Preliminary Thermal-Hydraulic Analyses for …oldweb.reak.bme.hu/fileadmin/user_upload/felhasznalok/...ThF4 molten salt [4] Based on the thermal power, desired inlet and outlet temperatures,

1 Copyright © 2013 by ASME

PRELIMINARY THERMAL-HYDRAULIC ANALYSES FOR DESIGNING AN EXPERIMENTAL MODEL OF A MOLTEN SALT REACTOR CONCEPT

Bog dán Yamaji Institute of Nuclear Techniques, Budapest Univ. of

Technology and Economics Budapest, Hungary

phone: +36 1 4632112, fax: + 36 1 4631954 e-mail: [email protected]

Attila Aszódi Institute of Nuclear Techniques, Budapest Univ. of

Technology and Economics Budapest, Hungary

phone: +36 1 4631989, fax: + 36 1 4631954 e-mail: [email protected]

ABSTRACT Based on the MSFR (Molten Salt Fast Reactor) reactor

concept presented within the framework of the EVOL (Evaluation and Viability of Liquid Fuel Fast Reactor System, EU FP7) international research project preliminary three-dimensional thermal-hydraulic analyses and the discussion of scaled experimental modelling will be presented.

The MSFR concept is a single region, homogeneous liquid fuelled fast reactor. The reactor concept uses fluoride-based molten salts with fissile uranium and/or thorium and other heavy nuclei content with the purpose of applying the thorium cycle and the burn-up of transuranic elements. The concept has a single region cylindrical core with sixteen radial inlet and outlet nozzles located at the bottom and top of the core. The external circuit (internal heat exchanger, pump, pipes) is broken up into sixteen identical modules distributed around the core.

Purpose of the three-dimensional computational fluid dynamics (CFD) calculations is to study the possibility of experimental investigation of the fluid flow in the core of the proposed MSFR concept using a scaled model and Particle Image Velocimetry (PIV) flow measurement technique.

First the main properties of the proposed MSFR concept are introduced, and the information on other experimental thermal-hydraulic modelling of different reactors, including MSRE (Molten Salt Reactor Experiment) are summarised.

With a scaled plexiglas MSFR model it would be possible to carry out flow field measurements under laboratory conditions using PIV method. Possible way of scaling are presented and a series of preliminary CFD calculations are discussed. Possibilities and limitations of such scaling and segmenting of a model and the use of water as substitute fluid for the experimental mock-up will be discussed.

Objectives of such a measurement series would be validation, benchmarking of CFD calculations and codes, application of CFD modelling experience in the detailed thermal-hydraulic design of the MSFR concept, possible measurements for the study of specific problems or phenomena, for example refinement of inlet geometry, effects of internal structures, coolant mixing.

BACKGROUND: DESCTRIPTION OF THE MSFR The MSFR concept is a single region homogeneous reactor with fluoride-based molten salt as primary coolant and fuel. The basic concept was introduced earlier with different nominal values concerning thermal power and inlet and outlet temperatures [1, 2, 3]. The parameters and design properties of the MSFR used in this paper are based on the EVOL MSFR benchmark proposal [4]. The geometry of the core is cylindrical, the fluoride-based fuel-coolant molten salt flows upward from the inlet nozzles located at the bottom of the core to the outlet nozzles at the top of the cylinder. Sixteen pair of inlet and outlet nozzles are placed at the perimeter of the core dividing it into sixteen identical segments. The liquid salt is transferred by pumps to internal/intermediate heat exchangers, bubble separators and then back to the inlet nozzles. The primary loops are connected to an intermediate heat transfer loop by heat exchangers. Below and above the core axial reflectors can be found, the core is surrounded by a fertile blanket (see Figure 1). In the present paper discussion focuses on the flow phenomena inside the core, external parts, such as pumps, the fertile blanket or the heat exchangers are not modelled.

Proceedings of the 2013 21st International Conference on Nuclear Engineering ICONE21

July 29 - August 2, 2013, Chengdu, China

ICONE21-16592


Figure 1: General layout of the MSFR concept [3, 4] Nominal thermal power of the MSFR concept is 3000 MW,

nominal inlet temperature is 650 °C and outlet temperature is 750 °C. Main thermal parameters of the MSFR concept are summarised in Table 1. The fuel-coolant molten salt composition is fluoride-based that includes 22.5% (mol%) actinide-fluorides of thorium, uranium, and/or other actinides. Physical properties of this composition are shown in Table 2. These values and functions are used in our further analyses.

Thermal power [MW] 3000

Electric power [MW] 1500

Initial fuel salt composition [mol%]

77.5% LiF – 22.5% (U/Th/Pu)F4

Initial blanket salt composition [mol%]

77.5% LiF – 22.5% ThF4

Inlet temperature [°C] 650

Outlet temperature [°C] 750

Core diameter and height [m] 2.255

Table 1: MSFR concept design parameters [4]

function of temperature value at 700 °C

Density [g/cm3] 4.094-8.82E-04×(T[K]-1008) 4.1249 Kinematic viscosity [m2/s]

5.54E-08×exp(3689/T[K]) 2.46E-06

Dynamic viscosity [kg/ms]

density×5.54E-08×exp(3689/T[K])

1.01E-04

Thermal conductivity [W/mK]

0.928 + 8,397E-05×T[K] 1.0097

Heat capacity [J/kgK]

(-1.111 + 0.00278×T[K]) × 1E+03

1594

Table 2. Physical properties of 77.5% LiF – 22.5% ThF4 molten salt [4]

Based on the thermal power, desired inlet and outlet

temperatures, with the physical properties of the proposed

molten salt taken into consideration general hydraulic properties of the MSFR core can be determined. Based on the core diameter flow in the core is highly turbulent, the Reynolds-number exceeds 1×106 (see Table 3). Based on the flow in the inlet nozzles, or considered as a jet flow, the flow is in the turbulent regime as the Reynolds-number is over 4.9×105.

Core thermal power, Q [W] 3E+09 Average temperature [°C] 700 Temperature difference (inlet-outlet), ∆T [°C]

100

Heat capacity, c [J/kgK] 1594 Density at 700 °C, ρ [kg/m3] 4125 Core mass flow rate, m=Q/(c× ∆T) [kg/s]

3E+09/(1594×100) = 18 820

Single inlet mass flow rate, m/16, [kg/s]

18 820 / 16 = 1176.33

Core diameter, D [m] 2.255 Core cross section, A [m2] 3.994 Core mean velocity, v=m/(A×ρ), [m/s]

1.14

Core Reynolds-numbercore = v×D/ν

1.05E+06

Inlet nozzle diameter, d [m] 0.3 Inlet nozzle average velocity, vin [m/s]

4.03

Inlet nozzle Reynolds-numberinlet = vin×d/ν

4.93E+05

Table 3: Nominal hydraulic properties of the MSFR concept

OPTIONS OF EXPERIMENTAL MODELLING OF MSFR

Fluid Selection for Experimental Modelling Objective of the planned experimental model is to

investigate thermal-hydraulic behaviour of the MSFR core under laboratory conditions using the Particle Image Velocimetry (PIV) measurement system installed at the Institute of Nuclear Techniques, Budapest University of Technology an Economics [5]. For working fluid water applied between room temperature and a maximum of about 60 °C is considered. Experience with water and its application for PIV with the suitable seeding particles suggests its use for the experiments, however other substitute liquids cannot be excluded from future considerations.

In case of water Table 4. shows the necessary flow rates at different temperatures considering the original MSFR geometry and Reynolds-number. One can see that in order to keep the nominal Reynolds-number very high flow rates would be needed even at higher temperatures (increased temperature reduces kinematic viscosity). In our investigation we discuss the possibility of reducing the size of the experimental model while maintaining the original Reynolds-number, and also investigate


the application of lower Reynolds-numbers that are still in the fully turbulent regime, namely greater than 1-2×104 [6].

Temp. [°C]

density [kg/m3]

Dynamic viscosity [kg/ms]

Kinematic viscosity [m2/s]

Average core velocity [m/s]

Mass flow rate [kg/s]

Volumetric flow rate [l/min]

20 998.16 0.001003 1.0048E-06 0.47 1869 112 347 50 988.02 0.000547 5.5363E-07 0.26 1019 61 899 80 973.18 0.000355 3.6509E-07 0.17 662 40 819

Table 4: Flow rates at different temperatures, water, Re=1.05×106

Scaled reactor models with water as working fluid Multiple different experimental systems are built and

operated worldwide to investigate thermal-hydraulics behaviour of nuclear power reactors. The ROCOM test facility [7] is a scaled plexiglas model of the Konvoi type pressurized water reactor, and is used to simulate mixing processes in the primary loop and the reactor pressure vessel. The Vattenfall test facility is a model of a three-loop Westinghouse PWR and made of plexiglas as well, while the Gidropress test facility is made of metal [8]. All three are 1:5 scaled models of power reactors and use water at around 20°C, except for the Vattenfall test facility, which uses water at 53.6 °C in order to achieve lower water viscosity [9]. With these three models the Reynolds number in the reactor pressure vessel model sections is around 105, which is two orders of magnitude lower than the nominal values at the nozzles or in the downcomer for the given power reactors (around 107). Purpose of these facilities is to experimentally investigate coolant mixing and to provide measurement data for validation of Computational Fluid Dynamics (CFD) codes and models.

In the frame of the Molten Salt Reactor Experiment (MSRE) programme at Oak Ridge National Laboratory during the design phase two mock-ups of the MSRE reactor were built for the investigations of coolant flow in the inlet region, in the downcomer and in the lower plenum [10]. The first was made of plastic, used water at 35 °C as working fluid and was scaled 1:5. This model was able to reproduce the Reynolds-number in the above mentioned regions of the MSRE reactor. The second mock-up was 1:1 scale and made of metal with plastic windows that allowed looking into the model. In case of this second model the Reynolds-number was not reproduced exactly but was high enough to be in the turbulent regime [11]. These examples suggest that geometrical scaling of the experimental model is plausible, and reduction of the Reynolds-number within a certain range can be acceptable if reproduction of the nominal value is not feasible.

CFD ANALYSES FOR THE SUPPORT OF EXPERIMENTAL MODELLING OF MSFR

Set of Analysed Cases A set of preliminary CFD calculations were carried out in

order to identify the possibilities of designing a scaled

experimental model of the MSFR with the purpose of carrying out PIV measurements. For the calculations the ANSYS CFX 13 three-dimensional CFD code was used. Based on the design concept of MSFR the geometry and the volumetric mesh of the CFD model was generated. With this geometry and the set of input data the following groups of cases were analysed (see Table 5):

• CFD modelling of the MSFR reactor concept defined with molten salt fluid properties, calculations with and without volumetric heat generation in the core. For heat generation uniform distribution and sin(z) axial distribution was considered.

• CFD modelling of possible water-based scaled experimental models: water at 20 °C as working fluid, full scale and downscaled geometry, quarter segment of the original geometry (full scale and scaled versions), at nominal and reduced inlet water flow rates corresponding to nominal and reduced Reynolds-numbers.

Inlet flow rates were set according to the nominal value (1.05×106), and in subsequent calculations flow rates corresponding to Reynolds-numbers of 6×105, 1.5×105 and 5×104 were defined. Not only the reduction of the inlet flow rates but the reduction of the size of the model was investigated. Therefore a set of calculations were carried out with 1:2 scaled models and 1:4.51 scaled geometries. In the latter case the core diameter was set to be 0.5 m. Another set of calculations were run in which we investigated, whether the modelling only one segment (corresponding to 90°) of the geometry could be used in the measurements. This segmentation would further reduce the necessary inflow and pumping power. With these calculations we analysed the effect of heat generation on the flow field inside the core, how the flow rate and size reduction affects the flow in the core region. Geometry and computational domain are shown in Figure 2. For the steady state calculations inlet boundary conditions were set as mass flow rates (accordingly to the desired Reynolds-number), k-ε turbulence model was applied. Tetrahedral volumetric mesh included 1.36 million elements, in case of the quarter segment models the element number was 0.36 million. For the scaled cases rescaled versions of the original geometry and mesh were used.

Full (360°) model1:1 full scaleRe=1 050 000 Re=600 000 Re=150 000 Re=50 000

MSFR molten saltHeat generation Sin(z) FU100_SLT_SIN_Re1000

Uniform FU100_SLT_UNI_Re1000No heat generation FU100_SLT_ZER_Re1000 FU100_SLT_ZER_Re0600FU100_SLT_ZER_Re0150FU100_SLT_ZER_Re0050Water FU100_WAT_Re1000 FU100_WAT_Re0600 FU100_WAT_Re0150 FU100_WAT_Re0050No heat generation 1:2 scaled (D=1.1275 m)

FU050_WAT_Re1000 FU050_WAT_Re0600 FU050_WAT_Re0150 FU050_WAT_Re00501:4.51 scaled (D=0,5 m)FU022_WAT_Re1000 FU022_WAT_Re0600 FU022_WAT_Re0150 FU022_WAT_Re0050Quarter segment (90°) model1:1 full scaleQ1100_WAT_Re1000 Q1100_WAT_Re0600 Q1100_WAT_Re0150 Q1100_WAT_Re00501:2 scaled (D=1.1275 m)Q1050_WAT_Re1000 Q1050_WAT_Re0600 Q1050_WAT_Re0150 Q1050_WAT_Re00501:4.51 scaled (D=0.5 m)Q1022_WAT_Re1000 Q1022_WAT_Re0600 Q1022_WAT_Re0150 Q1022_WAT_Re0050

Table 5. Set of CFD analyses – names correspond to geometry scaling and segmentation, fluid material, Reynolds-number, and modelling method of heat

generation


Figure 2: CFD model: dimensions and boundary

conditions (inlet: black, outlet: yellow); Front and side view with monitor lines (red)

MSFR Calculation Results (FU100_SLT series) A series of calculations were carried out using the molten

salt properties given in Table 2. In the core cylinder (see Figure 2) volumetric heat generation was defined for two cases: sinus axial profile along the vertical (z) axis (FU100_SLT_SIN) and uniform heat generation (FU100_SLT_UNI). To see the effect of neglecting heat generation and the reduction of inlet flow rates the previously defined lower Reynolds-numbers (see

Table 5) were set for the following calculations: FU100_SLT_ZER_Re1000, FU100_SLT_ZER_Re0600, FU100_SLT_ZER_Re0150 and FU100_SLT_ZER_Re0050.

Figure 3 shows the temperature and velocity distribution in the X-Z symmetry plane for calculation FU100_SLT_SIN_Re1000. A zone of higher temperature can be seen at the perimeter of the core, which is generated by the recirculation of the fuel salt near the wall. Qualitative and quantitative analyses (see Figure 4) of the flow field show that near the wall above the inlet nozzles downward flow develops, in the centre of the core the fluid passes with higher velocity.

Figure 3: Temperature and velocity distribution in the

X-Z symmetry plane (FU100_SLT_SIN_Re1500) Velocity and axial velocity component profiles were

extracted along monitor lines located at H/4, H/2 and 3H/4 (where H is the core height equals to 2.255 m) in two symmetry planes (shown in Figure 2). These velocity profiles show that the definition of heat generation does not influence the flow field in the core (see Figure 4a and 4b) and it is basically determined by forced convection. Therefore for further comparisons FU100_SLT_SIN_Re1000 and FU100_SLT_ZER_Re1000 were selected as reference results.


CS1 Velocity [m/s]

0

0,5

1

1,5

2

2,5

3

-1,5 -1 -0,5 0 0,5 1 1,5

FU100_SLT_SIN_Re1000

FU100_SLT_UNI_Re1000

FU100_SLT_ZER_Re1000

CS2 Velocity [m/s]

0

0,5

1

1,5

2

2,5

3

3,5

-1,5 -1 -0,5 0 0,5 1 1,5




CS3 Velocity [m/s]

0

0,5

1

1,5

2

2,5

3

-1,5 -1 -0,5 0 0,5 1 1,5




Figure 4a Velocity distributions along monitor lines

CS1 (H/4), CS2 (H/2) and CS3 (3H/4) Velocity distributions shown in Figure 5a and 5b indicate

that the reduction of the inlet flow rates to Re=6×105, 1.5×105 and 5×104 does not influence the shape of the velocity profiles and the general nature of flow inside the core. Values in Figure 5 are normalized to the maximum values of the given distribution. It is also visible that the recirculation region – with downward flow, i.e. negative axial component values – is limited within the outer 10 percent of the diameter of the core. These results suggest that in nominal case the forced flow rate defines the flow distribution within the core and heat generation modelling might be neglected. It is also suggested by these results that reduction of flow rate – keeping it in the highly turbulent regime – will not alter the results significantly.

CS1 Velocity W [m/s]

-1

-0,5

0

0,5

1

1,5

2

2,5

3

-1,5 -1 -0,5 0 0,5 1 1,5




CS2 W [m/s]

-1

-0,5

0

0,5

1

1,5

2

2,5

3

3,5

-1,5 -1 -0,5 0 0,5 1 1,5




CS3 W [m/s]

-1

-0,5

0

0,5

1

1,5

2

2,5

3

-1,5 -1 -0,5 0 0,5 1 1,5




Figure 4b Axial velocity component (W ) distributions

along monitor lines CS1 (H/4), CS2 (H/2) and CS3 (3H/4)

Our objective is not only to reduce the necessary inlet flow

rate (and pumping power) but to downscale the geometry and carry out experiments using water as working fluid. Calculations for these case are discussed in the following sections.


CS1 V/Vmax [m/s]

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5






CS2 V/Vmax [m/s]

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5






CS3 V/Vmax [m/s]

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5






Figure 5a: Normalized velocity distributions along monitor lines CS1 (H/4), CS2 (H/2) and CS3 (3H/4)

Results of Modelling with Water Substitution, Reduction of Flow Rate and Downscaling (FU100_WAT, FU050_WAT, FU022_WAT series)

In order to examine the feasibility of a resized water-based model of MSFR calculations were carried out simulating water as working fluid. Series of calculations were done for each size (1:1 original, 1:2 scaled and 1:4.51 scaled) with nominal and reduced Reynolds-number as described in Section “Set of Analysed Cases”. Selected results presented in Figure 6-8 indicate that by neglecting the volumetric heat generation in the core, using water as substitute fluid and the aforementioned reductions of inlet flow rate (FU100WAT series) will result in flow distributions very similar to the reference results. Values in Figure 6-8 are normalized to the maximum values (velocity, velocity component and radius) of the given distribution.

CS1 W/Wmax [m/s]

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5






CS2 W/Wmax [m/s]

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5






CS3 W/Wmax [m/s]

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5






Figure 5b: Normalized axial velocity component (W)

distributions along monitor lines CS1 (H/4), CS2 (H/2) and CS3 (3H/4)


CS2 V/Vmax [m/s]

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



FU100_WAT_Re1000

FU100_WAT_Re0600

FU100_WAT_Re0150

FU100_WAT_Re0050

CS2 W/Wmax [m/s]

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



FU100_WAT_Re1000

FU100_WAT_Re0600

FU100_WAT_Re0150

FU100_WAT_Re0050

Figure 6: Normalized velocity and axial velocity

component distribution, monitor line CS2 (H/2) (FU100_WAT series)

Y2 V/Vmax

0

0.2

0.4

0.6

0.8

1

1.2

-1.5 -1 -0.5 0 0.5 1 1.5



FU050_WAT_Re1000

FU050_WAT_Re0600

FU050_WAT_Re0150

FU050_WAT_Re0050

Y2 W/Wmax

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

-1.5 -1 -0.5 0 0.5 1 1.5



FU050_WAT_Re0600

FU050_WAT_Re1000

FU050_WAT_Re0150

FU050_WAT_Re0050

Figure 7: Normalized velocity and axial velocity

component distribution, monitor line Y2 (H/2, between nozzles) (FU050_WAT series)

CS3 V/Vmax

0

0.2

0.4

0.6

0.8

1

1.2

-1.5 -1 -0.5 0 0.5 1 1.5



FU022_WAT_Re1000

FU022_WAT_Re0600

FU022_WAT_Re0150

FU022_WAT_Re0050

CS3 W/Wmax

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

-1.5 -1 -0.5 0 0.5 1 1.5



FU022_WAT_Re1000

FU022_WAT_Re0600

FU022_WAT_Re0150

FU022_WAT_Re0050

Figure 8: Normalized velocity and axial velocity component distribution, monitor line CS3 (3H/4)

(FU022_WAT series)

Results of Quarter Segment Model with Water (Q1100_WAT, Q1050_WAT, Q1022_WAT series)

In order to further reduce the necessary water flow rates for a possible experiment calculations were carried out simulating only one quarter segment of the original geometry. Series of simulations were done for each size (1:1 original, 1:2 scaled and 1:4.51 scaled) with nominal and reduced Reynolds-numbers as described above. Results show that in the inner regions of the model, further away from the vertical bounding walls and away from the corner where the walls are connected flow behaviour is not affected by the walls. Influence of the corner is limited only to a narrow region near the corner. Wall effects can be neglected if we only investigate the flow between the four pair of nozzles, the similarity is even stronger when only the region between the two inner nozzle pairs is considered (Figure 9 a-d). Significant difference between the flow distribution in the quarter segment models and the full model only appear close to the vertical walls and their junction at the centre.


a,

b,

c,

d, Figure 9: Streamlines (a), axial velocity component (W) distributions at different elevations: b - H/4, x –

H/2, d – 3H/4 (Q1100_WAT_Re1000) (b-d) Selection of extracted calculation results are shown in

Figure 10-12, where values are normalized by the maximum

values and for scaled geometries radial distance is normalized by the radius of the core region. The results show that the difference in velocity values induced by the geometrical difference are very small and localized in a very narrow region.

CS3 V/Vmax

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



Q1100_WAT_Re1000

Q1100_WAT_Re0600

Q1100_WAT_Re0150

Q1100_WAT_Re0050

CS3 W/Wmax

-0,6

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



Q1100_WAT_Re1000

Q1100_WAT_Re0600

Q1100_WAT_Re0150

Q1100_WAT_Re0050

Figure 10: Normalized velocity and axial velocity component distribution, monitor line CS3 (3H/4)

(Q1100_WAT series)

CS2 V/Vmax

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



Q1050_WAT_Re1000

Q1050_WAT_Re0600

Q1050_WAT_Re0150

Q1050_WAT_Re0050

CS2 W/Wmax

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



Q1050_WAT_Re1000

Q1050_WAT_Re0600

Q1050_WAT_Re0150

Q1050_WAT_Re0050

Figure 11: Normalized velocity and axial velocity component distribution, monitor line CS2 (H/2)

(Q1050_WAT series)


CS2 V/Vmax

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



Q1022_WAT_Re1000

Q1022_WAT_Re0600

Q1022_WAT_Re0150

Q1022_WAT_Re0100

Q1022_WAT_Re0050

CS2 W/Wmax

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

-1,5 -1 -0,5 0 0,5 1 1,5



Q1022_WAT_Re1000

Q1022_WAT_Re0600Q1022_WAT_Re0150

Q1022_WAT_Re0100

Q1022_WAT_Re0050

Figure 12: Normalized velocity and axial velocity component distribution, monitor line CS2 (H/2)

(Q1022_WAT series) The general profile of the flow at different elevations are

well reproduced with the segmented geometry even at reduced dimensions (namely calculation series Q1050_WAT and Q1022_WAT) and inlet flow rates. The shapes of the velocity profiles are in good agreement, however it can be observed that the velocity (axial component) decrease toward the cylindrical wall of the core region increases with smaller models, but significant difference is limited to a region defined by about a maximum of 10% of the radius. It is also clear that the inner region closest to the corner (connections of vertical boundaries) does not represent well the flow behaviour of the original model, and will have to be neglected in case of scaled and segmented water based experiments

CONCLUSIONS Based on the properties of the single region homogeneous

molten salt reactor concept MSFR results of preliminary CFD calculations carried out for the support of the design of a possible small-scale experimental mock-up were presented and discussed. In the paper we summarised the main properties and limitations of existing scaled and water-based experimental facilities that model power reactors, including the mock-ups constructed during the design phase of the Molten Salt Reactor Experiment programme at Oak Ridge National Laboratory. Based on the experience with these test facilities CFD models were built and calculation results were presented in order to examine the possibility to build a water-based, scaled and segmented experimental model of the MSFR concept. The

purpose of such a mock-up would be to carry out PIV measurements to determine velocity distribution and flow behaviour in the core. On one hand it would help to validate and enhance CFD models built for the evaluation of the MSFR, and on the other hand it would help to identify important phenomena or particular regions of the concept where design improvements may be considered.

Results of the calculations presented in the paper indicate that a limited reduction of flow rates – maintaining it over Re~104 in order to stay in the fully turbulent regime – and the scaling of the geometry will not affect significantly the flow field in the core region, and velocity fields could be reproduced with water as fluid and with neglecting volumetric heat generation in the core. The selected results presented in the paper show that a quarter segment of the original geometry can be representative for the bulk of the flow, with marginal regions near walls to be left out of consideration.

Based on these analyses and considerations a scaled model of the MSFR is going to be designed and built, possibly with a diameter of the core region not greater than D=0.5 m and a reduced water flow rate corresponding to a Reynolds-number between 5×104-5×105. Particle image velocimetry measurements on this experimental model will provide the opportunity to validate CFD models aiming to analyse the MSFR design and to experimentally investigate specific problems or phenomena – inlet geometry, effect of internal structures, mixing – in the MSFR design.

ACKNOWLEDGMENTS The research leading to these results has received funding

from the European Community's Seventh Framework Programme under grant agreement n°249696 EVOL.

REFERENCES 1. S. Delpech, E. Merle-Lucotte, D. Heuer, M. Allibert, V.

Ghetta, C. Le-Brun, L. Mathieu, G. Picard: Reactor physics and reprocessing scheme for innovative molten salt reactor system, Journal of Fluorine Chemistry, 130 Issue 1, pp. 11-17 (2009)

2. S. Delpech, S. Jaskierowicz, G. Picard, E. Merle-Lucotte, D. Heuer, X. Doligez: Innovative nuclear system based on liquid fuel, Proceedings of ICAPP ’09, Tokyo, Japan, May 10-14, 2009 (2009)

3. E. Merle-Lucotte, X Doligez, D. Heuer, M. Allibert, V. Ghetta: Simulation tools and new developments of the molten salt fast reactor, Transactions of ENC 2010, 30 May - 2 June 2010, Barcelona, Spain (2010)

4. MOLTEN SALT FAST REACTOR Reference configuration – 15th of March, 2012, EVOL - Evaluation and Viability of Liquid Fuel Fast Reactor System, EU 7th Framework Programme

5. B. Yamaji, R. Szijártó, A. Aszódi: Investigation of Mixing and Thermal Stratification with PIV, Proceedings of the 8th International Topical Meeting on Nuclear Thermal-


Hydraulics, Operation and Safety (NUTHOS-8), October 10-14, 2010, Shanghai, China, N8P0302 (2010)

6. P. E. Dimotakis: The mixing transition in turbulent flows, Journal of Fluid Mechanics, 409, pp. 69-98 (2000)

7. T. Höhne, S. Kliem, U. Bieder: Modeling of a buoyancy-driven flow experiment at the ROCOM test facility using the CFD codes CFX-5 and Trio U, Nuclear Engineering and Design, 236, pp. 1309-1325 (2006)

8. S. Kliem, B. Hemström, Y. Bezrukov, T. Höhne, and U. Rohde: Comparative Evaluation of CoolantMixing Experiments at the ROCOM, Vattenfall, and Gidropress Test Facilities, Science and Technology of Nuclear Installations, Volume 2007, Article ID 25950, Hindawi (2007)

9. T. V. Dury, B. Hemström, V. Shepel: CFD Simulation of the Vattenfall 1/5th-Scale PWR Model for Boron Dilution Studies, Benchmarking of CFD Codes for Application to Nuclear Reactor Safety (CFD4NRS) Workshop Proceedings, Garching (Munich), Germany, 5-7 September 2006 OECD NEA (2006)

10. R. J. Kedl: Fluid Dynamic Studies of the Molten-Salt Reactor Experiment (MSRE) Core, ORNL-TM-3229, Oak Ridge National Laboratory (1970)

11. R. B. Briggs: Molten-Salt Reactor Program Semiannual Progress Report For Period Ending February, 28, 1962, ORNL-3282, Oak Ridge National Laboratory (1962)

Documents

Preliminary Thermal-Hydraulic Analyses for …oldweb.reak.bme.hu/fileadmin/user_upload/felhasznalok/...ThF4 molten salt [4] Based on the thermal power, desired inlet and outlet temperatures,