Modern nuclear data evaluation: straight from nuclear physics to applications Arjan Koning and Dimitri Rochman NRG Petten, the Netherlands ND-2010 April , Jeju, Korea

2 Contents Introduction: A new way of nuclear data evaluation Essence of software development and reproducibility -Example: TALYS code system Implications and possibilities: -Large scale nuclear data library production (TENDL) -“Total” Monte Carlo uncertainty propagation -Validation with integral measurements Conclusions

3 Objective There are limits in our knowledge of nuclear physics: Experimental possibilities and precision Theoretical nuclear structure and reaction models which requires and deserves everlasting support Mantra: let’s at least provide all nuclear physics knowledge we know up to now in a form ready for applications, while maximizing Completeness: no unnecessary omissions Quality: -no unnecessary approximations -with a quantitative measure about our knowledge (uncertainty information) -Satisfactory from differential and integral point of view

4 Nuclear data libraries: up to now During one nuclear data career: Evaluators make mistakes Evaluators improve their methods (less mistakes, more complete ENDF-6 files, covariance data, etc.) New experimental data and better nuclear model codes emerge Up to now, such progress has not been consistently implemented in isotopic data libraries. E.g. : 96-Cm-247 JAERI-ORNL EVAL-OCT05 R.Q. Wright, T.Nakagawa, T.Liu 96-Cm-248 HEDL,SRL,+ EVAL-APR78 Mann,Benjamin,Howerton, + 96-Cm-249 JAERI EVAL-OCT95 T.Nakagawa and T.Liu 96-Cm-250 JAERI EVAL-OCT95 T.Nakagawa and T.Liu 27-Co- 58 NEA RCOM-JUN83 Scientific Co-ordinating Group 27-Co- 58MNEA RCOM-JUN82 Scientific Co-ordination Group 27-Co- 59 ANL,ORNL EVAL-JUL89 A.Smith+,G.Desaussure+ 24-Cr- 50 LANL,ORNL EVAL-OCT97 S.Chiba,M.Chadwick,D.Hetrick and these data libraries are not reproducible

5 Nuclear data libraries: another way Nuclear data knowledge should no longer be assembled in an ENDF-6 nuclear data library, but one level deeper: Resonance parameters + uncertainties An error-free EXFOR database + selection of good data Nuclear models: -A robust, validated, multipurpose nuclear model code -A Reference Input Parameter Library (RIPL) -For important/measured nuclides: A set of adjusted model parameters + uncertainties If needed: per nuclide a script with other actions (copying parts of other libraries, direct inclusion of experiment, etc.) Store the above, and make sure that ENDF formatting, processing and integral validation become robust This yields entirely new possibilities

6 Resonance Parameters. TARES Experimental data (EXFOR) Nucl. model parameters TALYS TEFAL Output ENDF Gen. purpose file ENDF/EAF Activ. file NJOY PROC. CODE MCNP FIS- PACT Nuclear data scheme + covariances -K-eff -Neutron flux -Etc. -activation - transmutation Determ. code Other (ORIGEN) +Uncertainties +Covariances +(Co)variances +Covariances TASMAN Monte Carlo: 1000 TALYS runs

7 Loop over energies and isotopes PRE-EQUILIBRIUM Exciton model Partial densities Kalbach systematic Approx DSD Angular distributions Cluster emissions  emission Exciton model Hauser-Feshbach Fission  cascade Exclusive channels Recoils MULTIPLE EMISSIONSTRUCTURE Abundances Discrete levels Deformations Masses Level densities Resonances Fission parameters Radial matter dens. OPTICAL MODEL Phenomenologic Local or global Semi-Microscopic Tabulated (ECIS) DIRECT REACTION Spherical / DWBA Deformed / Coupled channel Giant Resonances Pickup, stripping, exchange Rotational Vibrational COMPOUND Hauser-Feshbach Fluctuations Fission  Emission Level densities GC + Ignatyuk Tabulated Superfluid Model INPUT projectile n element Fe mass 56 energy 1.2 TALYS code scheme OUTPUT Spectra Cross sections Fission yields DDX Ang. Distr. Astro rates Etc.

8 TALYS-1.2 Released December 21, 2009, see Use of TALYS increasing -Estimated users, publications Some recent improvements for TALYS-1.2: -Better fission + level density model (CEA Bruyeres-le- Chatel) -The option to easily/safely store the best input parameter set per nucleus (“best y”) -More flexibility for covariance development and adjustment to experimental data TALYS can be used for -In-depth nuclide/reaction analyses -Global multi-nuclide calculations -These two are now being merged

9 Resonance Parameters. TARES Experimental data (EXFOR) Nucl. model parameters TALYS TEFAL Output ENDF Gen. purpose file ENDF/EAF Activ. file NJOY PROC. CODE MCNP FIS- PACT Nuclear data scheme + covariances -K-eff -Neutron flux -Etc. -activation - transmutation Determ. code Other (ORIGEN) +Uncertainties +Covariances +(Co)variances +Covariances TASMAN Monte Carlo: 1000 TALYS runs

10 Uncertainties with Monte Carlo Standard procedure (Don Smith): Determine uncertainty range for each nuclear model parameter Perform K(=1000) TALYS calculations with all parameters randomly sampled around their central values Covariance matrix for cross sections i and j: Various refinements possible: -Reject outlying results (leads to parameter correlations) -More precise inclusion of experimental data (Unified MC, D. Smith, H. Leeb), backward-forward MC (E. Bauge), etc.

11 Uncertainties for Cu isotopes

12 Resonance Parameters. TARES Experimental data (EXFOR) Nucl. model parameters TALYS TEFAL Output ENDF Gen. purpose file ENDF/EAF Activ. file NJOY PROC. CODE MCNP FIS- PACT Nuclear data scheme + covariances -K-eff -Neutron flux -Etc. -activation - transmutation Determ. code Other (ORIGEN) +Uncertainties +Covariances +(Co)variances +Covariances TASMAN Monte Carlo: 1000 TALYS runs

13 Application 1: TENDL TALYS Evaluated Nuclear Data Library, n, p, d, t,h, a and g libraries in ENDF-6 format 2400 nuclides (all with lifetime > 1 sec.) up to 200 MeV Neutrons: complete covariance data (MF31-MF35) MCNP-libraries (n,p and d) and multi-group covariances (n only) Production time: 2 months (40 processors) Strategy: Always ensure completeness, global improvement in 2010, Extra effort for important nuclides, especially when high precision is required (e.g. actinides): adjusted parameters (data fitting). These input files per nuclide are stored for future use. All libraries are always reproducible from scratch The ENDF-6 libraries are created, not manually touched Zeroing in on the truth for the whole nuclide chart at once

14 TENDL users European Activation File (EAF): > 95% TALYS/TENDL based Fusion Evaluated Nuclear Data Library (FENDL) -Missing nuclides, high energies, covariances, protons and deuterons Joint Evaluated Fission and Fusion file (JEFF) -Missing nuclides (JEFF-3.2), protons and deuterons NEA Data Bank: Janis (E. Dupont, N. Soppera) IAEA: visualisation system (V. Zerkin) Fusion/IFMIF research (Sanz, Sauvan): protons and deuterons Many downloads from So far, TENDL is adopted “when nothing else exists”, but a lot of effort has been devoted to nuclide-by-nuclide neutron evaluations. We can, and will, be more ambitious!

15 TENDL: Complete ENDF-6 data libraries MF1: description and average fission quantities MF2: resonance data MF3: cross sections MF4: angular distributions MF5: energy spectra MF6: double-differential spectra, particle yields and residual products MF8-10: isomeric cross sections and ratios MF12-15: gamma yields, spectra and angular distributions MF31: covariances of average fission quantities (TENDL-2010) MF32: covariances of resonance parameters MF33: covariances of cross sections MF34: covariances of angular distributions MF35: covariances of fission neutron spectra (TENDL-2010) and particle spectra (TENDL-2011) MF40: covariances of isomeric data (TENDL-2011)

16 IAEA covariance visualisation system (V. Zerkin)

17

18

19

20

21 Quality of proton data (EXFOR vs MCNPX, A. Konobeyev, KIT) ENDF/B-VII-p (LA-150): nuclides TENDL-2009: 1170 nuclides (Chi-2) ( ) (H x F)

22 Other TENDL(-related) results -Astrophysical reaction rates for 4000 nuclides (Stephane Goriely) -Human-readable tables for normal nuclear physicists. Some possibilities: -Validation of the entire EXFOR -Complete table of all medical isotope production routes for all nuclides (“inverse search”) -Comparison with specific measurements -Etc.

23 Application 2: “Total” Monte Carlo Propagating covariance data is an approximation of true uncertainty propagation (especially regarding ENDF-6 format limitations) Covariance data requires extra processing and “satellite software” for application codes Alternative: Create an ENDF-6 file for each random sample and finish the entire physics-to-application loop. (Koning and Rochman, Ann Nuc En 35, 2024 (2008)

24 Resonance Parameters. TARES Experimental data (EXFOR) Nucl. model parameters TALYS TEFAL Output ENDF Gen. purpose file ENDF/EAF Activ. file NJOY PROC. CODE MCNP FIS- PACT Nuclear data scheme + covariances -K-eff -Neutron flux -Etc. -activation - transmutation Determ. code Other (ORIGEN) +Uncertainties +Covariances +(Co)variances +Covariances TASMAN Monte Carlo: 1000 TALYS runs

25 Resonance Parameters. TARES Experimental data (EXFOR) Nucl. model parameters TALYS TEFAL Output ENDF Gen. purpose file ENDF/EAF Activ. file NJOY PROC. CODE MCNP FIS- PACT Nuclear data scheme: Total Monte Carlo -K-eff -Neutron flux -Etc. - activation - transmutation Determ. code Other codes +Uncertainties +Covariances TASMAN Monte Carlo: 1000 runs of all codes

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49 Covariance versus Total Monte Carlo Advantages:Advantages: - Relatively quick- Exact - Use in sensitivity study- Requires only “main” software - Easier release (TENDL) Disadvantages:Disadvantages: - Approximative (cross-correlations)- (Computer) time consuming - No covariance for gamma production,- Backward (sensitivity) route DDX (MF36), etc. not obvious - Requires special processing - Requires covariance software for application codes

50 Application: criticality benchmarks Total of random ENDF-6 files Sometimes deviation from Gaussian shape Rochman, Koning, van der Marck Ann Nuc En 36, 810 (2009) Yields uncertainties on benchmarks

51 Application: SFR void coefficient KALIMER-600 Sodium Fast Reactor (Korea) Total Monte Carlo with MCNP Uncertainties due to Na alone: D. Rochman et al NIM A612, 374 (2010) Uncertainties due to major actinides: see D. Rochman presentation Extension to SFR burn-up underway

52

53 TMC: Other possibilities Random thermal scattering data libraries (?) Random decay data libraries Random fission yield libraries Normalization to experimental data or other nuclear data libraries at the basic input level (in progress) Optimization to integral benchmarks using e.g. simulated annealing (“search for the best random file”)

54 Conclusions The lifetime of the Nuclear Data Cycle (from basic data to applications and back) can be strongly reduced, even using existing formats and tools. The secret: Make everything reproducible from the start. Ingredients: -Selected experimental data from EXFOR (if available) -TALYS input parameters + uncertainties (or default) -Resonance parameters + uncertainties (if available) -Nuclide specific scripts (if needed) After some serious software development you can reproduce everything from that The first two applications -Talys Evaluated Nuclear Data Library (TENDL) -Total Monte Carlo uncertainty propagation

55 Conclusions The system is in place, from now on the main challenges are at the beginning of the cycle, e.g.: Better evaluations per nuclide Merge experimental + theoretical uncertainty methods (“Unified Monte Carlo”) Of course, this approach does not take away the need for progress in measurements, theory development, ENDF formatting, processing and validation, but any progress will have impact directly through the entire chain, and for all nuclides (reproducibility!)

56 Acknowledgements NRG Petten: Dimitri Rochman, Marieke Duijvestijn CEA-Bruyeres-le-Chatel: Stephane Hilaire, Eric Bauge, Pascal Romain Univ. Libre Bruxelles: Stephane Goriely CEA Saclay: Jacques Raynal IRMM Geel: Arjan Plompen IAEA: Roberto Capote JUKO research: Jura Kopecky All TALYS users for their feedback

57 The final spin-off: Is the challenge for ENDF/B-VII, JEFF, JENDL, CENDL and FENDL to stay ahead. We’ll meet again in ND-2013!