonable assumption that the mass ow rate of chalcopyrite in

The mass ow rate of chalcopyrite in the concentrate can be calculated


from the measured mass ow rate of solids in the concentrate (c) and
the copper concentrate grade (G). It is known that the recovery of hydrophobic minerals from the pulp phase is mainly by true otation, which
means that the collection of hydrophobic minerals by entrainment in
the pulp phase is limited. Thus, the bubble loading in froth phase (LB)
can be calculated from Eq. (5), where 2.89 is the ratio of the molecular
weight of chalcopyrite to the atomic weight of copper. As mentioned before, the bubble loading in the froth phase is double that in the pulp phase,
which is reected by the number 2 in Eq. (5). It should be pointed out
that this method may underestimate the absolute bubble loading (as

Table 2
Data required for calculation.
Test

R (%)

Jg
(cm/s)

D32
(cm)

vf
(cm/s)

h
(cm)

c
(g/s)

G (%)

Vw
(cm3/s)

P50
(m)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

93.73
89.90
98.30
95.51
93.83
93.56
93.53
94.38
94.40
94.42
95.44
94.57
78.51
92.35
95.05
95.48
91.87
94.52
85.35
81.38
90.72
85.26
91.63
92.11
76.62
93.04
89.05
94.37
95.89
97.26
93.99
89.71
92.87

1.4
1.0
1.8
1.8
1.0
1.8
1.0
1.8
1.4
1.4
1.8
1.4
1.0
1.8
1.0
1.8
1.8
1.4
1.4
1.0
1.0
1.0
1.4
1.4
0.6
1.4
1.4
1.4
1.4
2.2
1.4
1.4
1.4

0.043
0.048
0.053
0.051
0.048
0.051
0.048
0.053
0.043
0.043
0.051
0.043
0.048
0.053
0.048
0.051
0.053
0.043
0.043
0.048
0.048
0.048
0.043
0.043
0.058
0.043
0.042
0.043
0.057
0.048
0.043
0.043
0.043

5.54
3.10
4.87
4.95
3.93
5.31
4.05
5.67
5.48
5.35
5.12
5.65
3.93
5.39
3.80
5.30
4.40
5.98
6.15
1.99
3.01
2.16
2.96
6.22
1.57
5.96
6.02
5.69
5.72
6.64
6.00
4.74
5.24

2.30
2.40
2.80
2.50
1.40
2.20
1.90
2.10
2.00
2.00
1.90
1.90
2.10
2.30
1.80
2.10
2.60
2.00
2.10
1.80
2.90
1.30
0.80
2.10
1.20
2.30
1.90
3.10
2.10
2.50
2.20
1.50
2.90

14.82
7.63
19.82
11.45
10.47
14.05
11.58
13.68
13.50
16.99
13.75
15.49
10.57
12.65
11.82
9.02
4.17
16.01
16.84
1.97
4.79
4.79
5.46
17.35
5.19
16.66
15.78
16.44
12.78
17.37
12.19
7.27
11.45

8.15
11.67
4.56
9.25
3.42
2.08
2.82
2.65
7.61
8.30
3.04
6.17
3.83
10.98
12.74
10.95
6.91
6.50
5.30
9.57
15.97
21.31
12.71
4.37
13.34
6.78
7.71
12.38
13.17
6.58
11.34
1.74
6.53

56.03
28.75
59.06
36.33
32.21
47.80
42.19
52.50
50.34
53.58
44.99
50.22
36.03
44.19
33.81
31.49
23.51
56.09
56.43
11.34
18.85
12.21
21.15
58.68
14.93
56.20
59.06
58.06
43.06
63.15
48.17
37.35
49.11

10.21
8.44
10.64
8.91
5.54
11.79
10.17
9.73
10.97
14.08
9.18
9.64
8.90
13.19
18.86
14.14
2.52
12.08
9.83
3.69
9.16
26.01
10.06
13.20
18.11
12.38
11.67
13.58
15.57
12.66
10.93
3.52
4.62

C. Li et al. / Powder Technology 294 (2016) 5565

there is likely to be chalcopyrite drainage from the froth). However, it is


believed that this method provides a reasonable relative estimate of the
fraction of lamellae covered by solids between the tests.
LB 2 

2:89c  G
100  p

To simplify the analysis, the shape of the solid particles on the lamellae is assumed to be spherical. The mean particle size of the solids on the
lamellae (P50) is assumed to be the same as that of the sized concentrate. It is assumed that the particles exhibit a monolayer distribution
on the lamellae. Hence, the area of a particle occupying the lamella is
equal to its projected area. The fraction of the lamella covered by solids
() can be calculated by Eq. (6). Density in this equation is required to
convert the mass of chalcopyrite loaded on bubbles into a project area
(c = 4.18 g/cm3).
30; 000LB

2c  P50

3.3. Solids volume fraction in the plateau borders and vertices


The solids volume fraction in the plateau borders and vertices is
determined by the ratio of the volume of solid particles to the total
volume (including solids and water). It is equal to the ratio of the solids
volumetric ow rate (hydrophilic and hydrophobic particles) to the
total volumetric ow rate (hydrophilic and hydrophobic particles and
water) in the plateau borders and vertices.
The volumetric ow rate of hydrophobic particles in the plateau
borders and vertices is the overall volumetric ow rate of hydrophobic
particles in the nal concentrate minus the ow rate of the particles attached on bubble surfaces. The latter is directly determined by the bubble surface area ow rate overowing the concentrate lip and the
bubble loading. The bubble loading is obtained from Eq. (5): therefore,
the bubble surface area ow rate overowing the lip (f) is needed.
The volumetric ow rate of air overowing the lip (Va) can be determined by Eq. (7), where f is the gas hold-up in the froth, vf represents
the surface froth velocity over the lip, h is the froth height above the
lip, w is the lip width and k represents the velocity prole of the
overowing froth. When a linear velocity decrease with depth is assumed, the average velocity is half of the measured surface velocity
(k = 0.5) [28].
Va k  f  vf  h  w

the plateau borders (as silica is not hydrophobic). This can be calculated
based on the concentrate mass ow rate and the copper grade. The volumetric ow rate of silica in the concentrate (Vs) is determined by
Eq. (10) by considering the silica density (s = 2.65 g/cm3). The water
volumetric ow rate in the concentrate (Vw) is measured. Finally, the
solids concentration in the plateau borders (s) can be calculated using
Eq. (11).
Vs

c  1002:89G
100  s

10

Vs Vc
Vs Vc Vw

11

4. Results
4.1. Froth rheology
The measured froth rheology data were vane speed and torque as
shown in Fig. 5. This data was converted to shear stress and shear strain
to create the standard shear stress shear rate rheograms for each experiment using a method recently proposed for doing these calculations
when using a vane style rheometer surrounded by a tube [21].
The changing slope of the torque versus vane speed relationships
is an indication that the otation froths created in the test program
are exhibiting non-Newtonian shear-thinning behaviour. Thus the apparent viscosity of the froth increases as the shear rate applied decreases. As
the shear rate is expected to be related to froth velocity which is not constant throughout the froth phase, this poses a challenge when one wants
to compare the viscosity of different froths and how it changes with froth
properties. What viscosity should be compared?
The HerschelBulkley model (Eq. (12)) has been widely used to t
foam rheograms [2931]:
n
y  _

1 6V a

2 D32

Vc

2:89c  G100  LB  f
100  c

Ventura-Medina and Cilliers [11] reported that the water present in


the froth phase is mainly trapped in the plateau borders and vertices;
the contribution from the lamellae is negligible. Hence, the water ow
rate in the plateau borders and vertices is equal to that in the nal
concentrate. The silica ow rate in the concentrate is equal to that in

12

where is the shear stress, y is the yield stress, is the consistency


index, _ is the shear rate and n is the ow index (dimensionless). The
value of n indicates the deviation of the uid from Newtonian behaviour:
when n N 1, the uid is shear-thickening (i.e. dilatant); when n b 1, it is
shear-thinning (i.e. plastic or pseudo-plastic). When n = 1, the uid is
Newtonian, and the viscosity (i.e. ) is a constant [9].

The Sauter mean bubble diameter in the froth (D32 ) is obtained from
the surface froth image analysis. The bubble surface area ow rate
overowing the lip can be calculated using Eq. (8). The factor of 1/2 in
Eq. (8) corresponds to the sharing of each lamella by two neighbouring
bubbles. Hence the mass ow rate of chalcopyrite recovered in concentrate attached to lamellae (LB f) can be calculated. The volumetric ow
rate of chalcopyrite recovered by entrainment in the plateau borders
and vertices (Vc) can be calculated using Eq. (9).
f

59

Fig. 5. Froth rheology data measured in the 33 otation tests.

60

C. Li et al. / Powder Technology 294 (2016) 5565

Eq. (12) was used to t the data of shear rates and shear stresses collected from each test. These tted results show that the yield stress of
the froth in these experiments is very low, which may be a consequence
of the measurements being performed in a owing froth. Thus yield
stress in Eq. (12) can be made to equal zero without any signicant effect on the degree of t of the results. It was therefore removed from the
equation. In addition, it is apparent viscosity (), the ratio between the
shear stress and shear rate of a uid which is the most commonly
used rheological term to evaluate the rheology of non-Newtonian
ow. By dividing both sides of Eq. (9) by shear rate, an equation can
be developed relating the froth's apparent viscosity and the shear rate
being applied (Eq. (13)).
 _

n1

13

Bjrn, Monja, Karlsson, Ejlertsson and Svensson [32] has suggested


that the consistency index () in Eq. (13) can be used to compare the
viscosity of uids when the uids have similar ow indices (n).
The curves of apparent viscosity versus shear rate for all tests are
plotted in Fig. 6 on a loglog scale. As shear rate in the froth is less
than 4 s1 in all the tests [20], Fig. 6 only shows the froth ow curves
for shear rate values less than 4 s1. Apparent viscosity changes significantly with shear rate, conrming that the otation froths were shearthinning uids. The slopes of all the relationships, however, are very
similar, indicating that the ow index (n) may be able to be considered
a constant.
If the ow indexes can be considered constant, the variation of froth
rheology in these tests can be represented by the differences in the consistency indices (). In order to validate this hypothesis, the standard
error of the ow index for each test determined using Eq. (13) was evaluated by using Solver statistics. The ow index with its error at 95%
condence interval is plotted in Fig. 7, which shows clearly that the
ow indices are very similar. Paired t-tests were performed to test the
signicance of the observed difference between each pair of n values.
Eq. (14) was employed to calculate the critical difference for a pair of
n to be statistically different with 95% condence [33].
n1 n2
t q
SEn1 2 SEn2 2

14

Fig. 7. Flow indices calculated for each test showing their 95% condence interval.

where ni is the ow index and SEni is the standard error of ni. For the
whole set of n, using the calculated average standard error (SEn =
0.08) for the n value, Eq. (14) is changed to Eq. (15) as shown below:
n1 n2
t q :
2SEn 2

15

The t values were calculated in Excel using the function = TINV


(probability, degrees of freedom) (degrees of freedom = sample
size 2). Statistically, for 95% condence that the n values are signicantly different at the t-value of t95, 3 = 3.18, the critical difference for
ow index should be greater than 0.36. Fig. 7 shows that, in general,
the difference between each pair of ow indices is less than 0.36. Therefore, it may be concluded that there is no statistical difference between
the n values. As a result, the ow index can be treated as a constant, and
the froth rheology can be evaluated by using only the consistency index.
The effect of the froth properties estimated above on the froth rheology
in the 33 tests (as determined by the consistency index) is considered in
Section 5.1.
4.2. Calculated froth properties
A summary of the data measured in the otation experiments to
perform the calculations described in Eqs. (1) and (3) to (11) is given
in Table 2. A summary of the calculated froth properties for each experiment is shown in Table 3. As mentioned previously, there are seven repeat tests among the 33 otation experiments (i.e. Tests 1, 9, 10, 12, 18,
19, and 26). The variation of the calculated froth properties (CoV, the
ratio of the standard deviation to the mean) determined using these repeat tests is shown in Table 4. The standard deviation is generally less
than 10% of the mean value.
5. Discussion
5.1. Inuence of froth characteristics on froth rheology

Fig. 6. Froth rheograms determined for various tests [19].

5.1.1. Bubble size


Fig. 8 shows a negative relationship was observed between the consistency index obtained from the froth rheology model and the froth
bubble size (D32). There is scatter in this data, which is expected in a system where multiple froth characteristics are having a bearing on the result. According to the regression analysis there is a 99.98% condence
that the slope of this relationship is negative and non-zero. Therefore
it can be concluded statistically that bubble size is negatively correlated
with froth viscosity.
From a macroscopic viewpoint, with the air volume fraction being
above 0.95 in the froth in all the experiments (see Table 3), the otation

C. Li et al. / Powder Technology 294 (2016) 5565

61

Table 3
Summary of froth properties.
Test

D32 (cm)

(Pasn)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

0.9524
0.9646
0.9586
0.9736
0.9595
0.9662
0.9484
0.9646
0.9564
0.9531
0.9689
0.9564
0.9564
0.9698
0.9598
0.9789
0.9846
0.9527
0.9524
0.9863
0.9776
0.9844
0.9821
0.9499
0.9698
0.9517
0.9493
0.9495
0.9618
0.9643
0.9585
0.9682
0.956

0.91
0.89
0.81
0.86
1.00
1.23
0.94
1.09
0.92
0.97
1.15
0.97
0.80
0.86
0.82
0.93
0.76
0.84
0.84
0.78
0.71
1.01
1.07
0.94
0.90
0.87
0.88
0.82
0.93
1.00
0.86
1.15
0.87

0.014
0.0194
0.0095
0.0128
0.0119
0.0027
0.0059
0.0042
0.0111
0.0118
0.0049
0.0117
0.0084
0.0118
0.0147
0.0075
0.0128
0.0102
0.0107
0.0094
0.0154
0.0072
0.0081
0.0068
0.014
0.0108
0.0121
0.0177
0.0167
0.0074
0.0149
0.0043
0.0191

0.0835
0.0807
0.1075
0.0968
0.1058
0.0978
0.0913
0.0872
0.0851
0.0983
0.1004
0.0981
0.096
0.0871
0.1023
0.0872
0.0585
0.0911
0.0961
0.0556
0.0737
0.1028
0.0778
0.0962
0.1013
0.094
0.0847
0.0848
0.0878
0.0881
0.0774
0.0672
0.0756

1.47
2.04
1.80
1.92
1.02
2.02
1.44
1.07
1.39
1.40
1.35
1.24
1.73
1.10
1.77
1.22
1.26
1.46
1.48
1.59
1.84
0.64
0.57
1.02
1.66
1.27
0.98
1.65
1.38
1.43
1.30
0.88
1.58

froth structure can be considered to be similar to a dry foam. In two


phase aqueous foam studies, rheology is governed by the air volume
fraction and the bubble size [3437]. Air volume fraction has a positive
correlation with foam rheology, while bubble size inuences foam rheology negatively. The dominant factor determining foam rheology, transitions from air volume fraction to bubble size as foam dries. At a low air
volume fraction (Fig. 9a), isolated bubbles stay in a continuous liquid
phase. When a shearing force is applied on the foam, the resistance to
deformation is mainly determined by the bulk liquid rheology rather
than the foam texture [30]. As air volume fraction increases (Fig. 9b),
the foam regime changes from dispersed bubbles to compressed bubbles separated by the lamellae which form the plateau borders and vertices where the liquid is trapped. When an external shearing force is
applied on this type of foam, the bubbles are deformed and pass over
one another. The liquid trapped in the plateau borders is also sheared
during the ow. Therefore, the foam is expected to become more viscous and its rheological properties determined by the bulk liquid rheology and the bubble size [13]. When foam is very dry with a high air
volume fraction (Fig. 9c), the volume of liquid trapped in the plateau
borders and vertices is negligible, and bubble deformation and the
friction between adhered bubbles dene the foam rheology. In such
dry foam, the most important property affecting foam viscosity is the
bubble size [38].
From a microscopic viewpoint, the relationship in Fig. 8 can be explained in terms of the relative motion between neighbouring bubbles
Table 4
Statistics of the froth properties calculated using the repeat tests.
Froth property

D32 (cm)

(Pasn)

Mean
STDEV
CoV (%)

0.9536
0.0020
0.21

0.90
0.06
6.12

0.0115
0.0012
10.87

0.0923
0.0060
6.52

1.39
0.10
6.99

Fig. 8. Relationship between froth bubble size and the froth consistency index.

and bubble surface deformation. Froth ow is an irreversible process


which involves the rearrangement of bubbles. This leads to the appearance of local velocity gradients in the uid conned in the lamellae, as
the uid is sheared as shown in Fig. 10a [31]. Bikerman [12] noted
that the lamella was signicantly more viscous than the bulk liquid
viscosity. Therefore, the shearing friction in the lamellae signicantly
dissipates energy and contributes to the froth viscosity. The length of lamella in a given volume of froth is determined by the bubble size. A bigger mean bubble size in the froth indicates a lower total lamella length
and vice versa. This explains the negative correlation between bubble
size and froth viscosity.
Froth ow also involves bubble surface area change. In static circumstances, a lamella is in a force equilibrium composed of the pressure difference between neighbouring bubbles and the surface tension. In
sheared foam, the analysis of bubble dynamics has shown that the perpetual formation and disappearance of foam lms between colliding
bubbles leads to variation of the bubble surface area around its mean
value [39]. The viscous dissipation of energy due to surface area change
is related to the surface dilatational modulus which denes bubble surface rigidity. The bubble surface dilatational modulus is positively correlated with surface tension [40,41]. At liquidair interfaces, surface
tension results from the greater attraction of water molecules to one another than to the molecules in the air. The net effect is an inward force at
the surface which results in the elastic tendency of the liquid to acquire
the least surface area possible. Bubble deformation involves surface
stretch as shown in Fig. 10b: at each local point on the surface, the
stretch dissipates energy to offset the resistance resulting from surface
tension. The small bubbles possess high surface tension and, consequently, a higher surface dilatational modulus, which results in a greater
dissipation of energy during deformation. This is another reason why
froth viscosity is expected to increase as the size of bubbles decrease.
5.1.2. Fraction of lamellae covered by solids
Flotation froth is a three-phase regime involving the presence of
solid particles. The solid particles are distributed on the bubble surfaces
and likely have an effect on the froth rheology. Fig. 11 conrms this hypothesis, showing that the greater the coverage of the lamellae with
particles, the greater the consistency index and therefore the viscosity
of the froth. Regression analysis indicates that there is a 99.64% condence that the slope of this line is positive and non-zero.
The positive correlation between the fraction of lamellae covered by
solids and froth rheology is likely to be due to two reasons:
I. The presence of particles on a lamella likely roughens the bubble
surface as shown in Fig. 12. In the sheared froth, the relative motion
between neighbouring bubbles needs extra energy to overcome the
resistance resulting from the friction between the particles attached
on the bubble surfaces. The higher the proportion of the lamellae
covered, the greater the friction and the more energy is dissipated

62

C. Li et al. / Powder Technology 294 (2016) 5565

Fig. 9. Characteristics of a foam as the air fraction increases (a) low air fraction, (b) medium air fraction and (c) high air fraction.

by the relative motion; consequently, the more viscous is the froth.


Moreover, the particles on the lamellae will cluster together due to
hydrophobic attractive force [42] and Van der Waals' forces between
particles. When relative motion occurs between two neighbouring
bubbles, the particle clusters will be disrupted as the energy overcomes the bonds between particles. The phenomenon is also expected to dissipate shear energy.

II. The presence of hydrophobic particles on the lamellae can lower


surface tension and enhance froth stability [43,44]. A large number
of particles on a bubble surface stabilize a lm by forming a closely
packed monolayer. This prevents the interfaces from touching and
consequent coalescence between bubbles [45], resulting in small
bubble sizes. The probability for the bubbles to coalesce decreases
with an increase in the fraction of lamellae covered by solids, and,
consequently, the bubble size decreases. As discussed in the previous section, the froth becomes more viscous when bubble size is
decreased.

5.1.3. Solids volume fraction in plateau borders and vertices


Within the plateau borders and vertices which form between the
packed bubbles of a otation froth, water and suspended solids exist.
In aqueous suspensions, solids concentration by volume can have a considerable effect on bulk rheology and potentially the solids within the
plateau borders may also be having an effect on the froth rheology.
However, there was no trend observed between the froth viscosity (in
terms of the consistency index) and the estimated solids concentration
in the plateau borders and vertices (Fig. 13). Regression analysis estimates that there is a 95% condence that it is not sufcient to conclude
a correlation existing between the consistency index and the solid
fraction.
Now this observation may be a consequence of either the bubbles
being so tightly packed that the rheology of the suspension between
the bubbles plays no role or the percent solids in the water phase is
too low to be having an effect.
In a dilute system in which the solids volume fraction is less than
0.05, the collisions between particles can be ignored [8]. In such cases,
the inter-particle interaction contributes negligibly to the suspension
rheology. It is known that the suspension viscosity increases with increasing solids volume fraction [4648]. The presence of particles

distorts the ow eld and can therefore be expected to increase the energy dissipation during ow, and hence the viscosity. For such cases,
Einstein [49] developed a relationship between suspension viscosity
and the volume fraction of solids as shown in Eq. (16).
r 1 2:5s

16

where r is the relative viscosity dened as a ratio of suspension viscosity to the viscosity of suspension medium, and s is the volume fraction
of solids.
When the solids volume fraction increases to a moderate level of
about 10%, the average distance between particles becomes approximately equal to their average diameter. The interaction between pairs
of particles starts to inuence the suspension rheology by dissipating
shearing energy. The effect of particle interactions on suspension rheology is complex and can be generally expressed as a Taylor expansion in
powers of the solids volume fraction as shown in Eq. (17) [8], where ci
are expression coefcients.
r 1 2:5s c2 s 2 c3 s 3

17

Above a certain solids concentration, the suspension viscosity increases sharply due to the signicant energy dissipation resulting
from friction between particles. The suspension viscosity tends to innity when the solids volume fraction is close to the maximum where the
suspension stops owing and behaves like a solid body. In general, the
suspension viscosity increases slightly with increasing solid volume
fraction below solids concentration of about 0.50; above this value,
the suspension viscosity increases much more rapidly with increases
in solids concentration [8,46,50].
In this study, the solids concentration in the plateau borders was low
and varied in a narrow range between 5.56% and 10.75% (Fig. 13).
Therefore, it is believed that the contribution of solids concentration to
the viscosity of the suspension in the plateau borders and consequently
to the whole froth phase was low. This will result in any effect of the
solids volume fraction in the plateau borders on froth viscosity being
masked by the solids surface coverage and bubble size effects. Therefore
no clear trend was observed in Fig. 13.
5.2. Modelling froth rheology
According to the above analysis, bubble size and the fraction of lamellae covered by solids are the two important froth properties affecting the

Fig. 10. A schematic of the processes determining the lamella dynamics in sheared froth.

C. Li et al. / Powder Technology 294 (2016) 5565

Fig. 11. Effect of the fraction of lamellae covered by solids on the froth consistency index.

froth rheology. Based on these ndings, the literature was reviewed with
the objective of determining an appropriate model structure that could be
used to predict froth rheology based on the underpinning mechanisms.
Princen and Kiss [10] was found to have developed a model for foam
and highly concentrated emulsions that predicts apparent viscosity as a
function of the foam or emulsion properties.



y
32:0 f 0:73 0 Ca0:5
_

18

where y is yield stress, f is gas hold-up, 0 is the viscosity of the


Newtonian continuous phase and Ca is the capillary number.
Ca

0 R32 _

19

where R32 is the Sauter mean bubble radius and is the interfacial
tension. By substituting Eq. (19) into (18), gives:




 R32 _ 0:5
0
32:0 f 0:73
:
_
0

20

This is of the same form as the modied HerschelBulkley model


originally tted to the experimentally produced froth rheograms in
this study but includes parameters related to the properties of the
froth system its yield stress (when it is relevant), gas hold-up, the
bulk viscosity of the continuous phase which will exist in the plateau
borders, bubble size and interfacial tension.
Interestingly the ow index is assumed to equal 0.5, which is similar
to that observed in the otation froths produced in the experiments in
this study. A number of other researchers have also shown that the
ow index should equal 0.5 [10,30,5156]. The average value of the
ow indices of these 33 tests is 0.48 with the standard deviation being
0.07. A t-test was used to judge if the average value of the ow indices
is statistically different from 0.5. The standard t-value t95; 32 is 2.04
(32, degree of freedom). It is calculated that the t-value is 1.25 in this
study, which is less than the t-value of t95; 32. It is therefore statistically
valid to assume that the ow index is equal to 0.5. Hence, the froth
rheograms of these 33 tests were retted using a constant ow index

Fig. 12. Schematic of particles on bubble surfaces.

63

Fig. 13. Effect of solids concentration in the plateau borders and vertices on the
consistency index of the froth.

of 0.5. Fig. 14 shows the comparison between the apparent viscosities


calculated at various shear rates using the retted Eq. (13) against
those measured experimentally. The degree of t is acceptable with
the average deviation being 4.15% and the regression coefcient (R2)
being 0.92.
In this study, the otation froth exhibited no signicant yield stress.
The bulk viscosity of the continuous phase is expected to be close to the
liquid media and will not change as the proportion of solids in the water
in the plateau borders was reasonably low in all tests (Fig. 13). The interfacial tension, which determines the interfacial modulus, is likely to
be associated with the proportion of the lamellae which is covered by
the hydrophobic particles. Hence, Eq. (20) can be converted to Eq. (21).


 D32 _
f 0:73
2k0

!0:5
21

where k0 is the constant representing those parameters of the system


that should not change between the experiments.
Using the froth property data measured in this study (Table 3), the
apparent viscosity of the otation froth was tted to that predicted by
Eq. (21) by optimising the value of the constant, k0. k0 was found to
equal 1.22 PasNcm 1. It is therefore that Eq. (21) is converted to
Eq. (22). Fig. 15 shows a comparison between the experimental apparent viscosity and that predicted using the model at a number of the
lower shear rate measurements. The deviation of the prediction averages 6.23% and the correlation coefcient (R2) is 0.85. It therefore can

Fig. 14. Comparison between the experimental and calculated apparent viscosity when
n = 0.5.

64

C. Li et al. / Powder Technology 294 (2016) 5565

model. Interfacial surface tension which is an input to the model is likely


to be affected by frother type and dosage as well as the hydrophobicity
of the particles attached to the bubble lamellae.
Froth rheology is expected to affect the overall otation process by
affecting froth residence time. This will be investigated in future work.

Fig. 15. Comparison between the experimental apparent viscosity and that predicted at
various shear rates for all the tests.

be concluded that 85% of the deviation in the results can be accounted


for by the model, which is considered good. Most of the parameters
driving the froth rheology are being taken into consideration.


f 0:73

D32 _
2:44

!0:5
22

6. Conclusions
The effects of froth properties on froth rheology were investigated
under various otation conditions. The froth properties were calculated
based on several assumptions.
The froths exhibited shear-thinning behaviour that can be described
well using a power law model dened by a consistency index and a
ow index. No statistical difference was found between the tted ow
indices in the different tests. Therefore, the consistency index, which is independent of shear rate, was used to evaluate the froth rheology.
The understanding of the rheology of two-phase uids (i.e. aqueous
foams or suspensions) was consulted (and expanded upon) with the
objective of understanding the rheology of three-phase otation froths.
The results showed that the measured froth rheology could be related to
the estimated froth properties in each experiment. Bubble size had a
negative correlation with froth viscosity, which is in line with that observed in two phase dry foams. In a sheared froth, relative motion between bubbles occurs, which results in bubble deformation. The
energy dissipation rate is determined by the friction between bubbles
and the bubble surface dilatational modulus, both of which are negatively correlated to the bubble size. The fraction of lamellae covered
by solids was found to positively inuence the froth viscosity. The presence of solid particles stabilizes froth and consequently generates small
bubble sizes. In addition, the presence of particles on the lamellae will
be expected to increase the surface rigidity and bubble interfacial tension and therefore more energy will be dissipated as the bubbles are
sheared, increasing viscosity proportionally with bubble loading. No
clear relationship between froth viscosity and the solids volume fraction
in the plateau borders and vertices was observed and this was attributed to the percent of solids being low and not varying signicantly.
A model structure was developed which enables prediction of the
apparent viscosity of a three phase otation froth based on froth bubble
size, the proportion of the lamellae covered by particles and the shear
rate applied. This model was able to adequately predict the froth rheology measured in the tests performed in this study. It is hoped that this
model structure can be tted to any otation system to enable prediction of froth rheology. More work, however, is required to validate
this hypothesis. There may also be the potential to further rene this

Nomenclature
A
cell cross-sectional area (cm2)
Ca
capillary number
Sauter mean bubble diameter in the pulp (cm)
D32
D32
Sauter mean bubble diameter in the froth (cm)
G
copper concentrate grade (%)
h
froth height above discharge lip (cm)
supercial gas velocity (cm/s)
Jg
k
constant that accounts for the velocity prole of the
overowing froth
regression constant (PasNcm1)
k0
bubble loading in the froth (g/cm2)
LB
n
ow index ()
P50
mean particle diameter of chalcopyrite measured in the concentrate (m)
R
otation recovery (%)
Sauter mean bubble radius in the froth (cm)
R32
pulp recovery (%)
Rc
froth recovery (%)
Rf
bubble surface area ux in the pulp (cm2/cm2s)
Sb
velocity of the upper surface of the overowing froth (cm/s)
vf
volumetric ow rate of air overowing the lip (cm3/s)
Va
volumetric ow rate of air entering the cell (cm3/s)
VA
volumetric ow rate of chalcopyrite that is entrained in the
Vc
plateau borders and vertices (cm3/s)
concentrate slurry ow rate (cm3/s)
Vcs
volumetric ow rate of silica overowing the discharge lip
Vs
(cm3/s)
volumetric ow rate of water overowing the discharge lip
Vw
(cm3/s)
w
lip width (cm)

fraction of lamellae covered by solids


_
shear rate (s1)
gas hold-up in the froth
f
solids volume fraction in the plateau borders
s

consistency index (Pasn)


chalcopyrite density (g/cm3)
c
silica density (g/cm3)
s

interfacial tension (N/cm)

shear stress (Pa)


yield stress (Pa)
y
bubble surface area ow rate overowing the lip (cm2/s)
f
bubble surface area ow rate in the pulp (cm2/s)
p
solid ow rate of concentrate by mass (g/s)
c

apparent viscosity (Pas)


viscosity of the Newtonian continuous phase in a foam or
0
emulsion (Pas)
relative viscosity
r
Acknowledgments
The authors would like to thank the AMIRA P9P (013908) sponsors
for the project funding. The authors also gratefully acknowledge the
valuable help provided by Mr. Martin Harris from the University of
Cape Town for his advice on froth rheology as well as Professor Tim
Napier-Munn of the JKMRC who assisted with the statistical analysis
of the results, Dr. Sam Morar for analysing the froth images produced
in this work and Professor J.P. Franzidis for proof reading this paper.
The assistance of Jon Worth and others in the pilot plant who helped
set up the test rig is also gratefully acknowledged.

C. Li et al. / Powder Technology 294 (2016) 5565

Reference
[1] X. Zheng, J.P. Franzidis, E. Manlapig, Modelling of froth transportation in industrial
otation cells: part I. Development of froth transportation models for attached
particles, Miner. Eng. 17 (2004) 981988.
[2] M. Zanin, E. Wightman, S.R. Grano, J.P. Franzidis, Quantifying contributions to froth
stability in porphyry copper plants, Int. J. Miner. Process. 91 (2009) 1927.
[3] N. Barbian, J.J. Cilliers, S.H. Morar, D.J. Bradshaw, Froth imaging, air recovery and
bubble loading to describe otation bank performance, Int. J. Miner. Process. 84
(2007) 8188.
[4] J.G. Wiese, P.J. Harris, D.J. Bradshaw, The effect of increased frother dosage on froth
stability at high depressant dosages, Miner. Eng. 23 (2010) 10101017.
[5] S. Farrokhpay, The signicance of froth stability in mineral otationa review, Adv.
Colloid Interf. Sci. 166 (2011) 17.
[6] F.N. Shi, X.F. Zheng, The rheology of otation froths, Int. J. Miner. Process. 69 (2003)
115128.
[7] S. Farrokhpay, The importance of rheology in mineral otation: a review, Miner.
Eng. 3638 (2012) 272278.
[8] J. Mewis, N.J. Wagner, Colloidal Suspension Rheology, Cambridge University Press,
Cambridge, New York, 2012.
[9] H.A. Barnes, J.F. Hutton, K. Walters, An Introduction to Rheology, Elsevier, Amsterdam,
1989.
[10] H.M. Princen, A.D. Kiss, Rheology of foams and highly concentrated emulsions IV. An
experimental study of the shear viscosity and yield stress of concentrated emulsions, J. Colloid Interface Sci. 128 (1989) 176187.
[11] E. Ventura-Medina, J.J. Cilliers, A model to describe otation performance based on
physics of foams and froth image analysis, Int. J. Miner. Process. 67 (2002) 7999.
[12] J.J. Bikerman, Foams, Springer-Verlag, New York, 1973.
[13] D. Wang, Q. Hou, Y. Luo, Y. Zhu, H. Fan, Blocking ability and ow characteristics of
nitrogen foam stabilized with clay particles in porous media, J. Dispers. Sci. Technol.
36 (2014) 170176.
[14] P.C. Harris, V.G. Reidenbach, High-temperature rheological study of foam fracturing
uids, J. Pet. Technol. 39 (1987) 613619.
[15] A.B.J. Kroezen, G.J. Wassink, C.A.C. Schipper, The ow properties of foam, J. Soc. Dye.
Colour. 104 (1988) 393400.
[16] B. Herzhaft, Rheology of aqueous foamsa literature review of some experimental
works, Oil Gas Sci.Technol. 54 (1999) 587596.
[17] G.J. Hirasaki, J.B. Lawson, Mechanisms of foam ow in porous media: apparent
viscosity in smooth capillaries, Soc. Petrol. Eng. J. 25 (1985) 176190.
[18] M.A. Bos, T.V. Vliet, Interfacial rheological properties of adsorbed protein layers and
surfactants: a review, Adv. Colloid Interf. Sci. 91 (2001) 437471.
[19] M. He, Y. Wang, E. Forssberg, Parameter studies on the rheology of limestone
slurries, Int. J. Miner. Process. 78 (2006) 6377.
[20] C. Li, S. Farrokhpay, K. Runge, F. Shi, Determining the signicance of otation variables on froth rheology using a central composite rotatable design, Powder Technol.
287 (2016) 216225.
[21] C. Li, S. Farrokhpay, F. Shi, K. Runge, A novel approach to measure froth rheology in
otation, Miner. Eng. 71 (2015) 8996.
[22] S.H. Morar, M. Harris, D. Bradshaw, The use of machine vision to predict otation
performance, Miner. Eng. 3638 (2012) 3136.
[23] F. Contreras, J. Yianatos, L. Vinnett, On the froth transport modelling in industrial
otation cells, Miner. Eng. 41 (2013) 1724.
[24] J. Yianatos, F. Contreras, F. Daz, Gas holdup and RTD measurement in an industrial
otation cell, Miner. Eng. 23 (2010) 125130.
[25] M.A. Vera, J.P. Franzidis, E.V. Manlapig, Simultaneous determination of collection
zone rate constant and froth zone recovery in a mechanical otation environment,
Miner. Eng. 12 (1999) 11631176.
[26] J.A. Finch, G.S. Dobby, Clolumn Flotation, Pergamon Press, London, UK, 1990.
[27] B.k. Gorain, J.P. Franzidis, E.V. Manlapig, Studies on impeller type, impeller speed
and air ow rate in an industrial scale otation cell. Part 4: effect of bubble surface
area ux on otation performance, Miner. Eng. 10 (1997) 367379.
[28] J. Cilliers, R.A. Asplin, E.T. Woodburn, Kinetic otation modelling using froth imaging
data, in: J.A. Laskowshi, E.T. Woodburn (Eds.), Frothing in Flotation II, Gordon and
Breach Science Publishers, The Netherlands, 1998.
[29] A. Jsberg, P. Selenius, A. Koponen, Experimental results on the ow rheology of
ber-laden aqueous foams, Colloids Surf. A Physicochem. Eng. Asp. 473 (2015)
147155.

65

[30] B. Herzhaft, S. Kakadjian, M. Moan, Measurement and modeling of the ow behavior of


aqueous foams using a recirculating pipe rheometer, Colloids Surf. A Physicochem. Eng.
Asp. 263 (2005) 153164.
[31] N.D. Denkov, S. Tcholakova, K. Golemanov, K.P. Ananthpadmanabhan, A. Lips, The
role of surfactant type and bubble surface mobility in foam rheology, Soft Matter
5 (2009) 33893408.
[32] A. Bjrn, P. Monja, A. Karlsson, J. Ejlertsson, B. Svensson, Rheological characterization, Biogas, InTech, 2012.
[33] T.J. Napier-Munn, Statistical Methods for Mineral EngineersHow to Design Experiments and Analyse Data, 2 ed. Julius Kruttschnitt Mineral Research Centre, Brisbane,
Australia, 2014.
[34] S.J. Cox, A viscous froth model for dry foams in the Surface Evolver, Colloids Surf. A
Physicochem. Eng. Asp. 263 (2005) 8189.
[35] A.H. Falls, J.J. Musters, J. Ratulowskl, Apparent viscosity of foams in homogeneous
bead packs, Soc. Petrol. Eng. J. 4 (1989) 155164.
[36] P.C. Harris, Effects of texture on rheology of foam fracturing uids, Soc. Petrol. Eng. J.
4 (1989) 249257.
[37] S.A. Khan, C.A. Schnepper, R.C. Armstrong, Foam rheology: III. Measurement of shear
ow properties, J. Rheol. 32 (1988) 6992.
[38] A. Bronfort, H. Caps, Faraday instability at foamwater interface, Phys. Rev. E: Stat.
Phys., Plasmas, Fluids, 86 (2012) 15.
[39] S. Tcholakova, N.D. Denkov, K. Golemanov, K.P. Ananthapadmanabhan, A. Lips, Theoretical model of viscous friction inside steadily sheared foams and concentrated
emulsions, Phys. Rev. E: Stat. Phys., Plasmas, Fluids 78 (2008) 118.
[40] S. Costa, R. Hhler, S. Cohen-Addad, The coupling between foam viscoelasticity and
interfacial rheology, Soft Matter 9 (2013) 11001112.
[41] M. Karbaschi, M. Lot, J. Krgel, A. Javadi, D. Bastani, R. Miller, Rheology of interfacial
layers, Curr. Opin. Colloid Interface Sci. 19 (2014) 514519.
[42] G.V. Franks, H.H. Li, J.-P. O'Shea, G.G. Qiao, Temperature responsive polymers as
multiple function reagents in mineral processing, Adv. Powder Technol. 20 (2009)
273279.
[43] D. Tao, G.H. Luttrell, R.-H. Yoon, A parametric study of froth stability and its effect on
column otation of ne particles, Int. J. Miner. Process. 40 (2000) 2543.
[44] Z. Aktas, J.J. Cilliers, A.W. Banford, Dynamic froth stability: particle size, airow rate
and conditioning time effects, Int. J. Miner. Process. 87 (2008) 6571.
[45] G. Johansson, R.J. Pugh, The inuence of particle size and hydrophobicity on the stability of mineralized froths, Int. J. Miner. Process. 34 (1992) 121.
[46] P.K. Senapati, B.K. Mishra, A. Parida, Modeling of viscosity for power plant ash slurry
at higher concentrations: effect of solids volume fraction, particle size and hydrodynamic interactions, Powder Technol. 197 (2010) 18.
[47] A.A. Zaman, A.L. Fricke, Correlations for viscosity of kraft black liquors at low solids
concentrations, AICHE J. 40 (1994) 187192.
[48] S. Farrokhpay, G.E. Morris, D. Fornasiero, P. Self, Stabilisation of titania pigment particles with anionic polymeric dispersants, Powder Technol. 202 (2010) 143150.
[49] A. Einstein, Investigations on the Theory of the Brownian Movement, Dover, New
York, 1956.
[50] J.S. Chong, E.B. Christiansen, A.D. Baer, Rheology of concentrated suspensions, J.
Appl. Polym. Sci. 15 (1971) 20072021.
[51] S.P.L. Marze, D. Langevin, A. Saint-Jalmes, Aqueous foam slip and shear regimes
determined by rheometry and multiple light scattering, J. Rheol. 52 (2008)
10911111.
[52] S. Tcholakova, I. Lesov, K. Golemanov, N.D. Denkov, S. Judat, R. Engel, T. Danner,
Efcient emulsication of viscous oils at high drop volume fraction, Langmuir 27
(2011) 1478314796.
[53] R. Foudazi, I. Masalova, A.Y. Malkin, Flow behaviour of highly concentrated emulsions of supersaturated aqueous solution in oil, Rheol. Acta 50 (2010) 897907.
[54] G. Gutirrez, M. Matos, J.M. Benito, J. Coca, C. Pazos, Preparation of HIPEs with controlled droplet size containing lutein, Colloids Surf. A Physicochem. Eng. Asp. 442
(2014) 111122.
[55] B. Dollet, C. Raufaste, Rheology of aqueous foams, C. R. Phys. 15 (2014) 731747.
[56] G. Ovarlez, S. Rodts, A. Ragouilliaux, P. Coussot, J. Goyon, A. Colin, Wide-gap Couette
ows of dense emulsions: local concentration measurements, and comparison
between macroscopic and local constitutive law measurements through magnetic
resonance imaging, Phys. Rev. E: Stat. Phys., Plasmas, Fluids 78 (2008) 113.