close

Вход

Забыли?

вход по аккаунту

?

Theshaft torque change in a laboratory scraped surface heat exchanger used for making ice slurries.

код для вставкиСкачать
ASIA-PACIFIC JOURNAL OF CHEMICAL ENGINEERING
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
Published online 13 September 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI:10.1002/apj.073
Research Article
The shaft torque change in a laboratory scraped surface
heat exchanger used for making ice slurries
Frank G. F. Qin,1,2 Shashini Premathilaka,2 Xiao Dong Chen1,3 * and Kevin W. Free1,2
1
Freezcon Ltd., Auckland, New Zealand
Department of Chemical and Materials Engineering, The University of Auckland, Auckland City, New Zealand
3
Department of Chemical Engineering, Monash University, Clayton Campus, Victoria, Australia
2
Received 20 October 2006; Revised 18 June 2007; Accepted 22 June 2007
ABSTRACT: In this paper, extensive laboratory test results of a scraped surface heat exchanger for making ice slurries
are reported. The data have been analysed and interpreted. Some sensible conclusions have been obtained that should
provide an excellent platform for developing further fundamental understanding of the phenomena and establishing a
practical guide for design.  2007 Curtin University of Technology and John Wiley & Sons, Ltd.
KEYWORDS: scraped surface heat exchanger; shaft torque; ice slurry
INTRODUCTION
Scraped surface heat exchangers (SSHE) have been
used in a variety of engineering fields, such as in
the process of freeze concentration of aqueous solutions (Dass and Grenco, 1991). In recent years, using
SSHE to produce pumpable ice slurries as a secondary
refrigerant has aroused interest in the air-conditioning
industry because ice slurries have much higher thermal
capacity compared to conventional single-phase secondary coolants (Bel and Lallemand, 1999; Inada et al .,
2000; Knodela et al ., 2000; Tanino and Kozawa, 2001;
Kitanovski and Poredos, 2002; Saito, 2002; Ayel et al .,
2003). Moreover, ice slurries are an excellent medium
for cold storage using off-peak electricity, which is a
hot topic regarding the peak-load shift in energy industry. Of course, there are a number of ways to produce ice-containing fluids as a secondary refrigerant
besides using the sub-cooled SSHE (Tsuchida et al .,
2002; Ismail and Radwan, 2003; Meewisse and Ferreira,
2003). For instance, in a patented device, Wiegandt
et al . utilised the direct contact of butane with water to
partially freeze the water and produce ice slurries. The
butane vapour was continuously recovered and reused
(Wiegandt et al ., 1987). Fluidised bed was reported to
be used to produce ice slurries as well, where solid
particles were used in the flow of aqueous solution or
water to scratch the ice out of the cooling (tube) surface
(Habib and Farid, 2002; Meewisse and Ferreira, 2003).
*Correspondence to: Xiao Dong Chen, Department of Chemical Engineering, Monash University, Clayton Campus, Victoria, Australia. E-mail: dong.chen@eng.monash.edu.au
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Avalanche Ice Harvest system uses a simple shell-andtube heat exchanger to produce ice slurries from 7 wt%
ethylene glycol aqueous solution (Paul Mueller Company (U.S.), 2005).
However, using the sub-cooled SSHE still remains an
interesting method because of its reliability and safety
(Inada et al ., 2000; Ishikawa et al ., 2002; Yamada
et al ., 2002). A variety of SSHE configurations are currently available. The scraper can be constructed simply
as a straight-blade wiper or as a helical-blade auger (Bel
and Lallemand, 1999a,b). The processing surface can
be the exterior of the cooling tube or the interior of a
cooling jacket (Trommelen and Beek, 1971b; Ishikawa
et al ., 2002; Yamada et al ., 2002). The shaft torque of
the scraper is influenced by the increasing ice fraction
in the process fluid. The torque may also be affected by
the temperature of the cooling surface. The power consumption can be very high compared to a stirred tank
without a scraper (Trommelen and Beek, 1971b). This
makes the design and operation of an SSHE used for
freezing very different from an SSHE used for heating
or cooling (without phase change). A number of studies regarding the design and performance of SSHE have
been reported; however, the data regarding the torque
variation, especially when phase change occurs on the
scraped cooling surface, is still inadequate (Trommelen
and Boerema, 1966; Leung, 1967; Trommelen, 1967;
Trommelen and Beek, 1971a; Trommelen et al ., 1971;
Ben Lakhdar et al ., 2005).
In a previous study, we reported the characteristics of
the heat transfer and power consumption of a lab-scale
SSHE used for freezing aqueous solutions (Qin et al .,
2005). The current study will further investigate in more
Asia-Pacific Journal of Chemical Engineering
SHAFT TORQUE CHANGE IN A SCRAPED SURFACE HEAT EXCHANGER
detail the mechanisms of torque increase in a subcooled SSHE used as an ice slurry generator. Special
attention will be paid to three key aspects that may
influence the shaft torque: Eqn (1) the temperature (or
degree of suercooling) of the cooling surface, Eqn (3)
the ice content in the working fluid and Eqn (4) the
rotational speed of the scraper. Water and two aqueous
solutions were used as the working fluids in this study
and extensive experimental runs were conducted. A
number of conclusions were reached.
When an SSHE is used for making ice slurries, the
shaft torque required to drive the rotational scraper is
composed of three parts: the torque that overcomes the
friction between the blades and the cooling surface; the
torque that overcomes the resistance of the working
fluid; and the torque that overcomes the adhesive force
of ice on the cooling surface. The frictional fraction on
the smoothly machined cooling surface stays relatively
stable over a wide range of scraper rotational speeds.
The shaft torque due to the flow resistance is similar
to that due to the propeller agitator, and can therefore
be expressed as a function of Reynolds number in the
SSHE. The ice adhesive fraction is influenced by both
the ice content in the working fluid and the thickness of
the fouling layer of ice that develops on the cooling
surface. However, if the scraping action is applied
quickly to avoid the ice film from growing thicker,
then the thin film of ice would act as a lubricating
layer to reduce the scraping friction. Moreover, if the
ice is discharged out of the SSHE to maintain a low
ice content in the working fluid, the chance for ice
to adhere on the cooling surface will be effectively
reduced. Working on this manner, the shaft torque is
independent of the degree of supercooling of the cooling
surface.
EXPERIMENTAL
Apparatus
A schematic of the experimental set-up used for measuring the shaft torque is shown in Fig. 1. The SSHE
used in this study was constructed with two concentric
stainless steel tubes. The inner tube was used as a cylindrical vessel in which the ice slurry was formed. The
coolant, which was an ethylene glycol aqueous solution in about 50 wt%, was circulated between the two
tubes, thereby forming a cooling jacket. A rotational
scraper with two spring-forced polyethylene straight
blades was applied on the interior of the vessel. Its rotational speed and shaft torque were measured by a torque
sensor and transmitted to a computer for data acquisition (E302, Sensor Technology Ltd. UK). The power
consumption of the scraper was also automatically calculated according to the rotational speed and the torque
via the computer software. The shaft speed (rpm) of the
motor could be pre-set by a programmable speed controller (X302, PDL Electronics Ltd. NZ). A HAAKE
K20 Unit (Gebrüder HAAKE GmbH, Germany), which
was a multi-functional water bath with a built-in refrigerator and trance heater, circulated the coolant through
Figure 1. Schematic of the experimental rig (Torqsense transducer is used for measuring
and for transmitting the data for the torque measurement). This figure is available in
colour online at www.apjChemEng.com.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
619
FRANK G. F. QIN ET AL.
Asia-Pacific Journal of Chemical Engineering
scraping
blade
ice
inner cylinder
wall
outer cylinder wall
Tc2
Tb
Ts2
Ts1
Tb
thermal insulation
working
fluid
coolant
bulk
Ts1 scraped cooling surface temperature
Ts2 wall temperature on coolant side
Tc1 coolant inlet temperature
Tc2 coolant outlet temperature
Tc = (Tc1 + Tc2) / 2
T-type thermocouples
Tc1
Figure 2. Location of thermocouples in the SSHE. This figure is available in
colour online at www.apjChemEng.com.
the cooling jacket. The temperatures of the coolant inlet,
outlet, bulk solution in the SSHE and the cooling surface (inner cylinder wall) were measured with a set of
K-type thermocouples and transmitted to the computer
via a TC-801data logger (Pico Technology Ltd. UK) as
shown in Fig. 2.
An external loop, which circulated the working fluid
between the SSHE and a store cistern, was used to
control the ice content in the SSHE. This is shown in
the dotted region of Fig. 1.
Tf =−0.358C-0.0081C2−8×10−5C3
Rc=0.9996
-5
−10
−15
−20
−25
−30
0
5
10
15
20
25
30
35
40
Ethanol concentration, C (wt%)
Figure 3.
Freezing point of the ethanol aqueous
solution. This figure is available in colour online at
www.apjChemEng.com.
Materials and method
Three fluids, water, an ethanol aqueous solution (5 wt%
with freezing point at −2.1 ◦ C) and a sucrose solution
(10 wt % with freezing point at −0.65 ◦ C), were used
as the working fluids in this study. The reasons for
using the ethanol solution were the following: (1) only
a low concentration is needed to produce a significant
freezing point depression; (2) the solution viscosity
does not vary a great deal as the ethanol concentration
is changed, and therefore the torque increase can be
assumed to be due to the accumulation of the ice
content only. The working fluids were partially frozen
to produce ice slurries during the process. The freezing
point depression (FPD) of ethanol aqueous solution and
the refractive index are functions of the concentration.
These data are available in CRC handbook and other
literature (Lide, 1996). The expression of FPD versus
concentration is correlated in Eqns (1) and (2), and is
plotted in Fig. 3.
Tf = −0.358C − 0.008C 2 − 8 × 10−5 C 3 ,
(r 2 = 0.9996)
0
Freezing point, Tf (°C)
620
(1)
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
or inversely:
C = −2.588Tf − 0.086Tf 2 − 0.002Tf 3 ,
(r 2 = 0.9996)
(2)
where C is the weight concentration of ethanol, Tf the
freezing point of the solution and r the correlation coefficient. Thus, the concentration in the unfrozen solution
can be determined by measuring the bulk temperature,
or by measuring the refractive index with the Abbe
refractometer (Bellingham and Stanley Limited, UK).
Furthermore, the solid ice content in the ice slurry can
be calculated from the concentration change of ethanol
in the unfrozen solution of the ice slurry. This is shown
by Eqns (3) and (4):
C0
(3)
mice = m0 × 1 −
C
mice
C0
=
(4)
=1−
m0
C
where m0 is the total mass of the original solution, mice
is the ice mass formed in the SSHE, C0 is the original
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
SHAFT TORQUE CHANGE IN A SCRAPED SURFACE HEAT EXCHANGER
concentration of ethanol in the solution (wt%), C is the
ethanol concentration in the unfrozen solution after ice
is formed and is the mass fraction of ice in the ice
slurry.
When a force (F ) is needed to drive a scraping blade,
the torque accordingly is
M = bFR
scraper is ω (= 2π n/60), then the power consumption
can be written as:
P = ωM = ωbFR =
π nbFR
30
(6)
where n is the rotational speed of the scraper (rpm).
As mentioned in the introduction, when there is
neither ice in the working fluid nor ice crystallization on
the cooling surface, a base torque (Mb ) is still required
to spin the scraper. The base torque is composed of two
parts: (1) the torque that overcomes the friction between
the blades and the cooling surface (Mf ) and (2) the
(5)
where b is the blade number of the scraper, which is
equal to 2 in this study, and R is the inside radius of the
cylinder (Fig. 4). If the angular speed of the rotational
(a) Photograph of the rotor and scraper (looking from the bottom)
scraping blade
R
spring
ω
l=100 mm
α=42.7
α
Fn
Ft
F1=225 g (2.205 N)
Ft=152 g (1.49N)
Fn=165.4 g (1.621 N)
(b) Schematic diagram of the scraping blades and rotator (looking from the top)
Figure 4. The scraping blades and rotor: (a) Photograph of the rotor and
scraper (looking from the bottom); (b) Schematic diagram of the scraping
blades and rotor (looking from the top). This figure is available in colour
online at www.apjChemEng.com.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
621
FRANK G. F. QIN ET AL.
Asia-Pacific Journal of Chemical Engineering
The blade tension (F1 ) against the wall, which was
produced by a pair of spring coils, was measured
to be F1 = 225 g in the experimental SSHE (Fig. 4).
Considering the action angle of the blade and the wall,
the tension in the normal direction is Fn = F1 cos α;
in the tangential direction it is Ft = F1 sin α. Thus, the
friction that retards the scraping action is:
Ff = εbFn = 2εF1 cos α
(8)
where ε is the coefficient of friction on the cooling
surface and α is the angle between the blade and the
tangent line. Thus, Mf can be defined as:
Mf = 2εRF1 cos α
(9)
RESULTS AND DISCUSSION
Base torque of the scraper – before freezing
As mentioned previously, the shaft torque, which drives
the scraper before freezing in the SSHE, is the base
torque, and is a combination of the wall-blade friction
and the fluid resistance. The contribution of the friction
was determined using an empty SSHE in both dry and
wet conditions.
In the wet condition, a moist inside surface was
produced by filling the SSHE with water and emptying
it. The scraper was set at the desired speed, and once
the coolant had reached the set temperature of 0 ◦ C, the
shaft torque was recorded for 10 min. The experimental
result is shown in Fig. 5(a).
In the dry condition, the cooling surface was not
wetted and experiments were run as described for the
wet condition with a coolant temperature of 0 ◦ C. Shaft
torque variation for the dry condition is shown in
Fig. 5(b).
In the filled condition, the working fluid was expected
to add additional resistance to the movement of the
scraper, resulting in higher shaft torques. The torque for
overcoming the fluid resistance is the torque difference
of the filled condition and empty wet condition: Mr =
Mb − Mf . However Mb and Mf are so close in the current
SSHE that Mr is hardly measurable. This is shown
in Fig. 5(c) and (d), meaning that before freezing the
resistance of the fluid is negligible in the current scale
SSHE.
These results showed that the base torque seems to
be reasonably independent of the rotational speed of the
scraper, the fluid temperature and ethanol concentration,
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Base Torque (mNm)
(7)
140
120
100
80
60
40
20
0
1550
1600
1650
1700
1750
Time (s)
Base Torque (mNm)
Mb = Mf + Mr
(a) Plot of Base Torque vs Time at 200rpm for
Dry Condition
(b) Plot of Base Torque vs Time at 210rpm for Wet
Condition
100
80
60
40
20
0
1790
1840
1890
1940
1990
2040
Time (s)
(c) Plot of Base Torque vs Time at 200rpm for
Filled SSHE
Base Torque (mNm)
torque that overcomes the retardance of the working
fluid (Mr ). The base torque can be expressed as:
70
60
50
40
30
20
10
0
1525
1575
1625
1675
1725
Time (s)
(d) Average shafttorque vs rotational speed for dry,
wet and full SSHE
100
Base Torque (mNm)
622
Dry Condition
Wet Condition
Full SSHE
80
60
40
20
0
70
90 110 130 150 170 190 210 230 250 270 290 310
Rotational Speed (rpm)
The base torque in dry (a), wet (b) and filled
conditions of the SSHE (c) and the average torque (d) for
different rotational speeds. This figure is available in colour
online at www.apjChemEng.com.
Figure 5.
implying that the influence of flow resistance on the
torque seems to be insignificant before freezing.
The reason can be discussed as follows: The power
consumption is the product of the angular speed (ω =
2π n/60) and the shaft torque (M ):
P = ωM
(10)
Considering that the power consumption caused by
the flow resistance was the same as that due to the
propeller agitator, Trommelen et al . proposed a formula
that correlates the power number (NP ) with the stirring
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
SHAFT TORQUE CHANGE IN A SCRAPED SURFACE HEAT EXCHANGER
Reynolds number (Re) and blade number (b) of the
scraper (Trommelen and Boerema, 1966; Leung, 1967;
Trommelen and Beek, 1971a):
NP
= K Re −a b 0.59
LB
(11)
where LB is the blade length, b is the blade number and
a and K are correlation constants. NP is a dimensionless
number, i.e. power number:
NP =
P
ρn 3 dt 5
(12)
Therefore the shaft torque (M ) can be expressed as a
function of the stirring Reynolds number (Re).
When the rotational speed (n) of the scraper is varied
from 100 to 300 rpm, the stirring Reynolds number (Re)
in this study was about 103 –104 , which happened to be
in the range in between the transitional zone and the
fully developed turbulent zone. In this range the power
number tended to be a constant, i.e. Np was no longer a
function of the Reynolds number (McCabe et al ., 2001).
Therefore, the shaft torque became stable as well in this
range.
The torque for overcoming friction is the torque
measured in the empty, wet SSHE condition, which
approximates to 25 mN m. Thus the coefficient of
friction in the wet SSHE can be obtained using Eqn (9),
and is estimated as εw ≈ 0.1. The torque in the dry
condition was 65 mN m. As such, the coefficient of
friction in the dry SSHE was εd ≈ 0.28.
Influence of the ice content – after freezing
After freezing starts, however, the fluidity of the working fluid declines as the ice fraction increases in the
SSHE, resulting in a subsequent increase in the shaft
torque (Qin et al ., 2005). Meanwhile, the freezing
point depresses as the ethanol concentration increases
(Fig. 6). The ethanol concentration can be determined
by the freezing point in terms of Eqn (2), or by sampling
the solution every 5 min and calculating the ethanol
concentration (C ) using the refractive index. The ice
content is determined using Eqn (3).
Ice can be retained in the SSHE or removed from the
SSHE by using an external loop, in which the peristaltic
pump is switched on to create a circulation between the
SSHE and a 3-l cistern, as shown in the dotted region in
Fig. 1. The ice is continuously transmitted by the flow
current into the cistern and is kept there, but the ice-free
solution is fed back to the SSHE to maintain a low ice
content in it.
The results presented in Fig. 6 show a step increase
in the temperatures at the point where freezing starts.
However, the scraper torque at this time did not have
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
a step increase. In contrast, it reached its lowest value
at the onset of freezing and started to increase only as
the ice content increased further. Moreover, the torque
varied almost synchronously with the ice content. Once
the peristaltic pump was switched on, the torque started
to stabilise and even declined a little. At this stage,
the cooling surface temperature remained constant as
well, indicating that the fouling layer of ice on the
scraped cooling surface did not develop further. This
phenomenon was found to exist in the range of 100
to 400 rpm scraping speed and −3 to −10 ◦ C coolant
temperature setting.
These experiments raised the question of why the
shaft torque did not jump at the onset of freezing even
though the ice content and temperatures all underwent
a step increase. This may be explained in two aspects.
First, the new phase of ice appears as spreading thin
films (or patches) on the cooling surface rather than
suspension particles (Qin et al ., 2003a,b,c, 2004). The
thin film acts as a lubricating medium between the
scraping blade and cooling surface because of its pressand-melt nature (similar to skating on ice). Secondly,
the amount of ice formed at the onset of freezing is
not enough to significantly impact the fluidity of the
ice slurry. The mass of ice produced by nucleation
is determined by the degree of supercooling of the
solution, in which the sensible heat of the solution
equals the latent heat of freezing.
mo cp (Tf − T ) = mice H
(13)
where mo is the mass of the solution in the SSHE and
cp is the specific heat capacity of the solution, which is
approximately 3.98 kJ/kg ◦ C for 5 wt% ethanol aqueous
solution (Lide, 1995–1996). Tf and T are the freezing
point of the solution and the actual temperature, respectively, mice is the mass of ice formed by nucleation and
H is the latent heat of freezing of water, which is
334 kJ/kg. If the degree of supercooling of the ethanol
solution (Tf − T ) is 2.5 ◦ C, then the ice fraction produced by nucleation is
cps (Tf − T )
3.98 × 2.5
mi
=
≈ 0.03
=
ms
H
334
(14)
This means that after nucleation, the ice slurry
contains only about 3% ice. This small amount of ice
would perhaps be mostly in the form of thin films
adhering on the cooling surface and the small quantity
itself would not significantly change the fluidity of the
working fluid in the SSHE even though they are in the
suspension.
Experimental observation of the ice–liquid mixture
showed that the shape of ice particles newly scraped
off from the cooling surface had an irregular fragmental
appearance, as shown in Fig. 7(a). The microscopic
picture was taken with a bio-microscope in the closed
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
623
FRANK G. F. QIN ET AL.
Asia-Pacific Journal of Chemical Engineering
80
20
4
15
Temperature (°C)
10
40
2
20
0
0
Turn on the pump
−20
−2
−40
5
−60
−4
−80
Ice content
−6
0
0
200
400
600
Cooling Surface Temp
800
1000 1200
Time (s)
1400
Bulk InletTemp
Jacket Temp
Torque (mNm), Ice Content (%)
60
Torque
1600
−100
2000
1800
Torque
Ice Content
170
150
130
110
90
70
50
30
10
20
4
Torque
10
2
Temperature (°C)
Ice Content (wt %)
15
0
Torque (mNmI)
(a) 5% Ethanol solution at 100 rpm scraping speed
Turn on the pump
−2
5
−4
Ice content
0
−6
0
500
1000
Cooling Surface Temp
1500
2000
Time (s)
Jacket Temp
2500
Bulk Inlet Temp
3000
Torque
3500
Ice Content
20
4
Torque
15
Temperature (°C)
10
2
0
Turn on the pump
−2
5
−4
Ice content
0
−6
0
500
1000
1500
2000
2500
3000
170
150
130
110
90
70
50
30
10
−10
−30
−50
−70
−90
−110
−130
−150
−170
−190
3500
Torque (mNm)
(b) 5% Ethanol solution at 200 rpm scraping speed
Ice Content (%)
624
Time (s)
Cooling Surface Temp
Jacket Temp
Bulk Inlet Temp
Torque
Ice Content
(c) 5wt% ethanol solution at 250 rpm
Figure 6. Torque increase with ice content synchronously during the freezing period:
5% ethanol solution at (a) 100 rpm; (b) 200 rpm; (c) 250 rpm scraping speed. This
figure is available in colour online at www.apjChemEng.com.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
SHAFT TORQUE CHANGE IN A SCRAPED SURFACE HEAT EXCHANGER
somewhat like a brake shoe. This was the reason that
the driving shaft finally froze. Of course, the time to
freeze the blades was also affected by the shaft torque
of the driving motor.
Influence of the wall temperature – after
freezing
(a) Ice scraped off from the cooling surface
(b) Ice particles after ripening
Figure 7. Micrographs of ice produced by the SSHE: (a) Ice
scraped from the cooling surface; (b) Ice particles after
ripening.
external loop (not shown in the schematic diagram of
Fig. 1). Then the ice crystal magma started to ripen
and gradually became larger in crystal size (Smith and
Schwartzberg, 1985). Fusion and agglomeration of ice
may take place simultaneously, which may be helpful
to produce larger ice particles as shown in Fig. 7(b).
Small ice particles were not dispersed individually
in the solution. In contrast, they tended to adhere
together to form flocculates. When the ice content in
the slurry reached 15 wt% or more, all the separate
clusters of ice flocculates started to join and the working
fluid became a ‘uniform’ magma under agitation. The
fluidity/pumpability of the ice slurry rapidly declined
as the ice content increased further. At this time, the
ice magma was more like a rotational single entity,
a column of porous ice cake sliding over the cooling
surface, in which the convection and eddy currents
tended to disappear. The adhesive force of ice on
the cooling surface was getting stronger and stronger,
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
The wall supercooling is the driving force for ice
production. However, the degree of supercooling in the
bulk is extremely small, e.g. <0.2 ◦ C for most dilute
aqueous solutions (Huige, 1972; Omran and King,
1974; Stocking and King, 1976; Swinkels, 1985; Hartel
and Chung, 1993; Hartel and Espinel, 1993; Liang
et al ., 1999). In other words, ice is mainly produced
on the cooling surface or in the surface boundary layer.
This raised another question of whether a greater wall
supercooling degree would result in a higher shaft
torque of the scraper.
In studying the effect of wall supercooling on the
shaft torque, water and a 10 wt% sucrose solution were
used as the working fluids. In order to exclude the
influence of ice accumulation in the SSHE, the external
loop was still used to withdraw ice from the SSHE
to the cistern, in which a mixture of ice blocks and
the working fluid (water or aqueous sucrose solution)
was prepared beforehand to about 1 : 1 weight ratio.
The ice blocks were pre-produced with an icemaker in
about 10 g each. The existing ice blocks kept the ice
solution in thermal equilibrium, so as to maintain the
working fluid at the freezing point. The torque gain in
addition to the base torque (Mb ) should be attributed
to ice production (or the physical existence of ice) on
the cooling surface. In contrast, if no ice is adhering on
the scraped cooling surface to form the fouling layer,
the cooling surface temperature (Ts1 ) would be constant
and very close to 0 ◦ C.
The reason why pure water was used in this run
was that the ice formation would not result in freezing
point depression of the working fluid. The average wall
temperature (Tw ) could be controlled by pre-setting
the coolant temperature (Tc ), as shown by Eqn (12).
The wall temperature determined the ice formation
rate on the scraped cooling surface (Omran and King,
1974; Flesland, 1995; Qin et al ., 2003a,b). The wall
temperature of the jacket was defined as:
Ts1 + Ts2
(15)
2
where Ts1 is the temperature at the scraped cooling
surface and Ts2 is the surface temperature on the other
side of the wall (Fig. 2).
Similarly, the coolant temperature is the average of
the coolant inlet and outlet of the SSHE.
Tc1 + Tc2
(16)
Tc =
2
Tw =
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
625
626
FRANK G. F. QIN ET AL.
Asia-Pacific Journal of Chemical Engineering
Table 1 outlines the experiments that were carried out
at scraper speeds of 100, 150 and 200 rpm. The results
obtained and the concluding discussions will follow
from this outline.
In the experiment of freezing water, the coolant
temperature setting varied from −4 to −10 ◦ C. Some
selected trials are shown in Fig. 8(a)–(c). It was
expected that the ice production rate would increase
as the wall temperature decreased, provided the scraping speed was kept constant. However, the shaft torque
behaved in a similar manner for all degrees of supercooling, although the rate at which torque increased was
found to increase with the degree of supercooling. The
scraped cooling surface temperature was also found to
decline with freezing time, and it decreased faster with
lower temperature settings. Unbolting the cover of the
SSHE at the end of the runs and checking the inside
wall revealed that a firm ice layer with a smooth surface was adhering on the scraped cooling surface. The
ice scale was hard enough to resist the scraping action
of the blades, which were running at speeds of up to
200 rpm. The ice eventually pushed the spring-forced
blades back, causing them to stop rotating. Moreover,
there were no small ice particles coming out of the
SSHE in the cistern.
The temperature decline can be explained by ice
fouling on the cooling surface despite the continuous
scraping on the surface. It was assumed initially that
Table 1.
ments.
Different supercooling: outline of experi-
Scraper
rotational speed
100 rpm
150 rpm
200 rpm
Coolant
setting (◦ C)
−3.0
−4.0
−5.0
−5.5
−6.0
−6.5
−7.0
−7.5
−10.0
−3.0
−4.0
−5.0
−6.0
−7.0
−8.0
−10.0
−3.0
−4.0
−5.0
−6.0
−7.0
−8.0
−8.5
−10.0
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
under the scraping action, a fouling layer of ice would
not exist or would not develop significantly. This
assumption was proven wrong for the freezing of water.
One may also notice again that at the onset time of
freezing the torque decreased from the original base
torque to a lower value, which was about 10 mN m in
all the cases, despite the sudden appearance of ice on
the scraped cooling surface. This phenomenon became
more obvious when the coolant temperature setting was
lower than −4 ◦ C. The mechanism of this was discussed
earlier.
Using the sucrose aqueous solution (10 wt% in concentration) as the working fluid, in which the coolant
temperature setting varied from −3 to −10 ◦ C, showed
a significant change in the SSHE performance:
• The shaft torque increased gradually from the base
torque (∼20 mN m) after freezing of the solution,
but stabilised at 150 ± 60 mN m. A lower wall
temperature did not result in a greater shaft torque
at the end, but a high rate of torque increase;
• Ice particles were found coming out from the SSHE
into the cistern;
• The cooling wall and bulk temperatures both
remained constant after freezing started.
This phenomenon indicates that rather than forming
a fouling ice layer on the cooling surface, the ice was
scraped off continuously. One selected experiment in
this group is shown in Fig. 9.
Isikawa et al . also studied the scraping force needed
to remove a mushy layer of ice produced on a cooling
surface (Ishikawa et al ., 2002). It was found that the
force increased with the thickness of the ice layer.
A longer time interval of scraping (300–1000 s) and
a lower temperature would result in a greater force
needed. The time interval of scraping in this study was
much shorter (e.g.<1 s with 100–400 rpm rotational
speed). In such a short time the layer of ice could not
grow thick enough to generate sensible resistance to
the blades. In contrast, the thin layer of ice may instead
lubricate the movement of the blades.
Influence of the rotational speed
The SSHE performance during the freezing operation
was further studied by analysing the results obtained
for each coolant temperature setting at different speeds.
As shown earlier, the torque varied in a similar manner
for each coolant setting at a certain speed. For each
speed, the torque before freezing remained at values
close to the base torque, which was about 20 mN m in
the current SSHE.
Eventually, the movement of the blades stopped
(frozen), whereby the final torque reached average
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
150
130
110
90
70
50
30
10
Torque (mNm)
Temperature (°C)
(a)
n = 200rpm
Tset = –4.0 °C
270
250
230
210
190
170
150
130
110
90
70
50
30
10
0
Torque (mNm)
SHAFT TORQUE CHANGE IN A SCRAPED SURFACE HEAT EXCHANGER
310
290
270
250
230
210
190
170
150
130
110
90
70
50
30
10
0
Torque (mNm)
Asia-Pacific Journal of Chemical Engineering
0
0
−1
−2
−3
−3.5
500
1000
1500
2000
Time (s)
(a) Temperature and Torque vs Time for –4.0°C Water Stirred at 200 rpm
2500
3000
Temperature (°C)
(b)
n = 200 rpm
Tset = –7 °C
0
0
−1
−2
−3
200
400
600
800
Time (s)
(b) Temperature and Torque vs Time for –7.0°C Water Stirred at 200 rpm
1000
second
Temperature (°C)
(c)
n = 200 rpm
Tset = –10 °C
0
0
−1
−2
−3
100
200
300
400
Time (s)
500
600
700
(c) Temperature and Torque vs Time for –10°C Water Stirred at 200 rpm
Shaft torque;
Cooling surface temperature;
Jacket temperature;
Bulk temperature
Figure 8. Variation of shaft torque, cooling surface temperature and bulk temperature
with the freezing time in different coolant temperature settings (Tset ) for freezing
water: (a) −4.0 ◦ C, (b) −7.0 ◦ C (c) −10.0 ◦ C. This figure is available in colour online at
www.apjChemEng.com.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
627
FRANK G. F. QIN ET AL.
n = 220 rpm
Tcts = –4.0 °C
300
1800
200
1600
100
0
−100
−200
0
−300
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time Elapsed (s)
6
5
4
3
2
1
0
−1
−2
−3
−4
−5
Asia-Pacific Journal of Chemical Engineering
Torque (mNm)
Temperature (°c)
200rpm
150rpm
1400
100rpm
1200
1000
800
600
400
200
0
−12
Time (s)
−10
−8
Figure 9. Coolant temperature setting: −4.0 C. This figure
is available in colour online at www.apjChemEng.com.
values of 103, 160 and 218 mN m for speeds of 100, 150
and 200 rpm, respectively. This is shown in Fig. 10.
The final torque, at which the blades stopped because
of freezing, increased with the rotational speed as
shown in Fig. 11. This was because, first, in the
current apparatus the shaft torque provided by the motor
increased with the rotational speed, so it was able to
overcome a greater resistance and therefore operated
for a longer time before being frozen; second, a higher
rotational speed gave more scraping actions per unit
time, and hence is more effective to inhibit the ice
layer growth. These results are shown in Fig. 11, where
each marked node in the lines represents an individual
experiment with fixed rotational speed and coolant
temperature setting. For lower coolant temperature
setting, the time elapsed before freezing of the blades
became shorter.
CONCLUSIONS
The base torque of driving the SSHE is the measurement
of the combination of blade-wall friction and the flow
resistance of the working fluid. The coefficient of
friction between a polyethylene blade and the machined
−6
−4
−2
0
Coolant Setting (°C)
◦
Final Torque Reading (mNm)
628
Figure 11. Time elapse after the onset of freezing vs the
coolant temperature setting. This figure is available in colour
online at www.apjChemEng.com.
stainless steel wall was found to be 0.1 for the wet
surfaces and 0.28 for the dry surfaces in the current
apparatus. The shaft torque due to the flow resistance
was similar to that of the ordinary propeller agitator, and
varied with the Reynolds number in the laminar flow
regime, but was insensitive to Reynolds number change
in the turbulent flow regime. As such, in the turbulent
flow regime, the torque requirement became stable and
independent of the rotational speed of the scraper. In the
current SSHE, which uses a set of spring-forced straight
blades as the scraper, ice slurries can be produced with
various dilute aqueous solutions, provided the FPD of
the solution is larger than 0.5 ◦ C. However, water, or
very diluted solutions (e.g. FPD <0.3 ◦ C) will tend to
form a tough ice scaling on the scraped cooling surface,
which significantly reduces the heat transfer efficiency.
Ice slurry cannot be produced using such solutions.
Approximately 20–30 s after the onset of freezing (ice
nucleation), the shaft torque reached a minimum value,
and then started to increase slowly. This might be due
to the ice film formation on the cooling surface when
freezing started, which worked as a lubricating layer. As
the ice content increased in the SSHE, the fluidity of
250
200
150
100
50
0
0
50
100
150
200
250
Scraper Rotational Speed (rpm)
Figure 10. Final torque reading vs scraper rotational speed. This figure is
available in colour online at www.apjChemEng.com.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
SHAFT TORQUE CHANGE IN A SCRAPED SURFACE HEAT EXCHANGER
the ice slurry declined, and therefore the shaft torque
increased. The initial shaft torques are all the same
at the onset time of freezing even when a lower wall
temperature is used (from 0 to −10 ◦ C in this study).
But the wall temperature does affect the rate of torque
increase. Wall supercooling is the driving force of ice
crystallization. It was found that scraping was effective
in inhibiting ice fouling for the aqueous solutions tested,
but not for pure water. Consequently, if the ice is
removed from the SSHE to effectively maintain a low
content of ice, a lower wall temperature will not result
in a greater power consumption to drive the scraper.
The rotational speed of the scraper is also important.
A higher rotational scraping speed is helpful to inhibit
the development of ice fouling layer. Of course, this
would have to be balanced with the economy of power
consumption. Nevertheless, this rule of thumb may not
be applicable to water (or very dilute solutions with
FPD lower than 0.3 ◦ C) in the current experimental
SSHE, in which a pair of spring-forced straight blades
is used. Also, the high scraping speed may result in
heating effects, which would be undesirable.
NOMENCLATURE
a
b
Co
C
dt
F
F1
Fn
Ft
mice
mo
M
Mb
Mf
Mr
n
P
r
R
T
Tf
Tw
Ts1
Ts2
Tc
constant
blade number of the scraper
original weight concentration of the
working
weight concentration
diameter of the scraper
driving force that moves the scraping
blade
spring force applied on the cooling
surface
normal direction fraction of the
applied spring force
tangent fraction of the applied spring
force
mass of ice
mass of working fluid
torque
base torque
torque cause by surface friction
torque caused by flow resistance
rotational speed of the scraper
power consumption
correlation coefficient
inside radii of the SSHE
bulk temperature
freezing point
wall temperature
cooling surface temperature
the wall surface temperature on
coolant side
coolant temperature
wt %
wt %
m
N
N
N
N
kg
kg
Nm
Nm
Nm
Nm
rpm
W
m
K
K
K
K
K
K
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Tc1
Tc2
Tset
H
α
ε
ρ
ω
inlet temperature of the coolant
outlet temperature of the coolant
the coolant temperature of setting
latent heat of fusion of water
angle between the blade and the
tangent line of the cooling surface
coefficient of friction
mass fraction of ice
density of the working fluid
angular speed of the scraper
K
K
K
◦
kg m−3
arc
REFERENCES
Ayel V, Lottin O, Peerhossaini H. Rheology, flow behaviour and
heat transfer of ice slurries: a review of the state of the art. Int. J.
Refrig. 2003; 26: 95–107.
Bel O, Lallemand A. Study of a two phase secondary refrigerant 1:
intrinsic thermophysical properties of an ice slurry. Int. J. Refrig.
1999a; 22: 164–174.
Bel O, Lallemand A. Study of a two-phase secondary refrigerant 2:
experimental analysis of thermal and rheological behaviour. Int.
J. Refrig. 1999b; 22: 175–187.
Ben Lakhdar M, Cerecero R, Alvarez G, Guilpart J, Flick D,
Lallemand A. Heat transfer with freezing in a scraped surface
heat exchanger. Appl. Therm. Eng. 2005; 25: 45–60.
Dass JD. Current large-scale commercial application of freeze
concentration in the food industry. European Food and Drink
Review. Spring: 1991; 19–24.
Flesland O. Freeze concentration by layer crystallisation. Drying
Technology 1995; 13: 1713–1739.
Habib BA, Farid MM. Fluidised bed freeze concentration. In 9th
Asian Pacific Confederation of Chemical Engineering (APCChE)
and 30th Annual Australasian Chemical Engineering Conference
(CHEMECA). Christchurch, New Zealand (29 September –
3 October 2002).
Hartel RW, Chung MS. Contact nucleation of in fluid diry products.
J. Food Eng. 1993; 18: 281–296.
Hartel RW, Espinel LA. reeze concentration of skim milk. J. Food
Eng. 1993; 20: 101–120.
Huige NJJ. Nucleation and growth of ice crystals from water and
sugar solutions in continuous stirred tank crystallizers, PhD
Dissertation Thesis, University of Eindhoven, The Netherlands,
1972.
Inada T, Yabe A, Grandum S, Saito T. Control of molecular-level
ice crystallization using antifreeze protein and silane coupling
agent. Mater. Sci. Eng., A 2000; 292: 149–154.
Ishikawa M, Hirata T, Fujii T. Force estimation of mechanical
removing processes of mushy structure in an aqueous solution.
Int. J. Refrig. 2002; 25: 208–217.
Ismail KAR, Radwan MM. Modeling of ice crystal growth in
laminar falling films for the production of pumpable ice slurries.
Energy Conversion Management 2003; 44: 65–84.
Kitanovski A, Poredos A. Concentration distribution and viscosity
of ice-slurry in heterogeneous flow. Int. J. Refrig. 2002; 25:
827–835.
Knodela BD, Franceb DM, Choic US, Wambsganss MW. Heat
transfer and pressure drop in ice-water slurries. Appl. Therm. Eng.
2000; 20: 671–685.
Leung LS. Power consumption in a scraped-surface heat exchanger – comments by L.S. Leung. Trans. Inst. Chem. Eng. 1967;
45: T179–T181.
Liang B, Shi Y, Hartel RW. Freeze concentrating of aqueous
solutions with high efficiency. In IFT Annual Meeting, Chicago,
1999.
Lide DR (ed.). CRC Handbook of Chemistry and Physics (76th edn).
CRC Press: Cleveland, OH, 1995–1996.
McCabe WL, Smith JC, Harriott P. Unit Operations of Chemical
Engineering. McGraw-Hill, Inc.: Boston, MA, 2001.
Meewisse JW, Ferreira CAI. Validation of the use of heat transfer
models in liquid/solid fluidized beds for ice slurry generation. Int.
J. Heat Mass Transfer 2003; 46: 3683–3695.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
629
630
FRANK G. F. QIN ET AL.
Omran AM, King CJ. Kinetics of ice crystallization in sugar solution
and fruit juices. AIChE J. 1974; 20: 795–802.
Paul Mueller Company (U.S.). Avalanche Ice Harvest system. On
internet website: http://www.muel.com/products/thermalstorage
/avalanche.cfm thermal@muel.com (Last update 2005).
Qin FGF, Chen XD, Russell AB. Heat transfer at the subcooledscraped surface with/without phase change. AIChE J. 2003a; 49:
1947–1955.
Qin FGF, Russell AB, Chen XD, Robertson L. Ice fouling on
a subcooled metal surface examined by thermo-response and
electrical conductivity. J. Food Eng. 2003b; 59: 421–429.
Qin FGF, Zhao JC, Russell AB, Chen XD, Chen JJ, Robertson L.
Simulation and experiment of the unsteady heat transport in
the onset time of nucleation and crystallization of ice from
the subcooled solution. Int. J. Heat Mass Transfer 2003c; 46:
3221–3231.
Qin FGF, Chen XD, Farid MM. Growth kinetics of ice films
spreading on a subcooled solid surface. Sep. Purif. Technol. 2004;
39: 109–121.
Qin FGF, Chen XD, Ramachandra S, Free K. Heat transfer and
power consumption in a scraped-surface heat exchanger while
freezing aqueous solutions. Sep. Purif. Technol. 2006; 48:
150–158.
Saito A. Recent advances in research on cold thermal energy storage.
Int. J. Refrig. 2002; 25: 177–189.
Smith CE, Schwartzberg HG. Ice crystal size changes during
ripening in freeze concentration. Biotechnol. Process. 1985; 1:
111–120.
Stocking JH, King CJ. Secondary nucleation of ice in sugar solutions
and fruit juices. AIChE J. 1976; 22: 131–140.
 2007 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pacific Journal of Chemical Engineering
Swinkels WJ. Recent developments in freeze concentration. IDF
Seminar. IDF: Atlanta, GA, 1985; (in 8–9 October).
Tanino M, Kozawa Y. Ice-water two-phase flow behavior in ice heat
storage systems. Int. J. Refrig. 2001; 24: 639–651.
Trommelen AM. Heat transfer in a scraped-surface heat exchanger.
Trans. Inst. Chem. Eng. 1967; 45: T176–T178.
Trommelen AM, Boerema S. Power consumption in a scrapedsurface heat exchanger. Trans. Inst. Chem. Eng. 1966; 44:
T329–T334.
Trommelen AM, Beek WJ. Flow phenomena in a scraped-surface
heat exchanger (“Votator”-type). Chem. Eng. Sci. 1971a; 26:
1933–1942.
Trommelen AM, Beek WJ. The mechanism of power consumption
in a Votator-type scraped-surface heat exchanger. Chem. Eng. Sci.
1971b; 26: 1977–1986.
Trommelen AM, Beek WJ, Van De Westelaken HC. A mechanism
for heat transfer in a Votator-type scraped-surface heat exchanger.
Chem. Eng. Sci. 1971; 26: 1987–2001.
Tsuchida D, Kang C, Okada M, Matsumoto K, Kawagoe T. Ice
formation process by cooling water-oil emulsion with stirring in
a vessel. Int. J. Refrig. 2002; 25: 250–258.
Wiegandt HF, Madani A, Harriott P. Ice crystallization developments for the butane direct-contact process. Desalination 1987;
67: 107–126.
Yamada M, Fukusako S, Kawabe H. A quantitative evaluation of the
production performance of ice slurry by the oscillatory moving
cooled wall method. Int. J. Refrig. 2002; 25: 175–277.
Asia-Pac. J. Chem. Eng. 2007; 2: 618–630
DOI: 10.1002/apj
Документ
Категория
Без категории
Просмотров
2
Размер файла
738 Кб
Теги
exchanger, theshaft, change, torque, scraper, heat, ice, laboratory, slurries, surface, making, used
1/--страниц
Пожаловаться на содержимое документа