Group Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Title: 9.6.3 - Experimental Investigation of Vibration Effects on Air-Water Two-Phase Flow in an Annulus
ALL VOLUMES CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00102023/00235
 Material Information
Title: 9.6.3 - Experimental Investigation of Vibration Effects on Air-Water Two-Phase Flow in an Annulus Multiphase Flows with Heat and Mass Transfer
Series Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Physical Description: Conference Papers
Creator: Chen, S.W.
Hibiki, T.
Ishii, M.
Mori, M.
Watanabe, F.
Publisher: International Conference on Multiphase Flow (ICMF)
Publication Date: June 4, 2010
 Subjects
Subject: vibration
subcooled boiling two-phase flow
void fraction
annulus
 Notes
Abstract: In order to simulate the seismic vibration effect on the Boiling Water Reactor (BWR), vibration experiments of subcooled boiling two-phase flow in an annulus have been performed and visualized. The annulus test section was scaled down from a fuel assembly sub-channel of a BWR and attached to a vibration beam driven by an eccentric cam operating at low frequency to simulate seismic vibration effects on two-phase flow. The inner and outer diameters of the annulus test section are 19.1 and 38.1mm respectively, and the height of the test section is 2.75m. Subcooled boiling two-phase flow with a mass flux of 499.96 kg/m2s and inlet subcooling of 5.5 to 6.4oC were performed with eccentric cam rotation speed up to 270RPM and maximum displacement 19mm (d=±9.5mm). Vibration acceleration and area-averaged void fraction are acquired and analyzed to show the effects of low frequency vibration on subcooled boiling two-phase flow. Experimental results reveal that area-averaged void fraction will be affected by the variation of vibration frequency and displacement.
General Note: The International Conference on Multiphase Flow (ICMF) first was held in Tsukuba, Japan in 1991 and the second ICMF took place in Kyoto, Japan in 1995. During this conference, it was decided to establish an International Governing Board which oversees the major aspects of the conference and makes decisions about future conference locations. Due to the great importance of the field, it was furthermore decided to hold the conference every three years successively in Asia including Australia, Europe including Africa, Russia and the Near East and America. Hence, ICMF 1998 was held in Lyon, France, ICMF 2001 in New Orleans, USA, ICMF 2004 in Yokohama, Japan, and ICMF 2007 in Leipzig, Germany. ICMF-2010 is devoted to all aspects of Multiphase Flow. Researchers from all over the world gathered in order to introduce their recent advances in the field and thereby promote the exchange of new ideas, results and techniques. The conference is a key event in Multiphase Flow and supports the advancement of science in this very important field. The major research topics relevant for the conference are as follows: Bio-Fluid Dynamics; Boiling; Bubbly Flows; Cavitation; Colloidal and Suspension Dynamics; Collision, Agglomeration and Breakup; Computational Techniques for Multiphase Flows; Droplet Flows; Environmental and Geophysical Flows; Experimental Methods for Multiphase Flows; Fluidized and Circulating Fluidized Beds; Fluid Structure Interactions; Granular Media; Industrial Applications; Instabilities; Interfacial Flows; Micro and Nano-Scale Multiphase Flows; Microgravity in Two-Phase Flow; Multiphase Flows with Heat and Mass Transfer; Non-Newtonian Multiphase Flows; Particle-Laden Flows; Particle, Bubble and Drop Dynamics; Reactive Multiphase Flows
 Record Information
Bibliographic ID: UF00102023
Volume ID: VID00235
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: 963-Chen-ICMF2010.pdf

Full Text

Paper No 7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Experimental Investigation of Vibration Effects on Subcooled Boiling Two-Phase
Flow in an Annulus


S.W. Chen a, T. Hibiki a, M. Ishii a, M. Mori b, and F. Watanabe b

a School of Nuclear Engineering, Purdue University, West Lafayette, IN 47907-2017, USA
b Tokyo Electric Power Company Ltd, Tsurumi Ku, Yokohama, Kanagawa 230-8510, Japan
E-mail: swchen@purdue.edu, hibiki@purdue.edu and ishii@purdue.edu


Keywords: Vibration, subcooled boiling two-phase flow, void fraction, annulus




Abstract

In order to simulate the seismic vibration effect on the Boiling Water Reactor (BWR), vibration experiments of subcooled
boiling two-phase flow in an annulus have been performed and visualized. The annulus test section was scaled down from a
fuel assembly sub-channel of a BWR and attached to a vibration beam driven by an eccentric cam operating at low frequency
to simulate seismic vibration effects on two-phase flow. The inner and outer diameters of the annulus test section are 19.1 and
38.1mm respectively, and the height of the test section is 2.75m. Subcooled boiling two-phase flow with a mass flux of 499.96
kg/m2s and inlet subcooling of 5.5 to 6.40C were performed with eccentric cam rotation speed up to 270RPM and maximum
displacement 19mm (d=9.5mm). Vibration acceleration and area-averaged void fraction are acquired and analyzed to show
the effects of low frequency vibration on subcooled boiling two-phase flow. Experimental results reveal that area-averaged
void fraction will be affected by the variation of vibration frequency and displacement.


Introduction

In 1987 and 1993 earthquakes occurred in area of
Tohoku, Japan. The earthquakes caused nuclear reactor
scrams due to an increase in the neutron flux, which was
caused by a sudden increase in the reactivity inside the
reactors. In the 1993 earthquake a seismic intensity of 4,
lower than the scram criterion, resulted in a Boiling Water
Reactor (BWR) scram event about 0.5 seconds after the
seismic vibration (Japan Business, 1993). There are some
possible causes for the increase in the neutron flux and
reactor scram during an earthquake, which may result from
the effect of the vibrations on the gap of the fuel assemblies
or on subcooled boiling in the reactor core where changes in
void fraction may directly affect the thermal neutron
population, especially in earthquakes such as the recent very
powerful ones with intense seismic vibrations. For this
reason, the importance of the effects of seismic vibrations
on the thermal hydraulics in the reactor core should be
re-acknowledged.
Several researchers performed preliminary experiments
to verify the seismic vibration effects on adiabatic
two-phase flow and subcooled boiling flow. Nariai and
Tanaka (1994) carried out an experiment to study the effect
of an oscillating heater rod on void fraction of subcooled
boiling flow. The heater rod was oscillated by an eccentric
cam with various amplitudes and frequencies, and the
experimental results showed that the averaged void fraction
was drastically reduced when the oscillation frequency was
larger than 10Hz. Kawamura et al. (1996a) performed an
experiment to measure the thermal neutron flux variation


with an oscillating fuel assembly, and they claimed that the
neutron flux increased up to 20% when the simulated
seismic vibration was applied to the test assembly.
Kawamura et al. (1996b) also studied the effect of
horizontal excitation on bubble behavior in subcooled
boiling flow. It was found that the bubble growth and
collapse process was not affected by the horizontal
excitation. Shioyama and Ohtomi (1990) performed an
experiment to study pressure fluctuation and behavior of
vapor bubbles for a one-component two-phase flow
(Freon-113) in a vertical tube induced by longitudinal
excitation. They applied longitudinal excitation with
frequency of 5 to 50 Hz to a vertical pipe to produce
pressure fluctuation in subcooled boiling flow and they
found that the bubble size, thermal boundary layer and void
fraction were affected by the oscillations. Hibiki and Ishii
(1998) studied the effect of flow-induced vibration on local
flow parameters of two-phase flow and found that the
flow-induced vibration enhanced bubble coalescence
resulting in reduced interfacial area concentration and a
drastic change of void distribution from a wall-peaked to a
core-peaked distribution.
Very limited systematic experimental results and
models have been obtained to understand and predict the
effect of seismic vibration on two-phase flow behavior and
subcooled boiling heat transfer until now. With this
overview, the major task of this paper is to study the effects
of vibration on subcooled boiling flow characteristics.
Earthquakes usually result from the movement of
earth's crust or eruption of volcanoes, and they have existed
since the beginning of human history. With the increasing






Paper No


knowledge about the earthquake waves, one can analyze
seismic vibration with the earthquake wave form,
magnitude and intensity, which are related to the transport
conditions, energy and acting force (or acceleration). In
order to simulate the seismic vibration conditions, the
vibration characteristics of the test section, including
frequency, amplitude and acceleration, should describe the
earthquake vibration properties. Earthquake wave forms are
divided into four kinds: S-waves, P-waves, LR-waves and
LQ-waves, where S-waves and P-waves are body waves and
LR-waves and LQ-waves are surface waves. (Earthquake
Lab, 2004) Earthquake vibration frequency is usually less
than 20Hz, ranged from 0.1Hz to couple tens of Hz and
inaudible due to the low frequency. The vibration amplitude
may not be large and is usually in the order of millimeters.
However, it can result in much larger displacement for a tall
building. For the case of fuel rods in a nuclear power plant,
the maximum displacement of the rod top can be about the
order of 2cm. In order to relate the induced vibration waves
to real earthquake waves, earthquake intensity and
magnitude should be considered as well. The intensity
means the local condition or the local feeling of an
earthquake wave at a spot, and the magnitude refers to the
energy released in one earthquake incident. Therefore, for
an earthquake, there is only one magnitude value, and there
should be many intensity values distributed in different
regions. Table 1 shows the magnitude of earthquakes with
Richter magnitude scale. There are eight levels for this
magnitude scale, and fortunately the larger the magnitude of
an earthquake, the less frequent it will be. On the other hand,
Table 2 shows the intensity of earthquakes in terms of the
modified Mercalli scale. There are twelve scales for
distinguishing the intensity, and these scales can be related
to the human descriptions as well as specific quantities of
averaged peak velocity and averaged peak acceleration. For
example, an earthquake of intensity 6 at some region means
the people in that region feel a strong vibration with an
acceleration of 0.06g to 0.07g. (lg=9.8m/s2) (Hong, 2006;
Ministry of Energy, 2006; Bolt, 1993)


Table 1:
2006)


Richter magnitude scale (Ministry of Energy,


Number of earthquakes
Magnitude Number of earthquakes Typical intensity at epicenter
per year globally
>8.0 0.1 -0.2 XII
7.4-8.0 4 XI
7.0-7.3 15 X
6.2-6.9 100 VIII- IX
5.5-6.1 500 VII
4.9 5.4 1,400 VI
4.3-4.8 4,800 IV V
3.5-4.2 30,000 II III
2.0-3.4 800,000 I II


The strategies of this study are to perform boiling
water experiments with various vibration frequencies and
amplitudes under the different thermal-fluid conditions and
to analyze the effects of vibration on flow characteristics in
subcooled boiling two-phase flow and to develop the
physical understanding of them. In the following sections,
experimental facility design will be shown first, the test
matrix and test results will be presented next, and then the
analysis and discussions will be proposed to explain the
effect of vibration on subcooled two-phase flow with the
present data.


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

Table 2: Modified Mercalli scale (Ministry of Energy, 2006;
Bolt, 1993)
Intensity Description Average peak Average Peak Acceleration
velocity (cm/s) (g 9.8m/s2)
I Instrumental
II Feeble
III Slight
IV Moderate 1-2 0.015g-0.02g
V Rather strong 2-5 0.03g-0.04g
VI Strong 5-8 0.06g-0.07g
VII Very strong 8-12 0.10g-0.15g
VIII Destructive 20-30 0.25g-0.30g
IX Ruinous 45-55 0.50g-0.55g
X Disastrous More than 60 More than 0.60g
XI Very disastrous
XII Catastrophic


Nomenclature


a
D
Db
DH
Drod
d
f
G
g
h
hf,
Ahs,,b
ms
Nzu
Nsub
AP
q
Re
T
AT,,b
V
v
Vf
Vr
We
x
z


Acceleration [m/s2]
Diameter [m],
Bubble diameter [m]
Hydraulic diameter [m]
Rod diameter [m]
Displacement [m]
Frequency [Hz]
Mass flux [kg/m2s]
Gravitational acceleration [m/s2], Ig = 9.8m/s2
Height [m]; Enthalpy [J/kg],
Latent heat [J/kg]
Subcooling enthalpy [J/kg]
Motor speed[RPM]
Zuber number
Subcooling number
Differential pressure [kPa]
Heat flux [W/m2]
Reynolds number
Temperature [K]
Subcooling temperature difference [K]
Velocity [m/s]
Fluid velocity [m/s]
Relative velocity [m/s]
Weber number
Quality
Axial length [m]


Greek letters
a Void fraction
p Dynamic viscosity [kg/ms]
- 3.1416
p Density [kg/m3]
a- Surface tension [N/m1]


Subsripts
b I
eq I


Bubble
Equilibrium


f Liquid phase
fg Liquid-gas or phase change properties
g Gas phase
H Hydraulic diameter
R Ratio
r Relative value
sub Subcooling
Zu Zuber number






Paper No


Abbreviation
BWR Boiling water reactor
FFT Fast Fourier Transform
HP Horse power
LQ Earthquake LQ-wave form
LR Earthquake LR-wave form
P Earthquake P-wave form
S Earthquake S-wave form
Test # Test number (B001 B004)


Experimental Facility

Figure 1 shows the schematic of the vibration test loop.
This test loop system includes an annulus test section,
vibration module, compressed air source, water supply lines,
water pump, separation tank and main tank. In order to
simulate the subcooled boiling two-phase flow in a BWR
fuel assembly sub-channel, the annulus test section was
designed based on the scaling laws of geometrical similarity,
hydrodynamic similarity and thermal similarity. These
similarity criteria can be briefly summarized as the
following relations:
Geometrical Similarity:

S-1 (1)
DHR
Where DH and Db are the hydraulic diameter and bubble
diameter, respectively. The bubble diameter can be
approximated as the following relation:
D _
Db 2 gAp (2)

Where Db,max is the maximum distorted bubble limit given
by (Ishii and Zuber, 1979).
Hydrodynamic similarity:
The ratio of the relative velocity to the liquid velocity
should be scaled as:

=1 (3)
R
To estimate the liquid velocity in the model, we may
approximate the bubble rise velocity, v,, as (Ishii and Zuber,
1979):

v_ P (4)
P f
Thermal similarity:
Subcooled number, Nsub, and Zuber number (phase change
number), NZu, play an important role in the thermal
similarity criteria. The subcooled number is the ratio of
the subcooling to the latent heat as:

N p (5)

where Ahsub and hfg are the subcooling enthalpy and the
latent heat, respectively. The Zuber number is the ratio of
the heat flux used for phase change to the inlet subcooling
as:
N D= 4qL Ap (6)

where Lh is the heated length. From the steady state energy
equation balanced over the heated section using a control
volume analysis, Nsub and Nzu are related by:


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


P =N- NN (7)

Therefore, the similarity of the subcooling and Zuber
numbers yields:

P e (8)

This indicates that the vapor quality should be scaled by the
density ratio.


Separationtank


Annulus
Testston Vibration Beam
^ - -------.-.---- -
Accelerometer
S Slidngtrack and
Eccentric cam





Ball bearing


r Fixedl-Beam
inlet ton m
Vibration module


Drain
Figure 1: Schematic of the whole test system

Some of the important scaling criteria are highlighted
as described above, and detailed discussions on the scaling
criteria are found in the previous papers (Ishii and Kataoka,
1984; Kocamustafaogullari and Ishii, 1984). The loop
geometry and the thermal-hydraulic conditions in the
prototypic BWR and the scaled model are tabulated in Table
3. In the proposed test loop, the geometrical similarity is
almost preserved.

Table 3: Comparison of Prototype BWR and annulus test
section of the present study (Situ et al, 2004, 2005)
Quantity Prototype Model
Pressure (Mpa) 7.17 0.101
Saturated temperature (C) 287 100
Heater diameter (m) 0.0123 0.0191
Hydraulic diameter (m) 0.015 0.0191
Heated length (m) 3.81 1.73
Heater power (kW) 77.2 20
Heat flux (kW/m2) 526 193
Bubble size (mm) 3.18 5.01
Db/DH 0.212 0.263
Bubble rise velocity (m/s) 0.172 0.222
Liquid inlet velocity (m/s) 1.93 0.498 1.24
Reynolds number 2.24x105 3.22x104 8.05x104
Weber number 4.00 4.00
Subcooling number 0.65 5.99 15.0
Zuber number 4.72 7.61 -20.5
Inlet subcooling (C) 9.49 2 5

The annulus test section includes a liquid-gas inlet at
the bottom and the Pyrex glass tubes with instrumentation
ports on the upper region. A heater rod with 19.1mm (0.75")
diameter is located at the center of the test section to form a
heated annulus. The transparent Pyrex glass tubes with
inside diameter of 38.1mm (1.5") are introduced for flow
visualization, and the instrumentation port is made for
getting experimental data including temperature, pressure
and area averaged void fraction. Due to the heavy load of
instrumentation ports, each port is fixed by an aluminum
supporting structure, which is attached to the vibration beam.






Paper No


Figure 2 show the cross-sectional view of an
instrumentation port. The instrumentation port is made with
stainless steel (outside casing) and Teflon (inside insulation
material). The port height is 127mm (5"), and the inside
diameter is 38.1mm (1.5"). The pressure is measured
through a pressure tube; the void fraction is measured by the
ring-type impedance meter, and the temperature is measured
by a K-type thermal couple. The heater rod is fixed by four
positioning bolts at the downstream of the measurements.
Seven instrumentation ports are introduced to the annular
test section, and the first port (Port 1 at the bottom) and the
last port (Port 7 at the top) are served as inlet port and outlet
port. The axial locations of Port 2 through Port 6 from the
inlet port (Port 1) are in z=0.742m, 1.179m, 1.463m, 1.748m
and 2.032m, respectively.


Figure 2: Cross-sectional view of an instrumentation port

Figure 3 shows the schematic of the vibration module.
The vibration module includes a fixed I-beam, a moveable
aluminum beam (vibration beam), ball bearing, sliding track,
spring damper, 3-HP (horse power) motor, frequency
controller and an eccentric cam located at the top of the
vibration beam which is used to generate various vibration
conditions for simulating different kinds of seismic
vibration conditions. By changing the rotation speed of the
motor and the eccentricity of the eccentric cam, the
vibration module can provide variable vibration frequency
and displacement. Operation conditions of the system are
limited by the structure strength and safety design instead of
the motor capability.


Eccentric cam


ack Spring




RotationAxis


FixedI-Beam
Figure 3: Schematic of the vibration module


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

In order to ensure safe operation of the vibration
module, a simple calculation for vibration deformation and
natural frequency of the vibration beam was performed in
the design process. By taking into account all of the
geometries and material properties, the deformation and
vibration properties can be determined. The natural
frequency is designed as 73.4Hz, and the maximum
vibration displacement is set as 20mm with a safety factor
of 0.5 (i.e. the design limit should be at least two times
larger than the operation condition) for an allowable beam
deformation. However, the whole system is a complicated
combination of vibration module and annulus test section
with two-phase flow. The real vibration modes as well as the
transient variations may not be ideally predicted by simple
calculations, and therefore the true vibration conditions
should be recorded by an accelerometer. A triaxial
accelerometer (Dytran 7523A1) was then introduced to the
top of the vibration beam, which is capable of measuring 2g
(1g=9.8m/s2) acceleration. Vibration acceleration and
frequencies can be easily determined from the transient
acceleration data.
Figure 4 shows the schematic of data acquisition
process. The data logger is NI-SCXI 1000 with SCXI-1303
module and Labview 8 to get the analog data from
accelerometer, pressure transmitters, thermocouples and
impedance meters. Most of the analog data are acquired
with 700Hz frequency, and impedance signals are acquired
with 100Hz scanning frequency for each port due to
limitation of a channel switching process of the multiplexer.
A global measurement of void fraction can be shown by the
impedance signals. As discussed in the introduction, the
void fraction variation may be the major reason of the BWR
scram in 1993, and therefore the transient variations of void
fraction are important evidence in this experiment.


Figure 4: Schematic of the data acquisition process


Void Fraction Calibration and Uncertainty

In order to convert the impedance signals into void
fraction values, the impedance meters should be calibrated
by air-water experiments in pool condition. The annulus test
section is first filled with stagnant water, and then air is
injected into the water. By assuming no acceleration and
frictional pressure drops in this condition, there is only
gravitation pressure drop (hydrostatic pressure) inside the
test section. The pressure drop in the test section can be
expressed as following form:






Paper No


AP Apagh (9)
The void fraction can be determined by the differential
pressure. Through a set of calibration tests from low to high
gas flow condition and combined with the impedance
signals, the relations between impedance signals and void
fraction can be found for each impedance port. By
considering the error transport from each instrument
including the differential pressure transmitter error (0.075%),
distance error (0.033%) and density error (0.039%), the
overall error for void fraction measurement is about 5.35%.
In addition, the uncertainty of temperature measurement is
about 0.20C, and the nonlinearity of the accelerometer is
about 0.3%.


Test Matrix and Pre-Tests

Based on the capability and limitation of the test loop,
the test matrix of this study is shown in Table 4. Four sets of
subcooled boiling experiments were performed with
stationary and vibration conditions. Vibration conditions can
be also classified into two parts: small displacement tests
(displacement, d=0.8mm and motor speed, ms=O to
270rpm) as well as the large displacement (d=9.5mm and
ms=O to 90rpm). Tests B001 and B002 show the same
heating and flow conditions (heat flux, q"=67.6KW/m2 and
mass flux, G=499.96Kg/m2s) but different vibration
conditions (d=0.8mm and d=9.5mm). Similarly, tests
B003 and B004 show the same heating and flow conditions
(q"=96.5KW/m2 and G=499.96Kg/m2s) but different
vibration conditions (d=0.8mm and d=9.5mm). The
reason for lower frequency operation with higher
displacement as well as higher frequency operation with
smaller displacement is to prevent the glass tubes from
breaking. Table 5 shows the average acceleration of these
tests under different operation motor speed (ms) conditions.
In this study, the vibration acceleration is ranged from 0 to
0.653g (about 6.4 1in -), which already reaches the large
earthquake condition shown in Table 2 (intensity scale =X).

Table 4: Test matrix of the present study
ATsub Heat Flux q" Mass Flux G Displacement Motor Speed, ms
Test #
(C) (kW/m2) (Kg/m2s) d (mm) (RPM)
B001 5.9 67.6 499.96 +0.8 0, 90, 180, 270
B002 6.4 67.6 499.96 +9.5 0, 30, 90
B003 5.9 96.5 499.96 +9.5 0, 30, 90
B004 5.5 96.5 499.96 +0.8 0, 90, 180, 270

Table 5: Average acceleration (a) of each test
T st RP 30 90 180 270
B001 0.142g 0.275g 0.653g
B002 0.074g 0.189g
B003 0.074g 0.198g
B004 0.114g 0.278g 0.518g

In order to catch both the transient and steady state
phenomena, the experimental process can be separated into
three parts: (1) steady state in stationary condition (without
vibration), (2) start-up vibration transient, and (3) stop
vibration transient. In the beginning, the system was heated
up in a stationary condition. After the system reach the
steady state condition with desired inlet subcooling
condition, the first step can be finished by taking data for


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

the steady state condition for 5 seconds. The vibration
module then was turned on and started vibration with
specific rotation speed and displacement, and this is the
start-up vibration transient process. After the 65-second
vibration condition, the vibration module was then turned
off and kept recording data for 50 seconds, and this is the
third step for stop vibration transient. The two-phase flow
condition should recover in this period.
Before starting the experiments, some tests should be
carried out to verify the background noise and the flow
induced vibration in this facility. Background noise may
include the electrical noise and the background vibration,
and it can be tested with the completely shut-down system,
i.e. a stationary condition without any flow inside the
annulus test section. In addition, the flow induced vibration
means the system tremor of the boiling experiment with a
stationary condition, and this value can be record in the
experimental step (1): steady state in stationary condition as
mentioned above. The equivalent acceleration values of
background noise and the flow induced vibration are about
0.02g and 0.024 to 0.027g, respectively, and these values
can be considered the background error of acceleration
measurement.


Acceleration Signals and Frequency Analysis

As mentioned in the last paragraph, the experimental
process includes three steps: (1) steady state in stationary
condition, (2) start-up vibration transient, and (3) stop
vibration transient. Figures 5 and 6 show the transient
acceleration signals with time. The first five seconds in Fig.
5 show the steady state results; the later 55 seconds in Fig. 5
and the first 10 seconds in Fig. 6 show the start-up and
vibration transient; and the last 50 seconds in Fig. 6 show
the stop vibration transient results. As shown in Fig. 5 and
Fig. 6, the acceleration values are quite different with or
without vibration. For each vibration test, the only
difference of the signal plot is the magnitude of the
acceleration, and the vibration acceleration for each test has
been shown in Table 5.


u 0.5
o
S0.0

o -0.5


.~.1......


-1.0 ll

-1.5
0 10 20 30 40 50 60
Time, t [s]
Figure 5: Acceleration signals of steady state to start-up
vibration transient






Paper No


1.0


a 0.5

S0.0

S-0.5


! IF'1'7 1 j1T TI 1 W f'T I


-1.0 In "1


0 10 20 30 40 50 60
Time, t [s]
Figure 6: Acceleration signals of vibration to stop vibration
transient

With the acceleration magnitude, the vibration
frequency is also important information which can be
extracted from the transient acceleration signals. Although
the rotation frequency of the eccentric cam can be roughly
determined by the motor speed (ms, in RPM), the actual
vibration waves may contain some different modes due to a
complicated system motion. In order to analyze the real
vibration behaviour of the system, the FFT method (Fast
Fourier Transform method) is introduced in this study for
the acceleration signals. By performing the FFT, each
vibration mode hidden in the acceleration signals can be
shown. Figure 7 and Figure 8 shows the example figures of
stationary FFT and vibration FFT results of the experiments
B001. In stationary tests, there is no obvious dominant
frequency, and the amplitudes of the signals are pretty small,
which means these signals should come from the random
noise of the wire connections or background interference.
On the other hand, the dominant frequencies are quire
distinguished in vibration conditions not only in the peak
shapes but also in the magnitudes. In Figure 8, the operation
motor speed (ms) is 270RPM, and the 4.5Hz peak should
come from the motor rotation directly. Some other higher
modes may be the induced vibration of the system or some
periodical dents on the contact surfaces of cam wheel. Table
6 summaries the vibration frequencies determined from the
acceleration signals of each experiment by FFT method.
Most of these frequencies are based on the motor speed (ms)
and its higher modes. Interestingly, some higher frequencies
(>10Hz) can be found in B001 and B004 in 270RPM
vibration tests. Although in B002 and B003 90RPM tests a
10.5Hz peak was found, the amplitude of this peak is about
5 to 10 times smaller than those peaks found in the B001
and B004.

Table 6: Vibration frequencies (f) analyzed by FFT
T RPM) 30 90 180 270
Test #
B001 and B004 1.5Hz 3, 4.5, 4.5, 9,
(d=-0.8mm) 6Hz 13.5, 18Hz
B002 and B003 0.5, 5.5, 1.5, 3,4.5, 7.5,
(d=-9.5mm) 6, 6.5Hz 9, 10.5Hz


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

0.010 ..


0.008


S0.006

".
E 0.004


0.002


0.000
0 5 10 15 20
Frequency, f[Hz]
Figure 7: FFT frequency analysis of stationary signals


0.10


0.08


S0.06


0 0.04


0.02


000


0 5 10 15 20
Frequency,f [Hz]
Figure 8: FFT frequency analysis of vibration signals
(ms=270RPM, d=0.8mm)


Experimental Results and Discussion

A series of vibration tests were performed after setting
the subcooled boiling conditions to the steady state in the
annulus test section according to the properties in the test
matrix. Figure 9 and Figure 10 show the transient variation
of void fraction in Port 6 (z=2.032m) of the annulus test
section for B001 and B004 tests. When the vibration module
was operating in higher motor speed (ms=270RPM), the
void fraction in the test section drops down for these two
experiments. In Figure 9, a clear decreasing trend of void
fraction can be seen in B001 (q"=67.6kW/m2, d=0.8mm,
ms=270RPM); in Figure 10, the variation amplitude of void
fraction is much higher than that in Fig. 9 due to a higher
heat flux in B004 (q"=96.5kW/m2, d=+0.8mm,
ms=270RPM), and the averaged void fraction shows a
decreasing trend.


""AwILU





Paper No


0.3



0.2


7c
0
" 0.1


0 10 20 30 40 50 60
Time, t [s]
Figure 9: Transient variation of void fraction in Port 6 of
the annulus test section for B001 test (q"=67.6kW/m2,
d=0.8mm, ms=270RPM)


0 U.4

. 0.3

S0.2
>S


r


I.


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

d=0.8mm, ms=90, 180 and 270RPM). The void fraction
slightly increased for ms=180RPM conditions, and again a
decreasing trend of void fraction happen in ms=270RPM
condition.


g


0.00UU
ss. tr. vib. vib. tr. sp.
Time Period, [-]
Figure 11: Averaged void fraction variations (Port 6)
against experimental time steps (B001 tests, q"=67.6kW/m2,
d=0.8mm, ms=90, 180, 270RPM)


0.00
ss. tr. vib. vib. tr. sp.
Time Period, [-]
Figure 12: Averaged void fraction variations (Port 6)
against experimental time steps (B002 tests, q"=67.6kW/m2,
d=9.5mm, ms=30, 90RPM)


0.3


-
S0.2



S0.1



0.0


ss. tr. vib. vib. tr.
Time Period, [-]


Figure 13: Averaged void fraction variations (Port 6)
against experimental time steps (B003 tests, q"=96.5kW/m2,
d=9.5mm, ms=30, 90RPM)


0.05

0.04
i
0.03

- 0.02
0


0.1

<"\ -\


.u
0


10 20 30 40 50 60


Time, t [s]
Figure 10: Transient variation of void fraction in Port 6 of
the annulus test section for B004 test (q"=96.5kW/m2,
d=0.8mm, ms=270RPM)

By taking the average values of each experimental time
periods of test process as discussed in preceding paragraphs,
i.e. (1) steady state in a stationary condition, (2) start-up
vibration transient and (3) stop vibration transient, more
clear trends of void fraction variations in each experiment
can be seen. Figures 11 to 14 show the averaged void
fraction variations in Port 6 (z=2.032m) against
experimental time steps from steady state to vibration/stop
transients. Where ss., tr., vib. and sp. mean the stead state
stationary, transition, vibration and stop vibration conditions,
respectively. Figure 11 show the void fraction variations of
B001 (q"=67.6kW/m2) with vibration conditions including
d=+0.8mm, ms=90, 180 and 270RPM. The void fraction
almost keeps a constant for ms=90 and 180RPM conditions;
however, a drastic drop of void fraction happened in
ms=270RPM condition. Figure 12 and Fig. 13 show the
void fraction results of B002 (q"=67.6kW/m2, d=9.5mm,
ms=30 and 90RPM) and B003 (q"=96.5kW/m2, d=9.5mm,
ms=30 and 90RPM), and the void fraction in both tests
shows an increase trend in ms=90RPM conditions. Figure
14 shows the void fraction results of B004 (q"=96.5kW/m2,


*





--ms=90RPM
* ms=180RPM
ms-270RPM


.--'
------*----





--ms-30RPM
S--- ms-90RPM


:-

-- ms=30RPM
-- ms-90RPM


0.05

0.04
a
0.03

- 0.02

0.01


U I -


L .L IN I J .1 ,L Ilk.I,






Paper No


,-, '"---- ^s-----*-----*
0.3


7
0.2


u-
o 0.1
-ms=90RPM
-*-ms-180RPM
ms=270RPM
0.0
ss. tr. vib. vib. tr. sp.
Time Period, [-]
Figure 14: Averaged void fraction variations (Port 6)
against experimental time steps (B004 tests, q"=96.5kW/m2,
d=0.8mm, ms=90, 180, 270RPM)

The reasons for the increase and decrease of void
fraction mentioned above may be due to the following
reasons. In the lower frequency vibrations (low motor speed,
e.g. 30 and 90RPM), if the displacement is small (e.g.
d=+0.8mm), the acceleration is also small and there is
almost no effect on the void fraction distribution in the
annulus test section; however if the vibration displacement
is large (d=9.5mm), the acceleration becomes higher and
the vibration inertia results in a better heat transfer rate for
heater rod surface and a higher evaporation rate, and
therefore an increasing trend of void fraction can be seen in
Fig. 12 and Fig. 13. A similar situation happens in a higher
motor speed (ms=180RPM) with a small displacement (e.g.
d=+0.8mm), the void fraction increased slightly due to the
enhancement of heat transfer on the surface. On the other
hand, if the vibration frequency keep going up to the highest
motor speed (ms=270RPM, d=0.8mm), the void fraction
drops down drastically due to the disturbance of the thermal
boundary layer as well as the coalescence of vapor bubbles.
Actually, the acceleration is about 0.6g, and it almost
reaches the large earthquake condition. The acceleration is
much higher than those of the other tests, and the mixing of
the subcooled and saturated fluids will dominate the bubble
nucleation and growth rates. In addition, bubbles may
collapse due to vibration induced turbulence in the working
fluid. According to the test results done by Nariai et al. in
1994, the void fraction may drop down drastically if the
oscillation frequency of the heater rod is higher than 10Hz.
From the vibration frequency analysis by FFT shown in
Table 6, the resultant frequencies in vibration condition
ms=270RPM are 4.5, 9, 13.5 and 18Hz. Here two of the
vibration frequencies are higher than 10Hz, and this may be
related to some experimental similarities between the
present study and the Nariai's tests in 1994. In addition,
Hibiki and Ishii (1998) performed the two-phase flow
induced vibration experiments in a circular pipe, and a 10 %
reduction of void fraction can be seen in the adiabatic
two-phase flow. The frequencies of flow induced vibration
are usually shown in the similar region (<20Hz). Therefore,
both of the above literatures can support the reduction of
void fraction in high frequency vibration conditions of this
study, and this is one of the vibration effects on two-phase
flow.


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

Figure 15 and Figure 16 show the photos of the annulus
test section with stationary and vibration (d=0.8mm,
ms=270RPM) conditions for B001 (q"=67.6kW/m2) and
B004 (q"=96.5kW/m2) taken in between Port 4 and Port 5.
In Figure 15 (B001), the annulus test section was operated
with lower heat flux, and obviously the bubbles were almost
spherical. For the high motor speed vibration condition, a
small reduction in bubble size and number was found when
compared to the stationary condition. However in Figure 16
(B004), the annulus test section was loaded with higher heat
flux, and the bubbles were distorted and had more
interactions with one another. When the high motor speed
vibrations were applied to the test section, violent
interactions among bubbles as well as between liquid and
bubbles were found. Much more dispersed and distorted
bubbles were shown due to the high frequency vibration
effects. These photos show phenomenal evidence of the
void fraction reduction shown in Fig. 11 and Fig. 14 for
high motor speed (ms=270RPM) conditions.


Figure 15: Photos of the BOO1 test (q"=67.6kW/mZ) in
(a) stationary and (b) vibration condition (d=+0.8mm,
ms=270RPM)


Figure 16: Photos of the B004 test (q"=96.5kW/m2) in
(a) stationary and (b) vibration condition (d=0.8mm,
ms=270RPM)

In this study, the impedance void meters were utilized to
measure the area-averaged void fraction, and the global
variation trends of void fraction were acquired. However,
the distribution of local bubble sizes, locations and shapes
can only be recorded by the camera. In order to get more
detailed information about the two-phase flow under
seismic vibration conditions, a local measurement technique
by conductivity probes is required for the future verification
and identification of the vibration effects on two-phase
distribution and two-phase flow parameters.
Based on the current results, both the vibration
frequency and acceleration of the experiments in Fig. 15
and Fig. 16 were the highest, and these two effects may be
coupled with each other for the two-phase flow. Therefore,
more experiments are required to separate the vibration
effects of frequency and acceleration on the subcooled
boiling two-phase flow in an annulus.





7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


An annulus test section with vibration module was
designed, built and operated to simulate the seismic
vibration effects on subcooled boiling two-phase flow. The
geometries of the annulus test section were scaled down
from a prototypic BWR, and the earthquake vibration
properties including acceleration and frequency were taken
into account for this study. Some important results and
suggestions for future work are summarized as follows:
(1) The earthquake magnitude and intensity scale are
shown in Table 1 and Table 2. Based on these scales,
Table 4 shows the average acceleration values of each
vibration experiments, and the maximum values about
the ruinous and disastrous earthquake vibration
accelerations (0.5-0.6g) have been reached in this study.
(2) With the FFT method, the various modes of vibration
frequencies of each test can be revealed and are listed in
Table 6. The vibration frequencies can reach up to 13.5
and 18 Hz in tests B001 and B004 with motor speed
ms= 270RPM and displacement d=0.8mm conditions.
(3) Compared with the steady state stationary (no vibration)
conditions, the small displacement with low motor
speed vibration (d=0.8mm, ms=90RPM) conditions
result in almost no effect on void fraction.
(4) For large displacement vibration tests (d=9.5mm),
higher motor speed conditions (ms=90RPM) result in
an increasing trend on void fraction in both tests B002
(q"=67.6kW/m2) and B003 (q"=96.5kW/m2). The large
displacement may result in a heat transfer enhancement
and a higher void fraction condition.
(5) In the highest motor speed vibration experiments
(d=0.8mm, ms=270RPM), the void fraction results
show a decrease trend in both tests B001
(q"=67.6kW/m2) and B004 (q"=96.5kW/m2). This may
result from the disturbance of the thermal boundary
layer between subcooled and saturated fluids by high
liquid inertia (acceleration) or vibration induced
turbulence as well as the increase of coalescence rate of
bubbles by a high frequency vibration. Both Nariai's
experiments (1994) and Hibiki-Ishii's flow induced
vibration experiments (1998) support the reduction of
void fraction in this study.
(6) In order to verify and identify the seismic vibration
effects on two-phase flow in detail including the local
flow characteristics, phase distribution, bubble sizes,
locations, shapes, interface velocity and interfacial area
concentration, the local measurement experiments with
conductivity probes are suggested as the future work for
this study.
(7) In this study, both the acceleration and frequency are
the highest in the void fraction reduction condition.
More experiments are required to separate the effects of
these two variables.


Acknowledgements

This work was support by Tokyo Electric Power Company
(TEPCO), Japan. The authors would like to express their
sincere appreciation to TEPCO.


Bolt, Bruce A., Earthquake Intensity, (1993),
http://www.eas.slu.edu/Earthquake_ Center/mercalli.html

Earthquake Lab, Introduction to Earthquake Waves, (2'" I4),
lii p :. v cq cc. edu.tw /-ccusgm/ seismography.htm

Hibiki, T., and Ishii, M., Effect of flow-induced vibration on
local flow parameters of two-phase flow, Nuclear
Engineering and Design, 185, 113-125 (1998)

Ishii, M., and Kataoka, I., Scaling laws for
thermal-hydraulic system under single phase and two-phase
natural circulation, Nuclear Engineering and Design, 81,
411-425 (1984)

Ishii, M., and Zuber, N., Drag coefficient and relative
velocity in bubbly, droplet or particulate flows, AIChE
Journal, 25, 843-855 (1979)

Japan Business Press, August 21, (1993)

Kawamura, S., Koshi, Y, Hattori, K., Katayama, H.,
Fujimoto, S., and Kudo, Y., A study on neutron flux
transient in oscillating fuel assembly, 1996 Annual Spring
Meeting of Atomic Energy Society of Japan, No. A44,
Osaka Univ., Japan (1996)

Kawamura, S., Orii, A., Karasawa, H., Nishida, K., and
Soneda, H., Effect of horizontal excitation on bubble
behavior at subcooled temperature, 1996 Annual Spring
Meeting of Atomic Energy Society of Japan, No. C49,
Osaka Univ., Japan (1996)

Kocamustafaogullari, G., and Ishii, M., Scaling criteria for
two-phase flow loop and their application to conceptual 2 x
4 simulation loop design, Nuclear Technology, 65, 146-160
(1984)

Ministry of Energy, Mines and Petroleum Resources,
How big are earthquakes?, (2006)
ii p ..v.' cimpr gdov.bc.ca/Mining/GeolSurv/Surficial/quak
e/eq2.htm

Nariai H., and Tanaka, T., Void Fraction of Subcooled Flow
Boiling around Oscillating Heater Rod, 1994 Annual Spring
Meeting of Atomic Energy Society of Japan, No. J36, Univ.
of Tsukuba, Japan (1994)

National Science and Technology Center for Disaster
Reduction, (2'" i4)
http://ncdr.nat.gov.tw/chinese/result/naphmthesis/912625Z
94002.pdf

Shioyama, T., and Ohtomi, K., Pressure Fluctuation and
Behavior of Vapor Bubbles for a One-Component
Two-Phase Flow in a Vertical Tube, Induced by longitudinal
Excitation, H0977AA, Vol. 99 (1990)

Situ, R., Mi, Y, Ishii, M., and Mori, M., Photographic study
of bubble behaviors in forced convection subcooled boiling,
International Journal of Heat and Mass Transfer, 47,


Paper No


Conclusions


References






Paper No 7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

3659-3667 ,2'" I 4)

Situ, R., Hibiki, T., Sun, X., Mi, Y., and Ishii, M., Flow
structure of subcooled boiling flow in an internally heated
annulus, International Journal of Heat and Mass Transfer, 47,
5351-5364 2,, 14)

Situ, R., Hibiki, T., Sun, X., Mi, Y., and Ishii, M., Axial
development of subcooled boiling flow in an internally
heated annulus, Experiments in Fluids, 37, 589-603 2" i4)

Situ, R., Hibiki, T., Ishii, M. and Mori, M., Bubble lift-off
size in forced convective subcooled boiling flow,
International Journal of Heat and Mass Transfer, 48,
5536-5548 (2005)




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs