Droplet Entrainment From a

Shear-Driven Liquid Wall Film

in Inclined Ducts: Experimental
J. Ebner
e-mail: Joachim.Ebner@its.uni-karlsruhe.de
Study and Correlation
M. Gerendas Comparison
O. Schafer The primary objective of the present study is to clarify the droplet disintegration mecha-
nism and the film properties of liquid oil films driven by shear stress, which is induced by
S. Wittig a co-current gas flow. This work focuses on the flow behavior within the starting length of
the complex two-phase flow and the effect of inclination on the entrainment rate. Many
Lehrstuhl und Institut fur Thermische investigations have been performed in the past to determine the droplet entrainment in the
Stromungsmaschinen, gas core for fully developed flow conditions with respect to their relevance in pipes of
Universitat Karlsruhe (T.H.), power plants and various chemical engineering systems. In more recent work the effect of
Kaiserstrasse 12, inclination has been studied in detail. Nevertheless, a lack of knowledge can be realized
76128 Karlsruhe, Germany for droplet entrainment within the starting length of this complex flow type. Thus, funda-
mental experiments have been carried out to provide a data base for droplet entrainment
of liquid disintegrated from an oil film within its starting length at several inclination
angles of the flow. The experimental results have been compared with correlations from
literature. Additionally, the wall film thickness has been measured to allow a fully coupled
modeling of entrainment and liquid film properties depending on global flow parameters.
Based on film Reynolds number, Weber number, a dimensionless film flow length, and a
modified Froude number, taking into account the angle of inclination, correlations have
been developed, where those from literature are not applicable.
DOI: 10.1115/1.1476926

Introduction
A proper characterization of the air/oil flow in bearing compart-
ments and many other cavities of the so-called secondary air sys-
tem of modern aeroengines requires reliable models for droplet
entrainment from shear-driven oil films. In particular the develop-
ment of CFD-based design tools computational fluid dynamics
to simulate this complex type of flow requires detailed informa-
tion as shown by Glahn et al. 1. The pressure drop of the two-
phase flow as well as the heat transfer from the wall to the gas
core is strongly affected by the distribution of oil as wall film and
as droplets within the gas core 2,3. Thus, the thickness of the
liquid film becomes a vitally important parameter for a proper
description of the overall flow field.
A profound modeling of the droplet generation mechanism at
the gas-liquid interface, caused by the co-current air stream is
very demanding. For that reason, this type of flow is still under
research. Extensive investigations have been performed in the past
for fully developed two-phase flow with respect to their signifi-
cance, e.g., for air/water flow in pipe systems of nuclear power
plants or evaporators 4 8. In more recent work the effect of
inclination has been studied in detail 911. The present work
focuses on developing film flow, because the typical pipe lengths
within the secondary air system of aeroengines and many other
applications are by far too short to expect fully developed flow
conditions. Thus, the film flow length has to be taken into account.
Measurements of the entrainment fraction and the oil film
Contributed by the International Gas Turbine Institute IGTI of THE AMERICAN
SOCIETY OF MECHANICAL ENGINEERS for publication in the ASME JOURNAL OF
ENGINEERING FOR GAS TURBINES AND POWER. Paper presented at the Interna-
tional Gas Turbine and Aeroengine Congress and Exhibition, New Orleans, LA, June
4 7, 2001; Paper 01-GT-115. Manuscript received by IGTI, December 2000, final
revision, March 2001. Associate Editor: R. Natole.
thickness have been conducted for various air velocities, oil flow
rates, test section slopes, and film flow lengths. Thus, the major
parameters to describe the heat transfer and the propagation of the
liquid wall film are available for the air/oil flow under
investigation.
Experimental Setup and Operating Conditions
The measurements presented in this paper have been performed
using an unheated rectangular duct (30 mm50 mm) with only
the bottom wall wetted and a length of up to l30H900 mm.
The air pressure is p3 bar absolute and the oil has a kinematic
viscosity of f 5.2 mm2 /s DOW: DC 200 Fluid 5 cs. Thereby,
the film flows only on a width of b40 mm to minimize effects
due to a three-dimensional air flow in the duct. It flows in a height
adjustable groove such that it can be lowered a little bit more than
its film thickness with respect to the bottom wall of the duct. By
this method, the film remains in the restricted area and further-
more no point of discontinuity of the gas phase occurs at the start
of the film.
A schematic diagram of the test section is given in Fig. 1. The
air is supplied by a compressor at a constant pressure. A settling
chamber with a nozzle is placed upstream of the rectangular duct
to generate a constant velocity profile at the entrance. The oil film
is formed by a device consisting of a small slit, through which the
liquid flows into the test section at a shallow angle with respect to
the main stream direction. Therewith, a point of discontinuity is
avoided at the start of the film. For the same reason, the film flows
on a plate, which can be adjusted in its height. Thus, the cross
section of the gas flow remains constant changing the film load-
ing. Three different film flow lengths can be installed (l10
30H). The film draining is realized via a variable slit to adjust
it according to the present flow conditions. This procedure ensures

874 Vol. 124, OCTOBER 2002 Copyright 2002 by ASME Transactions of the ASME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab

 Fig. 1 Schematic of rectangular test section

the complete removal of the wall film with minimal air content for
all test conditions. The performance of the device can be observed
through a window.
Extensive measurements have been performed over a wide
range of air and oil flow rates. Table 1 gives a detailed overview
of the operating conditions and the fluid properties.
Table 1 Operating conditions of entrainment fraction and film
thickness measurements
Measurement Techniques
In the following a brief description of the measurement systems
used to analyze the complex two-phase flow will be given. First of
all, the mass flow rate measurement of the droplets in the gas core
and the liquid wall film to determine the entrainment fraction with
a high level of accuracy is described. The liquid flows in a closed
circuit. It is supplied to the test rig by a pump out of a reservoir
and is continuously drained by an adjustable slot. The functional-
ity of the device can be observed optically. Thus, the slot can be
adjusted according to the working conditions such that an addi-
tional drainage of droplets from the gas core can be avoided.
Between the outlet of the filter system downstream and the inlet of
the oil tank the liquid can be redirected by computer-controlled
valves onto a high-accuracy scale resolution: m0.1 g for a
well defined period of time. By this procedure the oil flow rate of
both, the droplets and the wall film can be determined with an
accuracy of m0.05 g/s.
The measurement of film thickness is very demanding. In the
past, several capacitive and inductive techniques have been ap-
plied e.g., Hewitt 12. A major restriction in the use of these
sensors are their large dimensions. In order to obtain a better
spatial resolution optical approaches have been suggested. An ad-
vanced system based on the absorption of infrared light is de-
scribed by Samenfink et al. 13. This technique is suitable for
water and alcohols due to the strong absorption within the fluid.
Unfortunately, nearly no absorption occurs for the oil under inves-

Fig. 2 Optical setup and the modified data acquisition of the nonintrusive
laser-based film thickness measurement technique

Journal of Engineering for Gas Turbines and Power OCTOBER 2002, Vol. 124 875

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab

 tigation. Due to the closed circuit of the liquid including several
filter systems the use of dye to increase absorption is not appro-
priate, because its concentration would not remain constant. Thus,
a laser-based system, a so-called laser focus displacement meter,
according to Takamasa et al. 14 is applied. The optical and elec-
tronic setup of the system is illustrated in Fig. 2.
The measurement principle is based on focusing the light cone
of a laser diode wavelength 670 nm onto the phase interface.
One of the focusing lenses oscillates due to electric attenuation,
shifting the spatial position of the focal point of the laser beam. At
the gas-liquid interface the beam is reflected to a great extent back
into the emitting optics. The reflected light is redirected through a
pinhole onto a photo detector by a half-transparent mirror. For one
cycle of oscillation the photo detector receives the maximum
amount of light in the time instant, when the focal point is directly
located on the phase interface as illustrated in Fig. 2. The relation
of the spatial position of the objective lens and the peak in light
intensity allows one to determine the distance of the film surface
from the sensor. Since the spacing of the sensor and the glass
window on which the film flows is known and remains constant,
the film thickness can be determined. The sample frequency of the
LFD sensor Keyence: model LT-8110 is f 1.5 kHz and the
range of the sensor is 0h2.1 mm in air. A spatial resolution of
h0.2 m is specified by the manufacturer. This resolution can
be obtained using the sensor with standard options, i.e., averaging
two or even more samples, thereby taking into account a decrease
in the data rate. In order to reach the maximal temporal resolution
to perform detailed measurements of the film structure, the analog
output of the LFD system has to be used. For this purpose, an
external trigger unit was built to generate a suitable pulse for the
analogue-to-digital conversion board of the data acquisition com-
puter. Due to this method the spatial resolution decreases to a
value of h2 m, which is very accurate for the measurement
of liquid films in the range of 0.1 mmh f 2 mm.
The refraction of the light at the glass/liquid interface as well as
the reflection of the light at the wavy surface of the film has to be
taken into account. According to Eq. 1 the measurement range of
the sensor increases by a factor of c(n,n f )1.404 due to the
refraction index of n f 1.390 of the oil under investigation. On
the other hand the half-angle of the aperture is reduced from a
value of /211.5 deg in air to f /28.25 deg in the liquid film.
h f h
tan /2
tan f /2
hc n,n f with f /2arcsinn
nf
sin /2 .
(1)
Thus, the film structure can be detected as long as the surface
angle of the film remains lower than f /28.0 deg. Previous in-
vestigations on film flows at the Institute of Thermal Turboma-
chinery Samenfink et al. 13,15 revealed that the surface angles
of shear-driven liquid films are mostly smaller. Therefore, the
LFD technique can be applied in order to perform spatial resolved
measurements of the film structure 16,17. In the case that the
maximal angle is exceeded and no measurement can be performed
the sensor gives a specific signal. This allows to discriminate
those cases from valid measurements. Since such strong reflec-
tions are rare the wave structure can be reconstructed easily by
interpolating the missing data points.
Theoretical Background
In this section dimensionless groups and fundamental equations
are provided to characterize the overall flow field. They are all
based on the forces acting in two-phase flows. In the following, a
brief description of the forces dominating the interaction of liquid
film and co-flowing air stream is given. Inertia F I , friction F F ,
and surface tension F S can be identified as vitally important pa-
rameters to characterize the droplet generation from shear-driven
liquid films Kataoka et al. 18. Due to the variation of the in-
clination of the flow and the large span of air velocities conducted,
876 Vol. 124, OCTOBER 2002
Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab
Table 2 Dimensionless groups used to characterize the two-
phase flow field
the film flow is not always dominated by shear stress at the phase
interface. Thus, gravity F G has to be taken into consideration
Spedding et al. 8.
Based on this system, four dimensionless groups can be de-
duced for the purpose to describe the film disintegration mecha-
nism: the film Reynolds number Re f , the Weber number We , the
Ohnesorge number On, and a modified Froude number Fr f . The
Froude number accounts for the inclination angle of the test sec-
tion, i.e., the gravitational force acting in film flow direction. The
superficial velocity u *f of the film is used in its definition, since no
information of the film velocity is available yet. The behavior of
the flow in the developing region is taken into consideration by a
dimensionless film length depending on the hydraulic diameter
of the duct D h 4A/ P. An overview of the definitions of all
groups to characterize the flow is given in Table 2.
Especially the studies of Ishii and co-workers are used quite
often for the prediction of the entrainment fraction. Those authors
summarized a large amount of droplet entrainment measurements
performed in the past. Based on these experimental data sets cor-
relations have been developed, to predict the droplet entrainment
fraction EF depending on the operating conditions. Those corre-
lations mainly focus on fully developed flow. Nevertheless, an
approach to describe the starting of the entrainment has been sug-
gested. Thus, the experimental results of the present study are
compared to the prediction according to Ishii and Mishima 19.
The prediction of the entrainment fraction EF according to Eq.
2 is done assuming that all liquid above the stability limit of the
film is disintegrated. Two different entrainment mechanisms are
introduced. The so-called wave undercut and shear off roll waves
depending on the flow conditions. This contemplation leads to a
equilibrium entrainment fraction EF cp. Eq. 3 of droplets in
the gas core and liquid at the wall for a fully developed flow. It
should be mentioned that the definition of both, the film Reynolds
number Re*f as well as the Weber number We* , differ from the
definitions used in this paper as shown in Eqs. 45. Thus,
these modified parameters are superscripted by an asterisk.
The following correlations have been introduced by Ishii and
Mishima 19:
V dr V dr
EF (2)
V f V dr V tot
EF tanh 7.25107 We* 1.25 Re*f 0.25 (3)
f u *f D h
Re*f (4)
f
We*
g u 2g D h
g
1/3
with u *f
V f
A bH
V f
and f g .
(5)
In order to describe the entrainment in the developing region an
exponential function has been suggested. A dimensionless length
Transactions of the ASME
 Fig. 3 Comparison of measured entrainment fraction with cor- Fig. 4 Comparison of measured entrainment fraction with cor-
relation according to Ishii and Mishima 19 film length l relation in entrance region Ishii and Mishima 19
30H

depending on film Reynolds number Re*f and a gas phase velocity separation of the parameters controlling the droplet disintegration
mechanism has been performed to derive singular relationships
u g* corrected for the droplet content are used. An overview of the for each. Thus, Figs. 6 8 give an overview of measured data sets
entrance correlation are given in Eq. 5. separating the parameters film loading Reynolds number, shear
EF 1e 10
5 2
EF stress Weber number, film length, and inclination. The angle
in Fig. 8 is defined positive for upwards directed flow according to

with
1
Re*f
and u g*
ug a force balance, which can be derived for a fluid element in the
film.

(6)
Dh u g* 4 g g 2/3

2g
The value of the constant in the relation has been derived based on
experimental data sets of Cousins et al. 20. A comparison of the
measured entrainment fraction of the present study and the half-
empirical approach is presented in the following.

Experimental Results
Entrainment Fraction. A comparison of measured and cal-
culated entrainment fraction is plotted in Fig. 3 in terms of the
dimensionless numbers described above. The diagram shows mea-
surements of entrainment fraction for horizontal flow. It displays
some shortcomings in the accuracy of the prediction for the
present application, especially taking into account the double-
logarithmic scaling. The calculated values according to the en-
trance correlation cp. Eq. 6 do not match the measured data.
On the other hand the comparison to fully developed flow condi-
tions appears reasonable. Thus, it can be assumed that the devia- Fig. 5 Entrainment function in entrance region Ishii and
tions occur due to an incorrect characterization of the entrance Mishima 19
effect.
In order to find out the origin of the discrepancies the entrance
effect is analyzed in detail. Therefore, the experimental results are
compared to data according to the exponential entrance function
cp. Eq. 6. A comparison of measured entrainment fractions and
data obtained by the correlation according to Ishii is given in Fig.
4. It can be seen clearly, that the run of the predicted curve does
not match the measured data.
Figure 5 presents a plot of the exponential function defined by
Eq. 6. The effect of the Reynolds number in the dimensionless
length is not described properly for the air/oil flow under inves-
tigation as evident in the diagram. The shape of the curve also
differs from the trend given by the measured entrainment frac-
tions. Thus, an adjustment of the model constants is not appropri-
ate in order to improve the predicting quality significantly.
Therefore, a new correlation to characterize the entrainment of
the developing air/oil flow has to be deduced. In order to obtain a
model with a high accuracy, a correlation based on the dimension- Fig. 6 Effect of Reynolds and Weber number on entrainment
less groups defined in Table 2 is presented in the following. A fraction film length l 30H ; inclination 0 deg

Journal of Engineering for Gas Turbines and Power OCTOBER 2002, Vol. 124 877

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab

 Fig. 7 Effect of film flow length on entrainment fraction Fig. 9 Comparison of measured entrainment fraction and cor-
relation film length l 20H ; inclination 0 deg

The correlation derived for the entrainment fraction by the

least-squares method reads as follows: Table 2, the experimental results are presented depending on film

loading, air flow rate, and film flow length. Figure 11 shows that
Re2f 2.25
the film thickness decreases with higher air velocities Weber
EF1.042107 We1.2 Re0.4
f 12.06110
0.8 4
. (7)
WeFr f number due to the rising shear stress at the gas/liquid interface.
The increase of shear accelerates the film and additionally in-
It has to be pointed out that the correlation has been derived from creases the droplet generation cp. Fig. 6. On the other hand no
all data sets performed within this study see Table 1. The calcu- significant effect of the film length can be observed at first sight.
lated values of the entrainment fraction according to Eq. 7 are This behavior can be explained with respect to the various inter-
also plotted in the diagrams. In order to give a measure for the locking effects. On the one hand the film loading decreases due to
overall prediction accuracy a mean relative error of EF/EF film disintegration. On the other hand the loss of momentum of
20% of the measured data and the entrainment fraction corre- the gas boundary layer leads to a decrease of the shear force. By
lation can be determined. Since no variation of the Ohnesorge this, the negligible influence of the length scale becomes evident.
number has been applied it should be mentioned that the applica-
bility of the formula is restricted to two-phase flows under com-
parable operating conditions.
The overall accuracy in predicting the entrainment fraction with
the newly developed correlation is also confirmed by the results
presented in Figs. 9 and 10. The tests have been performed over a
wide range of flow conditions which are typical for the secondary
air system of modern aero engines. Even though minor deviations
occur in some cases the correlation leads to a significant increase
of accuracy in predicting the entrainment of the air/oil flow under
investigation.
Film Thickness. Film thickness measurements have been
conducted for all operating conditions see Table 1. As mentioned
above, time-resolved film thickness measurements can be per-
formed with the LFD technique. Since modeling of liquid films is
usually done by an analysis based on mean values only time-
averaged film thickness are presented here. In order to deduce a
correlation depending on the dimensionless groups according to
Fig. 10 Comparison of measured entrainment fraction and
correlation film length l 30H ; inclination 20 deg

Fig. 11 Comparison of measured film thickness and correla-

Fig. 8 Effect of inclination angle on entrainment fraction tion for different film flow lengths

878 Vol. 124, OCTOBER 2002 Transactions of the ASME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab

 correlation has to be verified for other liquids if the properties
differ significantly from the oil used in the present study.
A global description of the air/oil flow requires not only de-
tailed information of the droplet phase. The heat transfer as well
as the pressure drop is also affected by the properties of the liquid
wall film. Thus, based on experimental data a correlation for the
mean film thickness is presented.

Acknowledgment
This work was supported by grant from the United Technology
Research Center UTRC, East Hartford U.S. which is gratefully
acknowledged. Special thanks are due to A. Glahn for the valuable
discussions on the subject and the excellent collaboration during
the investigation.

Nomenclature
Fig. 12 Comparison of measured film thickness and correla-
tion for different inclination angles AbH cross-section area m
b width of filmm
c conversion factor
D diameter m
The impact of the inclination on the film thickness is displayed f sample frequency 1/s
in Fig. 12. The effect is rather small as long as the slope of the EF entrainment fraction
duct is in the range of 20 deg 20 deg. Nevertheless, the F force N
inclination is taken into account in the correlation presented in the Fr Froude number
following. The film thickness decreases for a positive as well as h thickness, distance m
for a negative inclination angle. An explanation can be derived by H height of duct m
analyzing the gravity effect on the film propagation and the drop- l length m
let generation. In upwards directed flow the force of gravity de- LFD laser focus displacement meter
celerates the film. On the one hand, the film thickness rises up due m mass flow rate kg/s
to the conservation of mass. On the other hand, a thicker film n refraction index
exhibits a higher roughness, thereby leading to a higher interfacial On Ohnesorge number
shear stress. This results in an increase of film disintegration ac- P perimeter of duct m
cording to Fig. 8. Since both effects are of same magnitude, the p pressure Pa
dependence of the film thickness on the inclination angle becomes Re Reynolds number
clear. A similar contemplation can be conducted for downwards u velocity m/s
directed flow, respectively. u *f V f /A superficial velocity m/s
With respect to the findings presented above, the following cor- T temperature K
relation of the mean film thickness can be obtained:
V volume flow rate m3/s

h f 1.714We0.36 Re0.457
f
11.0961014
Re4f
Fr2f
0.375
mm . (8)
We
Greek Symbols
Weber number

The fraction term in the formula, taking into account the incli- aperture angle, angle of inclination deg
nation angle, confirms the considerations discussed above. The wavelength m
effect of the inclination is controlled both by the stability of the , dimensionless film length
film via the Reynolds number and by the gravity force via the kinematic viscosity mm2/s
Froude number. The predicted film thickness according to Eq. 8 surface tension N/m
are also presented in Figs. 11 and 12. The calculated curves match Indices
the measured values very well. A relative error lower than
dr droplet
h f /h f 10% of measured and predicted film thickness can be
f film
stated for all operating conditions.
F friction
g gas
Summary and Conclusions G gravity
h hydraulic
This paper supplies detailed information on the complex two- l inertia
phase flow. Extensive experimental work has been performed un- S surface tension
der operating conditions typical for the secondary air system of tot total
modern aero engines. Based on this experimental data and ad- fully developed
dressing the underlying physical relationships in terms of nondi- * modified parameter
mensional parameters, correlations for the entrained liquid and the
film thickness for air/oil flows have been developed with special References
emphasis to the starting length of the flow.
1 Glahn, A., Busam, S., Blair, M. F., Allard, K. L., and Wittig, S., 2000, Drop-
The study reveals that entrainment correlations from literature let Generation by Disintegration of Oil Films at the Rim of a Rotating Disk,
are not suitable for the present air/oil flow due to a lack in pre- 45th ASME Paper 2000-GT-279.
diction accuracy. In order to develop detailed models for CFD- 2 Glahn, A., Busam, S., and Wittig, S., 1997, Local and Mean Heat Transfer
based design tools a relationship based on dimensionless groups is Coefficients along the Internal Housing Walls of Aero Engine Bearing Cham-
bers, ASME Paper 97-GT-261.
provided. The accuracy of the correlation is excellent for all test 3 Busam, S., Glahn, A., and Wittig, S., 1999, Internal Bearing Chamber Wall
conditions performed. Due to its nondimensional form it can be Heat Transfer as a Function of Operating Conditions and Chamber Geometry,
scaled easily to similar applications. Nevertheless, the entrainment ASME J. Eng. Gas Turbines Power, 122, pp. 314 320.

Journal of Engineering for Gas Turbines and Power OCTOBER 2002, Vol. 124 879

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab

 4 Hewitt, G. F., and Taylor, N. S., 1970, Annular Two-Phase Flow, Pergamon
Press, Oxford, UK.
5 van Rossum, J. J., 1959, Experimental Investigation of Horizontal Film
Flow, Chem. Eng. Sci., 11, pp. 3552.
6 Ishii, M., and Grolmes, M. A., 1975, Inception Criteria for Droplet Entrain-
ment in Two-Phase Concurrent Film Flow, AIChE J., 212, pp. 308 318.
7 Woodmansee, D. E., and Hanratty, T. J., 1969, Mechanism for the Removal
of Droplets from a Liquid Surface by a Parallel Air Flow, Chem. Eng. Sci.,
24, pp. 299307.
8 Spedding, P. L., Watterson, J. K., Raghuanthan, S. R., and Ferguson, M. E. G.,
1998, Two-Phase Co-Current Flow in Inclined Pipe, Int. J. Heat Mass
Transf., 41, pp. 4205 4228.
9 Lee, C. K., and Mjani, A. M., 1995, Enhancement of Droplet Entrainment
From Liquid Film Flowing Over Horizontal Ribbed Surfaces, Proceedings
1st Int. Symp. on Two-Phase Modelling and Experimentation, G.-P. Celata and
R. K. Shah, eds, Edizione ETS, Pisa, 1, pp. 143148.
10 Azzopardi, B. J., 1997, Drops in Annular Two-Phase Flow, Int. J. Multi-
phase Flow, 23, pp. 153.
11 Azzopardi, B. J., Zaidi, S. H., and Jepson, D. M., 1997, Entrained Fraction in
Inclined Annular Gas/Liquid Flow, Proceedings of the ASME Fluid Eng Di-
vision, ASME, New York, 244, pp. 6976.
12 Hewitt, G. F., 1978, Measurement of Two Phase Flow Parameters, Academic
Press, London.
13 Samenfink, W., Elsaer, A., Wittig, S., and Dullenkopf, K., 1996, Internal
880 Vol. 124, OCTOBER 2002
Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/journals/jetpez/26816/ on 06/19/2017 Terms of Use: http://www.asme.org/ab
Transport Mechanisms of Shear-Driven Liquid Films, Proceedings of the
Eighth International Symposium on Applications of Laser Techniques to Fluid
Mechanics, Lisbon, Portugal, July 8 11.
14 Takamasa, T., Tamura, S., and Kobayashi, K., 1998, Interfacial Waves on a
Film Flowing Down Plate Wall in an Entry Region Measured With Laser
Focus Displacement Meters, 3rd International Conference on Multiphase
Flow, Lyon, France, June 8 12.
15 Samenfink, W., 1997, Fundamental Study of the Interaction of Droplets and
Shear-Driven Liquid Films, Ph.D. thesis, Institute of Thermal Turbomachin-
ery, University Karlsruhe, Germany.
16 Fujii, T., Asano, H., Takenaka, N., and Yamada, H., 1998, The Liquid Film
Structure in a Two-Phase Gas-Liquid Annular Flow, 3rd International Con-
ference on Multiphase Flow, Lyon, France, June 8 12.
17 Busam, S., Ebner, J., and Wittig, S., 2001, An Experimental Study of Liquid
Film Thickness in Annular Air/Oil Flow in a Vertical Pipe Using a Laser Focus
Displacement Meter, 46th ASME Paper 2001-GT-116.
18 Kataoka, I., Ishii, M., and Mishima, K., 1983, Generation and Size Distribu-
tion of Droplet in Annular Two-Phase Flow, Trans. ASME, 105, pp. 230238.
19 Ishii, M., and Mishima, K., 1989, Droplet Entrainment Correlation in Annu-
lar Two-Phase Flow, Int. J. Heat Mass Transf., 3210, pp. 18351845.
20 Cousins, L. B., Denton, W. H., and Hewitt, G. F., 1965, Liquid Mass Transfer
in Annular Two-Phase Flow, Symposium on Two-Phase Flow, Exeter, U.K., 1
Paper C4.
Transactions of the ASME