[American Institute of Aeronautics and Astronautics 10th AIAA/ASME Joint Thermophysics and Heat...
Transcript of [American Institute of Aeronautics and Astronautics 10th AIAA/ASME Joint Thermophysics and Heat...
`
Experimental study of natural frost formation on a
horizontal cylinder
M. Yaghoubi* H. Khoshnazar
Engineering School, Shiraz University
Shiraz, Iran
Abstract
In this study frost formation on a horizontal cylinder at free convection condition has been
experimentally investigated. The cold surface temperature of cylinder varied from -4 to -8 Co , the
humidity ratio of ambient is changed from 6.4 to 8.9 DAkgg and ambient temperature controlled
between 15 and 22 Co. The thickness of frost over cylinder and heat flux from the cold surface is
measured for average of eight hours. Smoke test is carried to illustrate cold plume how around cylinder
with the present of frost on the cylinder. Results showed that frost on the top surface of the horizontal
cylinder layer is thicker than the bottom layer. Heat flux at the beginning of frost formation is high and
then reduced to a constant value. Measured quantities were compared with results given by a simple
model based on similarity between heat and mass transfer and a correlation is developed to predict frost
thickness with respect to time for all the ranges of measurement.
Keywords: Frost deposition, Free convection, Horizontal cylinder.
10th AIAA/ASME Joint Thermophysics and Heat Transfer Conference28 June - 1 July 2010, Chicago, Illinois
AIAA 2010-4659
Copyright © 2010 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
Nomenclature
Specific heat capacity( 11 KkgJ ) Cp
Cylinder diameter ( m) D
Mass diffusivity of water vapor in air ( 12Sm ) vD
Fourier number ( 2tDDv ) Fo
Heat transfer coefficient( 12 KmW ) h
Mass transfer coefficient( 2mskg ) mh
Conductivity( 11 msW ) k
Lewis number Le
Mass flux ( 21 mskg ) .
m
Nusselt number Nu
Pressure( pa ) P
Heat flux(2mW ) 'q
Total energy transfered(W ) Q
Cylinder radius, ( m ) r
Rayleigh number Ra
Relative humidity( % ) RH
Frost thickness ( m ) fS
Dimensionless frost thickness ( DS f ) *
fS
Time ( s ) t
Temperature ( K ) T
Dimensionless temperature *T
Absolute humidity(1
DAkgkg ) w
Greeks
Density( 3mkg )
Estephan- Boltzmann( 428107.5 KmW )
Subscript
Air a
Average av
Conduction cond
Convection conv
Diffuse diff
Effective ef
frost surface fs
Latent lat
Water triple point tp
Wall w
1. Introduction
Frost begins to form when humid air comes into contact with a cold surface that is
kept below water freezing temperature. This process affects heat transfer rate
particularly in air-conditioners, refrigerators and heat pumps. In most cases, frost
formation is undesirable because it contributes to the heat transfer resistance and
pressure drop. Analyzing frost formation is not easy because it is a complicated
transient phenomenon in which a variety of heat and mass transfer mechanism occur
simultaneously. Recent articles [Ostine and Andersoon, 1991; lee and Ro, 2002; lee
and Ro, 2001; Cheng and Wu, 2003; Na and Webb, 2004; Salmanopur et al. 2008]
shows that frost formation during forced convection of humid air has been studied
extensively, while on the other hand, only a limited number of investigations deal
with heat-mass transfer during natural convection on a surface at subfreezing
temperatures. Barron and Han (1965) studied heat and mass transfer to a cryosurface
in free convection. Kennedy and Goodman (1974), studied frost formation on a
vertical surface in free convection condition. They found that humidity effect on frost
thickness growth is very small. Cremers and Mehra(1982) studied frost formation on
vertical cylinders in free convection and gave correlation for frost thickness. Transient
natural convection heat and mass transfer during refrigeration of air by horizontal
tubes is studied by Riedy (1983). They showed that the heat transfer coefficient
increases at the begging and reduces to a constant value. Chen and Oosthuizen (1994)
studied frost formation on a horizontal tube and shown that the effect of cylinder
surface temperature is more than ambient temperature and humidity on the frost
thickness. Fossa and Tanda (2002) studied free convection frost formation on a
vertical plate. Wu et.al(2007) studied early stage of frost formation on a horizontal
plate and showed that the surface temperature is the primary factor affecting the frost
crystal shape and relative humidity has less effect on the frost crystal shape. To the
authors' knowledge, there are few experimental study deals with natural convection
heat and mass transfer around horizontal cylinders. The present paper reports the
results of an experimental study of frost formation on a horizontal cylinder in natural
convection. The aim of this study is to find experimentally and theoretically the
effects of various environmental parameters on natural frost formation growth rate
and heat transfer rate during transient processing of frost formation on a horizontal
cylinder.
2. The apparatus and instrumentation
Schematic view of the experimental apparatus is shown in Fig. 1. The test section and
the measuring instruments were placed in a test room which could be regulated and
controlled in the range of desired values of air condition. The test room ambient
temperature is controlled by refrigeration cycle which is design and built in
laboratory. The secondary refrigerant (60/40 ethylene-glycol/water solution) is used
for control and keep the test section surface temperature at constant value. A cross
section of the test section is depicted in detail in Fig. 2. The test section lengths and
diameter is 300 mm and 80 mm. The outer and inner tube was constructed of
aluminum and between these two layers filled with an insulating material to provide a
measurable heat flux. The outer and inner diameter of cylinder was 80 and 17 mm.
The tube was cooled by internal circulation of the secondary refrigeration. Small holes
of 1.2 mm in diameter were drilled at four circumferential positions at 0, 90, 180,
and 270o
in the inner and outer tube to measure radial temperature distribution in the
cylinder. Two locations in longitudinal direction is chosen to calculate average heat
flux and minimized the error. We used total of 16 thermocouples to measure
temperature distribution in the hollow cylinder. Copper- constant thermocouples were
inserted in each hole. The thermocouples first calibrated at the required temperature
range with an uncertainty of Co1.0 . The frost thickness is measured with two
digital cameras (Cannon, power shot S3I), with estimated uncertainty 0.05 mm. Image
data from top and bottom surface of cylinder were acquired. Frost thickness is
measured every 30 min automatically and the other data were sampled every 2 min.
All data were saved in a computer for later analysis.
3. Experimental procedure and measurements
The experiments were conducted with variable ambient parameters and cold surface
temperature. The three dominant parameters are considered herein: The ambient
temperature varied in the range of 15-22 Co , the humidity ratio of air varied from 6.4
to 8.4 DAkgg , and the temperature of cold cylinder surface is changed from -8 to
-4 Co . Before cooling the test cylinder, the surface had been covered by a thin
polyethylene film so that water vapor could not condense on the test cylinder surface.
After test surface reached to the prescribed temperature, the test is started by taking
off the film. The duration of each test was 8 hr. The heat flux and frost was measured
at regular time intervals (typically 30 min) after starting the test.
The cylinder is used as a heat flux meter. Heat flux can be calculated from the radial
temperature difference between inner and outer tube with known thermal conductivity
of the insulator (0.1711KmW ). Because of high thermal conductivity of aluminum
(170 11KmW ), its thermal resistance is ignored. The maximum temperature
difference in longitude direction was less than 0.5 Co and hence its effect is
negligible. The average heat flux is measured with the outer surface of the cylinder at
the center of the longitudinal location using thermocouples. Based on the well –
known conduction equation, heat flux is given by the following equation:
(1)
Cold smoke is used for flow visualization because the frost surface temperature is low
and the smoke must not have any effects on frost surface. A thin sheet smoke which
DrrLn
TTkq
inout
inout
)(
)(2'
is produced from chemical materials that comprise of colidric acid ( ClH ) and
ammoniac ( 3NH ) is used for smoke test. Fig. 3 shows typical smoke test smoke test
that is conducted at Raleigh number 6102 ( CT oa 20 , CT o
s 6 ). Flow
visualization shows separation did not happened in the bottom of cylinder and flow
around the cylinder is laminar. The figure also shows that the cold plum velocity is so
small that frost roughness has not caused any turbulence or mixing of the boundary
layer.
6. Theoretical modeling
The frost formation process involves simultaneous heat and mass transfer during
varying thermo-physical properties. To simplify the analysis, following assumptions
are made.
1- Heat and mass transfer is one dimensional (only in radial direction).
2- Thermal system is assumed to be quasi-steady state.
3- Frost deposition is homogeneous over the cylinder surface.
4- The frost layer is to be characterized by average properties.
5- The air near the frost surface is saturated.
Heat that transfer from frost layer comprises, convection, radiation and phase change
that show with following equation.
)()(44
fsavsfsa TTfmLTThq (2)
Since the cylinder surface area ratio to the test room area is very small, radiation
factor f is assumed equal to the frost emmissivity, f .
Mass flux of water vapor transferring to the frost surface that is related to the heat
transfer coefficient is
)(.
fsam wwhm (3)
Mass transfer coefficient mh is related to heat transfer coefficient by the Lewis
number [Traybal, 1990].
pm CLe
hh
. (4)
Where Le is the Lewis number that is assumed to one for the system contains air and
water vapor, Cp is calculated at the average temperature, 2)( afsave TTT .
Convection coefficient kDNuh is derived from experimental results obtained by
Martyneko and Khramtsov( 2005) for heat transfer around horizontal cylinder in free
convection .
,125.0
,48.0
,85.0
,02.1
,675.0
333.0
25.0
188.0
148.0
058.0
RaNu
RaNu
RaNu
RaNu
RaNu
107
74
42
22
210
1010
1010
1010
1010
1010
Ra
Ra
Ra
Ra
Ra
(5)
The fluid properties such as density, viscosity and thermal volumetric expansion
coefficient in the Raleigh number are calculated at average temperature 2)( afsave TTT .
The heat conducted from the frost surface in the frost layer is,
wf
wfs
rrLn
TTLKq
2
(6)
Combination of Eqs. (2, 3 and 6) results,
)()()()(
)(2 44fsavsfsamfsa
wf
fswef TTfLwwhTThrrLn
TTLk
(7)
Where the latent heat of sublimation is (Fossa and Tanda, 2002)
fsvs TL 1951088.2 6 (8)
Absolute humidity calculated by,
)(
)(62198.0
s
s
PP
Pw (9)
The saturation pressure over ice for temperature range of KTK 15.27315.173 is (Mago,
and Sherif, 2005],
)()( 64
53
42
3210 TLndTdTdTdTdd
T
dPLn s (10)
Where
16
125
84
63
22
11
40
1041635019.0
109484024.0
1020747825.0
1062215701.0
109677843.0
10152305.5
1056745359.0
d
d
d
d
d
d
d
The saturation pressure over liquid water for temperature range of KTK 15.47315.273
is (Mago and Sherif, 2005)
)()( 53
42
3210 TLncTcTcTcc
T
cPLn s (11)
Where
16
46
14
13
11
40
1065459673.0
1014452093.0
1041764768.0
1048640239.0
105516256.0
1058002206.0
c
c
c
c
c
c
The effective thermal conductivity of frost, efk , is derived from experiential results is
(Fossa and Tanda, 2002),
963.0001202.0 fefk (12)
The above equation is valid for the temperature down to -22 Co and density up to 500
3mKg . It is derived by the following relation (Fossa and Tanda, 2002)
)]15.273(227.0exp[650 fsf T (13)
The deposition mass of frost per unit area of cylinder surface in a given time interval
is,
tmm.
(14)
With the above equation the frost thickness growth during t can be calculated,
ff
mS (15)
and the frost thickness at the new time can be calculated by,
fff SSS (16)
The calculation procedure can proceed as follows:
1- fsT is taken equal to wT at the initial time t=0;
2- A given time interval t is chosen;
3- Eqs. (3)-(5), (8) and (12)–(14) are used to determine .
m , mh , h , vsL efk , f and m .
4- Heat flux is evaluated buy Eq. (2).
5- The thickness fS of frost is evaluated by Eq. (15).
6- Frost temperature, fsT , is updated iteratively by solving Eq. (7).
7- The values of .
m , mh , h , vsL , efk , f and m are updated.
8- Heat flux is evaluated by Eq. (2).
9- Thickness fS at the updated time will be determined by Eq. (16).
7. Result and discussion
Experimental data for each run were stored in the computer as a function of time,
contains ambient temperature, humidity and surface temperature. Table 1 summarizes
the range of experimental conditions. For each experiment variation of frost layer
thickness with respect to time is plotted in Fig. 4. Frost thickness increases with
respect to time. This increase differs around cylinder primates. Frost layer on the top
are thicker than the bottom. This can be attributed to higher heat and mass transfer
coefficient on the top of cylinder ( o180 ) compare to the bottom of cylinder
( o0 ). At o180 , the thickness of boundary layer is very thin in comparison to
stagnation point o0 , which is high. At the beginning of the each experiment frost
thickness grow is rapid and then it reduces and became linear. Tanada an Fossa(2002)
showed that the surface temperature increases rapidly and then converge to a constant
value. Due to high temperature difference at the begging of each experiment, driving
force for mass transfer is high and then reduces. For this stage mass transfers
contribution is mostly on frost growth and less on diffusion insides the frost.
Figure 5 shows the effect of the cylinder surface temperature on the frost growth
layer. In this case all parameters except the cylinder surface temperature were
remained constant. A lower cylinder surface temperature induced a thicker frost layer.
Average frost thickness at top and bottom of cylinder surface are compared with
modeling results in Fig. 5. Results from the theoretical model tend to underestimate
frost thickness, especially at the higher cold plate temperature ( Co4 ). In Fig. 5
modeling shows that at the beginning frost thickness growth increases rapidly and
then became linear. The effect of air humidity on the frost growth layer is shown in
Fig.6. With increasing air humidity, the frost layer became thicker. As the humidity
ratio of the ambient air increases, the concentration gradient over the cylinder surface
and consequently the concentration driving force, which arises from the concentration
gradient, transports a greater amount of water vapor from the ambient air to the frost
layer. The effect of ambient air temperature on the frost growth layer at constant air
humidity ratio is presented in Fig. 7. With increasing ambient air temperature the frost
layer thickness become thinner. When the air temperature raises, the air – frost
interface temperature increases and it causes the air humidity ratio in the vicinity of
the frost surface to increase. It decreases the concentration driving force, which is the
difference between the humidity ratio of the air in the free stream and that of the air –
frost interface. This reduces frost deposition rate and consequently the frost layer
thickness as discussed (Lee and Ro, 2002). Lee and Ro(2001) showed that with
increasing ambient air temperature the mass transfer and the frost density increases.
The reason of increasing the frost density is melting of fragile frost surface and
diffuses of water to frost layer. However the mass transfer is increased but the amount
of increasing is low and its effect compared with frost density is minor and frost layer
thickness decrease.
For refrigeration and cooling application heat transfer from ambient air to tubes
containing cold fluid is important. The total heat flux during the frost formation is
combination of convection, radiation and phase change. Figure 8 shows variation of
heat flux with time for various cylinder surface temperatures. The heat flux increases
at the begging of the frost formation and then reduces to a constant value. The heat
transfer coefficient is; strongly dependent on the roughness of the frost surface and
flow conditions. The Stanton number for a rough surface is approximately 60% higher
than that for a smooth surface (Yun et. al, 2002). Hayashi divided the frost formation
process into three periods:” crystal growth period,” “frost layer growth period,” and
“frost layer full growth period” (Hayash et. al, 1977). The frost roughness had the
peak value in crystal growth period (Yun et. al, 2002). At this stage the roughness is
like a fin that increases the heat transfer area and its coefficient. Riedy(1983) reported
that heat transfer coefficient increased at about 10 to 15% at the begging of
experiment and then decreases. Such behavior is also observable in the present
experiments, Fig. 8. Theoretical model predicts heat flux well but it can not show the
raised of heat flux at the beginning of the experiment. The modeling is based on the
smooth cylinder surface and it can not induce frost surface roughness in its
computation. Fig. 8 also shows the effect of cooling surface temperature on heat flux.
Total heat flux was found to be higher at the lower cold surface. The effect of ambient
temperature on the heat flux is depicted Fig. 9. It shows that, with increase of ambient
temperature, heat flux increased. The effect of air humidity is illustrated in Fig. 10.
The effect of air humidity on the frost growth is less rather than the ambient
temperature and cooling surface effect. Its shows that, phase change has minor effect
on total heat flux rather than cooling cylinder surface temperature and ambient air
temperature. This is due to the fact that air humidity only affects the phase change
3464.2*9724.1706681.4*)1(104505.6 TwFoS f
heat transfer but cooling surface temperature affects convection, radiation and phase
change heat transfer.
Many correlation for frost layer thickness are reported by previous researchers
in force convection (Sengoupta et. al, 1989; Schneider, 1978; Yang and Lee, 2004).
In free convection Cramers and Mehra(1982) derived a correlation for frost thickness
on a vertical cylinder, but their correlation was obtained as a function of limited
parameters. In this study a dimensionless thickness is expressed as a function of
dimensionless frosting parameters by using dimensionless analysis. Variables for
dimensionless analysis are chosen as follow: ambient air temperature, cylinder surface
temperature, air absolute humidity, average frost thickness, triple point temperature of
water, mass diffusivity of water vapor in air, time and cylinder diameter.
For dimensional analysis parameters D , vD and sa TT are chosen as characteristic
parameters. The correlation for dimensionless thickness could be:
bcbf TwaFoS **
)1( (17)
Where dimensionless temperature and dimensionless time are:
wa
tpa
TT
TTT *
2D
tDFo v
(18)
By least square method using all the experimental data a dimensionless correlation for
frost thickness is derived as follows:
(19)
Figure11a and 11 b illustrate comparison between this correlation and experiment data
from Chen and Oosthuizen (1994). For the two range of dimensionless times the
present correlation for frost layer thickness agrees well with experiment data of Chen
and Oosthuizen(1994) for the specified condition.
.
9. Uncertainty analysis
Analysis of uncertainty is required in order to evaluate the accuracy of measurement
and proposed correlation. Lists the experimental values of different parameters and
associated uncertainties for proposed correlation originate in Table 2. The maximum
error uncertainties of measurements are determined according to tTble. 2 is less than
9%.
10. Conclusion
Based on the experiments carried out and laminar flow observed by smoke test the
following conclusions can be made.
1-Higher air humidity ratio and lower cooling surface of cylinder temperature resulted
in thicker frost layer, and greater quantity of heat flux.
2-As the ambient air temperature increased, the frost thickness decreases slightly but
the heat flux increases which means higher frost densities.
3-The frost layer on top of the cylinder is thicker than the thickness at bottom of the
cylinder.
4- The analytical heat and mass transfer model can predict well frost thickness for low
temperature, but the error at the beginning and for high temperature cylinder is large.
5- For practical applications, a correlation for frost growth over a horizontal cylinder
based on experiments is developed which shows good agreement with other
experimental studies.
Appendix A
The total uncertainty U comprises uncertainties of many components, influence on the
experiment [Teoch et al., 2002]. For a measurement M, whose results depend on
uncorrelated input estimates nxxx ...,, 21 , the standard uncertainty of the measurement
is obtained by appropriately combining the standard uncertainties of these input
estimates. The combined standard uncertainty of the estimate M denoted by U is
calculated from the following equations [Wang te al., 2004; Wang et al. 2007]
)....,,( 21 nxxxfM (20)
)()( 2
2
1
2
i
n
i i
xUx
fMU
(21)
Where f is the function of M in terms of input, estimates nxxx ...,, 21 , and each )( ixU is
a standard uncertainty of inputs.
By applying Eq. (19) to Eq. (21) average dimensionless frost thickness uncertainty is:
21
2*
2*
2*
2*
2*
*
tt
S
DD
Sw
w
ST
T
ST
T
S
S
f
ffw
w
fa
a
f
f
(22)
References
[1] Barron, R.F., Han, L.S., 1965. Heat and mass transfer to a cryosurface in free
convection. Journal of Heat Transfer. 17, 499-506
[2] Chen, J., Oosthuizen. Frost formation on a horizontal cylinder with free
convection, M.S thesis, Queen’s university, Kingston, Canada,1994.
[3] Cheng, C.H., Wu, K.H., 2003. Observations of early-stage frost formation on a
cold plate in atmospheric air flow. Journal of Heat Transfer. 125, 95-101.
[4] Cremers, C.J., Mehr, V.K., 1982. Frost formation on vertical cylinders in free
convection. Journal of Heat Transfer. 104, 3-7.
[5] Fossa, M., Tanda, G., 2002. Study of frost formation on a vertical plate.
Experimental Thermal and Fluid Science. 26, 661-668.
[6] Hayashi, Y., Aoki, A., Adachi, S., Hori, K., 1977. Study of frost properties
correlating with frost formation types. Int. J. Heat Transfer. 99, 239-245.
[7] Kennedy, L.A., Goodman, J., 1974. Free convection heat and mass transfer under
conditions of frost deposition. Int. J. Heat and Mass Transfer. 17, 477-484.
[8] lee, Y.B., Ro, S.T., 2001. An experimental study of Frost formation on a
horizontal cylinder under cross flow. International Journal of Refrigeration. 24, 468 –
474.
[9] lee, Y.B., Ro, S.T., 2002. Frost formation on a vertical plate in simultaneously
developing flow. Experimental Thermal and Fluid Science. 26, 939 – 945.
[10] Mago, J., Sherif, S.A., 2005. Heat and mass transfer on a cylinder surface in
cross flow under supersaturated frosting condition. International Journal of
Refrigeration. 28, 538-899.
[12] Martyneko, O.G., Khramtsov, P.P., Free-convective heat transfer. Springer,
2005.
[13] Na, B., Webb, R.I., 2004. Mass transfer on and within a frost layer. International
Journal of Heat and Mass Transfer. 47, 899-911.
[14] Ostine, R., Andersoon, S., 1991. Frost growth parameters in a forced air stream.
International Journal of Heat and Mass Transfer. 34, 1009-1017.
[15] Riedy, M.K., 1983. Transient natural convection heat and mass transfer during
refrigeration of air by horizontal tubes. Int. J. Heat and Mass Transfer. 17, 863-867.
[16] Salamanpour, M., Nourani Zonouz O., and Yaghoubi, M., 2008, Analysis of
frost formation through a two- dimensional duct with turbulent flow. Heat Transfer
Research. 39, No. 4, pp.35-58,
[17] Sengoupta, S., Sharif, A.A., Wong, E.Y., 1989. Frost deposition on a cylinder
in cross flow. Int. J. energy Res. 22, 615-622
[18] Schneider, h.W., 1978. Equation of the growth rate of frost forming on cooled
surface. International Journal of Heat and Mass Transfer. 21, 1019-1024.
[19] Teoch, P.L., Shirinzadeh, B., Foong, C.W., Alici, G., 2002. The measurement
uncertainties inthe laser Interferometry-based sensing and tracking technique.
Measurement. 32, 135–150.
[20] R. Traybal, R.E., 1990. mass transfer operations. third edition. McGraw.
[21] Yang, D.R., Lee, K.S., 2004. Dimensionless correlations of frost properties on a
cold plate. International Journal of Refrigeration. 27, 89-96
[22] Yun, R., Kim, Y., Min, M.K., 2002. Modeling of frost growth and frost
properties with airflow over flat plate. International Journal of Refrigeration. 25, 362-
371.
[23] Wang, C.C., Huang, R.T., Sheu, W.j., Chang, Y.J., 2004. Some observations of
the frost formation in free convection: with and without the presence of electric field,
International Journal of Heat and Mass Transfer. 47, 3491-3505.
[24] Wang, X., Bibeau, E., Naterer, G.F., 2007. Experimental correlation of force
convection heat transfer from a NACA air foil, Experimental Thermal and Fluid
Science. 31, 1073-1084.
[25] Wu, X., Dai, W., Xu, W., Tang, L., 2007. Mesoscale investigation of frost
formation on a cold surface. Experimental Thermal and Fluid Science. 31, 1043-1048.
2-a- Longitudinal view of test tube
2-b- Cross section of tube with location of thermocouples
Fig. 2.Test cylinder
0
0.5
1
1.5
2
2.5
0 100 200 300 400 500 600Time(min)
Fro
st th
ickn
ess(
mm
)
Bottom
Top
Fig. 4. Variation of frost thickness on top and bottom of cylinder with
time( CT oa 22 , DAkggw 4.6 CT o
s 6 )
Bottom
fsT
aTTop
0
1
2
3
4
5
6
0 100 200 300 400 500 600Time(min)
Fro
st th
ickn
ess(
mm
)
Fig. 5. Variation of average frost growth layer for different cylinder surface , ,
DAkggw 9.8 ) CT oa 18 temperature (
CT os 8
CT os 6
CT os 4
: Modeling
Symbols: Experiment
0
1
2
3
4
5
6
0 100 200 300 400 500 600Time (min)
Fro
st la
yer
thic
knes
s (m
m)
Top layer thichness
Fig. 6. Effect of ambient air humidity on the frost layer growth ( CT oa 18 ,
CT os 8 , DAkggw 1.8 )
DAkggw 9.8DAkggw 1.8DAkggw 4.6
0
1
2
3
4
5
0 100 200 300 400 500 600Time(min)
Fro
st th
ickn
ess(
mm
)
Fig. 7. Effect of the ambient air temperature on the frost growth layer
( DAkggw 1.8 , CT os 8 ) .
CT oa 22
CT oa 18
100
200
300
400
500
0 100 200 300 400 500 600Time(min)
Hea
t flu
x(W
/m^2
)
Fig. 8. Effect of cooling surface temperature on the heat flux from air to cylinder
surface.
DAkggw 1.8
CT oa 18
CT os 8
CT os 6
CT os 4
: Modeling
Symbols: Experiment
0
100
200
300
400
0 100 200 300 400 500 600Time(min)
Hea
t flu
x(w
/m^2
)
Fig. 9. Effect of ambient air temperature on the heat flux from ambient air to
cylinder surface.
CT oa 18
CT oa 22
DAkggw 1.8
CT os 4
100
200
300
400
0 100 200 300 400 500 600Time (min)
Hea
t flu
x(W
/m^2
)
Fig. 10. Effect of air humidity on the heat flux from ambient air to cylinder surface.
DAkggw 2.7
DAkggw 9.8
DAkggw 1.8CT o
a 18
CT os 8
0
0.01
0.02
0.03
0.04
0 1 2 3 4 5
11-a
0
0.02
0.04
0.06
0.08
0 5 10 15 20 25
11- b
Fig. 11. Comparison of dimensionless frost growth prediction with other
measurement
CT oa 28.25
CT os 75.11
%18.48RHmmDcylinder 224
Fo
*
fS
Relation (19)
Chen and Oosthuizen (1994)
CT oa 9.28
CT os 18.6
%9.47RHmmDcylinder 158
Fo
*
fS
Chen and Oosthuizen (1994)
Relation (19)
Fig. 1. Schematic diagram of the test facilities
Fig. 2.Test cylinder
2-a- Longitudinal view of test tube
2-b- Cross section of tube with location of thermocouples
Fig. 3. Smoke test with cold plume flows
Fig. 4. Variation of frost thickness on top and bottom of cylinder with
time( CT oa 22 , DAkggw 4.6 CT o
s 6 )
Fig. 5. Variation of average frost growth layer for different cylinder surface
temperature
CT oa 18 , DAkggw 9.8
Fig. 6. Effect of ambient air humidity on the frost layer growth ( CT oa 18 ,
CT os 8 , DAkggw 1.8 )
Fig. 7. Effect of the ambient air temperature on the frost growth layer
( DAkggw 1.8 , CT os 8 )
Fig. 8. Effect of cooling surface temperature on the heat flux from air to cylinder
surface.
Fig. 9. Effect of ambient air temperature on the heat flux from ambient air to
cylinder surface
Fig. 10. Effect of air humidity on the heat flux from ambient air to cylinder surface
Table 1.Experimental condition carried out
Table 1.Experimental condition carried out
Parameter Range
Air humidity ratio( DAkgg ) 6.4 - 8.9
Surface temperature of cylinder ( Co ) -8.0 - -4.0
Ambient air temperature ( Co ) 18.0 - 22.0
Table. 2. Values for the parameters of Eq. (19)
*
*
f
i
i
f
S
x
x
S
ix ix Parameter
1.2
0.25
0.5
0.5
291.15
295.15
Ambient temperature
( Co)
0.417
0.2
0.5
0.5
-4
8-
Cylinder surface
temperature Co
0.00033
0.02
80
Cylinder diameter
mm
2.3
0.78
10
10
28800
28800
Time (S)
1.6
3.3
1
1
RH=69
RH=39
Humidity (%)