Development of three-dimensional hot pool model in a system analysis code for pool-type FBR

Nuclear Engineering and Design 256 (2013) 264–273

Contents lists available at SciVerse ScienceDirect

Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes

Development of three-dimensional hot pool model in a system analysis code for pool-type FBR Danting Sui a,∗ , Daogang Lu a , Lixia Ren b , Yizhe Liu b a b

School of Nuclear Science and Engineering, North China Electric Power University, Beijing 102206, China China Institute of Atomic Energy, Beijing 102413, China

h i g h l i g h t s    

A 3-D hot pool analysis code with porous medium model was developed. The coupling between 3-D model and system analysis code was finished. The coupled code was used to analyze T-H behavior in upper plenum of MONJU. Complex flow field in CEFR hot pool was analyzed with the coupled code.

a r t i c l e

i n f o

Article history: Received 6 January 2012 Received in revised form 24 July 2012 Accepted 1 August 2012

a b s t r a c t A three-dimensional hot pool analysis code with SIMPLE algorithm was developed based on staggered grid, which can predict thermal-hydraulic characteristic in hot pool of pool-type liquid metal fast breeder reactor (LMFBR) under Cartesian and cylindrical coordinate systems. After being incorporated into system analysis code for pool-type fast reactor in China (SAC-CFR), the coupled code was used to analyze the thermal-hydraulic behavior in the upper plenum of fast breeder reactor “MONJU” during reactor scram transient. A basic agreement was obtained, which means the present model is effective. And then the coupled code with newly developed porous medium model was used to analyze the flow field in China Experimental Fast Reactor hot pool under steady-state operation condition. The distribution characteristic of flow field in hot pool showed the effectiveness of porous medium model, which formed preparations for further development of passive residual heat removal system in-vessel. © 2012 Elsevier B.V. All rights reserved.

1. Introduction The sodium in pool-type liquid metal fast reactor provides vital function of removing reactor generated heat. Accurate prediction of coolant thermal-hydraulic characteristic in hot pool cannot only evaluate the performance of key component in hot pool properly, but also improve the system analysis ability through more accurate intermediate heat exchanger (IHX) inlet temperature. In advanced FBR, the directly in-vessel decay heat remove system (DIDHRS) is being considered to be used to improve inherent safety, which makes it impossible to analyze the complicated phenomenon caused by DIDHRS with traditional 1D or 2D system analysis code. Therefore, the coupling of system code and 3D thermal-hydraulic analysis code is always an important challenge to take into account local 3D effects on global system behavior. Considering that the hot pool is arranged with more complex key

∗ Corresponding author. Tel.: +86 10 51963824; fax: +86 10 51963351. E-mail addresses: [email protected] (D. Sui), [email protected] (D. Lu), [email protected] (L. Ren), [email protected] (Y. Liu). 0029-5493/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.nucengdes.2012.08.032

components and filled with higher-temperature coolant than cold pool, 3D thermal-hydraulic characteristic in hot pool is a research focus in this study. Many researches on thermal-hydraulic characteristic in hot pool have been carried out currently in some countries and organizations. In the aspect of experiment, many kinds of in and out of core experiment have been performed for their specified research purpose in America, France, Japan, Russia, etc. (Hofmann and Essig, 1993; Kasinathan, 1993; Ieda, 1993). In the aspect of theory, besides the commercial computational fluid dynamics software such as CFX, FLUENT, many three-dimensional thermal-hydraulic analysis codes have been specially developed for evaluating the thermalhydraulic performance in fast reactor, such as COMMIX (Chien et al., 1993) series developed by ANL, AQUA (Muramastsu et al., 1987) developed in PNC, TRIO U (Tenchine et al., 2012)in France, FASTOR3D (Degui et al., 1998) and DHRSC (Yishao et al., 1991) in China. During transient analysis, all these analysis codes can only predict the thermal-hydraulic performance in particular component such as hot pool at a certain boundary condition. As to the coupling of system code and 3D analysis code, many researches have been performed by organizations and researchers. For instance in Europe,

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

the Trio U has been coupled to CATHARE by NURISP (Emonot et al., 2012) and THINS (Xu et al., 2010) teams, and the related application, verification and validation are in progress. Also, the coupling used for PWR is also a research focus in recent years. In Japan, two-dimensional upper plenum model was incorporated into a one-dimensional system analysis code named SSC-L (Mochizuki, 2007, 2010). From the aspect of fast reactor sustainable development in China, it is the only choice to develop system analysis code and 3D analysis code specialized for CEFR and next generation demonstration fast reactor. In the present study, a three-dimensional hot pool analysis code with SIMPLE algorithm was developed based on staggered grid under Cartesian and cylindrical coordinate systems. After being incorporated into System Analysis Code for China Fast Reactor (SAC-CFR), the coupled code was used to analyze the thermalhydraulic characteristic in upper plenum of MONJU during the scram transient starting from 45% thermal power operation. And then the coupled code with newly incorporated porous medium model was used to analyze the flow field in China Experimental Fast Reactor (CEFR) hot pool under steady-state operation condition. 2. Governing equations and discretization The general form of governing equations in Cartesian and cylindrical coordinate systems can be written as the following: 1 ∂(rJx ) ∂ 1 ∂Jy ∂Jz + + + = S r ∂x r ∂y ∂t ∂z Jx = u − 

1 ∂ r ∂y

(3)

V

and directional surface porosity  x ,  y and  z can be expressed as: i =

Af A

, where i = x, y, z

(4)

Volume porosity is defined as the ratio of the volume occupied by fluid in a control volume Vf to the total control volume V. The directional surface porosity is similarly defined as the ratio of the area available Af for fluid flow through a control surface to the total control surface area A. It should be noted that volume porosity and directional surface porosity are used to model the flow space reduction for the presence of solid structure. Distributed heat source and distributed resistance are used to model the effect of the presence of solid structure on momentum and energy exchange, which are classified into the source terms. The finite volume method based on staggered grid is chosen as the discretization method. The discretized equations are derived by integrating the governing equations over a control volume. The power-law scheme was used to construct the discretized equation. The power-law expressions for aE can be written as (Patankar, 1980):

⎧ 0, ⎪ ⎪ ⎪ ⎪ ⎨ (1 − 0.1P e )5 ,

aE = De ⎪ (1 + 0.1P

P e =

e )

5

P e > 10 0 ≤ P e ≤ 10

− P e ,

−P e ,

⎞

(5)

−10 ≤ P e ≤ 0 P e < −10

Fe uıxe = De

e

(6)

Fe = (u)e × Ae

⎜ ∂p ⎟ ⎜ − + gx ⎟ ⎜ ∂x ⎟ ⎜ ⎟ ⎜ ∂p ⎟ ⎜ + gy ⎟ , S = ⎜− ⎟ ∂y ⎜ ⎟ ⎜ ∂p ⎟ ⎜ − − gz ⎟ ⎝ ∂z ⎠

⎛

0

⎜ ∂p v2 2 ∂v u ⎜ − + − 2 · − 2 ⎜ ∂r r ∂ r r ⎜ ⎜ 1 ∂p uv 2 ∂vr v S=⎜ ⎜ − r ∂ − r −  r + r 2 · ∂ ⎜ ⎜ ∂p ⎜ −g − ⎝ ∂z

Q

⎞ ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎠

Q

The conservation equations incorporating porous medium formulation are based on averaging local volume. The averaging local volume is similarly defined as those in continuum medium, and provides spatially smoothed equations solved with the general solution method. After incorporating porous medium model, the governing equations can be written as: V

Vf

aE is the neighboring coefficient, representing the convective and diffusion effect from the neighboring node. P e is defined as the ratio of convective mass flux to diffusion conductance at cell faces.



Value of r is 1 for Cartesian coordinate and remains as r for cylindrical coordinate. For the continuity equation, momentum equation and energy equation,  represents 1, u, v, w, and h respectively. For Cartesian coordinate, three directions are denoted as x, y and z. For cylindrical coordinate, three directions are denoted as r,  and z. The source terms in Cartesian and cylindrical coordinate are: 0

V =

⎪ ⎪ ⎪ ⎩

∂ Jz = w −  ∂z

⎛

Specially, volume porosity  V can be defined as:

∂ ∂x

 Jy = v − 

(1)

265

∂ 1 ∂(rx Jx ) 1 ∂(y Jy ) ∂(z Jz ) + + + = S r r ∂y ∂t ∂x ∂z

Jx , Jy and Jz have the same expressions with previous ones.

(2)

De =

(7)

e × Ae

  ıx

(8)

e

where Fe indicates the strength of convection, while De is the diffusion conductance. , u, e and Ae is the density, velocity, dynamic viscosity and flow area of control volume respectively. It should be noted that, whereas D always remains positive, F could take either positive of negative values depending on the direction of flow. Take the cylindrical coordinate system for example to perform the discretization of governing equation. The discretized form of momentum equations in three directions obtained through the conservation of total flux over a control volume can be written as: aP ϕP = aE ϕE + aW ϕW + aN ϕN + aS ϕS + aT ϕT + aB ϕB + b

(9)

The expressions of b in three directions are: For the radial direction, b = SU1 + SU2 − (pP − pW )r ϕ z + SU1 =

u0 V t

(10)

1 2  rϕ z[v(i, j, k) + v(i, j + 1, k) + v(i − 1, j, k) + v(i − 1, j + 1, k)] 16

SU2 = − r z [v(i, j + 1, k) + v(i − 1, j + 1, k) − v(i, j, k) − v(i − 1, j, k)] r

For the axial direction, b = −(pP − pB )r r ϕ +

w0 V + g V t

(11)

266

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

Table 1 Three types of temperature boundary condition and implementation methods in the code. Temperature boundary condition

First-type boundary condition temperature on the boundary is specified TW

Heat flux coming into boundary-adjacent control volume Implementation method in code abd a aP b

q=

b c

− TP )

0

ıx

TW ×

Su c a

(TW ıx

ıx

Second-type boundary condition, heat flux on the boundary is specified q

Third-type boundary condition, convective heat transfer coefficient and coolant temperature are specified

q

q=

0 –

0

(Tf −TP ) 1/h+ıx/

1 (1/h+ıx/ ) Tf

q

(1/h+ıx/ )

abd neighboring coefficient from the boundary. aP additional term added to the former value of aP . Su additional term added to the former value of Su .

of the inlet plane to the nodal values just upstream and downstream of the outlet plane.

For the circumferential direction, b = SU1 − (pP − pS ) r z + SU1 =

v0 V t

(12)

(i, 1, k) = (i, nj − 1, k)

(15)

r z r

(i, nj, k) = (i, 2, k)

(16)

[u(i, j, k) + u(i + 1, j, k) − u(i, j − 1, k) − u(i + 1, j − 1, k)]

3.4. Reference pressure The pressure field obtained by solving the pressure correction equation does not give the absolute pressure. It is common practice to fix the absolute pressure at one inlet node and set the pressure correction to zero at that node. Having specified a reference value, the absolute pressure field inside the domain can be obtained.

3. Boundary conditions and solution of discretized equations 3.1. Inlet and outlet boundary conditions

3.5. Temperature boundary conditions The distributions of all flow variables are specified at inlet boundaries. Calculation of velocity at the outlet plane by local mass continuity over control volume gives, u∗ (i + 1, j, k) = u(i, j, k) + +

Sx (v(i, j, k) − v(i, j + 1, k)) Sy

Sx (w(i, j, k) − w(i, j, k + 1)) Sz

(13)

Assumed that the fluid model is subcooled single-phase liquid metal sodium, and the flow is incompressible flow. Then, the velocities are corrected with the mass conservation over the whole computational domain (Wenquan, 2009). The final outlet plane velocities with the continuity correction are given by: u(i, j, k) = u∗ (i, j, k) ×



FLOWIN u∗ (i, j, k) × Sx

(14) × (i, j, k)

outlet

The relation between the neighboring nodes is constructed by the variable flux transferred at the cell faces. According to this theory, heat flux coming into the computational cell from the boundary under three types of temperature boundary conditions and the implementation methods in the code are shown in Table 1. It can be seen from the table that the coding method of implementing three types of boundary conditions is to split the heat flux to two parts: one is constant part classed into the source term, the other is variable part. 3.6. Solution of discretized equations Three matrix solvers (Qingyang et al., 2001) are employed to solve the discretized equations. They are Gauss–Seidel iteration method, successive overrelaxation method, iterative method based on tri-diagonal matrix algorithm (TDMA) and cyclic tridiagonal matrix algorithm (CTDMA). The iteration equation for the Gauss–Seidel method is as following:

FLOWIN is the mass flux coming into the domain. (k)

3.2. Wall boundary conditions

xi

=

i−1    −aij j=1

The slip condition is the appropriate condition for the velocity components at solid walls in this case. The normal component of velocity can simply be set to zero at the boundary no matter under slip or non-slip condition. For all other variables, special sources are constructed by considering the diffusion effect, while neglecting the convective effect from the variables at solid walls.

aii

(k)

xj

To apply cyclic boundary conditions we need to set the flux of all flow variables leaving the outlet cyclic boundary equal to the flux entering the inlet cyclic boundary. This is achieved by equating the values of each variable at the nodes just upstream and downstream

n    −aij j=i+1

aii

(k−1)

xj

+

bi aii

i = 1, 2, . . . , n (17)

The iteration equation for successive overrelaxation method are as following: (k)

xi

⎡ i−1  n      −aij −aij (k−1) (k) (k−1) =x + ˛⎣ x + x i

j=1

3.3. Cyclic boundary condition

+

aii

j

j=i+1

aii

j

⎤ b + i⎦ aii

i = 1, 2, ..., n

˛ is the relaxation parameter.

(18)

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

267

Fig. 1. Schematic view of MONJU plant (Mochizuki, 2007).

4. Thermal-hydraulic analysis in MONJU upper plenum during 45% thermal power scram transient 4.1. Description of the test A turbine trip test at 45% thermal power of MONJU (Mochizuki, 2007, 2010) was conducted in 1995 to investigate the capability of ACS in an actual situation. The reactor was scrammed by the signal of turbine trip. Main pumps in the primary and secondary loops tripped at the same time of reactor shutdown. Pony motors took over operation when flow rates in primary and secondary loops were approximately 10% and 8% respectively. The schematic view of the plant system is shown in Fig. 1. It consists of three loops: the primary loop, the secondary loop and the water-steam loop. The steady-state operation parameters under 45% thermal power are listed in Table 2 together with those calculated by SAC-CFR. It can be seen that the value differs between the two columns; this is because some parameters are estimated when have not been found in public literature. The geometry of the upper plenum is shown in Fig. 2 (Yamaguchi and Ohshima, 1988). The System Analysis Code for China Fast Reactor (SAC-CFR) was developed specialized for China fast reactor to predict the plant response during operational and accidental transients. After incorporating three-dimensional hot pool analysis model, the newly

developed SAC-CFR can calculate the thermal-hydraulic response accurately in hot pool and the interaction with system response. When coupling, SAC-CFR calculates the system response and an overlapping method are used to couple the newly developed 3D code with the system code. System code gives boundary such as coolant temperature and mass flow rate at core outlet to the 3D analysis code, and feedbacks from 3D calculation such as coolant temperature and mass flow rate at hot pool outlet are exerted on system calculation. In present analysis, computation is only

Table 2 Comparison of initialization state for 45%-power scram transient between simulation and reported results.

Flow rate in primary loop (kg/s) Flow rate in secondary loop (kg/s) R/V outlet temperature (◦ C) R/V inlet temperature (◦ C) IHX inlet temperature in secondary loop (◦ C) IHX outlet temperature in secondary loop (◦ C)

MONJU

SAC-CFR

689 411 486 364 282 486

688.15 427.8 486.09 364.0 288.0 483.0

Fig. 2. Geometry of upper plenum (Yamaguchi and Ohshima, 1988).

268

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

Fig. 3. Mass flow rate and inlet temperature of IHX in secondary loop.

performed on the reactor, the primary loop and the intermediate heat exchanger. Assuming the upper core structure to be solid structure for simplification, and the porous medium method is not being incorporated here. The boundary conditions showed in Fig. 3 are IHX inlet temperature and mass flow rate in the secondary loop, which correspond to the data measured experimentally. 4.2. Results analysis The computed mass flow rate in primary loop is shown in Fig. 4 together with the experimentally measured one. Fig. 4 also shows decay heat ratio after scram (Xirong and Zhennian, 1988). The flow pattern in upper plenum under steady state operation is shown in Figs. 5–9. It can be seen that the flow pattern at horizontal cross-section of different heights are basically symmetrical. Core outlet jet changes flow direction after impinging on upper core structure, and then goes up on a slant. After impinging on the reactor vessel, main flow goes downward to form a strong vortex, while little goes upwards to form a weak vortex. The lower vortex can enhance the mixing of coolant from different core channels to form a more uniform temperature field, while the upper vortex has a weak disturbance on the coolant mixing. So the coolant

Fig. 4. Mass flow rate in primary loop and decay heat ratio after scram.

Fig. 5. Flow pattern at outlet longitudinal section of upper plenum.

in the upper part of upper plenum has a slower thermal-hydraulic response than in the lower part, which is a major factor causing thermal stratification in upper plenum. For the thermal liner was neglected for simplification when modeled, the phenomenon of fluid going over the top of the liner cannot be predicted in detail. The position of interface is where the largest temperature gradient exists. And the value of temperature gradient is an important parameter to assess the thermal striping of component structure. The maximum temperature gradient and its position at different moment during the scram transient are listed in Table 3. The computed axial temperature distribution near the thermal liner at different moment is shown in Figs. 10 and 11. From the table and the figures, some thermal stratification characteristics can be found. At

Fig. 6. Flow pattern at non-outlet longitudinal section of upper plenum.

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

269

Table 3 Characteristics of thermal stratification in upper plenum. Time (s)

Maximum temperature gradient (◦ C/m)

Elevation of the maximum temperature gradient from core outlet (m)

0 30.25 60.25 120.25 180.25 480.25 2000.25 4800.25 5400.25 6200.25 8000.25 8600.25 9000.25

0 0 49.83640 84.30382 84.07354 93.24695 141.8174 136.9282 123.8378 116.9868 95.62407 96.72161 98.00254

– – 0.823 0.823 0.823 1.235 1.235 1.235 1.235 1.235 2.161 2.161 2.161

Fig. 7. Flow pattern at horizontal cross-section of upper plenum outlet nozzle height.

Fig. 10. Axial temperature distribution near the thermal liner during 0–480 s after scram.

Fig. 8. Flow pattern at horizontal cross-section below upper plenum outlet nozzle height.

Fig. 9. Flow pattern at horizontal cross-section above upper plenum outlet nozzle (Z = 2.9).

Fig. 11. Axial temperature distribution near the thermal liner during 0–9000 s after scram.

270

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

Fig. 12. Axial temperature distribution near the thermal liner under steady state operation.

the initial moment, the stratified interface between the hot and cold coolants rises gradually. After the initial moment, the interface is fixed at a position. After a long term, the interface rises to another height and fixed at the position. The exact location of interface may be related to the mesh size, but the trend of interface is also useful to locate the structure where temperature fluctuation occurs. The value of maximum temperature gradient grows gradually. After the peak value is reached, it then decreases gradually. The computed axial temperature distribution near the thermal liner at different moment is shown in Figs. 12–17 together with the measured one (Doi et al., 1996). It can be seen that the computed temperature distribution is basically in accordance with the measured one except for those at the initial moment. It is also clear that both the computed and the measured ones show the steep temperature gradient, which is to say that the thermal stratification phenomenon is well simulated. The reason why the computed temperature is higher than the measured one near the thermal liner can be summarized as follows. Firstly, the uniform axial temperature distribution was obtained at the initial moment, while the measured temperature is lower than the computed one

Fig. 14. Axial temperature distribution near the thermal liner 120 s after scram.

Fig. 15. Axial temperature distribution near the thermal liner 180 s after scram.

Fig. 16. Axial temperature distribution near the thermal liner 240 s after scram. Fig. 13. Axial temperature distribution near the thermal liner 60 s after scram.

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

Fig. 17. Axial temperature distribution near the thermal liner 300 s after scram.

for the thermocouple was possible placed near the outlet of the bypass channel. Secondly, strong turbulence exists in the upper plenum during the scram transient. The lack of turbulence model results in the incomplete mix between the core outlet coolant and the existed coolant in the upper plenum. Thirdly, since no coolant flows over the top of the liner for the thermal liner was neglected when modeled, the coolant near the thermal liner with higher temperature cannot mix with the coolant with lower temperature from core during the scram transient. Besides, the core outlet

Fig. 18. Configuration scheme of CEFR hot pool.

271

Fig. 19. Flow pattern at horizontal cross-section below outlet height.

jet has been weakened to some extent for the mesh size is not small enough. Near the outlet nozzle height, temperature difference in circumferential direction is most obvious. At this height, coolant temperature at outlet nozzle azimuth is a little higher than at the non-outlet nozzle azimuth. And this difference becomes smaller as the height increases or decreases. The reason can be summarized as follows: Since a strong vortex exists before the coolant reaches outlet nozzle, the main coolant reaches non-outlet nozzle azimuth ahead of outlet nozzle azimuth, which causes the coolant temperature is higher at outlet nozzle azimuth. While at the non-outlet nozzle height, coolant flow is weak, so the temperature difference in circumferential direction is not obvious. It is also can be seen that the absolute value of temperature difference in circumferential direction is not so large for the symmetry of the structure. If the system layout in upper plenum is non-symmetric, it can be predicted that the difference in circumferential direction will be considerable.

Fig. 20. Flow pattern at horizontal cross-section of outlet height.

272

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

Fig. 22. Flow pattern at longitudinal section at the azimuth of DHX. Fig. 21. Flow pattern at longitudinal section at the azimuth of IHX.

5. Flow field analysis in CEFR hot pool under 75% power steady-state operation condition To analyze the fluid dynamic and thermal characteristics of sodium pool involving multiple components or complex geometry accurately, porous medium method is introduced to model the intermediate heat exchanger, primary pump, decay heat exchanger, and radial shielding. The schematic view of CEFR hot pool is shown in Fig. 18. It can be seen that key components includes four intermediate heat exchangers, two primary pumps, two decay heat exchangers, and so on. Component arrangement and main flow path can refer to references (Yijun, 2003; Zhijie et al., 1998). The simplifications made to the hot pool model can be summarized as:

Also from the flow pattern, it can be seen that the solid structure of IHX, pump, and direct reactor auxiliary heat exchanger (DHX) are successfully simulated with porous medium model. Flow patterns at longitudinal section of different azimuths are shown in Figs. 21–23, which locates at the same azimuth of IHX, pump, and DHX respectively. It can be seen that the flow through the shielding columns is strictly symmetrical for the assumption of shielding columns distributed uniformly. The core baffle located at the inner side of shielding columns prevents the coolant from flowing to the shielding material surrounding the core, which reduces the heat loss and pressure loss. For the diversion of upper core structure and inner baffle, coolant from the core exit flow to the IHX inlet hole through the opening on the outer baffle with strong regularity. Also from the above flow fields, it can be seen that the

(1) The hot pool is modeled to 360◦ in circumferential direction under the cylindrical coordinate system. Intermediate heat exchanger, pump, decay heat exchanger occupy one or some control volumes according to their actual sizes. (2) Supposed the four-layer radial shielding columns distributed uniformly, and then the directional surface porosity and volume porosity kept same for the corresponding control volumes. (3) The openings facing the IHX on the core baffle are the flow channels between the inner and outer pool. (4) Fuel handling structure is neglected when modeling. (5) The openings on the primary side of IHX are outlet boundary for the hot pool model, while the rest parts are treated as solid structure. (6) The solid structure of IHX, pump, and decay heat exchanger are modeled with porous medium method. The newly coupled code with porous medium model are used to analyze the flow filed in China Experimental Fast Reactor hot pool under 75% power steady-state operation condition. The flow patterns are shown in Figs. 19–23. Horizontal flow fields shown in Figs. 19 and 20 are at different height. One is below the outlet boundary height on the primary side of IHX, and the other keeps at the same height. It can be seen that the flow fields at different heights are basically symmetrical for the key components are mirror-symmetrical arranged in hot pool.

Fig. 23. Flow pattern at longitudinal section at the azimuth of pump.

D. Sui et al. / Nuclear Engineering and Design 256 (2013) 264–273

structural characteristic of outer core baffle is successfully simulated with porous medium method and appropriate boundary conditions, which is the opening of the core baffle facing to the IHX and decay heat exchanger but not towards the pump. The influence of the solid structure such as IHX, pump and DHX on the flow field can also be seen from the longitudinal flow patterns. The distribution characteristic of flow field in hot pool showed the effectiveness of porous medium model, which formed preparations for further development of passive residual heat removal system in-vessel. 6. Conclusions A three-dimensional hot pool analysis code with SIMPLE algorithm was developed based on staggered grid, which can predict thermal-hydraulic characteristic in hot pool of pool-type LMFBR under Cartesian and cylindrical coordinate systems. And the coupling between newly developed 3-D hot pool model and system analysis code was finished. After being incorporated into System Analysis Code for China Fast Reactor (SAC-CFR), the coupled code was used to analyze the thermal-hydraulic behavior in the upper plenum of fast breeder reactor “MONJU” during reactor scram test. A basic agreement between the computational results and experimental data was demonstrated, which means the present model is effective. Also, from analytical results, some characteristics of thermal stratification in the upper plenum were also discovered. After incorporating porous medium model into hot pool analysis model, the newly developed code was used to analyze flow field in China Experimental Fast Reactor (CEFR) hot pool under 75% power steady-state operation condition. The distribution characteristic of flow field in hot pool showed the effectiveness of porous medium model, SAC-CFR with three-dimensional hot pool analysis model is now ready to simulate passive residual heat removal under natural convection for CEFR after incorporating interassembly model. References Chien, T.H., Domanus, H.M., Sha, W.T., 1993. COMMIX-PPC: A Three-Dimensional Transient Multicomponent Computer Program for Analyzing Performance

273

of Power Plant Condensers, Volume 1: Equations and Numerics, ANL-92/2. Argonne National Laboratory, USA. Degui, L., Shitong, S., Wanchen, L., 1998. Calibration of FASTOR-3D for calculation of heat transfer and flow in sodium pool of fast breeder reactor. J. Shanghai Jiaotong Univ. 32 (1), 115–118. Doi, Y., Muramatsu, T., Kunogi, K., 1996. Thermal Stratification Tests in MONJU Upper Plenum – Temperature Distribution Under Normal and Scram Conditions with 40% Power Operation , PNC TN9410 96-117, Japan. Emonot, P., Perdu, F., Baviere, R., et al., 2012. Coupling CATHARE and Trio U Using ICoCo. NURISP, Karlsruhe. Hofmann, F., Essig, C., 1993. Investigation on natural convection decay heat removal for the EFR Status of the program. In: Specialists’ Meeting on Evaluation of Decay Heat Removal by Natural Convection, IWGFR/88, Japan, p. 1993. Ieda, Y., 1993. Strategy of experimental studies in PNC on natural circulation decay heat removal. In: Specialists’ Meeting on Evaluation of Decay Heat Removal by Natural Convection, IWGFR/88, Japan, p. 1993. Kasinathan, N., 1993. Analysis of decay heat removal by natural convection in PFBR. In: Specialists’ Meeting on Evaluation of Decay Heat Removal by Natural Convection, IWGFR/88, Japan, p. 1993. Muramastsu, T., Maedawa, I., Matsumoto, M., et al., 1987. Single-Phase MultiDimensional Thermal-Hydraulic Analysis Code: Aqua-Program Description , PNC N9520 87-013, Japan. Mochizuki, H., 2007. Verification of NETFLOW code using plant data of sodium cooled reactor and facility. Nucl. Eng. Des. 237 (1), 87–93. Mochizuki, H., 2010. Development of the plant dynamics analysis code NETFLOW++[J]. Nucl. Eng. Des. 240 (3), 577–587. Patankar, S., 1980. Numerical Heat Transfer and Fluid Flow. John Benjamins Publishing Co., USA. Qingyang, L., Nengchao, W., Dayi, L., 2001. Numerical Analysis, 4th ed. Tsinghua University Press, Beijing. Tenchine, D., Barthel, V., Bieder, U., et al., 2012. Status of TRIO U code for sodium cooled fast reactor. Nucl. Eng. Dept. 242, 307–315. Wenquan, T., 2009. Numerical Heat Transfer, 2nd ed. Xi’an Jiaotong University Press, Xi’an. Xirong, S., Zhennian, Z., 1988. Decay heat character of fast test reactor after shutdown. Nucl. Power Eng. 9 (3), 90–94. Xu, C., Class, A., Meloni, P., et al., 2010. THINS project presentation. In: Seventh Framework Programme of EUROATOM for Nuclear Research and Training Activities (2007–2011). Yamaguchi, A., Ohshima, H., 1988. Key Technology Design Study of a Large LMFBR – Improvement of Reactivity Feedback Modeling in SSC-L and Analysis of Plant Thermal Hydraulic Behavior During ATWS Accident , PNC N9410 88-006, Japan. Yijun, X., 2003. 3-Dimensional Thermal-hydarulic Analysis of Sodium Pool of CEFR. China Institute of Atomic Energy, Beijing. Yishao, C., Fan, J., Haosong, Y., 1991. A computer code DHRSC for natural convection and decay heat removal in experimental fast reactor. Atom. Energy Sci. Technol. 25 (5), 85–89. Zhijie, L., Yuxin, R., Zhaohui, Y., et al., 1998. 3-D thermal-hydraulic analysis of hot plenum of CEFR. Chin. J. Nucl. Sci. Eng. 18 (2), 125–131.