7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010












Experimental study on boiling flow of liquid nitrogen in
inclined tube- liquid slug length distribution and velocity



Wang Jing, Wang Shu Hua and Zhang Hua
Institute of Engineering Thermo-physics, Shanghai Jiaotong University
800 Dongchuan Road, Shanghai, 200240, China

San,-'i sjtu Ldu ela

keywords: slug length, cryogenic two-phase flow, nitrogen


Abstract. An experimental study was carried out to understand the phenomena of the boiling flow of liquid
nitrogen in an inclined tube with closed bottom by using a high speed motion analyzer. The experimental tube is
0.018 m ID and 1.0 m in length. The range of the inclination angle is 45-90 from the horizontal. The liquid slug
length and velocity were studied. The mean liquid slug lengths increased first, and then decreased with
decreasing 0, maximum at 600, which showed the Taylor bubble was easier to coalescence from vertical to
inclined, but coalescence lessening at 450. The standard deviations of liquid slug lengths increase with increasing
x/D at all inclination angles. The liquid slug propagation velocity increases first, and then decreases with
decreasing 0, maximum at 0=600, and is almost stable at 50D and 55D positions. These conclusions provide a
basis for further study of the cryogenic two-phase slug flow.


INTRODUCTION

Gas-liquid slug flow is highly complex with
an inherent unsteady behavior. It is
characterized by long bullet-shaped bubbles
separated by liquid slugs that may be aerated
by small dispersed bubbles. Slug flow is found
in many industrial applications one of which is
transport and handling of cryogenic fluids.
With the development of current aerospace
technology, cryogenic propellants are
increasingly used in the missile industry. In
cryogenic engineering, superheating always
exists in conveyor and storage system of
cryogenic liquid. So cryogenic two-phase flow
is unavoidable The propagation and storage
of cryogenic liquids make many problems,
such as stratification, geysering and rollover
(Hands, 1988). Phenomena allied to geysering
can cause high transient pressures and vapor


flow rates, in some cases large enough to
damage equipment. These bring new
challenges on the application of the multi-
phase flow theory in cryogenic engineering.
Many researches were carried out to
understand the liquid slug lengths which used
normal atmospheric temperature liquid such as
air-water and air- kerosene, and were carried
out mainly for horizontal or slightly inclined
slug flow and for vertical flow in developed
slug flow (Nydal et al.1992; Brill et al.1981;
Nicholson et al.1978; Andreussi et al. 1993;
Cook et al. 2000; Griffith et al. 1961; Bemicot
et al. 1989; van Hout et al. 1992; Costigan et al.
1997; van Hout et al. 2001; Mao et al.1989;
Felizola et al. 1995; van Hout et al.2003). The
liquid slug lengths distribution can be
described by positively skewed distributions,
such as the log-normal, the gamma, or the
inverse Gaussian [Nydal et al.1992; Brill et







7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010


al.1981; van Hout et al. 2001; van Hout et
al.2003).
On the liquid slug lengths, the mean and
maximum lengths were focused on study. For
the horizontal tubes, the range of the mean
liquid slug lengths varies from 10 to 100D
(Nydal et al.1992; Brill et al.1981; Nicholson
et al. 1978; Andreussi et al. 1993). For the
vertical tubes, the normalized mean liquid slug
lengths vary between 10 and 20D with
standard deviations between 30% and 50%
(Griffith et al. 1961; Bemicot et al. 1989; van
Hout et al. 1992; Costigan et al. 1997; van
Hout et al. 2001). For the inclined tubes, the
mean liquid slug length has a minimum of
about 16D at 600 for air-kerosene slug flow
(Felizola et al. 1995). R.van Hout, et al (2003)
studied the evolution of the mean liquid slug
lengths along the tubes at various inclination
angles for air-water slug flow. The result
shows the mean liquid slug lengths decrease
with decreasing inclination angles.
Cryogenic vapor-liquid slug flow is seldom
studied in inclined tube with closed bottom.
Compared with normal atmospheric
temperature liquid, cryogenic liquid has high
compressibility, low density difference
between vapor and liquid and low latent heat
of vaporization. There are large differences on
bubble motion in cryogenic two-phase flow
and normal temperature two-phase flow.
The purpose of the present study is to
investigate experimentally the distributions of
the liquid slug length and the liquid slug
velocity in inclined tube with closed bottom.
The liquid nitrogen is used as working
medium.

EXPERIMENTAL APPARATUS AND
PROCESS

Experimental Set-up

Fig. 1 shows the schematic diagram of the
experimental apparatus. The main body of the
experiment, which is made of double layer
Pyrex glass, includes a 0.4 m long stock tank
with inner diameter 0.1 m, a 1.0 m long test
sections with inner diameters, D, 0.018m. The
main body of the experiment can be rotated
around its axis and fixed at 40-900 inclination


angles from the horizontal. The vacuum
interlayer is 0.021 m, which is vacuumized by
vacuum pump to serve as the thermal
insulation to decrease the convection heat
transfer. The degree of vacuum in vacuum
interlayer is 6x 10-2 Pa. The test part as a whole,
only the upper end pipeline connected, so the
heat leakage caused by the heat conduction is
very small. The vacuum interlayer is
vacuumized, so convection heat transfer is also
very small. The main heat leakage on the
pipeline is the radiation heat transfer.
The liquid nitrogen stored in a Dewar, which
is heated by the electric heating rod, is
supplied to the test section with the help of the
high-pressure nitrogen gas in Dewar. The
heating is controlled by a power Switch. The
power switch is off when the liquid level of the
upper tank is about 1.18 m, and is on when the
liquid level of the upper tank is about 1.16 m.

High Speed Motion Analyzer

Fig. 2 is the schematic design of image
processing system. The high speed motion
analyzer (REDLAKE Motion-ProX3, 1280
x1024 pixels resolution, 1000 frames/s with
the full resolution) is employed in the
experiment, together with a lens (AI NIKKOR
50/F1.2S).In the experiment, 512x512 pixels
resolution is used with 1000 frames/s. The
recorded images are transmitted to the
computer for further analysis. Two photoflood
lamps are used as light source, whose power is
1000W.

Experimental Condition

During the experiment, the range of
inclination angles is 40-900. The positions of
20D-55D from the bottom of tube are
measured by using high speed motion
analyzer.

Image Processing

The local propagation slug velocity of the
bubble interface is calculated as the shift of the
corresponding interface, divided by the time
elapsed between the frames:







7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010


XU x1
U nAt
sonst n t


Where, x, is the bubble's tail position of
Frame n x2 is the bubble's tail position of

Frame n2, see Fig. 3, n = n, n,, At = ( f is

the Frame rate).
The liquid slug length is determined by
multiplying the residence time of the liquid
slug over the measured position by the liquid
slug velocity.

RESULTS

Liquid Slug Video Images Along The Tube
at Various Inclination Angles

Fig.4 presents the evolution of the liquid
slug along the tube at various inclination
angles. The left of the video images (Fig.4) are
the bottom of the leading nitrogen Taylor
bubbles, and the right are the nose of the
trailing nitrogen Taylor bubbles. Fig.4
indicates that small bubbles of liquid slug
regions are more and more near the tube upper
with decreasing angles.

Liquid Slug Velocity

The statistics data of liquid slug propagation
velocity are listed in Tab. 1. Tab. 1 shows that
liquid slug propagation velocity <
UALs >(m/s)increases first, and then decreases
with decreasing 0, maximum at 0=600, and is
almost stable at 50D and 55D positions. It
should be noted that the standard deviation of
ULLs(m/s) can approach about 0.10. This
indicates small variability between individual
slugs.

Liquid Slug Length Distributions

The histograms showing the distribution of
liquid slug lengths are given in Fig.5. In
general, the mean and the mode of the length
distributions increase along the tube and the
liquid slug length distributions are left-skewed
in all cases.


The effect of the inclination angle on the
measured liquid slug length distributions at
different locations along the tube is shown in
Fig. 5. The mean and the most mode liquid
slug lengths increased first, and then decreased
with decreasing 0, maximum at 0=600, and the
distributions become more inhomogeneous
with decreasing 0.

Mean Liquid Slug Lengths and Standard
Deviation

Fig. 6 present the evolution of the
dimensionless mean liquid slug lengths as a
function of inclination angle. In all cases, the
values of Ls,,me are 3.5-5.5D at x/D=20 and
increase to 7.5-9.5D at x/D=50, x the axial
pipe distance measured from the bottom of the
tube.The evolution of the slug length along the
tube is hardly affected by the inclination angle,
while the mean liquid slug lengths increased
first, and then decreased with decreasing 0,
maximum at 0=600, which shows Taylor
bubble is easier to coalescence from vertical to
inclined, but coalescence lessening at 450.
Similarly to Lsea,,, the values of LB,,ea depend
on the tube inclination as well, with the longest
bubbles observed at 600.
Fig. 7 presents the evolution of the
dimensionless liquid slug lengths standard
deviations as a function of inclination angle.
Liquid slug lengths standard deviation
increases with increasing x/D in all cases
which shows liquid slug length distribution is
inhomogeneous with increasing x/D.

The Log-normal Distributions of The
Liquid Slug Length

Fig.5 shows the liquid slug length
distributions are left-skewed in all cases. The
log-normal shape is fitted to the measured
distributions and is depicted in Fig.5 as a solid
line. The probability density function of the
log-normal distribution is

2

1 Ls 1- D (2)
f (). D exp







7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010


where Ls is the values of the liquid slug lengths
along the tube at various inclination angles,
y > 0, A > 0 .The parameters A and in Eq. (2)
are given in Tab. 2 for all cases.

SUMMARY AND CONCLUSIONS

An experimental study of the evolution of
continuous liquid nitrogen boiling slug flow
along inclined tube with internal diameters
0.018 m is presented. The hydrodynamic and
statistical parameters include liquid slug length
distributions and instantaneous velocity of
liquid slug. The measurements were carried
out by high speed motion analyzer at different
positions along the tube.
Through images analyzing, the small
bubbles of liquid slug regions were more and
more near the tube upper with decreasing
angles.
Liquid slug propagation velocity increases
first, and then decreases with decreasing 0,
maximum at 0=600, and is almost stable at 50D
and 55D positions. The standard deviation of


U[As(m/s) can approach about 0.10. This
indicates small variability between individual
slugs.
For all cases, measured length distributions
were well described by the log-normal shape.
The mean and the most mode liquid slug
lengths increased first, and then decreased with
decreasing 0, and the distributions became
more inhomogeneous with decreasing 0,
maximum at 0=600.
In all cases, the values ofLsean,, were 3.5-
5.5D at x/D=20 and increased to 7.5-9.5D at
x/D=50.The evolution of the slug length along
the tube was hardly affected by the inclination
angle, while the mean liquid slug lengths
increased first, and then decreased with
decreasing 0, maximum at 0=600, which
showed Taylor bubble was easier to
coalescence from vertical to inclined, but
coalescence lessening at 450. Liquid slug
lengths standard deviation increased with
increasing x/D in all cases which showed
liquid slug length distribution was more
inhomogeneous with increasing x/D.


Tab. 1 Liquid slug velocity and standard deviation along the tube at various angles

x/D parameters 900 800 700 600 450 400
20 0.30 0.33 0.34 0.44 0.35 0.33
S.D. 0.10 0.11 0.09 0.12 0.13 0.11
30 0.33 0.36 0.40 0.46 0.36 0.35
S.D. 0.09 0.14 0.11 0.13 0.10 0.13
40 0.34 0.36 0.39 0.43 0.39 0.39
S.D. 0.10 0.12 0.11 0.12 0.11 0.10
50 0.38 0.37 0.38 0.39 0.34 0.30
S.D. 0.13 0.09 0.11 0.10 0.12 0.08
55 0.39 0.38 0.38 0.39 0.35 0.31
S.D. 0.09 0.10 0.09 0.09 0.11 0.12
Tab. 2 Parameters a and b of log-normal fit
angles parameters 20D 30D 40D 50D
90o 0.56 0.62 0.57 0.67
S1.34 1.38 1.66 1.77
80 A 0.48 0.8 0.29 0.5
S1.42 1.64 1.58 1.76
70 A 1.11 0.5 0.35 0.36
S1.51 1.55 1.57 2.06
60 A 0.43 0.52 0.4 0.23
S1.77 1.74 1.9 1.76
450 1.07 0.27 0.43 0.36
j1.29 1.51 1.79 1.76






7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010


1. Power supply 2. Nitrogen dewar 3. Electric
heating rod 4. Power cord 5. Liquid nitrogen
delivery tube 6. Ball valve 7. Flexible Tube 8.
Upper tank 9. Vacuum valve 10. Test section 11.
Vacuum tube 12. Vacuum pump
Fig. 1 Experimental apparatus


3

DIT


1. Computer 2. High speed motion analyzer 3.
Light 4. Screen 5. Taylor bubble 6. Liquid slug
7. Vacuum interlayer
Fig. 2 The schematic design of Image
processing system


Frame 5


P


Frame 35



I .,
Fig. 3 Determination of the liquid slug length
by image processing. Example for 0=45,
frame numbers 5 and 35. Frame rate =1000fps


20D 30D 40D 50D


700nfl



450

Bottom ofleading bubble Liquidslug Nose oftrailingbubble
Fig. 4 Liquid slug images along the tube at
various inclination angles








7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010


VoL Io =0o







01 0G 8
-s c
O0


0o1 4 8 -P2a



0 4 8 P 15 2 0 4 8 2
Lsp, LsD


x/D-20
6=7(


S1
.a.

I4 8 12p L 6 s 0 4 8 Ls ,D1


vc 20
0-4
o=48












1 4 8 12 15 ) 2L
LS,


48P 15 2
LS,


0



0 4 8P 15 2
Lsp


x/D-40
0 8


0



4 8 12 16 2) 2
Lsp


0 4 0 4
LSD0 LD ^^ 0






) a o






S4 8 15 B 21 0 4 8 P 15 2) 2 4
1LS, LS4,


LSp


x/D-50
x/D-D=5xI/x50 o 0 0-6 x O50
o -d o=8s =7 0 =4
















Fig. 5 Liquid slug length distribution along the tube at various inclination angles
Fig. 5 Liquid slug length distribution along the tube at various inclination angles







7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30-Jane4,2010


12
11
10
9
Q 8
7
6
5
4
3


4
0a

0


20 30 40 50 60


Fig. 6 Liquid slug mean length along the tube
at various inclination angles


ACKNOWLEDGEMENTS

This research is supported by National
Natural Science Foundation of China
(50476015) and National high-tech research
development plan (863) Project of China
(2006AA09Z333).

REFERENCES

Andreussi, P. et al. (1993). "Void distribution
in slug flow," Int. J. Multiphase Flow, 19(5),
pp.817-828.

Bemicot, M. et al. (1989). Slug length
distribution in two-phase transportation
systems," '89 Proceedings of the fourth
international conference on multiphase flow,
Nice, France, BHRA,Cranfield, Beds ,pp.485-
493.

Brill, J. P. et al.(1981). "Analysis of two-phase
tests in large-diameter flow lines in prudhoe
bay field," Society of Petroleum Engineers
Journal, 271(6),pp.363-378.

Cook, M. et al. (2000). "Slug length prediction
in near horizontal gas-liquid intermittent
flow," Chemical Engineering Science,
55(11),pp.2009-2018.

Costigan, G. et al. (1997). "Slug flow regime
identification from dynamic void fraction
measurements in vertical air-water flows," Int.
J. Multiphase Flow, 23(2), pp.263-282.


E 900
o 800
A 700
v 600
< 450 V
A a


20 30 40 50 60
x/D
Fig. 7 Liquid slug lengths standard
deviation along the tube at various
inclination angles

Felizola, H. et al. (1995). "A unified model for
slug flow in upward inclined pipes," Journal of
Energy Resources Technology, 117(1),pp.7-12.

Griffith, P. et al. (1961). "Two-phase flow,"
Journal of Heat Transfer, 83(3),pp. 307-320.

Hands,B.A.(1988)"Problems due to
superheating of cryogenic liquids,"
Cryogenics, 28(12),pp.823-829.

Nicholson, M. K. et al.(1978). Intermittent
two phase flow in horizontal pipes: Predictive
models," The Canadian Journal of Chemical
Engineering, 56(12),pp. 653-663.

Nydal, O. J. et al.(1992). "Statistical
characterization of slug flow in horizontal
pipes," Int. J. Multiphase Flow, 18(3),pp.439-
453.

van Hout, R. et al. (1992). "Spatial distribution
of void fraction within the liquid slug and
some other related slug parameters," Int. J.
Multiphase Flow, 18(6), pp.831-845.

van Hout, R. et al. (2001). "Evolution of
statistical parameters of gas-liquid slug flow
along vertical pipes," Int. J. Multiphase Flow,
27(9), pp.1579-1602.

van Hout, R. et al.(2003). "Evolution of
hydrodynamic and statistical parameters of
gas-liquid slug flow along inclined pipes," Int.
J. Multiphase Flow, 29(1),pp.115-133.