close

Вход

Забыли?

вход по аккаунту

?

Minimum Wetting Rates for Falling Films on Stainless Steel.

код для вставкиСкачать
Dev. Chem. Eng. Mineral Process. 14(1/2), pp. 153-162, 2006.
Minimum Wetting Rates for Falling Films
on Stainless Steel
K.R. Morison” and G. Tandon
Dept of Chemical and Process Engineering, University of Canterbury,
Private Bag 4800, Christchurch, New Zealand
The minimum wetting rate is defined as the minimum mass flowrate per unit
circumference that is required to maintain of complete falling film of liquid on a
surface. In this work minimum wetting rates in a 48 mm i d . stainless steel tube were
determined jor water, 50% sucrose solution, and reconstituted skim milk under
isothermal and heat transfer conditions. It was found that initial distribution of the
liquid at the top of the tube was critically important. The minimum wetting rates
under isothermal and heat transfer conditions ranged from 0.22 kg m-’s-lfor 50%
sucrose at 20% to 0.12 kg m-’s-’ for water az 70°C. Nearly all rates decreased with
temperature. Equations in the literature did not accurately predict the results
obtained.
Introduction
Falling film evaporators are used when the fluids are temperature-sensitive, a short
residence time is required, and a low-pressure drop is necessary. Falling film
evaporators are extensively used in the food industry for their ability to handle heat
sensitive materials, e.g. fruit and vegetable juices [ 13. The dairy industry uses falling
film evaporators for concentrating milk before spray drying to produce milk powder.
In falling film evaporators (see Figure 1) the feed enters at the top of the vessel, it
is distributed so that it flows evenly down all the tubes as a film, and the vapour and
concentrate leave from the bottom. A complete film should be maintained inside the
tubes at all times. Film breakdown will decrease the efficiency of the process and may
cause excessive fouling [2]. The thinning of the film to the point where it cannot
completely wet the tube surface is termed film breakdown. The minimum flowrate
required in order to maintain a complete film (normally designated r,,,,,,)is
determined from the minimum wetting rate which is defined as the mass flowrate per
unit perimeter of the tube. There are two types of minimurn wetting rate. One type is
for a liquid that is wetting an initially dry surface, and the other is a complete film
where the flowrate is decreased to the point of film breakdown [3]. This project was
concerned with the first type of wetting.
* Author for correspondence
(ken.morison@canterbury.ac.nz).
IS3
K.R. Morison and G. Tandon
Feed
Steam
___,
Steany
chest
Condensate
F
T
Concentrate
Figure 1. A falling film evaporator.
Approximate
streamlines
Boundary of
. liquid film
Figure 2. Film breakdown [4/.
Hartley and Murgatroyd [4)presented some of the earliest work done in the field
of film breakdown. They developed Equation (1) below for stability of a dry patch
under isothermal conditions, by applying a force balance at the top of the dry patch
(see Figure 2) for laminar isothermal flow.
where cr is surface tension; B is contact angle; p is viscosity; and p is density.
I54
Minimum Wetting Rates for Falling Films on Stainless Steel
Their work was extended by many others, including Hoke and Chen [5] who
included terms for heat transfer. They obtained Equation (2) which is solved for the
is obtained using
film thickness, &,,, from which the minimum wetting rate, TI,
Equation (3). When heat transfer occurs, then Equation (4) is solved for the film
thickness with heat transfer, &,n,HT.
where d is tube diameter; k is thermal conductivity; q is heat flux; and L is tube
length. Numerous other equations exist in the literature. Munakata et al. [6]
investigated the effect of viscosity and added another term to Equation (1) to take into
account the reaction force to induce horizontal viscous flow around the dry patch.
Very little experimental data for liquids on stainless steel has been published.
Paramalingham et al. [2] obtained minimum wetting rates of 0.222 kg m-'s-l for water
and 0.123 kg m-'s-l for 40% whole milk concentrate both at 20°C. Hobler and Czajka
[7] reported wetting rates of 0.17 rt 0.05 kg m-ls-' for water at 10 to 1 4 T , as well as
rates for various glyceroVwater mixtures. Munakata et al. [6] report values from about
0.065 kg m-'s-' to 0.1 kg m-ls-' for water, and other values for glyceroywater
mixtures. They found that wetting rates were strongly affected by the type of
distribution system used but they did not discuss the variability in their results. The
range of their values, and hence the range in the parameters for their proposed
equation, makes it difficult to draw clear conclusions from their work.
Experimental Methods
The apparatus used is shown in Figure 3, and comprises an evaporator tube 1 m long
with inner diameter 47.6 mm inside an annulus water jacket with an outer diameter of
72 mm. The evaporator tube and the other tubing in the apparatus were #304 stainless
steel. Resistance thermometers (Pt100) were attached under insulation at the outlet of
the water jacket and the inlet of the feed.
Several different distributor designs were tested. The initial designs were based on
an annular gap between a central cylinder and the tube wall, and a horizontal gap
between the distributor and the tube sheet. Other designs used flow over a plastic,
stainless steel or frosted glass annular weir. The final distributor was made from
unglazed ceramic (see Figure 4), and sat at the top of the evaporator tube in order to
distribute the feed and form a film.
K.R. Morison and G. Tandon
-
Distributor
Rotameter
preheater
Gear pump
Figure 3, Apparatus for determination of minimum wetting rate.
Figure 4. Unglazed ceramic distributor with 36 holes of 1.5 mm.
Before every run, the evaporator tube was cleaned of experimental liquid with a
sequence of 2% sodium hydroxide, a rinse, 2% nitric acid, and several f m l rinses.
For a particular temperature, the feed was heated separately in a flask on a heating
mantle and fed into the feed tank. The preheaters were adjusted to 2°C higher than the
156
Minimum Wetting Rates for Falling Films on Stainless Steel
required temperatures of the feed and the annulus water. The initial gear pump speed
was set to give a flowrate of about 0.7 Llmin, and then was slowly increased by as
little as 0.020 Wmin every 2 minutes to ensure that the feed and the tubing heated up
to the required temperatures. The speed was increased and the minimum wetting rate
was determined from a calibration equation for the pump. All of the runs reported in
this paper were conducted at atmospheric pressure with the top lid removed to allow
observation of the wetting patterns. The concentration of the 50% sugar solution was
determined using a refractometer, and the concentrations of 10% and 40% milk were
determined by oven drymg.
Contact angle was measured by photographing a sessile drop and applying the
axisymmetric drop shape analysis [ 8 ] . Surface tension was measured using the Du
Nouy ring method with a Fisher Tensiometer Model 20, and viscosity was measured
using a Haake concentric cylinder viscometer. Further details are given by Tandon
[91.
Minimum wetting rates were determined for distilled water, aqueous 50% sucrose
solution (domestic sugar), and aqueous 10% and 40% reconstituted slum milk (“lowheat”, 54.1% lactose, 33.4% protein, 7.9% minerals, 3.8% moisture and 0.8% fat,
NZMP).
Observations and Results
Several different distributor designs were tested. The first was based on an annular
gap between the evaporator tube and the distributor, but a variation of less than
0.1 mm in the gap caused poor distribution
was 0.28 kg m-ls-l for water at
20°C). Weir-type overflow distributors were made using plastic, glass and stainless
steel. However at low flowrates, surface tension held the liquid back everywhere
except a single point (rh 0.20 kg m-ls-l). The final design (see Figure 4) was wetted
very easily by the solutions and gave consistently good distribution ( r
0.14 kg m-ls-’). Liquid flowed from the outside, through the holes, and down the
inside.
(rh
Figure 5. Observedflow patterns at about 90% of the minimum wetting rate.
157
~
K.R. Morison and G. Tandon
0.23
1
0 50% sugar
40% milk
X 10% milk
A water
0.21
ul
E
?
0.19
3
0.17
g
5
0.15
.-E
--.
...
C
3
A=...
0.13
'"-A
0.11
10
20
30
40
50
60-
70
Feed Temperature, "C
Figure 6. Minimum wetting rates of all the liquids under isothermal conditions.
0.21
0'23
B
I
0.19
50% sugar
ai
Bp
0.17 -
'
0.15 -
f
0.13 -
g
9
40% mil
I
ul
10% mil
water
0.11
10
20
30
40
50
60
70
Feed Temperature, ' C
Figure 7. Minimum wetting rates with a 5°C temperature difference across the tube
I S8
Minimum Wetting Rates for Falling Films on Stainless Steel
It took approximately 20 minutes to determine the minimum wetting rate for a
particular condition. At the start of a run the flowrate was low with a wetting rate of
0.08 kg m-ls-'. At this flowrate one or more fat rivulets were seen, and when the
flowrate was increased a film formed for a few millimetres below the top of the
stainless steel evaporator tube (see Figure 5 ) . The final stage of complete wetting
often occurred over a period of about 2 minutes with no further increase in flowrate.
The film never broke up further down the tube.
Minimum wetting rates were determined without heat transfer for temperatures
from 20°C to 70°C for the test liquids (see Figure 6). The feed inlet temperature was
generally maintained within *0.4"C of the annulus heated water. Repeated readings
were obtained for all the experimental liquids used, and the repeatability was typically
*0.01 kg m-ls-'.
Minimum wetting rates were also determined under heat transfer conditions, with
a temperature difference of 5°C f 2°C between the outer annulus heating liquid and
the experimental liquid. The minimum wetting rates for distilled water decreased with
increasing feed temperature up to 50"C, after which there was an increase in
minimum wetting rate (see Figure 7). At feed temperatures above 65"C,as the
distilled water film moved down to wet the surface, the advancement was hindered by
the drying out of the film. For sucrose a slight increase in the minimum wetting rate
was seen only for a feed temperature of 65°C. Minimum wetting rates of aqueous
10% and 40% reconstituted skim milk decreased over the entire temperature range.
The initial reading of aqueous 40% reconstituted skim milk in Figure 7 is very high,
since foaming of milk was observed and the foam blocked a number of holes on one
side of the distributor.
To understand the effect of temperature difference on minimum wetting rate,
minimum wetting rates were measured for water at a constant feed temperature of
50°C with an increasing temperature difference between the water jacket and the feed.
The minimum wetting rate (see Figure 8) showed a general increase with increasing
temperature difference. The equations of Hoke and Chen [ 5 ] were then applied, using
an overall heat transfer coefficient of 320 W m-' K ' that was determined to obtain a
slope similar to the data. The experimental heat transfer coefficient was not measured
in this work, but was calculated independently to be 350 W m-' K". A contact angle
of 74" was used for the prediction.
Discussion
It was concluded that the distributor needs to have better wetting characteristics, and
hence a lower minimum wetting rate, than the evaporator tube. In assessing the
effectiveness of the distributor we relied on observations and comparisons with
previous designs. Commercial-scale evaporators have much simpler distribution
systems that might not be as effective as the system used in this work, and thus their
minimum wetting rates might be much higher than measured here.
The wetting rates we obtained were a little lower than those of Paramalingham et
al. [2] and Hobler and Czajka [7]. However our rates were up to twice as high, but
much less variable, than those of Munakata et al. [6]. Their low wetting rates suggest
159
K.R.Morison and G.Tandon
0.20 1
0.19
ln
.
E
2
0.18
0.17
.-P
P
0.13
0
5
10
15
20
Temperature Difference, "C
Figure 8. Minimum wetting rates for distilled water at 50°C for increasing different
temperature diflerences, with the prediction (line) from Equations (2) - (4).
that wetting rates lower than those presented here are possible. If the flow is steady
and the system uncontaminated, it is expected that the lowest wetting rates obtained
would be the most accurate.
The experimental minimum wetting rates were compared with those predicted by
Equations ( 2 ) and (3) under isothermal conditions. The experimental predictions are
lower than the theoretical predictions as seen in Figure 9, especially for 50% sucrose
solutions. It is not clear if the theoretical basis for Equation (2) is applicable in this
case, but there is insufficient experimental data to develop an alternative theoretical
basis. Numerical experimentation using these equations showed that the high
viscosity of 50% sucrose solutions was the main contributor to the much higher
predicted values. However, it was observed that the sucrose solution film was much
more coherent and seemed less inclined to split in order to move around a dry patch.
This is consistent with the approach taken by Munakata et al. [6]who found that high
viscosity fluids had lower minimum wetting rates.
Equations (2) to (4) were successfully used to show the effect of temperature
difference on wetting rate as given in Figure 8. These results indicate that the main
cause of the temperature difference effect was the Marangoni effect, that was included
in the analysis used to obtain these equations. The Marangoni effect arises from the
surface tension gradient caused by temperature differences at the surface.
I60
Minimum Wetting Rates for Falling Films on Stainless Steel
0.33 0 50% sugar
0.31 ul
$,
-
x
0 40% milk
A water
0.29 0.27 0.25 -
0.21
0.17
0.13
0.11 7
10
20
30
40
50
60
70
Feed Temperature, 'C
Figure 9. Experimental minimum wetting rates (symbols), with predictions (lines)
from Equations (2) and (3) under isothermal conditions.
Conclusions
The apparatus was successfully used to measure minimum wetting rates for distilled
water, 50% aqueous sugar solution, and 10% and 40% aqueous reconstituted skim
milk. A considerable amount of time was spent in achieving a good distribution
system, and it was shown that a good distribution system gives lower wetting rates.
Good distribution was achieved when a very wettable material (unglazed ceramic)
was used. Increases in temperature difference increased the minimum wetting rates
for water, thus following the theoretical predictions.
References
1. Maroulis, Z.B.,
and Saravacos, G.D. 2003. Food Process Design, Marcel Dekker Inc., New York.
2. Paramalingam, S., Winchester, J., and Marsh, C. 2000. On the fouling of falling film evaporators due
to film break-up. Trnns IChemE Pnrt C,Food Bioprod. Proc., 78(C2) 79-84.
3. Watanabe, K., Munakata, T I and Matsuda, A. 1975. Minimum wetting rate on wetted-wall column in
the absence of mass and heat transfer, J. Chem. Eng. Japan, 8,75-77.
4. Hartley, D.E., and Murgatroyd, W. 1964. Criteria for the break-up of thin liquid layers flowing
isothermally over solid surfaces, Int. J. Hent Moss Transfer,7 , 1003-1015.
5 . Hoke, B.C., and Chen, J.C. 1992. Thermo capillary breakdown of falling film liquid films, Ind. Eng.
Chem. Res., 31,668-694.
6. Munakata, T., Watanabe, K., and Miyashita, K. 1975. Minimum wetting rate on wetted-wall column correlation over wide range of liquid viscosity, J. Chem. Eng. Jnpnn, 8,440-444.
K.R. Morison and G.Tandon
7 . Hobler, T., and Czajka, J. 1968. Minimum wetting of a flat surface (in Polish), Chemio Sfosowana, 2B,
169-186.
8. Lahooti, S., Del Rio, O.I., Neumann, A.W., and Cheng, P. 1996. Axisymmetric drop shape analysis,
Applied Sufloce Ihermodynamics, Eds: Neumann, A.W., and Spelt, J.K., Surfactant Science Series, 63,
pp.441-465, Marcel Dekker Inc., New York.
9. Tandon, G. 2004. Experimental Determinafion of Minimum Wetting Rates in Falling Film Evaporator,
M. Eng. Thesis, Department of Chemical and Process Engineering, University or Canterbury,
Christchurch, New Zealand.
I62
Документ
Категория
Без категории
Просмотров
0
Размер файла
515 Кб
Теги
steel, films, minimum, wetting, fallin, rate, stainless
1/--страниц
Пожаловаться на содержимое документа