Multi-physics (SP3) requires a software environment that supports code coupling: SALOME. First presentation given by. Nicolas Crouzet Overview on SP5

Transcription

1 Multi-physics (SP3) requires a software environment that supports code coupling: SALOME First presentation given by Nicolas Crouzet Overview on SP5 1

2 NURISP SP3 Multi-Physics CEA, HZDR, GRS, KIT-G, KIT-U, PSI, KTH, UPM, INRNE, NRI, IRSN Martin A. Zimmermann SP3 Coordinator 2

3 Acknowledgments CEA GRS HZDR IRNRE IRSN KIT-G KIT-U KTH NRI PSI UPM Roland Baviere, Philippe Emonot, Stéphanie Zimmer Gilardi Angel Papukchiev Sören Kliem, Andre Gommlich Nikola Kolvev Jean-Marc Ricaud Javier Jiménez Escalante, Armando Gomez, Victor Sanchez- Espinoza Florin Badea Ivan Gajev, Tomasz Kozlowski Jan Hadek, Ladislav Vyskocil Annalisa Manera, Omar Zerkak, Martin Zimmermann Diana Cuervo 3

4 Introduction Members of SP3 opted for Overview presentation 3 highlight presentations 4

5 Main Goals of NURESIM SP3-Project Develop common integrated platform for multi-scale and multi-physics analysis of LWR s (BWR, PWR, VVER) using for neutronics (nodal) 2-gr diffusion deterministic (simplified) transport (2D/3D), and for TH FLICA-4 Support integration of more detailed physics in each scientific domain 5

6 Objectives Develop and demonstrate beyond the state-of-the-art simulation capabilities for improved understanding of safety margins Integrate a range of thermal-hydraulic, core physics (neutronics) and fuel thermo-mechanics codes and solvers within the SALOME platform to form an integrated European platform with state-of-the-art code system to support safety analysis of current and evolving LWRs Drive the future platform development of a high-fidelity tool for reference-type solutions, initially for asymmetric PWR transients with strong 3D-effects on mixing in parts of the system using (among others) CFD, thermo-mechanics 6

7 Code Coupling NURISP-SP3 (CEA) NURESIM NURISP FLICA HotZone Sample of coupling interface evolution CRONOS API-c FLICA API-f DRACCAR API-d CATHAR E-2 ICoCo TRIO-U ICoCo CATHAR E-3 ICoCo-2 DRACCAR ICoCo-2 SALOME Method calls Data exchanges 7

8 Structure of SP3 Multi-Physics WP3.0 Status and limits of current methods for plant analysis WP3.1 Coupling Schemes for PWR and VVER applications WP3.2 Coupling Schemes for BWR applications WP3.3 Development of new coupling schemes WP3.4 Education and Training 8

9 Objective of WP3.1 Focus (spatial) analysis detail to regions of interest Two-level coupling Fine detail in region of interest (hot assembly/ies) Minimal data generation/transfer/storage Minimal computing effort Adaptive coupling and switching algorithm Use higher-order model one where needed Demonstrated for application of core kinetics 9

10 WP3.1: PWR coupling schemes Details of WP1 See highlight presentation by Sören Kliem (HZDR) Two level coupling for core modeling 10

11 WP3.1: VVER coupling schemes Codes adaptation to the new schemes and coupling with FLICA4 COBAYA3 and CRONOS2 coupling schemes (YACS) adaptation to the new FLICA4 API and unstructured meshes generation with FLICA4 Nodal level Coupled nodal CRONOS2/FLICA4 and COBAYA3/FLICA4 steady state and transient solutions compared to each other and to independent couplings of COBAYA3/COBRA3 and DYN3D-FLOCAL. The variety of couplings allowed separation of the effects of mixing models, node subdivision in neutronics and the coupling schemes Pin level: Development of non-conforming meshes coupling schemes for hexagonal geometry Standalone N and TH core models subsets including 19-pin cluster and fuel assembly tested in structured and unstructured meshes Coupled N and TH core models First, coupled 3D N/TH results for 19 pins and hot assembly analyzed in steady state using COBAYA3 N/TH simulation with the internal TH module. Pin-by-pin solutions in 2, 4 and 8 neutron energy groups were compared. Extending the simulation to a mini-core of 7 assemblies is underway. Testing of COBAYA3/FLICA4 and CRONOS2/FLIC4A coupling at the pin-cell level is ongoing. 11

12 WP3.1: VVER nodal level results Fission power, MW VVER MSLB: Time history of fission power in Scenario 2 (nodal level) DYN3D/FLOCAL COBAYA3/FLICA4 COBAYA3/COBRA3c Time, s VVER MSLB: COBAYA3 6N /FLICA4 predicted 3D power distribution at time of highest return to power (elevation 3.0 m) 12

13 Fxyz WP3.1: VVER MSLB nodal results DYN3D/FLOCAL COBAYA3/FLICA4 COBAYA3/COBRA3c Time, s Time history of F xyz in Scenario 2. Impact of higher resolution: TH with 1 node/hexagon DYN3D with 1 node/hexagon, COBAYA3 with 6 nodes/hexagon, 13

14 WP3.1: Adaptive coupling & switching algorithms Report on state of the art of adaptive coupling and switching algorithms for neutron kinetic models of different levels of detail Automatic adaptive coupling & switching algorithms implemented in DYN3D: DYN3D_MG (Multi-Group) / PK (Point Kinetics) / FLOCAL (core TH) Testing of the algorithms on two benchmarks: Boron Dilution and Rod Movement 14

15 Nuclear power (MW) Nuclear Power (MW) WP3.1: DYN3D_MG (Multi-Group) / PK (Point Kinetics) FLOCAL (core TH) Boron Dilution Power 4000 Rod Movement DYN3D_MG: 3D DYN3D_MG: 3D - PK DYN3D_MG: PK D PK DYN3D_MG: 3D DYN3D_MG: 3D - PK DYN3D_MG: PK D PK PK 3D Time (s) Time (s) Results produced by the automatic dimensionally adaptive switching algorithm (3D - PK) retain the accuracy of the 3D reference calculations; even for difficult transients (i.e. Rod Movement), where simple PK fails 15

16 Dynamic reactivity ($) WP3.1: DYN3D_MG (Multi-Group) / PK (Point Kinetics) FLOCAL (core TH) Dynamic reactivity ($) 0.8 Boron Dilution Dynamic Reactivity 0.2 Rod Movement 0.6 DYN3D_MG: 3D DYN3D_MG: 3D - PK DYN3D_MG: PK DYN3D_MG: 3D DYN3D_MG: 3D - PK DYN3D_MG: PK Time (s) Time (s) automatic dimensionally adaptive switching algorithms (3D - PK) require only 30 to 70% of the CPU time needed by the 3D reference calculations while maintaining the 3D-accuracy 16

17 Structure of SP3 Multi-Physics WP3.0 Status and limits of current methods for plant analysis WP3.1 Coupling Schemes for PWR and VVER applications WP3.2 Coupling Schemes for BWR applications WP3.3 Development of new coupling schemes WP3.4 Education and Training 17

18 Objective of WP3.2 Improve NURESIM coupling schemes in timedomain 18

19 WP3.2: Outcome of the S-O-A Review Most of current code coupling schemes based on Operator Splitting (OS) Fixed-Point-Iteration (FPI) at end of time-step Suggested improvements Improved OS through temporal extrapolation Fixed-Point-Iteration at deeper (Outer) iteration levels Newton-method-based iterative schemes Conditions for the implementation phase Implementation of most advanced schemes (FPI, Newton) requires access to the source codes Improvements without access to source codes still possible (improved OS, externalized adaptive time-step control) 19

20 STANDARD OPERATOR SPLITTING (NURESIM V1.0) For each time step n 1) Compute 2) Compute x n C A t n, x n 1, n n n 1 y CB t, y, y x n n 1 t n-1 x n-1 y n-1 y n-1 End time step Advantage : Minimum number of code iterations (1 temporal integration per code per time-step) C A C B Disadvantage : Explicit coupling field for the leading code (C A ) (stability conditioned to time-step size) Limited accuracy x n t n x n y n 20

21 WP3.2: Deviations from original plans Original plan BWR Turbine Trip transient Very large transient problem with memory overflows restricting minimum possible time-step and sometimes leading to dubious results Long computation times required (2-phase flow) not amenable to detailed time-step convergence studies Switch to a more adequate test case PWR Minicore Rod Ejection Accident (REA) Exemplary transient with very strong coupling between the physics (core power from N-K solver and fuel temperature from T-H solver) Verified to be converged in time 21

22 WP3.2: FUEL TEMPERATURE ACCURACY (t END ) Bottle neck : Heat transfer in fuel rods (1 st order Backward Euler in T-H solver) Zerkak, Gajev et al. NURETH-14 (2011) 22

23 WP3.2: ACCURACY AS FUNCTION OF CPU SPEED-UP (ADAPTIVE TIME-STEP ALGORITHM) FUEL TEMPERATURE POWER Zerkak, Gajev et al. NURETH-14 (2011) 23

24 WP3.2 Conclusions High-fidelity simulation methods in NURESIM will require adequate temporal coupling to avoid deteriorating the accuracy of the individual codes, and to optimize the CPU use Accuracy vs. efficiency of Operator Splitting coupling schemes can be improved with limited effort Temporal extrapolation of coupling fields Externalized adaptive time-step control based on all different physics of the problem Further verification is needed for more challenging two-phase flow problems More stable (e.g. semi-implicit FPI) and accurate (e.g. Newton-based) coupling schemes could be implemented without significant intrusion into the codes structures 24

25 Structure of SP3 Multi-Physics WP3.0 Status and limits of current methods for plant analysis WP3.1 Coupling Schemes for PWR and VVER applications WP3.2 Coupling Schemes for BWR applications WP3.3 Development of new coupling schemes WP3.4 Education and Training 25

26 Objective of WP3.3 Develop new coupling interface ICOCO-2 Integrate new coupling with new physics into NURESIM platform Thermo-mechanics Thermal-hydraulics (Fuel behavior code DRACCAR) Integrate multi-scale capability for thermalhydraulics CFD System code 26

27 WP3.3 See details in 2 highlight presentations by Philippe Emonot (CEA) CFD-System code coupling and Jean-Marc Riccaud (IRSN) presented by Pierre Ruyer (IRSN) DRACCAR TH coupling 27

28 28

29 Published Results 1. A. M. Gomez Torres, V. H. Sanchez Espinoza, S. Kliem, A. Gommlich, U. Rohde, Integration of DYN3D inside the NURESIM Platform, 17th Pacific Basin Nuclear Conference, Cancún, Q.R., México, October 24-30, A. Papukchiev, G. Lerchl, J. Weis, M. Scheuerer and H. Austregesilo, DEVELOPMENT OF A COUPLED 1D-3D THERMAL-HYDRAULIC CODE FOR NUCLEAR POWER PLANT SIMULATION AND ITS APPLICATION TO A PRESSURIZED THERMAL SHOCK SCENARIO IN PWR Proc. Int. Top. Meetg. Nuclear Reactor Thermalhydraulics, NURETH-14, Toronto, Canada, September 25-30, O. Zerkak, I. Gajev, A. Manera, T. Kozlowski, A. Gommlich, S. Zimmer, S. Kliem, N. Crouzet, M.A. Zimmermann, Revisiting Temporal Accuracy in Neutronics/T-H Code Coupling using the NURESIM LWR Simulation Platform, Proc. Int. Top. Meetg. Nuclear Reactor Thermalhydraulics, NURETH-14, Toronto, Canada, September 25-30, I. Spasov, T.Tzanov, N.Kolev (INRNE), J.Hadek(NRI), DYN3D/FLOCAL vs. COBAYA3/FLICA4 solutions of the VVER MSLB benchmark Proceeding AER 2011 Conference, Dresden, September S. Kliem, A. Gommlich, A. Grahn, U. Rohde, J. Schiitze, Th. Frank, A. Gomez and V. Sanchez, Development of multi-physics code systems based on the reactor dynamics code DYN3D, KERNTECHNIK 76 (2011) J. Jiménez, M. A. Zimmermann, O. Zerkak, Ph. Emonot, Presentation of the Multi-Physics Subproject of the NURISP European Project, presented at ANNUAL MEETING OF THE SPANISH NUCLEAR SOCIETY,

30 Education 4 PhD s UPM Javier Jimenez (now PostDoc KIT-G) KIT-G Armando Gomez KTH Ivan Gajev INRNE Ivan Spasov, Nonka Zheleva 30

31 Meetings 5 (+1) SP3 meetings (Madrid [UPM, Nov 2008]) Villigen (PSI, Nov 2009) Grenoble (CEA, May 2010) Karlsruhe (KIT-G, Nov 2010) Stockholm (KTH, June 2011) Villigen (PSI, Jan 2012) 31

32 Summary Created an efficient coupling community Developed coupling expertise Successfully included new physics domains Ready for launching into NURESAFE with interesting multiphysics applications 32