Numerical investigation of Fluid – Structure interaction (FSI) on sodium leakage accident at prototype fast breeder reactor monju

Nuclear Science and Technology, Vol.10, No. 2 (2020), pp. 15-25  
Numerical Investigation of Fluid – Structure Interaction (FSI) on  
Sodium Leakage Accident at Prototype Fast Breeder Reactor Monju  
Hoang Tan Hung, Bui Thi Hoa, Hoang Minh Giang  
Institute for Nuclear Science and Technology, 179 Hoang Quoc Viet, Nghia Do, Cau Giay, Ha Noi  
(Received 28 February 2020, accepted 21 April 2020)  
Abstract: This study presents a numerical investigation of fluid-structure interaction about sodium  
leakage accident at prototype fast breeder reactor Monju. The dynamic forces of sodium liquid flow  
acting on the thermowell are the main cause of fatigue failure, which leads to the occurrence of  
thermowell cracking due to intense stresses and high fatigue-cycle. Since the location and magnitude  
of these stresses are unknown, an analysis of the vibrational characteristics and stresses caused by FSI  
was performed by using ANSYS to prevent similar accidents in the future. The most dangerous case  
was found for FSI analysis by comparisons between the natural frequency of thermowell and vortex  
shedding frequency at several operating conditions. The results showed that a stress concentration  
location is similar to the location of broken thermowell, and the amplitude of stress is large enough for  
fatigue damage.  
Keywords: Monju, Sodium leakage accident, FSI, ANSYS.  
was to study the interaction between water and  
circular cylinder structure under the  
operational conditions of the Monju reactor.  
I. INTRODUCTION  
Monju is a fast breeder reactor (FBR)  
that has been developed in Japan for a long –  
term research, development, and utilization  
program in nuclear energy. On 8th December  
1995, the reactor suffered a serious accident  
with sodium leaked through a temperature  
sensor due to a break of thermowell inside a  
secondary pipe resulting from intense vibration.  
A study on sodium leakage at Monju was  
performed in [1] to investigate the causes and  
consequences of the accident. From the results,  
it was concluded that the breakage of the  
thermowell was caused by high cycle fatigue  
of the well tube tip due to Vortex-Induced  
Vibration (VIV) phenomenon.  
On the other hand, to prevent the sodium  
leakage accident, the improvement of  
thermowell design was obtained by means of  
an analytical study using the finite element  
method (FEM) [3]. Various designs of  
thermowell were taken into account. The  
results showed that the design of thermowell  
with a continuous low-gradient slope and no  
neck could avoid the effects of VIV.  
It can be seen that, after the sodium  
leakage accident at the Monju reactor, many  
studies were performed to find the cause of the  
accident and also to prevent the accident from  
happening again. This prevention could be  
done by the thermowell design and its working  
condition improvement. However, the number  
of studies on thermowell stress concentration  
prediction under the operational condition still  
limited. Therefore, it is necessary to perform a  
In the study of Ref. [2], it was found that  
vortex-induced vibration depends on the  
specific range of reduced velocity. Therefore, a  
range of the reduced velocity was created using  
the experimental facility to avoid that  
phenomenon. The purpose of this experiment  
©2020 Vietnam Atomic Energy Society and Vietnam Atomic Energy Institute  
NUMERICAL INVESTIGATION OF FLUID – STRUCTURE INTERACTION (FSI)…  
study on the prediction of the stress  
concentration of thermowell. On the other hand,  
the Monju accident happened due to the VIV  
phenomenon as a consequence of FSI. One  
study on the consequence of FSI using ANSYS  
has done [4] to predict thermowell stress  
concentration on the steam temperature sensor.  
Base on this paper, our work focus on ANSYS  
simulation to predict the stress concentration of  
Monju’s thermowell.  
generating capacity of 280 MW. Liquid  
sodium is used in the primary and secondary  
cooling systems. An accident occurred due to  
the failure of a thermowell located in the  
secondary cooling system in the course of  
power  
thermowell was broken apart by high cycle  
fatigue failure. Figure shows the  
buildup  
test  
programs.  
The  
1
configuration of thermowell before and after  
the accident. The tip of the well-tube was  
lost and the thermocouple was bent  
approximately 45 degrees from the original  
position. Thermowell is a type of step-shank  
with its whole length of 185.5 mm and its  
thickness of 3 mm as shown in Figure 1. It is  
made of 304 stainless steel, which is 8000  
kg/m3 in density, Young modulus, E, is in the  
range of 193 – 200 GPa, and the Poisson  
ratio equal to 0.29 [5].  
In this study, following the study as  
mentioned in [4], an analysis of FSI between  
sodium liquid and Monju’s thermowell will be  
performed by using ANSYS. Two approaches  
of one-way FSI and two-way FSI are applied to  
investigate the VIV phenomenon to find the  
stress concentration location and the stress  
amplitude acting on Monju’s thermowell.  
II. EXPERIMENTAL FACILITY AND  
CALCULATING METHODS  
From the operation history of the  
secondary loop of the Monju fast breeder  
reactor [6] and thermodynamic and transport  
properties of sodium liquid [7], the  
characteristics of sodium liquid flow depend on  
the history of the secondary heat transport  
system operation is presented in Table I.  
A. Overview of Monju facility  
The Monju fast-breeder reactor is a  
sodium-cooled reactor operating at a thermal  
output of 714 MW and has an electrical  
Fig. 1. Thermowell configuration before and after the accident [8]  
Table I. Characteristics of sodium liquid flow in the secondary heat transport system  
Temperature  
(   
Velocity  
(m/s)  
5.20  
2.08  
2.08  
Dynamic viscosity  
Density (kg/m3)  
Reynold number  
(10-4 Pa.s)  
4.52  
Case 1  
Case 2  
Case 3  
Case 4  
200  
200  
450  
485  
903.0352  
903.0352  
846.2525  
838.1427  
1.04x105  
4.15x104  
6.92x104  
7.23x104  
4.52  
2.55  
2.41  
2.08  
16  
HOANG TAN HUNG et al.  
phenomenon could lead to large amplitude  
B. Calculation Method  
The VIV phenomenon is the main cause  
vibrations of structure, which in turn causes  
fatigue damage. However, this situation occurs  
in a range of frequencies, not a single value of  
frequency as the normal resonance. The  
following condition is given to avoid fatigue  
damage [10]:  
of damaging the thermowell of the Monju  
reactor. Vortices were formed and detached  
periodically from either side of cylindrical  
bodies causes an unsteady oscillating flow. An  
asymmetrical flow pattern develops around the  
body because of the periodic shedding and  
changes in the pressure distribution behind the  
body. Vortex shedding occurs at a certain  
frequency, which is called vortex shedding  
frequency (fS). This frequency depends on flow  
velocity U and the cylinder diameter D,  
Strouhal number (St) and formulated as [9]:  
fS < 0.4 fN  
(2)  
Because the Monju reactor was operated  
under several conditions, it is necessary to find  
the most dangerous case to perform FSI  
analysis by using the criteria of equation (2).  
Firstly, Modal analysis was applied to  
determine the natural frequency (fN) and the  
mode shape of the structure. Secondly, 2D  
flow analysis in Fluent was conducted to  
determine the vortex shedding frequency of  
different flow conditions. A comparison  
between the natural frequency of structure (fN)  
and vortex shedding frequency (fS) could show  
the most dangerous case to perform FSI  
analysis. Finally, the FSI calculation was  
performed to analyze the location and value of  
stress concentration on the thermowell tube.  
The value of stress concentration was  
compared with the endurance limit of material  
(SE). The diagram of the analysis process is  
shown in Figure 2.  
    
 
     
(1)  
 
As a result of the periodic change of the  
vortex shedding, the pressure distribution on  
the cylinder outer surface may also change  
periodically, thereby generating a periodic  
force on the cylinder surface. Each cylindrical  
body has its natural frequencies (fN). If the  
vortex shedding frequency (fS) is close to one  
of the natural frequencies (fN), resonance could  
have happened. Then, as a result, the  
cylindrical bodies start to vibrate with the same  
frequency as the vortex shedding frequency,  
called the lock-in phenomenon. This lock-in  
Fig. 2. Analysis process  
17  
NUMERICAL INVESTIGATION OF FLUID – STRUCTURE INTERACTION (FSI)…  
it may lead to the resonance which causes the  
failure of the structure. Therefore, it is an  
essential process to find out the natural  
frequencies and mode shapes of a structure to  
find the most dangerous case of flow operation  
of the Monju reactor.  
1. Modal analysis  
The vibrational characteristics of the  
structure were determined using Modal  
analysis. These characteristics which contain  
natural frequencies and mode shapes of the  
structure are important in designing a  
structure subjected to a dynamic load. The  
basic equation used in the Modal analysis for  
free, undamped vibration which is given [11]  
as followings:  
2. 2D Fluent analysis  
The 2D Fluent analysis was performed  
with a typical cross-section of thermowell.  
2D model of thermowell and mesh is showed  
in Figure 3. As it can be seen in Table I, all  
of the cases have Reynold numbers greater  
 
([ ]  
 
[ ]{  
}
{ }  
        =   
(3)  
 
 
[ ]  
[ ]  
Where   and   are the stiffness and  
than 104 so the standard k -   (k is the  
,
mass matrix;   is a mode shape vector  
turbulence kinetic energy (m2/s2) and   is  
the specific rate of dissipation (m2/s3))  
turbulence model for near-wall interactions  
model was chosen for simulation. By using a  
Fast Fourier Transformation (FFT) of  
velocity, the vortex shedding frequency of  
different flow conditions was found and  
compared with natural frequencies from  
Modal analysis.  
 
(Eigenvector) of mode i and   is a natural  
 
frequency of mode i.  
Dynamic response of structures could  
occur under the harmonically varying load,  
whose frequencies are used to determine  
whether the resonance occurs or not. If the  
frequency of vortex shedding (fs) is close to  
one of the natural frequencies (fi) of structures,  
D1 = 0.01 m = D  
D2 = 0.14 m = 14 D  
D3 = 0.16 m = 16 D  
D4 = 0.15 m = 15 D  
Fig. 3. Model of 2D Fluent analysis at the typical cross-section of thermowell  
the fluid and solid domain should be  
3. FSI analysis  
From the previous results, the most  
continuous. A set of coupling boundary  
conditions, therefore, is introduced. The  
dangerous case was selected for FSI analysis.  
An interface surface between the solid and  
the fluid domain always existed in FSI  
analysis. At this interface, both the governing  
equations and the boundary conditions from  
interface  
conditions  
of  
displacement  
continuity, Eq. 4, and traction equilibrium  
(momentum transfer), Eq. 5 must be satisfied  
along the fluid-structure boundary [9].  
18  
HOANG TAN HUNG et al.  
df = ds  
Pf = Ps  
(4)  
(5)  
data is transferred from Fluent solver to the  
Structure solver. In contrast, the two-way FSI  
approach uses both the traction equilibrium, Eq. 5,  
and the displacement continuity condition, Eq. 4.  
Through the two-way FSI process using both  
Eq. 4 and 5, displacement values are updated to  
the fluid mesh.  
Where df and ds are the deformations of  
fluid and structure domain; Pf and Ps are the  
pressure of the fluid and structure domain.  
The analysis in this study was divided  
into two simulations, one-way FSI and two-  
way FSI, to compare the structural response to  
static and dynamic load.  
C. Verification of the Modal and FSI  
calculation models  
To demonstrate further the validity of  
the Modal and FSI calculation models with  
oscillation problems, the oscillation of an  
exemplary vertical plate in a cavity filled  
with fluid was investigated. Figure 4 shows  
a schematic diagram of the oscillation  
problem [12].  
In the one-way FSI simulation, the focus  
is on the effect of sodium flow to the  
thermowell, so the pressure load data only  
transferred from Fluent solver to the Structure  
solver as mentioned in equation 5. However,  
the solid displacement is not transferred back  
to the Fluent solver. Therefore, pressure data  
taken from the Fluent solver is imported to the  
Structure solver.  
A thin plate is anchored to the bottom  
of a closed cavity filled with fluid. There is no  
friction between the plate and the side of the  
cavity. During the first 0.5 seconds, a  
uniformly distributed load of 30 N/m is  
applied to distort it. Once this load is released,  
the plate oscillates back and forth to regain its  
equilibrium, and the surrounding air damps  
this oscillation. The flexible plate has  
Young’s modulus of E = 2.5 MPa, a Poisson’s  
ratio of   = 0.35, and a density of 2550 kg/m3.  
The one-way FSI simulation is one-way  
data transport. Nevertheless, the two-way FSI  
simulation is more complicated because of  
their interaction between both solvers. The  
motion of thermowell simulation is solved in  
the Transient structure, while the motion of  
flow is solved in Fluent. The System Coupling  
controls the solution process as well as data  
transfers between the two analysis systems.  
Force data from the motion of flow is received  
by the Transient structure as mentioned in  
equation 5. After finishing the Structural  
solving process, the displacement data from the  
motion of thermowell is received by the Fluent  
as mentioned in equation 4 then the Fluent  
solves the fluid behavior over time. As a  
consequence, the mesh is deformed according  
to the response of the structure. Displacement  
values are interpolated to the fluid mesh which  
results in deformation of the fluid domain.  
The simulation with fluid density is   =  
 
1kg/m3, and three different dynamic viscosities  
of the fluid   = 0.2, 1.0, and 5.0 Pa.s are  
considered for the flow conditions.  
In summary, the one-way FSI approach  
Fig. 4. Geometry of oscillating plate  
uses only the traction equilibrium, Eq. 5, and the  
(not drawn to scale)  
19  
NUMERICAL INVESTIGATION OF FLUID – STRUCTURE INTERACTION (FSI)…  
In Modal analysis, a fixed support  
constraint is needed to hold the bottom of the  
thin plate in place and the set up to find the  
first natural frequency and mode shape of the  
plate. The result of Modal analysis showed that  
the value of the first natural frequency of the  
plate is 0.317Hz.  
exist slight discrepancies in the frequency and  
the amplitude of oscillation. A different mesh  
model could be the leading cause of these  
discrepancies. Sodium liquid is the fluid domain  
with specific viscosity and density in FSI  
analysis of Monju’s thermowell. If its density  
and viscosity change, its behavior could be  
different. In four different cases listing in Table  
I, lower viscosity of sodium liquid in the range  
of 2.41 – 4.52×10-4 Pa.s causes a lower damping  
coefficient; it may take a longer time for the  
plate to reach the equilibrium state.  
An interface between the solid domain  
and the fluid domain was created for data  
transferring, as mentioned in the previous  
section of FSI analysis. Displacement values of  
the plate are interpolated to the fluid mesh using  
two-way FSI. Figure 5 (a, b, c) shows the  
horizontal displacement of the free end of the  
plate comparisons between our calculation  
From the analytical calculation, the first  
natural frequency of the simplified of a  
clamped beam is defined as follows [12]:  
results with that of Gl  
viscosities of fluid. Our results have the same  
tendency as Gl ck: the higher the viscosity of  
̈ck [12] at three different  
 
   
           
 
(6)  
√
 
 
 
    
̈
the fluid, the faster the plate is damped and  
reaches the equilibrium state. However, there  
Where I is the polar moment of inertia  
and m is the total mass.  
(a)   = 0.2 Pa.s  
(b)   = 1.0 Pa.s  
(c)   = 5.0 Pa.s  
Fig. 5. Comparison of the displacements of the free end of the beam form three different fluid viscosities  
with Gl  
̈
ck et al (2001) [12]  
20  
HOANG TAN HUNG et al.  
However, our result only showed the  
Therefore, the Modal and FSI calculation  
models can be applied for fundamental analysis  
of the oscillation of thermowell.  
displacement of the plate in time. Therefore,  
FFT is used to find the frequency of the plate to  
compare it with the analytical calculations.  
Applying the FFT of the displacements to the  
case of the dynamic viscosity of 0.2 Pa.s, the  
first natural frequency of the plate can be found,  
III. RESULTS AND DISCUSSION  
Because the lock-in can occur when the  
frequency of vortex shedding (fs) is close to  
one of the natural frequencies (fi) of structures,  
the Modal simulation set up to find the first six  
natural frequencies of thermowell. Table II  
shows the results of Modal analysis, the value  
of frequencies are nearly equal in pair because  
of the symmetry characteristic of the geometry  
model of y and z-direction.  
f= 0.319 Hz. The analytical value of   = 0.300  
 
Hz is approximately equal to the simulation  
value of Modal analysis 0.317 Hz (5.6%  
difference) and FSI analysis 0.319 Hz (6.3 %  
difference). It can be seen that the characteristic  
of the oscillation of the plate is nearly the same  
with the thermowell on the Monju reactor.  
Table II. Natural frequency of thermowell  
Mode  
1
2
3
4
5
6
Natural  
frequency  
fN (Hz)  
261.15 261.22 1611.70 1612.00 4327.20 4327.50  
Four cases of 2D Fluent analysis were  
performed to find the frequency of vortex  
shedding (fs) of each flow condition with a  
typical cross-section of thermowell. The  
velocity of sodium flow is 5.20 m/s and 2.08  
m/s, corresponding to the 100% and 40% flow  
operation. Figure 6 shows that the vortex  
shedding appears behind the thermowell, and  
the change of drag and lift forces coefficient is  
periodical in Case 1. The frequency of vortex  
shedding (fs) of each case is found by using  
FFT. A detailed comparison between the  
average natural frequency of structure (fN) and  
vortex frequency fo fluid (fS) for Mode 1 and  
Mode 2 is shown in Table III. From Eq. 2, it can  
be seen that the safe cases could be reached if  
the frequency ratio, f, is smaller than 0.40, and it  
could be dangerous in reverse situations. Except  
for Mode 1 and 2, it can be seen that f is much  
smaller than 0.40 in other modes with higher  
frequencies. The most dangerous case could  
occur in Mode 1 and 2. Therefore, higher  
natural frequency comparisons are not necessary.  
The results show that the most dangerous case  
of flow condition can lead to fatigue damage is  
Case 1 with ratio f = fS/ fN = 0.65. The obtained  
result in Case 1 was in good agreement with that  
in [1].  
Table III. Comparison of natural frequency of thermowell and vortex shedding frequency  
Vortex shedding  
frequency fS (Hz)  
Average natural frequency fN  
(Mode 1 and mode 2) (Hz)  
Ratio f = fS/ fN  
Case 1  
Case 2  
Case 3  
Case 4  
170.94  
64.10  
64.10  
64.10  
261.185  
261.185  
261.185  
261.185  
0.65  
0.25  
0.25  
0.25  
21  
NUMERICAL INVESTIGATION OF FLUID – STRUCTURE INTERACTION (FSI)…  
a) Appearance of vortex shedding behind the  
typical cross-section of thermowell  
b) Lift coefficient and drag coefficient variations  
of thermowell  
Fig. 6. 2D Fluent vortex shedding (case 1)  
According to the previous result, the  
selected Case 1 was chosen for FSI analysis to  
find the location and value of stress  
concentration on the thermowell tube. Figure 7  
shows the vortex shedding appears behind the  
thermowell in both scenarios of FSI analysis.  
The changes in the diameter from the root to  
the top of the thermowell resulted in the  
differences in size and time of vortex  
shedding’s appearance. In the case of one-way  
FSI, the flow pressure will be mapped on the  
deformation and stress concentration. However,  
no reversed feedback process on the flow field  
could occur. In the case of two-way FSI, the  
deformation information will be feedback to  
the flow field to calculate the influence of  
deformation on the flow. Because of the  
intensive interaction between sodium flow and  
thermowell, the lift force and drag force are  
changed. Their change makes the thermowell  
oscillated around the equilibrium position. As a  
result, the thermowell was bent by these forces  
that similars to thermowell configuration [8].  
thermowell  
surface  
to  
calculate  
the  
a) One-way FSI  
b) Two-way FSI  
c) Thermowell configuration [8]  
Fig. 7. Appearance of vortex shedding behind the thermowell  
Figure 8 shows that the location of stress  
value of stress concentration is very different  
between the two types of FSI analysis. The  
maximum equivalent stress of 18.6 MPa in  
one-way FSI is much smaller than that in the  
case of two-way FSI. The results show a big  
difference between the structural response of  
concentration on the thermowell in FSI  
analysis is found in the place of the notch  
which is the same broken location of  
thermowell in the accident [8] for both cases  
one-way FSI and two-way FSI. However, the  
22  
HOANG TAN HUNG et al.  
static and dynamic load. The one-way FSI  
load in two-way FSI in which the resonance  
could lead the system to oscillate with larger  
amplitude and maximum stress. In one-way  
FSI simulation, the fluid domain was not  
updated when the thermowell deformation  
happened. Differences in the pressure field  
lead to more conservative deflection values,  
whereas two-way FSI simulations updated the  
fluid domain related to the thermowell  
deformation at each time step. However, this  
approach needs a higher computational cost.  
could not lead to fatigue damage because the  
endurance limit of structural steel is 86.2 MPa.  
On the other hand, in two-way FSI, the  
absolute value of the maximum equivalent stress  
curve shows almost harmonic behavior (Figure  
9). Although the time of simulation is short,  
which is enough time to observe the thermowell  
oscillating a few times. It is seen that only the  
first thermowell oscillating period is unstable, but  
the maximum equivalent stress of other periods is  
still greater than the endurance limit which could  
cause fatigue damage. The obtained result of  
two-way FSI gives a similar phenomenon that  
happened in the Monju accident.  
Although the calculation value of  
maximum equivalent stress two-way FSI is  
higher than the endurance limit, this study does  
not provide a complete picture of the  
assessments due to limitations in the  
computational resources. Therefore, further  
analysis for observing the behavior of stress  
concentration on thermowell need to be done  
in the future.  
The difference of results between one-  
way FSI and two-way FSI can be caused by  
pressure load. The static load in one-way FSI  
could not have resonance effects like dynamic  
a) One-way FSI  
b) Two-way FSI  
Fig. 8. Stress concentration on thermowell  
Fig. 9. Maximum equivalent stress (two-way FSI)  
23  
NUMERICAL INVESTIGATION OF FLUID – STRUCTURE INTERACTION (FSI)…  
VIETNAM ATOMIC ENERGY INSTITUTE  
(VINATOM) through the R&D Project: "Fluid –  
Structure Interaction (FSI) of sodium leakage  
incident at prototype fast breeder reactor Monju  
by ANSYS".  
IV. CONCLUSIONS  
In this study, two approaches of one-way  
FSI and two-way FSI using for analysis were  
performed  
to  
investigate  
the  
VIV  
phenomenon and the stress concentration  
location and the stress amplitude acting on  
Monju’s thermowell. The flow condition  
with maximum stress was chosen for FSI  
analysis to find the most severe case leading  
to thermowell fatigue damage. The obtained  
result shows that fatigue damage initiated at  
an early stage of the 100% flow operation.  
REFERENCES  
[1]. H. Mikami, A. Shono and H. Hiroi, “Sodium  
Leak at Monju (I) - Cause and Consequences”,  
Tech. Committee Meeting on Evaluation of  
Radioactive Materials Release and Sodium  
Fires in Fast Reactors, Oaria, Ibaraki, Japan,  
International Atomic Energy Agency, Vienna,  
Austria, 1996.  
- In both approaches, the vortex  
shedding appears behind the thermowell. The  
changes in the diameter from the root to the top  
of the thermowell resulted in the differences in  
size and time of vortex shedding’s appearance.  
[2]. Masaya Kondo, Yoshinari Anoda, “Water  
Flow Simulation Test on Flow – Induced  
Oscillation of Thermowell in Prototype Fast  
Breeder Reactor MONJU”, JSME International  
Journal, Serires B, Vol.4, No.4, 2001.  
- The one-way FSI approach predicts  
well the stress concentration location. However,  
the value of the maximum equivalent stress at  
that location is not predicted with enough  
accuracy. In opposite, for the two-way FSI, it  
could predict both the stress concentration  
location and the value of the maximum  
equivalent stress at that location. Therefore, a  
two-way FSI approach for the evaluation of the  
stress on the thermowell of the Monju reactor  
is suitable.  
[3]. Hideki Aoi, Yue Zhao, Tamotsu Hayase,  
“Improvement of the Damaged Thermowell of  
the Monju Fast-Breeder Reactor Based on  
Vibration Analysis”, The Society for Modeling  
and Simulation International, SIMULATION,  
Vol. 78, Issue 8 , p. 504-511, 2002.  
[4]. Sebastian Kornet, Daniel Sławiński, Paweł  
Ziółkowski, Janusz Badur, “Analysis of  
unsteady flow forces on the thermowell of  
steam temperature sensor”, Transactions of the  
Institute of Fluid-Flow Machinery, No. 129, p.  
25–49, 2015.  
- Fatigue damage could occur in the  
condition of the maximum value of stress getting  
higher than the endurance limit. This condition is  
reached in our analysis using two-way FSI.  
However, it is needed a longer time of simulation  
to observe the change in the stress amplitude for  
the overall fatigue damage process.  
[5]. AISI Type 304 Stainless Steel, [online]  
Available at:  
[6]. Masaki Morishita, Yusaku Wada, “Fatigue  
analysis of thermowell due to flow – induced  
vibration”, Flow – Induced Vibration and  
ACKNOWLEDGMENTS  
The authors of this paper wish to express  
Transient Thermal  
– Hydraulics, ASME  
their appreciation for the financial support from  
International, p. 120, 1998.  
24  
HOANG TAN HUNG et al.  
[10].Brauschke D., Wiklund D., Kitzman A., Zulic  
[7]. J.K.Fink, L.Leibowitz, “Thermodynamic and  
Transport Properties of Sodium Liquid and  
Vapor”, Argonne National Laboratory, pp. 86-  
95 and p. 207, 1995.  
D., “Thermowell calculations”, Emerson  
Process Management, White Paper 00840–  
0200–2654, Rev AC, p. 11, March 2014.  
[11].ANSYS Tutorials & Training Materials, “Fluid  
[8]. A.Miyakawa, H.Maeda, Y.Kami, K.Ito,  
“Sodium Leakage Experience at the Prototype  
FBR Monju”, Unusual Occurrences during  
LMFR Operation, IAEA-TECDOC-1180,  
Vienna (Austria), p. 43 -56, 1998.  
Structure  
Interaction  
FSI  
with  
ANSYS Fluent 17.0” and “ANSYS  
Mechanical Linear and Nonlinear Dynamic”,  
ANSYS Inc, 2016.  
[12].Gl ̈ck, M., Breuer, M., Durst, F., Halfmann, A.,  
[9] Anton Gustafsson, “Analysis of Vortex-  
Induced Vibrations of Rises”, Master’s thesis  
in Applied Mechanics, Chalmers University Of  
Technology, Sweden 2012.  
Rank, E., “Computation of fluid–structure  
interaction on lightweight structures”, Journal  
of Wind Engineering and Industrial  
Aerodynamics 89, p. 1351–1368, 2001.  
25  
pdf 11 trang yennguyen 12460 Free
Bạn đang xem tài liệu "Numerical investigation of Fluid – Structure interaction (FSI) on sodium leakage accident at prototype fast breeder reactor monju", để tải tài liệu gốc về máy hãy click vào nút Download ở trên

File đính kèm:

  • pdfnumerical_investigation_of_fluid_structure_interaction_fsi_o.pdf