See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/295563299

The effect of impeller type on the mixing time of the non-Newtonian fluids in
stirred tanks

Article in Indian Journal of Science and Technology · January 2016

CITATIONS READS

4 3,301

3 authors:

Hani Taleshi Jafarsadegh Moghaddas

University of Canterbury Sahand University of Technology
7 PUBLICATIONS 45 CITATIONS 94 PUBLICATIONS 1,113 CITATIONS

SEE PROFILE SEE PROFILE

Pariya Noeparvar
University of Canterbury
3 PUBLICATIONS 8 CITATIONS

SEE PROFILE

All content following this page was uploaded by Hani Taleshi on 24 February 2016.

The user has requested enhancement of the downloaded file.

 ISSN : 0974 - 7443 Volume 11 Issue 4

chemical technology
An Indian Journal
Full Paper
CTAIJ 11(4) 2016 [123-132]

The effect of impeller type on the mixing time of the non-Newtonian

fluids in stirred tanks
Hani Taleshi Ahangari*, Pariya Noeparvar, Jafar Sadegh Moghaddas
Sahand University of Technology, Transport Phenomena Research Center, Chemical Engineering Faculty, Tabriz, (IRAN)
E-mail: hanitaleshi@gmail.com; paria1219@gmail.com; jafar.moghaddas@sut.ac.ir

ABSTRACT KEYWORDS
In the following paper, experimental data represent mixing time in a stirred Mixing time;
tank. Electrical conductivity methods are used to obtain results of mixing Concave impeller;
time. Weight percent of CMC solution, which indicates pseudo plastic Speed;
behavior of the fluid, was deemed 0.5%. What is more, it was used as Aeration;
continuous phase and air as the dispersed phase. The effect of blade angle Non-Newtonian fluids.
and length were investigated with 7 concave impellers. The experimental
results demonstrate that the mixing time will decrease by increasing rota-
tion speed. The impeller No. 6 is favorable to reduce mixing time in con-
stant power consumption because of the maximum angle between the other
impellers. Consequently, an equation to predict mixing time by using four
independent parameters is suggested; moreover, this equation can be uti-
lized as a predictor relation in comparison with the experimental data.
 2016 Trade Science Inc. - INDIA

INTRODUCTION applied in food industries such as yoghurt and soup.

Shear stress behaviors of these fluids are not
Mixing is a combination of two or more differ- investigable with the relations of Newtonians flu-
ent substances which results ideal homogeneous ids; in other words it is difficult to predict interac-
physical and chemical products. This is a technique tions. The few available researches have focused
used widely within industries such as paper, plas- on macroscopic aspects of the flow; therefore, re-
tics, ceramics, foods, medicines and so forth[1]. Mix- search regarding fluid foundations is serious need
ing time is a significant determinant for phase mix- in determining fluid properties (Rheology)[3,4]. Stirred
tures as well as parameters in phenomena interpre- tanks are more applicable than any other mixer sys-
tation. The methods of mixing time measurement are tems; furthermore, tanks with stirrers have the larg-
divided in two groups: physical and chemical. The est surface of contact and are the best facilities for
comparison of mixing time highly depends on the mixing in non-Newtonian fluids[5]. Mixers in the tanks
definitions, methods of measurement, non-homoge- create a series of cuts and flow lines in the liquids.
neous, types of probe and tools by which the tracer Effects in transfer of momentum and layers of liquid
is injected and its location[2]. An important part of movements in sections of the tank eventuate in com-
industry products consists of non-Newtonian fluids bining of liquids into a single mixture[6, 7].
124 The effect of impeller type on the mixing time of the non-Newtonian fluids in stirred tanks CTAIJ, 11(4) 2016

Full Paper
The power law shows the behavior of the pseudo 1  
plastic fluid (Eq. 1): W   3
 0.0638 (V ) 2  (4)
1.622  T 
  K n (1)
Where  , K,  and n are shear stress, power law Where  W , T,  and V are shear stress at the wall,
constant or consistency index, shear rate and power tank diameter, fluid density and tank volume, respec-
law exponent or flow behavior index, respectively. tively. It is worth to note that the contribution of the
In which relates the shear stress in the fluid to the pressure can be ignored compared to the torque. Thus
shear rate being exerted on it. Dynamic viscosity the estimated shear stress on the wall is calculated
relation is indicated by Eq. 2: by Eq. 5[8]:

 A   K  n1 1 
(2) W    (5)
 1.622  T 3 
Where  A is apparent viscosity of the fluid[8]. Bird By respecting Eq. 2, the Reynolds number should
et al. predict an equation for the shear rate on the be considered by Eq. 6[8-10]:
wall of a baffled tank as a function of the torque on ND 2 ND 2
Re   (6)
the mixer shaft as shown by Eq.3:  K n1
 v  Where Re, N and D are Reynolds number, impeller
    R   dS   Rpbaff dA (3)
S  r W A
rotational speed and impeller diameter, respectively.
Where  , R, r, S, P baff, and A are torque on the shaft, If an analogy is used between the pressure loss
tank radius, radius, wall area, pressure on the baffles characteristic of the Non-Newtonian fluid and the
and area of baffles, respectively. The shear stress power characteristic of a stirrer, Eq. 7 is achieved[3, 8, 9]:
on the wall in a tank can be estimated by assuming  N 2 n D 2   n  n 
the constant shear rate on the wall and applying the Re    8   (7)
 K  6n  2  
appropriate integration limits followed by Eq. 4:

Figure 1 : Tank experimental system

chemical technology
An Indian Journal
CTAIJ, 11(4) 2016 Hani Taleshi Ahangari et al. 125

Full Paper
Equipment and method The aqueous solution in this research is of 0.5%
Geometric characteristics and experimental sys- in weigh of CMC. The 0.5% solution and measure-
tem of the tank are shown in Figure 1 and TABLE 1, ments of viscosities
[5]
include the values of K=0.66
respectively. and n=o.66 105.18 g of CMC should be solved in
Figure 2 illustrates the shape of a concave im- water based on of the stirred tank volume. A con-
peller and the specifications of seven impellers are ductivity meter was used to measure mixing time
indicated by TABLE 2. and conductivity. A multi meter was also consid-
In the event of two phases, gas flow rate and ered to change the electrical conductivity data to digi-
distributed inside the liquid phase are measured by tal ones. 30 ml of saturated KCl solution was uti-
rotameter and gas sparger with the diameter of 7.5 lized as a tracer acted as a pulse at each step. The
cm with 31 holes each (1 mm diameter), respec- mixing time should be taken to inject the tracer until
tively. The behavior of the non-Newtonian fluid can the contents of the tank reaches the favored degree
be determined by Rheometer; subsequently a shear of homogeneous is recorded. The degree of homog-
stress is provided by exerting a shear rate on the enous is explained by the Eq. 8:
solution. Based on these data, K and n in the Eq. 1 Y  (C i C 0 ) /(C   C 0 ) (8)
are achieved. TABLE 3 shows the results of Rhe- Where C i , C 0 , C  and Y are initial, final, moment
ometer data in different weight percent solutions. concentration of bulk mixture and degree of unifor-
TABLE 1 : Tank geometric characteristics
Geometric characteristics
30Cm Tank internal diameter (T)
30Cm Liquid height (H)
3T Agitator diameter (D)
10 T Baffles width (B)
100 T Baffles thickness (h)
0.5 T Distance between agitator and bottom of tank (C1)
0.5 T Distance between agitator and liquid level (C)
- Distance between two agitator (C)
- Distance between upper agitator and liquid surface (C2)
- Distance between lower agitator and liquid surface (C1)

Figure 2 : Shape of a concave impeller

chemical technology
An Indian Journal
126 The effect of impeller type on the mixing time of the non-Newtonian fluids in stirred tanks CTAIJ, 11(4) 2016

Full Paper
TABLE 2 : specifications of seven impellers (m)
Impeller 1 Impeller 2 Impeller 3 Impeller 4 Impeller 5 Impeller 6 Impeller 7
Blade height (w) 0.025 0.0 25 0.025 0.0125 0.0375 0.050 0.025
Differences between blades (B-C) 0.00375 0.00 625 0.0125 0.00625 0.00625 0 0
Blades angle (Degree) 40 40 40 25 50 55 40
Blades thickness(t) 0.002 0.0 02 0.002 0.002 0.002 0.002 0.002
Impeller diameter (D) 0.1 0.1 0.1 0.1 0.1 0.1 0.1

TABLE 3 : Samples of rheology results [4]

0.4 wt% 0.5 wt% 0.6 wt% 0.7 wt%
Shear stress Shear Shear Shear Shear
Shear stress (pa) Shear stress (pa) Shear stress (pa)
(pa) Rate (1/s) Rate (1/s) Rate (1/s) Rate (1/s)
3.63 75 4.13 50 4.79 40 7.15 40
3.8 80 4.7 60 5.64 50 8.57 50
3.97 85 5.29 70 6.5 60 9.72 60
4.13 90 5.84 80 7.23 70 10.8 70
4.29 95 6.35 90 7.92 80 11.9 80
4.44 100 6.85 100 8.56 90 12.8 90
5.95 150 8.92 150 9.19 100 13.8 100
7.45 200 10.9 200 14.6 200 20.8 200
10.2 300 14.4 300 18.9 300 25.8 300
12.6 400 17.5 400 22.5 400 29.3 400
14.9 500 20.3 500 25.8 500 33.2 500
17 600 23 600 28.8 600 37 600
19 700 25.4 700 31.6 700 40 700
20.9 800 27.5 800 34.5 800 43.7 800
23.2 900 29.5 900 37.2 900 46.8 900
25.2 1000 32.1 1000 39.7 1000 49.8 1000
27.2 1100 34.3 1100 42 1100 52.8 1100
29.1 1200 36.3 1200 44.3 1200 55.7 1200
30.8 1290 37.9 1290 46.4 1290 58.1 1290
k= 0.14 k= 0.66 k= 0.90 k= 1.01
n= 0.75 n= 0.66 n= 0.64 n= 0.58
mity, respectively. Reaching the proper mixing, for 4 according to the used fluid, speeds and related
instance, Y=1 in a limit time is almost impossible. equations.
A favored degree of uniformity is in the range of At first, the effect of rotation speed on mixing
Y=0.9-0.99, and hence there are various explana- time was examined for each of the 7 impellers. The
tions of mixing time due to this[10, 11]. In this experi- rotation speeds of 200 and 400 and 600 rpm were
ment the degree of homogeneity for mixing time is tested in this study. As shown in Figures 3, 4 and 5,
equal to Y=0.95. the mixing times for the impellers were measured at
different gas flow rate.
RESULTS AND DISCUSSION It can be depicted that high rotation speeds is
directly responsible for short mixing time in all
Effect of rotation speed cases. Rotational speed increases the power con-
Reynolds number, torque on the shaft, shear stress sumption, causing the increase of stress rates. The
on the wall and shear rate were shown in the TABLE increase in shear stress is independent to the type of
chemical technology
An Indian Journal
CTAIJ, 11(4) 2016 Hani Taleshi Ahangari et al. 127

Full Paper
TABLE 4 : Reynolds number at different speed for the 0.5 weight percent solution

N(rpm) N(rps) Re  W W
100 1.66 69.5 5.326 121.592 2706.2
200 3.33 175.57 21.434 489.33 22313.7
300 5 301.7 48.325 1103.259 76477
400 6.66 441.4 85.739 1957.42 182313
500 8.33 594.4 134.128 3062.14 359148
600 10 757.9 193.3 4413 624802

Figure 3 : Mixing time versus rotation speed for 7 impellers at Q=0 lit/hr

Figure 4 : Mixing time versus rotation speed at Q=100 lit/hr

Figure 5 : Mixing time versus rotation speed at Q=200 lit/hr

chemical technology
An Indian Journal
128 The effect of impeller type on the mixing time of the non-Newtonian fluids in stirred tanks CTAIJ, 11(4) 2016

Full Paper
impeller and causes to lead the rotation of the fluid 200, 400 and 600 rpm are shown for each of the
in axial and radial directions. individual impellers in Figures 6, 7 and 8.
Flow convection increases the Kinetic Energy At the presence of gas phase, different results
of Eddy resulting in mixing time decrease. Mixing occurred due to the complexity of two phase sys-
time, in terms of velocity, have a sharp slope at first tems and the effects of various parameters on the
and then decrease, so that at the speeds of 600 rpm distribution of bubbles. One of these parameters is
and higher. The mixing time was not significantly the preferable impeller which was tested in this
different and power consumption will increase in- study. In general, the effect of aeration on mixing
dependently at this operation speed, subsequently, time changes in 4 cases as below.
the fluid will behave as Newtonian fluids. It was 1. Mixing time initially increased, and then de-
observed that the impellers numbers 1, 2 and 7 will creased with aeration.
have the same result because of their equal blade 2. Mixing time increases with aeration.
height, angle; moreover, slight difference between 3. Mixing time initially decreased and then in-
them is due to the values in different blades for all creased with aeration.
rotation speeds. The impeller number 3 has the maxi- 4. Mixing time is independent of aeration.
mum blade difference but its blade height and angle As a rule of thumb it can be concluded that the
is the same with impeller numbers 1 and 2, so the first case occurs when the impeller is unable to ef-
blade difference only causes to mixing time becom- fect the gas distribution and causes the impeller
ing longer. The impeller numbers 5 and especially 6 power to reduce. There is a critical value for this
have the minimal mixing time in the constant power situation which after this, aeration helps the mixing
consumption due to the maximum angle and height (for example impellers No.2 and No.3). The second
between the blades. These impellers pump a higher case is a subset of the first one which the impeller
flow in a rotation, thereby creating holes in the back operates weaker in gas distribution or disturbance
of the blades. This causes the rotation speeds there- is not to that much that gas distribution is effective
fore increase whilst mixing time decreased. Hence (for example impeller No.4). The third case occurs
it was clear that with increasing the rotation speed, for the impellers which are larger in size because
mixing time was almost equal to that obtained by they reach to a maximum pneumatic mix suddenly
different impellers and is almost independent of the (for example impellers No.5 and No.6). The fourth
type of impeller. case depends on the operating conditions and such a
situation is usually observed at high speeds.
Effect of aeration The existence of gas phase in one hand causes
As gas is a dispersed phase, the effect of gas on the increase of pneumatic mixing in the system. That
mixing time is expressed in terms of gas flow rate. leads to shorter mixing time, but on the other hand
Effects of aeration on mixing time at the speeds of mechanical mixing can be reduced by gas, resulting

Figure 6 : Mixing time versus gas flow rate at 200 rpm

chemical technology
An Indian Journal
CTAIJ, 11(4) 2016 Hani Taleshi Ahangari et al. 129

Full Paper

Figure 7 : Mixing time versus gas flow rate at 400 rpm

Figure 8 : Mixing time versus gas flow rate at 600 rpm

Figure 9 : Effect of blade angle on mixing time

in a higher mixing time. It is important to realize that ing by impeller and reach to the liquid surface. By
the two factors are most effective at any test. There- increasing the speed to 600 rpm, it was found that the
fore, a different trend is observed for each impeller gas flow rate has little effect on mixing time. At me-
at any speed. At low speeds, the impeller No. 4 is dium to high speeds due to the prominent role of the
weak in the gas distribution. A major part of bubbles rotation speed in creating turbulence flow, the gas flow
reduce the mechanical mixing power without break- rate does not significantly reduce mixing time.
chemical technology
An Indian Journal
130 The effect of impeller type on the mixing time of the non-Newtonian fluids in stirred tanks CTAIJ, 11(4) 2016

Full Paper

Figure 10 : Blade length effects on mixing time in different aerated intensity

Figure 11 : Experimental and calculated mixing time at 200 rpm

Figure 12 : Experimental and calculated mixing time in presence of gas at 200 rpm

comes shorter. By increasing the rotation speed, ef-

EFFECT OF GEOMETRY
fects of blade angel decrease and nearly for all im-
pellers, one mixing time was obtained.
Blade angel
Blade length
The effect of blade angle on mixing time is shown
in Figure 9 for 25, 40 and 50 degrees. The effects of blade length on mixing time at dif-
For each of the cases, by increasing the blade ferent aeration rate, for the angle of 40 degrees and
angle, the flow number will rise and the impeller the speed of 200 rpm is shown in Figure 10. It can be
pumps increase flow to the rotation; consequently, illustrated that with increasing the blade length, the
homogeneity is obtained earlier and mixing time be- ability of impellers for breaking bubbles are reduced.
chemical technology
An Indian Journal
CTAIJ, 11(4) 2016 Hani Taleshi Ahangari et al. 131

Full Paper
These bubbles decrease the impellers mechanical sult between them and Eq. 9 is appropriate for pre-
power without participating in flow patterns. dicting mixing time.

CALCULATING MIXING TIME ACKNOWLEDGMENT

According to the experimental data and linear The authors would like to thank the technical
regressions, calculating mixing time is suggested staffs of Sahand University of Technology.
based on Eq. 9:
   N  200   
t  150  0.281   0.221 
   25  
    0.156  
 0.042  Q 
   0.04  L    (9)
FUNDING
   400     30    200 

Where t, è, L and Q are mixing time, blade angle, Transport Phenomena Research Center and the
blade length and gas flow rate, respectively. management of Sahand University of Technology are
Figures 11 and 12 depict the difference between the funding sources for this project.
experimental mixing time and calculated mixing time.
NOMENCLATURE
CONCLUSIONS

In this research the gas-liquid macro mixing in a  shear stress, N 2

m
stirred tank of specific geometry has been investi- power law co nstant or consistency index,
gated with solution of 0.5% in weigh of CMC as the K
pa  s n
continuous phase and air as the dispersed phases 1
 shear rate, s
with 7 different concave impellers. The mixing time
A apparent viscosity, pa. s
has been determined by means of electrical conduc-
 torque on the shaft, N.m
tivity method. The following findings were con-
R tank radius, m
cluded: 2
S wall area, m
The most efficient factor in reducing the mixing
Pbaff pressure on the baffles, pa
time is the rotation speed that by increasing dead
zones which are affected by circulation will be lost A area of baffles, m2
and homogeneity occurs earlier and mixing time gets W shear stress at the wall, N
m2
shorter. It occurred for all 7 impellers. Decreasing T tank diameter, m
mixing time with increasing rotation speed was fast  fluid density, Kg 3
m
at first but then the curve grew less steep. 3
V tank volume, m
Increasing the rotation speed will also increase
D impeller diameter, m
the power consumptions. In this study by respecting
Ci mol
these factors the best speed is between 400 and 600 initial concentration of bulk mixture, lit
rpm. C0 mol
The type of impeller is the most important factor moment concentration of bulk mixture, lit
in a constant power consumption that can reduce the C mol
final concentration of bulk mixture, lit
mixing time. Because of the maximum blade angle t mixing time, sec
and height, the impeller number 6 is the best and  blade angle, degree
impeller number 4 is the worst. L blade length, m
Existence of gas phase has different effects on Q lit
gas flow rate, hr
mixing time that depends on the parameters such as
rotation speed, kind of impeller and the gas flow rate. Re Reynolds number
Calculated mixing time compared with the ex- N impeller rotational speed, rp m
n power law expo nent or flow behavior index
perimental data shows that there is acceptable re-

chemical technology
An Indian Journal
 132 The effect of impeller type on the mixing time of the non-Newtonian fluids in stirred tanks CTAIJ, 11(4) 2016

Full Paper
REFERENCES [7] G.Ascanio, B.Castro, E.Galino; “Measurement of
power consumption in stirred vessels”, A-Review,
[1] F.A.Holland, R.Bragg; “Fluid flow for chemical en- Chem.Eng.Res.and Des., DOI: 10.1205/
gineers”, Edward Arnold, London, (1973). cerd.82.9.1282.44164, 82(9), 1282-1290 (2002).
[2] L.Manna; “Comparison between physical and [8] E.L.Paul, Atiemo A.Obeng, S.M.Kresta; “Handbook
chemical methods for the measurement of mixing of industrial mixing, science and practice”, ISBN,
times, “Chem.Eng.Journal, DOI: 10.1016/S1385- 978-0-471-26919-9 (2004).
8947(97)00059-4, 67(7), 167-173 (1997). [9] M.Zlokarnik; “Stirring : Theory and Practice”, DOI:
[3] R.P.Chhabra, J.F.Richardson; “Non-newtonian flow 10.1002/9783527612703.ch01, (2001).
in the process industries”, Butterworth-Heinemann, [10] Q.Zhang, Y.Yong, Z.S.Mao, C.Yang, C.Zhao; “Ex-
ISBN: 978-0-7506-3770-1, (1999). perimental determination and numerical simulation
[4] P.E.Arratia, T.Shinbrot, M.M.Alvarez, F.J.Muzzio; of mixing time in a gas-liquid stirred tank”,
“Mixing of non-newtonian fluids in steadily forced Chem.Eng.Sci., DOI: 10.1016/j.ces.2009.03.030,
systems”, Physical Review Letters, DOI: 10.1103/ 64(12), 2926-2933 (2009).
PhysRevLett.94.084501, 94, 1-3 (2005). [11] V.Roussinova, S.M.Kresta; “Comparison of con-
[5] J.Eghbali; “A mixing study of non-Newtonian fluids tinuous blend time and residence time distribution
in stirred tanks”, M.Sc.thesis, Sahand University of models for a stirred tank”, J.Ind.Eng.Chem.Res.,
Technology, Tabriz, Iran, (2011). 47(10), 3532-3539 (2008).
[6] M.Zlokarnik; “Stirring: theory and practice”, Wiley-
VCH Verlag GmbH, (2001).

chemical technology
An Indian Journal

View publication stats