STH-CFD Codes Coupled Calculations Applied to HLM Loop and ...

Research ArticleSTH-CFD Codes Coupled Calculations Applied toHLM Loop and Pool Systems

M Angelucci1 D Martelli1 G Barone1 I Di Piazza2 and N Forgione1

1Dipartimento di Ingegneria Civile e Industriale University of Pisa Pisa Italy2Italian National Agency for New Technologies Energy and Sustainable Economic DevelopmentC R ENEA Brasimone Camugnano Italy

Correspondence should be addressed to M Angelucci morenaangelucciforunipiit

Received 26 June 2017 Accepted 29 August 2017 Published 9 October 2017

Academic Editor Tomasz Kozlowski

Copyright copy 2017 M Angelucci et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

This work describes the coupling methodology between a modified version of RELAP5Mod33 and ANSYS Fluent CFD codedeveloped at the University of Pisa The described coupling procedure can be classified as ldquotwo-wayrdquo nonoverlapping ldquoonlinerdquocoupling In this work a semi-implicit numerical scheme has been implemented giving greater stability to the simulations AMATLAB scriptmanages both the codes oversees the reading andwriting of the boundary conditions at the interfaces and handlesthe exchange of data A new tool was used to control the Fluent session allowing a reduction of the time required for the exchange ofdataThe coupling tool was used to simulate a loop system (NACIE facility) and a pool system (CIRCE facility) both working withLead Bismuth Eutectic and located at ENEA Brasimone Research Centre Some modifications in the coupling procedure turnedout to be necessary to apply the methodology in the pool system In this paper the comparison between the obtained couplednumerical results and the experimental data is presented The good agreement between experiments and calculations evinces thecapability of the coupled calculation to model correctly the involved phenomena

1 Introduction

Thermal-hydraulic analyses are a fundamental issue in thedevelopment design and licensing of nuclear power plants(NPPs) The investigation of the plant performance duringaccidental conditions has always been one of the mainconcerns of nuclear safety SystemThermal-Hydraulic (STH)codes such as RELAP5 CATHARE ATHLET and TRACEare widely used in the assessment and improvement of thesafety aspects of existing and new NPPs The Best-Estimate(BE) analysis with STH codes allows the simulation ofaccidental transients as realistically as possible and is usedto assess the values of safety margins BE system codes arecurrently the only ones accepted for licensing purposes pro-vided that they are properly qualified and used together witha methodology to evaluate uncertainties

STH codes are based on two-phase flow one-dimensional(1D) equations where mass momentum and energy con-servation equations are solved for both the liquid andvapor phases separately Several constitutive equations and

empirical correlations are required to close the set of balanceequations [1] The optimization and the improvement ofthe tools used for thermal-hydraulic safety analysis are anongoing process and a permanent objective of the researchefforts One of the major limits of the STH codes is relatedto its one-dimensional approach which prevents the studyof more complex 3D phenomena In this context the useof computational fluid dynamic (CFD) codes has increasedin the last fifteen years CFD codes provide more detailedinformation since the discretization of the domain is highlyrefined They also include models for simulating turbulenceheat transfer multiphase flows and chemical reactionsNevertheless the complexity of the used models and thehigh spatial resolution results in higher computational effortscompared with those required by the STH codes whenapplied to an equivalent domain

In this scenario coupling techniques between STH andCFD codes for thermal-hydraulic analysis purposes areemerging worldwide The main advantage of this method-ology is the possibility of modelling multiscale phenomena

HindawiScience and Technology of Nuclear InstallationsVolume 2017 Article ID 1936894 13 pageshttpsdoiorg10115520171936894

2 Science and Technology of Nuclear Installations

A B

Geometrical domain

CFD domain STH domain

x1 x2 xp

xp+1 xp+2 xq

xq+1 xq+2 xr

A1B1

A2 B2

AmBm

Figure 1 Simplified sketch of a ldquononoverlappingrdquo ldquotwo-waysrdquo coupled scheme

at different level of detail The CFD code should be used toanalyse a smaller part of the domain where 3D effects aresignificant andor detailed flow information is demandedThe STH code should be applied to the portions of the systemcharacterized by 1D components (ie pipes) and to modelmultiphase andor multicomponent flow

First developments and applications of coupling method-ology can be found in Korea in the works of Jeong et al[2 3] and in USA in the works of Aumiller et al [4 5] Morerecent works on coupled calculation applied to water systemscan be found in literature [6ndash9] Moreover couplingmethod-ologies found wide application in GEN-IV systems thermal-hydraulic analysis as in [10] where coupled simulation wasperformed on a gas-cooled reactor In Europe the FrenchAtomic Energy Commission (CEA) developed a couplingtool between the 3D CFD code TRIO U with the BE-STH code CATHARE with applications to PHENIX NaturalCirculationTest [11] GRS Society for Plant andReactor Safety(Germany) worked on the coupling of the 1D System CodeATHLET with the 3D CFD Software Package ANSYS CFX[12] while KTH Royal Institute of Technology developed acoupling methodology between RELAP5 and the CFD codeSTAR-CCM+ In the context of the 7th Framework EU Pro-gram THINS project GRS and KTH applied the developedtools to the Heavy Liquid Metal (HLM) system TALL-3DThe comparison of the two coupling approaches is reportedin [13] At the Belgian Nuclear Research Centre (SCKsdotCEN)a numerical algorithm of a STHCFD coupling method wasdeveloped for multiscale transient simulations of pool-typereactors [14]The application of STHcodes toGEN-IV systemrequired the implementation of thermophysical propertiesand correlations for new kind of fluids (eg liquid metals)Balestra et al at the University of Rome in collaborationwith the Idaho National Laboratory (INL) developed a toolfor the generation of new properties binary file needed bythe code RELAP5-3D [15] and implemented thermophysicalproperties for LBE and Lead recommended in [16]

At the University of Pisa (UniPi) the relationship usedby the RELAP5Mod33 to generate the table of lead LBElead-lithium and sodium properties were modified [17] inagreement with the ones suggested in [18] and then taken asthe reference in [16] Further specific convective heat transfercorrelations for LMs have been implemented inside the codeThis modified version of RELAP5Mod33 was adopted at theUniversity of Pisa in order to develop an in-house couplingmethodology between the STH code and Fluent commercialCFD code since 2012 The first release of the coupling toolused an explicit scheme for the advancement in time Thistool was applied to the NACIE facility with good agreementof the obtained results against experimental data [19 20] Inthe present work the improvements of the coupling method-ology developed atUniPi such as the semi-implicit numericalscheme are presented in Section 2 The coupling tool hasbeen applied to NACIE and CIRCE respectively a loop and apool facility designed and built at the Italian National Agencyfor New Technologies Energy and Sustainable EconomicDevelopment (ENEA) in Brasimone RC both using a HLMas working fluid The computational domain of the twofacilities and the main results of the coupled calculationare presented respectively in Sections 3 and 4 The mainconclusions of this work are summarized in Section 5

2 Coupling Procedure

The coupling approach consists in a ldquononoverlappingrdquo ldquotwo-wayrdquo coupling scheme In a ldquononoverlappingrdquo strategy theoverall domain is divided into some regions modelled usingthe CFD approach (119860) and other regions modelled using asystem code (119861) similar to the simplified scheme reportedin Figure 1 From the geometrical domain subdivision oneor more ldquocouplingrdquo interfaces (Π) are conveniently selectedhere the thermofluid-dynamics data are transferred from theCFD code to the STH-code-region and vice versa (ldquotwo-way couplingrdquo) Several and different physical parameters

Science and Technology of Nuclear Installations 3

Read CFD resultsWrite RELAP5 input

Read RELAP output

End of transient

End of run

Yes

No

n = n + 1

Start of run n = 0 t0 = 0

n = 1

tn = tnminus1 + Δt

CFD run tnminus1 rarr tn

RELAP5 run tnminus1 rarr tn

(a)

n = n + 1

j = j + 1

Read CFD resultsWrite RELAP5 BC

Read RELAP results

End of transient

End of run

Yes

No

No

Yes

n = 1

0nxi = ＞efault values

j = 1

tn = tnminus1 + Δt



forallvariable ijnxi minus

jminus1n xi

jnxi

lt i

Start of run n = 0 j = 0 t0 = 0

(b)

Figure 2 Logic of the explicit (a) and semi-implicit (b) numerical schemes

(eg mass flow velocity pressure and temperature) can bepassed at each interface Π The number and the type ofexchanged quantities are not necessary the same for eachinterface (referring to Figure 1 119901 not necessarily equal to 119902minus119901andor 119903 minus 119902)

If the set of CFD and STHbalance equations weremergedinto a single system and solved by a unique code a so-called ldquomonolithic solutionrdquo would be implemented Yet thismethod requires major modifications to the source codes(if available) Alternatively a ldquopartitioned solutionrdquo can beadopted where each code solves independently its own setof equations and a coupling interface is required to exchangethermal-hydraulic variables at the interfaces In this workthe ldquopartitioned solutionrdquo approach was chosen and a thirdsoftware MATLAB was selected to manage the couplinginterfaces An appropriate MATLAB script handles the codessynchronization drives the execution of both the solversenables the exchange and processes the data when required

In the previous work of Martelli et al [20] an explicitnumerical scheme was implemented In Figure 2 the logicof the explicit (a) scheme compared to the semi-implicit

(b) scheme is illustrated Regarding the explicit scheme atthe beginning of the run (119899 = 1) the Fluent (mastercode) advances first from the initialization time t0 = 0 upto the end of the first time step (Δt) Suitable CFD resultsare used as BCs for a RELAP5 simulation Accordingly theRELAP5 (slave code) computation follows and its resultsare temporarily saved in MATLAB in order to be usedas Fluent BC in the following time step At the end ofboth the runs (Fluent and RELAP5) the coupled calculationtransient time advances (119899 = 119899 + 1) The main drawbackof the explicit method is that the conservation of massmomentum and energy at the interfaces is not necessarilyguaranteed and very small time step could be mandatoryto avoid numerical instabilities The time step is usuallychosen considering the involved physical phenomena alwaysrespecting the CourantndashFriedrichsndashLewy (CFL) conditionfor the stability The chosen value is used as time step in theCFD calculation and as duration of each RELAP5 transientcalculation To reduce the computational cost and improvethe numerical stability a semi-implicit numerical schemewasdeveloped and implemented in the present work Since the


Fluent

MATLAB

RELAP5

CFDresultstxt

TUI commands

CORBA

inter

face

RELAP5executionRELAP5

outrst

Read CFDresults

Write R5inputRead R5

results

Check ofconvergence

Selection offluent file Write coupled

calculation output every time step (txt)

Save data after kexternal iterations

(n minus 1 or n)

Figure 3 Coupled code management through the MATLAB software

two codes solve independently their own set of equations(ldquopartitioned solutionrdquo) it is not possible to implement afully implicit numerical scheme which needs the solutionof a single system of discretized equations Nevertheless itis common practice to refer to this semi-implicit method asldquoimplicitrdquo as it could be done later on In the semi-implicitscheme (Figure 2(b)) a generic external iteration (119899) beginswith the time step advancement Δt The choice of the timestep is mainly guided by the CFD portion of the calculationconsidering the physical phenomena involvedThe same timestep is used in RELAP5 as total duration of each transientcalculation An inner iteration (119895 = 1) also starts withthe calculation of the master code Fluent The CFD resultsare set as BCs for the STH code RELAP5 which runs forthe same time step and its results are read by MATLABand stored in a ldquotxtrdquo file At this point the convergenceis checked by comparing the updated value of all thermal-hydraulic quantities xi at each interface Π with the results ofthe previous iteration stored in MATLAB If the convergencecriterion is satisfied a new external iteration 119899 + 1 beginsand the updated RELAP5 results are transmitted to a newFluent calculation (from 119899 to 119899 + 1) Otherwise a new inneriteration is performed (119895 = 119895 + 1) and the last RELAP5results are transmitted again to the Fluent calculation relatedto the time step 119899 The convergence parameter 120576i is usuallythe same for each type of physical parameter (mass flowvelocity temperature pressure etc) and is kept constantduring the whole simulation This approach assures thecontinuity at the physical interfaces and the conservationof the main physical parameters (mass momentum andenergy) In fact each code is verified independently whereas

the succession of inner iterations ensures the coherencebetween the information transferred from the CFD outletboundary to the RELAP5 inlet boundary and the informationtransferred from theRELAP5 outlet boundary to theRELAP5inlet boundary

A new logic developed to manage the Fluent code fromMATLAB is also illustrated in the present work The basicidea of the new methodology is to execute the Fluent codedirectly from MATLAB at the beginning of the simulationin Server mode (ldquoFluent as a Serverrdquo) on the Local PCused The use of CORBA add-ons for Fluent [21] allowsthe establishment of an interface between MATLAB andldquoFluent as a Serverrdquo so that it is possible to manage theFluent session through MATLAB Text User Interface (TUI)specific commands This method brought about a greatadvantage in terms of global computational time with respectto the method used before (about 4 times faster) avoidingfurthermore the use of a parallelized User Define Function(UDF) and journal files to exchange data between Fluent andRELAP5

Following the scheme of Figure 3 after the CFD com-putation the MATLAB script is in charge of reading theresults necessary as BCs for RELAP5 The above-mentioneddata are written in the input file for RELAP5 calculationwhich is then launched Similar to what is described beforethe results from RELAP5 computation are read through theMATLAB script and the information needed to be exchangedwith Fluent is extracted Then the MATLAB process checksthe convergence and consequently a new inner or externaliteration occurs RELAP5 and Fluent files are saved andoverwritten at the end of each inner iteration whereas


Expansionvessel HX

FPS

Figure 4 Isometric view of NACIE primary loop

converged calculations and main results are saved at the endof each time step (outside the inner iterations)

3 Application to Loop Facility

31 NACIE Facility NACIE [22] is a Lead Bismuth Eutectic(LBE) loop-type facility It consists of a rectangular loopmade of two vertical stainless steel pipes (2510158401015840 Sch 40) 75mlong acting as riser and downcomer connected with twohorizontal pipes 1m long The layout of the facility is shownin Figure 4 A heat source is installed in the bottom partof the riser and consists of two electrical heated rods withnominal thermal power of about 43 kW A waterLBE ldquotubein tuberdquo counterflow heat exchanger (HX) is placed on theupper part of the downcomer The secondary side is fed bywater at low pressure The facility can work in both naturaland gas enhanced circulation conditions Indeed argon canbe injected downstream the heat source through a 10mmdiameter pipe housed inside the riser from the expansionvessel

The computational domain for RELAP5Fluent coupledcalculations of theNACIE facility already described and usedin previous work [20] is depicted in Figure 5 It is dividedinto two nonoverlapping regions the CFD code was usedto simulate the Fuel Pin Simulator (FPS) representing theheat source whereas the remaining parts of the loop and theHX (both primary and secondary sides) were modelled withthe STH code RELAP5 Referring to Figure 5(a) Pipe-130represents the riser while TDV-400 TDJ-405 and Branch-125 at bottom of Pipe-130 simulate the injection of the argonwhich promote the circulation of the LBE The expansionvessel (from Pipe-146 to Pipe-156) allows the separationbetween argon and LBE so that the LBE can flow throughthe horizontal pipe and then through the HX (Pipe-180) and

the downcomer (Pipe-200 and Pipe-206) The secondary sidewater system was modelled by means of TDV-500 TDJ-505 the HX secondary side annular zone (Annulus-510) andTDV-520

TheFPSwasmodelled by the Fluent codewith both a sim-plified 2D axial-symmetric domain and a 3D model shownrespectively in Figures 6(a) and 6(b) The 2D geometricalmodel was discretized by a structured mesh composed of7668 rectangular cells uniformly distributed in both the axialand radial coordinates Tomodel the FPS form losses mainlydue to spacer grids a constant pressure drop coefficient wasassumed in the 2D domain For this purpose five distinctinterior faces were set as ldquoporous jumprdquo with each onecharacterized by an equivalent constant local pressure dropcoefficient of 07 The 3D geometrical domain was modelledusing a symmetry plane as shown in Figure 6(b) Theelectrical pins were not included in the domain and heatflux boundary condition (BC) was used at the pin walls Thedomain was discretized using 141045 hexahedral elementswith refinement close to pins wall near the inlet and outletsections One interior face was set as a porous jump and anequivalent constant coefficient of concentrated pressure dropequal to 05 was set in order to simulate the pressure drop dueto the spacer grid not simulated in the CFD model

Concerning the coupled calculation the exchange of datatakes place at the inlet and outlet of the FPS following thescheme displayed in Figure 5(b) Fluent gives the mass flow2and temperature119879

2 at the outlet of the FPS to RELAP5 in

TDJ-115 andTDV-112 the pressure1199011 obtained from the inlet

section of the CFD domain is set in TDV-110 in the RELAP5domain Similarly the CFD BCs at inlet section that is massflow

1and temperature 119879

1 are obtained from the values

computed from RELAP5 in the last volume of Pipe-100 whilethe outlet surface BC pressure 119901

2is obtained from the value

computed in the first volume of Pipe-120

32 Test Matrix The application to NACIE facility ofRELAP5CFD coupled calculations concerned an isothermaltest in gas enhanced circulation (Test 206) During theexperimental test the FPS was switched off and the HX wasempty and nonactive The LBE temperature remained in therange of 200ndash250∘C The experimental test matrix is shownin Table 1 which reports the adopted boundary conditionsTest 206 started with LBE at rest in the facility (0 kgs massflow rate) then the gas mass flow controller was switchedon to a nominal value of 2Nlmin (Step 1) The gas flowrate was kept constant for the time needed to have a steadystate statistically significant After that the other steps camein succession increasing or decreasing the Ar flow rate withthe order delineated in Table 1

33 Results The coupled calculation started with the facilityat rest and no gas injected in the riser After the injection ofargon was activated (flow rate to 2Nlmin) the LBE massflow rate increased up to a value of about 77 kgs as shownin the first step of Figure 7(a) Increasing the argon flow rate(4-5-6-8-10Nlmin) the LBE mass flow rate increased upto a maximum of about 14 kgs In the second part of thetest gas injection was decreased symmetrically with respect


PIPE 100

TDJ-115

160

Expansionvessel

210

125

130

146

148

150

156

152

170

172

180

510

TDV-500(water in)

HX

200

206

TDJ-505

TDJ-405

TDV-520(water out)

TDV-320(argon out)

TDV-400(argon in)

TDV112

Reference cell forpressure data needed for fluent outlet bc

Reference junction forflowrate data needed

for inlet fluent bcTDV110

Reference sectionfor pressure data neededfor RELAP5 outlet bc

Reference sectionfor flowrate andtemperature data neededfor RELAP5 inlet bc

Reference cellfor temperature dataneeded for inlet fluent bc

2D3D domain used for fluent code

in the coupled simulationsPart of the nodalization of the primary loop used for RELAP5 code in the coupled simulations

PIPE 120

0765m

15 m

5m

75 m

005 m

105 m 11 m

1m

(a)

TDJ-115

TDJ-405

PIPE 120

PIPE 100210

206

200

TDV-400(argon in)

125

TDV110

TDV112

p2

p1

T2

m2

m1 T11m

11 m

(b)

Figure 5 NACIE loop computational domain (a) codes exchanged data (b)

Inletsection

11 mOutletsection

x

for RELAP5 outletbc data

Reference sectionfor RELAP5 inletbc data

Reference section

(a)

Support rod

Electric pins

000

2500

5000

7500

10000 (mm)

(b)

Figure 6 2D (a) and 3D (b) CFD domains of the Fuel Pin Simulator (FPS)


Table 1 Boundary conditions of Test 206 in NACIE facility

TLBE FPS power HX power Nominal Ar flow rate [Nlmin]

200ndash250∘C 0 kW 0kW

Step 1 20 Step 7 80Step 2 40 Step 8 60Step 3 50 Step 9 50Step 4 60 Step 10 40Step 5 80 Step 11 20Step 6 100

0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5

Coupled 2DCoupled 3D

(a)

ExperimentalRELAP5

Coupled 2D explicitCoupled 2D implicit

0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)

Figure 7 Time trend of the LBE mass flow rate for Test 206 in NACIE facility

to the increasing ramp and the LBE flow rate followed thesame trend As it can be noticed in Figure 7(a) calculatedLBE mass flow rate tended to overestimate the experimentaldata Further a good agreement was found between thecoupled code simulations with the 2D and 3D CFD domainFigure 7(b) shows the comparison between results obtainedadopting both the explicit and implicit coupling schemeusing the same time step (0005 s) and the 2D CFD domainThe results in terms of LBE mass flow rate and pressuredifferences between inlet and outlet section of the FPSshowed differences lower than 1 Nevertheless the use of animplicit scheme allows larger time steps and tends to be morestable than explicit scheme A sensitivity analysis showed thata time step of 0025 s (five times greater than the one adoptedfor the explicit scheme) still allowed us to obtain good resultswithout losing accuracy

The overall result of the isothermal test is reported inFigure 8 in terms of calculated LBEmass flow rate versus theexperimental value averaged over each step of the test Mostof the numerical results stay within an error band of 10withrespect to the experimental results up to a maximum of 12of discrepancy

In this application the main advantage of the coupledcalculation was the possibility of calculating local values inthe portion of the domain modelled with the CFD code Forinstance the 3D CFD model of the NACIE FPS allowed us

0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16

Experimental LBE flow rate (kgs)

Calc

ulat

ed L

BE fl

ow ra

te (k

gs)

RELAP5 Coupled 2D explicitCoupled 2D implicit


+10

minus10

Figure 8 Average values of the LBE mass flow rate for Test F400 inNACIE facility

to analyse the velocity field inside the test section Figure 9(a)shows the contour plot of the velocitymagnitude between theexit of the pin bundle and the outlet section during Step 6


114e + 00111e + 00107e + 00103e + 00991e minus 01953e minus 01915e minus 01877e minus 01839e minus 01801e minus 01763e minus 01724e minus 01686e minus 01648e minus 01610e minus 01572e minus 01534e minus 01496e minus 01458e minus 01419e minus 01381e minus 01343e minus 01305e minus 01267e minus 01229e minus 01191e minus 01153e minus 01114e minus 01763e minus 02381e minus 02000e + 00

(a)


(b)

Figure 9 Velocity contour plot [ms] (a) and vector velocity coloured by z-velocity (b) in the 3D CFD domain during Step 6 of Test 206

of Test 206 The region downward the two pins is evidencedby the stagnant and low velocity LBE caused by the geo-metrical discontinuity The latter region is also highlightedin Figure 9(b) where the velocity vectors coloured by axialvelocity (along 119911-axis) are reported

4 Application to Pool Facility

41 CIRCE Facility CIRCE is a liquid metal-cooled poolfacility realized and located at ENEA Brasimone ResearchCentre It consists of a cylindrical main vessel which can hostseveral kinds of test sections usually hung from a boltedvessel head The Integral Circulation Experiment (ICE) testsection was designed in order to simulate the primary systemof a HLM reactor It is composed of a Venturi-nozzle flowmeter placed downstream the inlet and connected with theheat source (HS) made of 37 pins arranged in a hexagonallattice The upper section of the HS is hydraulically linked tothe fitting volume From this latter a riser leads themain flowpath towards the separator A nozzle is installed in the lowersection of the riser to inject the argon which enhances thecirculation of the primary coolant The separator allows theseparation of the Argon from the LBE coming from the riserThe main flow path then goes in the heat exchanger (HX)Another component is the dead volume placed above theHSwhich accommodates the power supply cables Furthermorea Decay Heat Removal (DHR) system hydraulically decou-pled from the primary system is also installed in the pool Asketch of the facility is displayed in Figure 10 More details onthe CIRCE-ICE facility and its instrumentation can be foundin [23]

The computational domain of the CIRCE-ICE facility wasdecomposed in two regions themain pool and theDHRweremodelled with ANSYS Fluent CFD code the ICE test sectionand the water secondary system of the HX were modelledwith the STH code RELAP5Mod33 Regarding the CFD

Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter

Figure 10 Layout of the CIRCE facility and ICE test section

model of the pool both a simplified 2D and complete 3Dgeometrical domains were developed The 2D geometricalmodel shown in Figure 11(a) was initially developed withthe aim of assessing the methodology employing a limitedthe computational power This model assumes the DHRrsquosaxis as axis of symmetry and the flow area in each sectionis preserved in agreement with the actual facility geometryThe grid is composed of about 242500 elements mainlystructured the first cell thickness close to the wall is suchthat 119910+ gt 30 Nevertheless the laminar model was adoptedfor the wall treatment in this work The 3D CFD domain ofthe pool and the DHR is shown in Figure 11(b) It represents


Cover gas Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230

tubes tubes tubes54 7

P1 T1

m1

(b)

Figure 11 CIRCE-ICE computational domain for coupled calculations with 2D (a) and 3D (b) CFD models

the ICE test section and the DHR The grid is composed of1450000 cells bothwith tetrahedral and hexahedral elementsThe cover gas above the LBEpoolwas considered applying theVOF mathematical model in Fluent RELAP5Mod33 wasadopted for simulating the ICE test section Figure 11 showsthe nodalization of the test section which models the FPS(Pipe-060 represents the active length) the fitting volumewith Pipe-090 the riser (Pipe-130) the gas injection systemthanks to TDV-003 TDJ-004 and Branch-100 the separator(Pipe-135) and the HX (the primary side is Pipe-170) Thesecondary side of the HX was modelled with three series ofpipes comprised between TDV-005 and TDV-006

Concerning the coupling strategy the application to thepool system CIRCE differs from the application to the loopNACIE The main difference involves the thermodynamicvariables exchanged at the interfaces between the domainsThe scheme used for NACIE (Figure 5) is unstable whenapplied to a pool system since a small variation of the pres-sure given by RELAP5 to Fluent at the outlet of the pool (testsection inlet) resulted in very large variation of the LBEmassflow rate entering the pool (Fluent domain inlet) A variationin the pressure boundary condition at the pool outlet causesa variation in the LBE level in the pool even small leveldifference needs to be adjusted through significant LBE massflow rate at the pool inlet mainly due to the large dimensionof the pool Therefore a new scheme was used according toFigure 11 In this case Fluent gives to RELAP5 the pressuresat both the inlet and outlet sections of ICE whereas RELAP5gives the mass flow at the inlet and outlet of the pool (CFDdomain)The temperature at the exit of the HX is given to theFluent pool inlet by RELAP5 and vice versa Fluent gives thetemperature at the inlet of ICE to TDV-010 of RELAP5 Thisstrategy yielded a stable behaviour to the simulation of large

pool-type facility like CIRCE ensuring that the depth fromthe free surface of the two interfaces is the same for both thedomains so that the pressure due to the column of fluid iscorrectly considered in both the codes

42 Test Matrix The RELAP5CFD developed model of theCIRCE facility was employed to simulate the experimentaltest performed at the ENEA Brasimone and representativeof an isothermal test with gas enhanced circulation Theexperimental test (Test F400) is composed of several stepscharacterized by different amount of argon injected insidethe riser of ICE test section During the test the FPS and theHXwere not activated and the LBE temperature was uniformat about 280∘C The boundary conditions of Test F400 aresummarized in Table 2

43 Results In the coupled simulation the argon was in-jected in the RELAP5 model through the TDJ-004 (seeFigure 11) The gas flow rate was set as boundary conditionin agreement with the nominal experimental value Figures12 and 13 report the main results of this test Figure 12displays the time trend of the LBE mass flow rate for both 2Dand 3D coupled calculations The results are compared withthe experimental values and with the RELAP5 standalonecalculations The comparison is quite satisfying in all casesbut the computationwith 3DCFDdomain bettermatches theexperimental results In addition the 3D model provided amore realistic three-dimensional flow pattern inside the poolSome spikes appeared in the coupled calculations whereasthey did not show up in RELAP5 standalone simulation Thecause of these spikes is related to the choice of the time steptoo large time step can cause the loss of accuracy in thenumerical results In this case the consequence was some


Table 2 Boundary conditions of Test F400 in CIRCE facility

TLBE FPS power HX power Nominal Ar flow rate [Nls]

280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5


Figure 12 Time trend of the LBE mass flow rate for Test F400 inCIRCE facility

oscillations at the beginning of the rapid variation of thegas flow which takes place in few seconds Figure 13 showsthe average value of the LBE mass flow rate for each stepcomputed by the RELAP5 standalone and by the coupledsimulations as function of the experimental values All thenumerical results are set within the error band of 40 asshown in Figure 13

The main LBE flow path inside the pool can be examinedfrom the CFD-modelled portion of the domain Figure 14(a)illustrates the contour plot and vectors of the velocity magni-tude during Step 8 (argon flow rate 5Nls) in the 2D modelIt should be pointed out that the vectors length is normalizedand not proportional to its module for graphical reasonsThemain flow tends to go downward from the heat exchangeroutlet (pool inlet) toward the inlet of the test section (pooloutlet) The LBE enters from the heat exchanger outlet (poolinlet) which is modelled through a thin annular section Atthe inlet the fluid spreads in the large volume of the poolcreating a recirculation zone around it The main flow tendsto go down towards the inlet of the test section (pool outlet)at low velocity due to the large flow area In the lower plenumthe LBE is sucked by the test section inlet There the velocitymagnitude increases up to 08ms (full scale range in thezoomed region around the pool outlet) since the flow areareduces considerably The fluid remains almost stagnant inthe upper part of the pool Some similar considerations can

50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)

RELAP5Coupled 2DCoupled 3D

+4

minus4

Figure 13 Average values of the LBEmass flow rate for Test F400 inCIRCE facility

be deduced analysing the CFD results obtained with the 3Ddomain shown in Figure 14(b) The LBE enters from thepool inlet heads for the lower plenum and spreads in thebottom of the pool The velocity increases at the pool outletdue to the small cross-section area achieving a maximumvalue of about 065ms However the range in Figure 14(b)was set to 0ndash01ms to better appreciate the velocity variationin the lower part of the pool Furthermore some LBE tendsto go from the pool bottom upward and to recirculate inthe zone surrounding the FPS pipe The LBE in the upperpart of the pool is almost stagnant Vectors in Figure 14(b)have an arbitrary length to visualize more clearly the flowdirection Comparing Figures 14(a) and 14(b) it can benoticed that the main flow pattern from the inlet to theoutlet is represented in both cases with comparable velocitymagnitude Nevertheless some secondary flows as well asrecirculation zones are considerably affected by the geometryof themodel and just the 3Dmodel is able to predict themainLBE stream realistically

5 Conclusions and Perspectives

This work illustrated the main improvements implementedin the coupling methodology already established at the Uni-versity of Pisa between a modified version of RELAP5 code


Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)

Figure 14 Velocity magnitude field in the 2D (a) and 3D (b) CFD domain of the CIRCE pool during Step 8 of Test F400

and ANSYS CFD code A semi-implicit coupling scheme wasdeveloped in which the two codes exchange data withinthe same time step until a specified convergence criterionis satisfied The main advantages introduced by the implicitscheme are represented by an improved stability of the cal-culation and consequently the possibility of adopting largertime steps (compared to the previously developed explicitscheme) which leads to a computational time reductionwhile preserving the results accuracy A further improvementconsisted in the introduction of Fluent CORBA add-onsto manage more efficiently the Fluent session inside theMATLAB platform through the TUI specific commands

The updated coupling tool was used to simulate a loopsystem (NACIE facility) and a pool system (CIRCE facility)The application to a pool system required a different setof thermodynamic variables exchanged at the interfacesIndeed the methodology used for a loop-type system isunstable when applied to pool-type system A new stableapproach was found and employed in this work

Two experimental tests (Test 206 inNACIE andTest F400in CIRCE) performed for the hydraulic characterization ofthe facilities were reproduced employing the coupling tool

to assess the methodology Results related to the simulationof Test 206 were quite satisfying even though it tended tooverestimate the experimental results with 12 discrepancymaximum In the author opinion this discrepancy is linkedto inconsistencies in the modelling of the spacer gridswhich were not directly modelled but taken into accountthrough local pressure drop coefficients The comparison ofnumerical results with experimental data was even moresatisfactory in the simulation of Test F400 in CIRCE Inthis case the numerical results approximate the experimentaldata within an error band 4 of the average experimentalvalue

The main achievement in the use of the developedcoupling tool relies on the possibility of exploiting simulta-neously the benefits of multiple codes The required level ofdetail andor the need of specific physical models determinethe use of CFD or STH codes in different parts of thegeometrical domainTheRELAP5multiphase andmulticom-ponent models are more suitable to investigate two-phaseflow regimes (eg simulation of the gas-induced flow or thechange of phase in the heat exchanger of the secondary watersystem) and in general large portions of the system domain


On the other hand the CFD code is the appropriate toolto reproduce complex 3D thermohydraulic phenomena Inparticular this work focused on the analysis of the flow pathand velocity distribution for both NACIE (pin bundle) andCIRCE (pool) CFDdomains emphasizing the stagnant zonesand recirculation areas The positive assessment reported forthis coupling methodology will permit further utilization tomore complex nonisothermal conditions More specificallythe modelling of heat source andor heat sink will provideamong others accurate information on temperatures dis-tribution heat transfer and thermal stratification phenom-ena

Notations

Symbols

1 Mass flow rate at inlet section [kgs]2 Mass flow rate at outlet section [kgs]1199011 Pressure at inlet section [Pa]1199012 Pressure at outlet section [Pa]1198791 Temperature at inlet section [∘C]1198792 Temperature at outlet section [∘C]

Acronyms

ATHLET Analysis of Thermal Hydraulics ofLEaks and Transients

BC Boundary conditionBE Best-EstimateBWR Boiling water reactorCATHARE Code for analysis of thermal

hydraulics during an accident ofreactor and safety evaluation

CFD Computational fluid dynamicCFL CourantndashFriedrichsndashLewyCIRCE CIRculation EutecticDHR Decay heat removal systemFPS Fuel Pin SimulatorHLM Heavy Liquid MetalHS Heat sourceHX Heat exchangerICE Integral Circulation ExperimentLBE Lead Bismuth EutecticNACIE NAtural CIrculation ExperimentNPP Nuclear power plantPC Personal computerPWR Pressurized water reactorRELAP Reactor loss of coolant analysis

programSTH SystemThermal HydraulicTDJ Time-Dependent JunctionTDV Time-dependent volumeTHINS Thermal Hydraulics of Innovative

Nuclear SystemTRACE TRACRELAP advanced

computational engineTUI Text User InterfaceUDF User defined function

Conflicts of Interest

The authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

This work was performed in the frame of and supportedby the Euratom Seventh Framework Program CollaborativeProject Thermal-Hydraulics of Innovative Nuclear Systems(no 249337) and by the Programmatic Agreement (AdP)between the Italian Ministry of the Economic Development(MSE) and ENEAThe authors gratefully acknowledge ENEABrasimone RC for the concession of experimental data ofNACIE and CIRCE facilities

References

[1] A Petruzzi and F DrsquoAuria ldquoThermal-hydraulic system codes innulcear reactor safety and qualification proceduresrdquo Science andTechnology of Nuclear Installations vol 2008 Article ID 46079516 pages 2008

[2] J-J Jeong S K Sim C H Ban and C E Park ldquoAssessmentof the COBRARELAP5 code using the LOFT L2-3 large-breakloss-of-coolant experimentrdquo Annals of Nuclear Energy vol 24no 14 pp 1171ndash1182 1997

[3] J-J Jeong K S Ha B D Chung and W J Lee ldquoDevelopmentof a multi-dimensional thermal-hydraulic system code MARS131rdquo Annals of Nuclear Energy vol 26 no 18 pp 1611ndash16421999

[4] D L Aumiller E T Tomlinson and R C Bauer ldquoA CoupledRELAP5-3DCFD methodology with a proof-of-principle cal-culationrdquo Nuclear Engineering and Design vol 205 no 1-2 pp83ndash90 2001

[5] D L Aumiller E T Tomlinson and W L Weaver ldquoAn inte-grated RELAP5-3D and multiphase CFD code system utilizinga semi-implicit coupling techniquerdquo Nuclear Engineering andDesign vol 216 no 1-3 pp 77ndash87 2002

[6] T Watanabe Y Anoda and M Takano ldquoSystem-CFD coupledsimulations of flow instability in steam generator U tubesrdquoAnnals of Nuclear Energy vol 70 pp 141ndash146 2014

[7] W Li X Wu D Zhang G Su W Tian and S Qiu ldquoPrelim-inary study of coupling CFD code FLUENT and system codeRELAP5rdquo Annals of Nuclear Energy vol 73 pp 96ndash107 2014

[8] T Feng D Zhang P Song et al ldquoNumerical research onwater hammer phenomenon of parallel pump-valve system bycoupling FLUENT with RELAP5rdquo Annals of Nuclear Energyvol 109 pp 318ndash326 2017

[9] T P Grunloh and A Manera ldquoA novel domain overlappingstrategy for the multiscale coupling of CFD with 1D systemcodes with applications to transient flowsrdquo Annals of NuclearEnergy vol 90 pp 422ndash432 2016

[10] Y Yan and R Uddin ldquoCoupled CFD-System-code Simulationof a Gas Cooled Reactorrdquo in Proceedings of the Internationalconference onMathematics and Computational Methods Appliedto Nuclear Science and Engineering Rio de Janeiro Brazil May2011

[11] D Pialla D Tenchine S Li et al ldquoOverview of the system aloneand systemCFD coupled calculations of the PHENIX NaturalCirculation Test within the THINS projectrdquo Nuclear Engineer-ing and Design vol 290 pp 78ndash86 2015


[12] A Papukchiev G Lerchl C Waata and T Frank ldquoExtension ofthe Simulation Capabilities of the 1D System Code ATHLET byCoupling with the 3D CFD Software Package ANSYS CFXrdquo inProceedings of NURETH13 Kanazawa City Ishikawa PrefectureJapan 2009

[13] A Papukchiev M Jeltsov K Koop P Kudinov and G LerchlldquoComparison of different coupling CFD-STH approaches forpre-test analysis of a TALL-3D experimentrdquo Nuclear Engineer-ing and Design vol 290 pp 135ndash143 2015

[14] A Toti J Vierendeels and F Belloni ldquoImproved numericalalgorithm and experimental validation of a system thermal-hydraulicCFD coupling method for multi-scale transient sim-ulations of pool-type reactorsrdquo Annals of Nuclear Energy vol103 pp 36ndash48 2017

[15] P Balestra F Giannetti G Caruso and A Alfonsi ldquoNewRELAP5-3D lead and LBE thermophysical properties imple-mentation for safety analysis of Gen IV reactorsrdquo Science andTechnology of Nuclear Installations vol 2016 Article ID1687946 15 pages 2016

[16] Nuclear Energy Agency ldquoHandbook on Lead-bismuth Eutec-tic Alloy and Lead Propertiesrdquo In Materials CompatibilityThermal-hydraulics and Technologies OECDNEA 2015

[17] G Barone N Forgione D Martelli and W Ambrosini Systemcodes and a CFD codes applied to loop- and pool-type exper-imental facilitiesrdquo CERSE-UNIPI RL 15302013Adp MSE-ENEA LP2C1 2013 2013

[18] V Sobolev Database of thermophysical properties of liquidmetal coolants for GEN-IVrdquo Scientific report of the Belgiannuclear research centre SCK∙CEN-BLG-1069 November 2010

[19] D Martelli N Forgione G Barone A Del Nevo I Di Piazzaand M Tarantino ldquoCoupled simulations of natural and forcedcirculation tests in Nacie facility using relap5 and Ansys fluentcodesrdquo in Proceedings of the 2014 22nd International Conferenceon Nuclear Engineering ICONE 2014 Prague Czech RepublicJuly 2014

[20] D Martelli N Forgione G Barone and I di Piazza ldquoCoupledsimulations of the NACIE facility using RELAP5 and ANSYSFLUENT codesrdquoAnnals of Nuclear Energy vol 101 pp 408ndash4182017

[21] ANSYS Fluent as a Server Userrsquos Guide ANSYS Inc Release150 November 2013

[22] M Tarantino D Bernardi G Coccoluto et al ldquoIcone18 - 2996natural and gas enhanced circulation tests in the nacie heavyliquid metal looprdquo in Proceedings of the 18th International Con-ference on Nuclear Engineering ICONE18 pp 933ndash942 XirsquoanChina May 2010

[23] M Tarantino D Martelli G Barone I Di Piazza and N For-gione ldquoMixed convection and stratification phenomena in aheavy liquid metal poolrdquo Nuclear Engineering and Design vol286 pp 261ndash277 2015

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of


Journal ofPetroleum Engineering


Industrial EngineeringJournal of


Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of


Journal of


Renewable Energy

Submit your manuscripts athttpswwwhindawicom


StructuresJournal of

International Journal of

RotatingMachinery


EnergyJournal of


Hindawi Publishing Corporation httpwwwhindawicom

Journal of

Volume 201Hindawi Publishing Corporation httpwwwhindawicom Volume 201

International Journal ofInternational Journal of


Nuclear InstallationsScience and Technology of


Solar EnergyJournal of


Wind EnergyJournal of


Nuclear EnergyInternational Journal of


High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


A B

Geometrical domain

CFD domain STH domain

x1 x2 xp

xp+1 xp+2 xq

xq+1 xq+2 xr

A1B1

A2 B2

AmBm

Figure 1 Simplified sketch of a ldquononoverlappingrdquo ldquotwo-waysrdquo coupled scheme

at different level of detail The CFD code should be used toanalyse a smaller part of the domain where 3D effects aresignificant andor detailed flow information is demandedThe STH code should be applied to the portions of the systemcharacterized by 1D components (ie pipes) and to modelmultiphase andor multicomponent flow

First developments and applications of coupling method-ology can be found in Korea in the works of Jeong et al[2 3] and in USA in the works of Aumiller et al [4 5] Morerecent works on coupled calculation applied to water systemscan be found in literature [6ndash9] Moreover couplingmethod-ologies found wide application in GEN-IV systems thermal-hydraulic analysis as in [10] where coupled simulation wasperformed on a gas-cooled reactor In Europe the FrenchAtomic Energy Commission (CEA) developed a couplingtool between the 3D CFD code TRIO U with the BE-STH code CATHARE with applications to PHENIX NaturalCirculationTest [11] GRS Society for Plant andReactor Safety(Germany) worked on the coupling of the 1D System CodeATHLET with the 3D CFD Software Package ANSYS CFX[12] while KTH Royal Institute of Technology developed acoupling methodology between RELAP5 and the CFD codeSTAR-CCM+ In the context of the 7th Framework EU Pro-gram THINS project GRS and KTH applied the developedtools to the Heavy Liquid Metal (HLM) system TALL-3DThe comparison of the two coupling approaches is reportedin [13] At the Belgian Nuclear Research Centre (SCKsdotCEN)a numerical algorithm of a STHCFD coupling method wasdeveloped for multiscale transient simulations of pool-typereactors [14]The application of STHcodes toGEN-IV systemrequired the implementation of thermophysical propertiesand correlations for new kind of fluids (eg liquid metals)Balestra et al at the University of Rome in collaborationwith the Idaho National Laboratory (INL) developed a toolfor the generation of new properties binary file needed bythe code RELAP5-3D [15] and implemented thermophysicalproperties for LBE and Lead recommended in [16]

At the University of Pisa (UniPi) the relationship usedby the RELAP5Mod33 to generate the table of lead LBElead-lithium and sodium properties were modified [17] inagreement with the ones suggested in [18] and then taken asthe reference in [16] Further specific convective heat transfercorrelations for LMs have been implemented inside the codeThis modified version of RELAP5Mod33 was adopted at theUniversity of Pisa in order to develop an in-house couplingmethodology between the STH code and Fluent commercialCFD code since 2012 The first release of the coupling toolused an explicit scheme for the advancement in time Thistool was applied to the NACIE facility with good agreementof the obtained results against experimental data [19 20] Inthe present work the improvements of the coupling method-ology developed atUniPi such as the semi-implicit numericalscheme are presented in Section 2 The coupling tool hasbeen applied to NACIE and CIRCE respectively a loop and apool facility designed and built at the Italian National Agencyfor New Technologies Energy and Sustainable EconomicDevelopment (ENEA) in Brasimone RC both using a HLMas working fluid The computational domain of the twofacilities and the main results of the coupled calculationare presented respectively in Sections 3 and 4 The mainconclusions of this work are summarized in Section 5

2 Coupling Procedure

The coupling approach consists in a ldquononoverlappingrdquo ldquotwo-wayrdquo coupling scheme In a ldquononoverlappingrdquo strategy theoverall domain is divided into some regions modelled usingthe CFD approach (119860) and other regions modelled using asystem code (119861) similar to the simplified scheme reportedin Figure 1 From the geometrical domain subdivision oneor more ldquocouplingrdquo interfaces (Π) are conveniently selectedhere the thermofluid-dynamics data are transferred from theCFD code to the STH-code-region and vice versa (ldquotwo-way couplingrdquo) Several and different physical parameters



Read RELAP output

End of transient

End of run

Yes

No

n = n + 1


n = 1

tn = tnminus1 + Δt



(a)

n = n + 1

j = j + 1


Read RELAP results

End of transient

End of run

Yes

No

No

Yes

n = 1


j = 1

tn = tnminus1 + Δt




jminus1n xi

jnxi

lt i


(b)







Fluent

MATLAB

RELAP5

CFDresultstxt

TUI commands

CORBA

inter

face


outrst

Read CFDresults


results

Check ofconvergence




(n minus 1 or n)







Expansionvessel HX

FPS




















PIPE 100

TDJ-115

160

Expansionvessel

210

125

130

146

148

150

156

152

170

172

180

510

TDV-500(water in)

HX

200

206

TDJ-505

TDJ-405

TDV-520(water out)

TDV-320(argon out)

TDV-400(argon in)

TDV112









PIPE 120

0765m

15 m

5m

75 m

005 m

105 m 11 m

1m

(a)

TDJ-115

TDJ-405

PIPE 120

PIPE 100210

206

200

TDV-400(argon in)

125

TDV110

TDV112

p2

p1

T2

m2

m1 T11m

11 m

(b)


Inletsection

11 mOutletsection

x



Reference section

(a)

Support rod

Electric pins

000

2500

5000

7500

10000 (mm)

(b)







0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5


(a)

ExperimentalRELAP5


0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)





0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)



+10

minus10





(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of
















Read RELAP output

End of transient

End of run

Yes

No

n = n + 1


n = 1

tn = tnminus1 + Δt



(a)

n = n + 1

j = j + 1


Read RELAP results

End of transient

End of run

Yes

No

No

Yes

n = 1


j = 1

tn = tnminus1 + Δt




jminus1n xi

jnxi

lt i


(b)







Fluent

MATLAB

RELAP5

CFDresultstxt

TUI commands

CORBA

inter

face


outrst

Read CFDresults


results

Check ofconvergence




(n minus 1 or n)







Expansionvessel HX

FPS




















PIPE 100

TDJ-115

160

Expansionvessel

210

125

130

146

148

150

156

152

170

172

180

510

TDV-500(water in)

HX

200

206

TDJ-505

TDJ-405

TDV-520(water out)

TDV-320(argon out)

TDV-400(argon in)

TDV112









PIPE 120

0765m

15 m

5m

75 m

005 m

105 m 11 m

1m

(a)

TDJ-115

TDJ-405

PIPE 120

PIPE 100210

206

200

TDV-400(argon in)

125

TDV110

TDV112

p2

p1

T2

m2

m1 T11m

11 m

(b)


Inletsection

11 mOutletsection

x



Reference section

(a)

Support rod

Electric pins

000

2500

5000

7500

10000 (mm)

(b)







0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5


(a)

ExperimentalRELAP5


0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)





0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)



+10

minus10





(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of















Fluent

MATLAB

RELAP5

CFDresultstxt

TUI commands

CORBA

inter

face


outrst

Read CFDresults


results

Check ofconvergence




(n minus 1 or n)







Expansionvessel HX

FPS




















PIPE 100

TDJ-115

160

Expansionvessel

210

125

130

146

148

150

156

152

170

172

180

510

TDV-500(water in)

HX

200

206

TDJ-505

TDJ-405

TDV-520(water out)

TDV-320(argon out)

TDV-400(argon in)

TDV112









PIPE 120

0765m

15 m

5m

75 m

005 m

105 m 11 m

1m

(a)

TDJ-115

TDJ-405

PIPE 120

PIPE 100210

206

200

TDV-400(argon in)

125

TDV110

TDV112

p2

p1

T2

m2

m1 T11m

11 m

(b)


Inletsection

11 mOutletsection

x



Reference section

(a)

Support rod

Electric pins

000

2500

5000

7500

10000 (mm)

(b)







0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5


(a)

ExperimentalRELAP5


0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)





0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)



+10

minus10





(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of















Expansionvessel HX

FPS




















PIPE 100

TDJ-115

160

Expansionvessel

210

125

130

146

148

150

156

152

170

172

180

510

TDV-500(water in)

HX

200

206

TDJ-505

TDJ-405

TDV-520(water out)

TDV-320(argon out)

TDV-400(argon in)

TDV112









PIPE 120

0765m

15 m

5m

75 m

005 m

105 m 11 m

1m

(a)

TDJ-115

TDJ-405

PIPE 120

PIPE 100210

206

200

TDV-400(argon in)

125

TDV110

TDV112

p2

p1

T2

m2

m1 T11m

11 m

(b)


Inletsection

11 mOutletsection

x



Reference section

(a)

Support rod

Electric pins

000

2500

5000

7500

10000 (mm)

(b)







0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5


(a)

ExperimentalRELAP5


0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)





0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)



+10

minus10





(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of















PIPE 100

TDJ-115

160

Expansionvessel

210

125

130

146

148

150

156

152

170

172

180

510

TDV-500(water in)

HX

200

206

TDJ-505

TDJ-405

TDV-520(water out)

TDV-320(argon out)

TDV-400(argon in)

TDV112









PIPE 120

0765m

15 m

5m

75 m

005 m

105 m 11 m

1m

(a)

TDJ-115

TDJ-405

PIPE 120

PIPE 100210

206

200

TDV-400(argon in)

125

TDV110

TDV112

p2

p1

T2

m2

m1 T11m

11 m

(b)


Inletsection

11 mOutletsection

x



Reference section

(a)

Support rod

Electric pins

000

2500

5000

7500

10000 (mm)

(b)







0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5


(a)

ExperimentalRELAP5


0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)





0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)



+10

minus10





(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of



















0 1 2 3 4 5 6 7 8 90

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

ExperimentalRELAP5


(a)

ExperimentalRELAP5


0 1 2 3 4 5 6 7 80

2

4

6

8

10

12

14

16

18

20

Time (h)

LBE

mas

s flow

rate

(kg

s)

(b)





0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)



+10

minus10





(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of
















(a)


(b)






Deadvolume

DHRHX

CIRCEmain vessel

(S100)

Downcomer

Feedingconduit

Flowmeter





water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of
















water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

T2

P2

m2

m1

30tubes tubes tubes

54 7

P1 T1

(a)

Equalpressures 5

water

6steam

HX(91 tubes)TDV

3TDV190

TDV151

TDV10

P2

T2m230


P1 T1

m1

(b)










280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of

















280∘C 0 kW 0kW


0 05 1 15 2 25 3 35 4 45 5 5540

45

50

55

60

65

70

75

Time (h)

LBE

mas

s flow

(kg

s)

ExperimentalRELAP5





50 52 54 56 58 60 62 64 66 6850

52

54

56

58

60

62

64

66

68

70


Calc

ulat

ed L

BE fl

ow ra

te (k

gs)


+4

minus4






Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of















Velocity range

Velocity

Poolinlet

Pooloutlet

(m Ｍminus1)

Z X

Y

0 1500

22500750

3000 (m)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(0ndash08 ms)

(a)

Poolinlet

Pooloutlet

Z

X

Y

0

1250

2500

3750

5000 (m)

Velocity

(m Ｍminus1)

1000e minus 001

9000e minus 002

8000e minus 002

7000e minus 002

6000e minus 002

5000e minus 002

4000e minus 002

3000e minus 002

2000e minus 002

1000e minus 002

0000e + 000

(b)









Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of
















Notations

Symbols


Acronyms










Acknowledgments


References



























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of





























FuelsJournal of







Advances in



Journal of


Renewable Energy





RotatingMachinery


EnergyJournal of



Journal of














STH-CFD Codes Coupled Calculations Applied to HLM Loop and ...

Documents

Transcript of STH-CFD Codes Coupled Calculations Applied to HLM Loop and ...