close

Вход

Забыли?

вход по аккаунту

?

ICONE25-67931

код для вставкиСкачать
Proceedings of the 2017 25th International Conference on Nuclear Engineering
ICONE25
July 2-6, 2017, Shanghai, China
ICONE25-67931
BEHAVIOR OF SIMULATED MOLTEN METAL
AT CHANNEL MODELED BOILING WATER REACTORS
Yutaro Hihara, Kota Matsuura, Hideaki Monji, Yutaka Abe, Akiko Kaneko
University of Tsukuba
Tsukuba, Ibaraki305-8573, Japan
s1720965@s.tsukuba.ac.jp
Hiroyuki Yoshida, Susumu Yamashita
Japan Atomic Energy Agency
2-4, Shirakata, Tokai, Naka, Ibaraki319-1195, Japan
structure of the lower part in the pressure vessel of a BWR as
investigated in the study. The data of the study is provided to
examine the numerical simulation code.
The objective flow is a molten flow of a control rod. In
order to do an experiment to observe and measure the flow in
detail, a simulation experiment was done with water as a
simulated molten material and a transparent acrylic flow
channel. The flow channel as a test section in the experiment
was designed modelling the structure around the control rod.
In the experiment, a high-speed video camera, LIF and a
laser displacement meter were used to measure the behavior and
the thickness of the liquid film. The behavior of a liquid film
was observed by an imaging technic. Furthermore, the liquid
film thickness and its velocity were obtained and compared with
the numerical simulation results.
ABSTRACT
When a severe accident occurs, decommissioning work
becomes important task. In the decommissioning work after
the severe accident, establishing the way to estimate the
sedimentation place of molten debris is important. However,
the technique to estimate exactly sedimentation place has not
been enough. Therefore, the detailed and phenomenological
numerical simulation code named JUPITER for predicting the
molten core behavior is under development. The comparison
between experimental and numerical results is necessary to
clarify the validity of the numerical analysis code. This study
provides the experimental data for a BWR to examine the
numerical simulation code in order to contribute to progress of
the decommissioning work.
INTRODUCTION
In decommissioning work of a nuclear power plant after a
severe accident, it is important to identify the sedimentation
place of molten materials. The identification of the relocation
course of the molten materials in the nuclear reactor structure is
necessary to grasp the sedimentation place, but there are few
knowledges about the relocation behavior of the molten material
flow in the structure of the nuclear reactor. However, it is
difficult to investigate experimentally the relocation behavior of
the molten material flow in the nuclear plant. Therefore,
numerical codes which can simulate the molten material flow in
the nuclear plant are desired. The numerical simulation cord
"JUPITOR" is one of such numerical codes simulating the
molten material flow in the nuclear plant (1, 2).
The numerical simulation code is useful to estimate the
molten material flow in the nuclear power plant but its ability
should be examined comparing with experimental results. In
order to obtain the experimental results compared with the
results by numerical simulation code, a flow in a modelled
EXPERIMENT
Experimental Apparatus
The flow channel in the experimental apparatus has a modeled
and simple structure of the lower part of the pressure vessel.
Figure 1 shows images of a target part of the lower structure in
the pressure vessel, which contains a fuel aggregate, a control
rod, a fuel support part and a control rod speed limiter. In the
1
Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/25/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig.1 Test section
Fig. 3 Measurement method and position
overflow tank from the upper tank. The part of the supplied
liquid flows to the lower tank while the liquid level at the
overflow tank is kept at constant. Therefore, the inlet condition
of the test section can be supposed as a constant pressure.
The size of the narrow channel in the test section is 100 mm ×
6 mm × 860 mm and the test section is made of transparent
acrylic material for flow visualization and image processing.
The angle of the slope part is 30 degrees from vertically.
Experimental Conditions
In the experiment, the working fluid was water. Water was
used as the working fluid. In this experiment, the effect of heat
is not considered, only the behavior of the fluid is the
measurement object of this experiment. Therefore, water is
suitable for visualization and easy to treat in the experiment.
The experiment was done under the atmospheric pressure and
the water temperature was in the range from 17 °C to 20 °C.
The flow was unsteady in the experiment, because the water
begin to flow after opening the shutter value. The flow,
however, becomes to be steady after the tip of the water flow
reached to the end of the flow channel. In the experiment the
measurement was done for a steady state.
Fig.2 Experimental apparatus
study, the molten material flow is focused. Therefore, a test
section as the modelled flow channel consists of a narrow part,
an extended part and a successive slop part as shown in the right
figure of Fig. 1. The narrow part in the test section is a model
of the flow channel between the fuel aggregate and the control
rod. The extended part is a model of the fuel support part, the
slope part corresponds to the control rod speed limiter.
Figure 2 shows the experimental apparatus. It consists of an
upper tank, an overflow tank, the test section, a test tank. There
is a shutter valve between the overflow tank and the test section.
After opening the shutter valve, the liquid in the overflow tank
flows down into the test section. The liquid is supplied for the
Measurement Methods
2
Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/25/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
(a) Expand part
(b) Slope part
Fig. 5 Flow images at test section
Fig. 4 Setup of equipment for LIF and PIV
liquid film increases at the both edges of the flow channel, based
on the flow image.
In the experiment, the flow state was observed and measured
by using optical techniques. The thickness of the liquid film at
the slope part was measured by a laser displacement meter.
Figure 3 shows the setup of the equipment and the measurement
points on the centerline of the flow channel. The thickness was
taken at every 20 ms and the spatial resolution was 0.1µm.
The thickness of the liquid film was also measured by LIF
(laser induced fluorescence) as shown in Fig. 4. A laser sheet
was irradiated from the back side of the channel and not affected
by the shape of the liquid film surface. When LIF was used,
Rhodamine B was mixed in the water as a fluorescent dye. The
sampling frequency of the liquid film image was 80 Hz.
By suing the same setup of LIF, the flow velocity in the
liquid film was measured by PIV (Particle image velocimetry).
The florescent tracer particle was mixed in the water for PIV. The
tracer particle was of about 10 m diameter and its specific
gravity was 1.5. By using an optical filter in front of the video
camera, a reflection of the laser light on the flow channel wall
can be cut and only the tracer particle image was clearly
obtained.
Thickness of Liquid Film
Figure 6 shows the distribution of the liquid film thickness along
the flow channel at the slope part. The abscissa of the figure is
the distance along the slope from the upper end of the slope.
The liquid film thickness was a time average of thickness
measured by the laser displacement meter and its measurement
points were on the centerline of the flow channel as shown in
Fig. 3.
The liquid film thickness changes with time due to the
surface wave as shown in Fig. 5. An error bar shows intensity
of a liquid thickness fluctuation and its one side length is 2
where  is the standard deviation of the data. Based on Fig. 6,
the liquid film thickness was almost constant along the slope and
about 4 mm and its fluctuation intensity is also almost same.
RESULTS
Flow Images
Figure 5 shows the flow images at the steady state flow.
Figure 5 (a) shows the flow around the expand part. Before the
expand part, the narrow channel was filled by the water. At the
expand part, there is a liquid film on a vertical wall of the flow
channel. Surface waves was observed on the liquid film.
Surface waves seems to have two-dimensional structure.
Figure 5 (b) shows the liquid film flow at the slope part. The
liquid film also seems to have two-dimensional surface wave.
In both of the expand and the slope parts, the thickness of the
Fig. 6 Thickness for main stream direction
3
Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/25/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 7 Time change of the thickness
Figure 7 shows that the liquid film thickness fluctuated with time
at the center of the slope part.
The liquid film thickness was also measured by LIF.
Figures 8 and 9 shows the results by LIF on the centerline of the
channel at the expand part and the slope part, respectively.
Figure 8 (a) shows a binary image based on the flow image by
LIF.
The flow image by LIF is a liquid phase distribution
because the liquid contains the fluorescent dye. Originally, the
image by LIF is gray level corresponding to the brightness of the
point on the image but the phase is liquid or air. Therefore, the
banalization of image corresponding to the liquid or the air was
done to decide the liquid phase distribution. Based on the
liquid phase distribution, the thickness of the liquid film was
measured.
Figure 8 (b) shows the distribution of the liquid film
thickness. The axis of ordinate in the figure is the height of the
position where the origin of the height is at the bottom end of the
vertical channel. The film thickness decreases along the
channel from the exit of the narrow channel. The broken line
shows the estimated value based on the empirical equation
shown by Eq. (1) , where =0° corresponding to be vertical.
Eq. (1) gives the thickness of the developed liquid film. In the
experiment, the flow is under developing, because the flow
length is not enough. Therefore, the liquid film thickness will
be small downstream if the flow channel is long enough.
Figure 9 also shows (a) LIF image and (b) liquid film
thickness at the slope part. The origin of the location is the exit
of the flow channel and the axis of ordinate in the figure is the
height of the position. The liquid film thickness decreases
along the slope. The broken line also shows the estimated
value based on the modified Brauer’s equation considering the
inclination of the plane, shown by Eq. (1), where =30°. In this
region, also, the thickness by Eq. (1) is larger than the
experimental results. At the slope part, the film thickness was
about 4 mm. The measurement result by the laser displacement
meter was also 4 mm as shown Fig. 6. This fact supports the
correction of the LIF result.
(b) LIF image
(b) liquid film thickness
Fig. 8 Liquid film thickness at the expand part
(a) LIF image
(b) liquid film thickness
Fig.9 Liquid film thickness at the slop part
 = 0.302 (
3 2
 cos 
1
3
8
(1)
) Re15
4
(2)
Re = 
Where  denotes the liquid film thickness and g, the gravitational
accretion, , the viscos coefficient of the liquid. The Reynolds
number, Re, is given by Eq. (2), where Q,  and L denote the
flow rate, the density of the liquid and the wetted perimeter,
respectively.
4
Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/25/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Fig. 10 velocity distribution along the plate
(a) Liquid film by LIF
(b) Numerical simulation
Fig. 12 Flow at the expand part
Fig11 Flow channel and height
Liquid Film Velocity
Figure 10 shows the velocity distribution of the liquid film.
The ordinate is the height of the measurement point and the
origin is the exit of the flow channel as shown in Fig. 11. In
Fig. 10, the location of 200 mm corresponds to the connection
between the expand part and the slope.
The special resolution of PIV was not enough to get the
velocity distribution in the liquid film. Therefore, the time
average of the depthwise average velocity was obtained and
plotted in Fig. 10. The error bar denotes the fluctuation
intensity of the velocity, 2, where  is the standard deviation.
The velocity change was not detected along the flow direction.
(b) Liquid film by LIF
(b) Numerical simulation
Fig. 13 Flow at the slope part
result has different tendency from the result of the experiment in
some points. The velocity of the simulation does not match that
of the experiment. The velocity is constant at any part of the
test section by the result of experiment. On the other hand, the
velocity decreases at slope part by the result of the numerical
simulation. Many droplets, however, can be seen in the
numerical simulation results while such many droplets were not
visualized in images by LIF. Because the detection of droplets
Comparison with Numerical Simulation
The numerical calculation was done by JUPITER under the
same condition for the same flow channel. Figure 12 shows the
result at the expand part, and Fig. 13, at the slope part. The
numerical simulation shows the similar flow pattern. Figure 14
shows the liquid film velocity of the numerical simulation. The
5
Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/25/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
simulation. The flow state by the numerical simulation was
different from the experimental result.
Especially, may
droplets was generated in the numerical calculation but they
were not observed in the experiment. At the point, there is
difference between the experiment and the numerical
calculation. Therefore, further study will be done to exam both
results.
REFERENCES
[1] Matsuura, K. Monji, H. et al., 「 Experimental and
Numerical Simulation on Hydraulic Behavior of Molten
Flow in the Lower Part of Reactor Core 」 , The 24th
International Conference on Nuclear Engineering
(ICONE24), 2016.
[2] Matsuura, K. Monji, H. et al., 「Experimental Study on
Interface Behavior of Simulated Molten Fluid in Relocation
Process 」 , 10th Japan-Korea Symposium on Nuclear
Thermal Hydraulics and Safety (NTHAS10), 2016
[3] Yamashita, S. et al., Development of Numerical Simulation
Method for Melt Relocation Behavior in Nuclear Reactors:
Analysis of Relocation Behavior for Molten Materials with a
Simulated Decay Heat Model, Proceedings of 22nd
International Conference on Nuclear Engineering (ICONE
22), Prague, Czech Republic, 2014, ICONE22-30972
[4] Yamashita, S. et al., Development of Numerical Simulation
Method for Melt Relocation Behavior in Nuclear Reactors:
Relocation Behavior in a Simplified Core Structures,
Proceedings of 23rd International Conference on Nuclear
Engineering (ICONE 23), Chiba, Japan, 2015, ICONE231581
[5] Yamashita, S. et al., Development of Numerical Simulation
Method for Melt Relocation Behavior in Nuclear Reactors:
Validation of Applicability for Actual Core Support
Structures, Proceedings of 24th International Conference on
Nuclear Engineering (ICONE 24), Charlotte, North Carolina,
USA, 2016, ICONE24-60453
Fig. 14 Liquid film velocity of the numerical analysis.
by LIF is difficult, we cannot conclude the error of the numerical
simulation. It need that
further research of flow by both numerical simulation and
experiment.
CONCLUSION
The experiment was done using water in the flow cannel
modeling the structure around the control rod. The flow
channel consists of the narrow vertical channel, the expand part
and the slope. The behavior and the thickness of the liquid film
were mainly measured by image processing in the expand and
the slope parts.
The liquid film flow was two-dimensional wavy flow and
its thickness decreased along the flow channel but was larger
than that the estimated value by the empirical equation. The
liquid flow velocity was also obtained but the detailed
distribution along the depthwise direction was not clear because
the liquid film was thin. The velocity of the liquid film are
almost same between the experiment and the numerical
6
Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/25/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Документ
Категория
Без категории
Просмотров
0
Размер файла
925 Кб
Теги
icone25, 67931
1/--страниц
Пожаловаться на содержимое документа