Production and Testing of Multi-group Nuclear Data from the ENDF/B VII.1 Library Alberto Milocco on leave from Institute Jožef Stefan, Jamova cesta 39 1000, Ljubljana, Slovenia alberto.milocco@ijs.si Andrej Trkov, Ivan Alexander Kodeli Institute Jožef Stefan andrej.trkov@ijs.si, ivan.kodeli@ijs.si ABSTRAC On December 2011 the ENDF/B-VII.1 library has been released. The ENDF/B-VII.1 library includes a re-evaluation of the neutron interaction data for Manganese-55. Many application files presently include nuclear data for Mn-55 that rely on previous evaluations. This is the case of the ADS-2.0 nuclear data library, available from the International Atomic Energy Agency (IAEA) for the calculations of accelerator-driven systems (ADS) and based on the ENDF/B-VII.0 evaluated nuclear data library. The new Mn-55 nuclear data were processed at IJS with the NJOY system in the ADS 421-group and the VITAMIN-J 175-group energy structures. The multi-group data were processed through TRANSX and ANISN for the simulation of a simple integral experiment with a Manganese shell that was performed at the OKTAVIAN facility. The deterministic calculation has been cross-validated with MCNP calculations adopting consistent nuclear data. Afterwards, the ANISN model has served for comparison with the 199-group cross sections from JEFF-3.1 library processed at ENEA-Bologna, the VITAMIN-J 175-group cross sections from the JENDL-3.3 processed at JAERI & SAEI and the neutron 150-group KAFAX libraries based on ENDF/B-VII.0 nuclear data and produced at KAERI. The method described allows assessing the quality of the computational models and hence of the Mn-55 multigroup files from the ENDF/B-VII.1 library. 1 INTRODUCTION The release of the ENDF/B-VII.1 nuclear data library occurred in December 2011 [1]. The new version of the library includes a brand new evaluation of the neutron-interaction cross sections for the Manganese-55. It is the result of a broad collaboration among NNDC [2], IAEA-NDS [3], ORNL [4] and IJS-F8 [5]. The new Mn-55 evaluation is considered here with the aim to provide further feedback on the quality of the cross sections. The beta versions of the Mn-55 evaluation have already been addressed in recent papers. The resonance parameters file has been tested by integral experiments involving Mn-55 activation foils measurements. The new evaluation has allowed a clear improvement in the simulation of the reaction rates measured by activation foils. Evidence has been given through 305.1 305.2 the FNG experiments on the ITER shield [7] and on the Helium Cooled Lithium Lead (HCLL) blanket [8]. Moreover, testing of the cross sections in the resolved resonance region was done by modelling the Grenoble lead-slowing-down benchmark [8], showing excellent agreement between calculated and measured spectra of the capture reaction rate. At higher energies, the validation of the nuclear data can be properly achieved with the neutron leakage spectrum measurement performed in 1987 at the Japanese OKTAVIAN facility with a ~14 MeV D-T (Deuteron-Triton) neutron source and a ~ 30 cm thick Manganese shell. The experiment has already been adopted for tuning the inelastic cross sections [7]. In the same paper, one can find a description of the evaluation performed with the code EMPIRE in the fast energy region. This time, the OKTAVIAN experiment is used for testing multigroup cross sections with the deterministic discrete-ordinates code ANISN [9]. First, the original evaluation file in ENDF-6 format [10] has been processed with the codes NJOY [11] to produce the 421- group file in MATXS format for Accelerator Driven System (ADS) applications and also the 175group file (VITAMIN-J structure). Then, the TRANSX code [12] has been used to prepare the cross sections for ANISN. The approximations intrinsic to the ANISN analysis are assessed by means of MCNP models [13] that may be accurate to a greater extent. In this way, the deficiencies in the experimental specifications, affecting the quality of the experiment itself, are spotted out and the validity of the ANISN model can be established. Subsequently, the ANISN results obtained with the ENDF/B-VII.1 evaluation are further compared with multigroup libraries produced in other laboratories and based on the ENDF/B-VII.0, JENDL3.3 and JEFF3.1 evaluations. The conclusions provide some general statements on the accuracy of the ENDF/B-VII.1 nuclear data for Mn-55 after processing the application files and indicate further activities. 2 SIMULATION OF THE OKTAVIAN EXPERIMENT Experiments performed at the OKTAVIAN facility were compiled and included in the ‘Shielding Integral Benchmark Archive Database – SINBAD –’ [14]. The experiment with a Manganese shell is going to be included in a next SINBAD release. At the OKTAVIAN facility, Deuteron – Tritium (D-T) source neutrons were produced at the centre of material shells and neutron leakage spectra were measured with the Time-of-Flight (TOF) technique. The samples were spherical shells with a cavity and a duct to accommodate the Tritium target and the ion tube. The deuterons were accelerated up to 250 keV. Recent OKTAVIAN measurements (including the Manganese case) have been performed with the detector system at 55° with respect to the deuteron beam line (Fig. 1). 2.1 Preparation of the application files The multigroup application files (MATXS format) with the ADS 421-group structure and with the VITAMIN-J 175-group structure have been generated from the ENDF/B-VII.1 Manganese data by NJOY-99.259 at the Jozef Stefan Institute (IJS). The self shielding tables have been prepared from 0.001 to infinite dilution. The weighting function is represented by a thermal+1/E+fission+fusion shape. For the rest, the same NJOY input was used as for the production of the FENDL-2.1 MATXS library [15]. An independent preparation of the Mn-55 continuous cross section file for the MCNP code (ACE format) has also been pursued. 305.3 Figure 1: Experimental set-up for the 55° TOF experiments at OKTAVIAN. The TRANSX input includes the following physics settings: a) mean chord length = 40 cm (~ 4V/S), b) temperature = 300 K; c) order of Legendre expansion = 5; d) inconsistent-P transport correction for higher order Legendre expansion 2.2 ANISN model ANISN calculates the spectrum of the neutron positive current out of the 30 cm thick spherical shell containing pure manganese. At the centre there is a void spherical cavity of 2.8 cm radius and an isotropic source with the energy distribution experimentally specified. The problem is solved in the P5/S16 approximation (5th Legendre order, 16-angles quadrature). The radial nodes in the shell are distant 3 cm, which is about the mean free path of a 14 MeV neutron in Manganese. At first instance, the calculation has been performed with the VITAMIN-J group structure. The ANISN modelling approximations can be assessed by comparison with three MCNP models (Fig. 2): 1. A very simple MCNP model, which has spherical symmetry, experimental but isotropic source spectrum and calculates the current at the outer surface of the sphere. Except for the cross sections used (continuous versus multigroup) and the method of solution (Monte Carlo versus deterministic), this MCNP model is equivalent to the ANISN model concerning the geometry, the material composition, the source and the output specifications (‘ANISN EQUIVALENT’ label in Fig. 2). 2. Simple but usually quite accurate MCNP models have been developed for the analysis of OKTAVIAN experiments. They allow quick calculations (hence the label ‘ROUTINE’ in Fig. 2). The ‘routine’ MCNP model includes without approximations the sphere cavity, the vessel and the sample composition with impurities. The anisotropy of the neutron source is taken into account. The TOF spectrum is calculated by a point detector in void at the position of the measurement. The experimental neutron source spectrum is converted into TOF spectrum and is used to broaden the calculated TOF spectrum. At the end, the broadened spectrum is converted into the energy domain and compared with the experimental energy spectrum. The Fig. 2 shows improved results around 12 MeV because of improved source specifications. Also, this model usually benefits from taking into account the multi-scattering effect thanks to the time domain analysis. 305.4 3. A detailed model of the experimental set-up has been developed to include the source assembly (e.g. ion beam tube), the detector system (pre-collimator, collimator, detector and shield). The source term is modelled by a source routine that precisely takes into account the deuteron slowing down and the D-T reactions in the target. It was already pointed out [15] that the OKTAVIAN specifications for the some experimental components are not of high quality, especially with reference to the calculations below the MeV region. It can be seen from Fig. 2 that the ANISN model is in fair agreement with the measured leakage spectrum, at least above 1 MeV. Concerning the bump occurring at 1 MeV, it is worthwhile to notice that in the original ENDF file the resonance region extends up to this energy. The data in multigroup form were checked and the fully-shielded capture cross section near 1 MeV was found to be nearly 30 % smaller than the unshielded one. Clearly there is a discontinuity in the self-shielding effect as introduced in the MATXS file from the ENDF/B-VII.1 file through TRANSX. The step-effect is less pronounced in the Monte Carlo calculations performed in the time-domain due to the smearing effect of multiple scattering. The apparently-better agreement of the ANISN model with experimental data is a coincidence due to the lacking self-shielding correction above 1 MeV. With some caution in the interpretation of the resonance effects, the ANISN model might be considered cross-validated through careful assessment of the MCNP models. Figure 2: Results of the simulation of the neutron leakage spectrum from a ~ 30 cm thick Manganese shell with mono-dimensional ANISN and MCNP models, an accurate, even still simple MCNP model and a detailed tri-dimensional MCNP model. 2.3 Multigroup Libraries The Mn-55 multigroup cross sections produced at IJS (ADS and VITAMIN-J group structures) from the ENDF/B-VII.1 evaluation are benchmarked with other three multigroup files with the same ANISN model. These are: a) the 199-group cross sections from JEFF-3.1 library processed at ENEA-Bologna with NJOY-99.160bo b) the VITAMIN-J 175-group cross sections from the JENDL-3.3 processed at JAERI (‘Japan Atomic Energy Research Institute’) & SAEI (‘Sumitomo Atomic Energy Industries’ c) the neutron 150-group cross sections processed at KAERI from the ENDF/B-VII.0 library 305.5 Fig. 3 shows the neutron current calculations with these libraries. In the inelastic region interesting differences emerge between the ENDF/B-VII.1 results and those from the other libraries. The ADS energy-group structure has been pursued also with the Manganese nuclear data from the ENDF/B-VII.0 and JEFF-3.1. The results are presented in Fig. 4. By comparison between the different calculations, the influence of the nuclear data can be qualitatively assessed. Note the structure in the calculated spectrum in the 100 keV region based on ENDF/B-VII.0 data, which originates from the structure in the cross sections. This structure is not present in ENDF/B-VII.1 data and is an item for consideration for the improvement of this library. It is also evident from Fig. 2 that extending the unresolved resonance region to higher energies would destroy the apparently good agreement between the ENDF/B-VII.1 results and the measurements, but this would be rectified by a more accurate model of the experiment. Figure 3: Simulation with Mn-55 data from different multigroup libraries and multigroup files from the Manganese data in the ENDF/B-VII.1 library Figure 4: ANISN calculations in the ADS group structure with different source libraries 305.6 3 CONCLUSIONS The paper presents an early attempt to produce application files for deterministic calculations from the ENDF/B-VII.1 nuclear data library. In particular, the multigroup energy structure for ADS application has been addressed. Attention has been paid to the sources of uncertainty due to the OKTAVIAN experimental specifications and modelling. The deterministic model for the ANISN code has been validated for quick routine calculations by comparison with more detailed Monte Carlo calculations and the associated biases determined. The simulations with the same model but nuclear data from different libraries allow a qualitative assessment of the differences in the results due to the cross sections. The main conclusion related to the ENDF/B-VII.1 evaluation of Mn-55 is the need to extend the unresolved resonance region to about 4 MeV. The inclusion of the structure in the cross sections in the unresolved resonance range similar to that in the ENDF/B-VII.0 library should be considered. It is noticed that the ENDF/B-VII.1 data for Mn-55 include the evaluation of the covariances. Future activities should involve their use to precisely estimate the quality of the nuclear data uncertainty. This could be done starting from the present deterministic calculations and models. REFERENCES [1] M.B. Chadwick, M. Herman, P. Obložinský, M.E. Dunn, Y. Danon, A.C. Kahler, D.L. Smith, B. Pritychenko, G. Arbanas, R. Arcilla, R. Brewer, D.A. Brown, R. Capote, A.D. Carlson, Y.S. Cho, H. Derrien, K. Guber, G.M. Hale, S. Hoblit, S. Holloway, T.D. Johnson, T. Kawano, B.C. Kiedrowski, H. Kim, S. Kunieda, N.M. Larson, L. Leal, J.P. Lestone, R.C. Little, E.A. McCutchan, R.E. MacFarlane, M. MacInnes, C.M. Mattoon, R.D. McKnight, S.F. Mughabghab, G.P.A. Nobre, G. Palmiotti, A. Palumbo, M.T. Pigni, V.G. Pronyaev, R.O. Sayer, A.A. Sonzogni, N.C. Summers, P. Talou, I.J. Thompson, A. Trkov, R.L. Vogt, S.C. van der Marck, A. Wallner, M.C. White, D. Wiarda, P.G. Young, “ENDF/B-VII.1 Nuclear Data for Science and Technology, Cross Sections, Covariances, Fission Product Yields and Decay Data”, Nuclear Data Sheets, Volume 112, Issue 12, December 2011, Pages 2887-2996. [2] “National Nuclear Data Center” http://www.nndc.bnl.gov/ [3] “International Atomic Energy Agency, Nuclear Data Section” http://www-nds.iaea.org/ [4] “Oak Ridge National Laboratory” http://www.ornl.gov/ [5] “Jožef Stefan Institute, Reactor Physics Department” http://www.rcp.ijs.si/main/en/index.html [6] H. Derrien, L. C. Leal, N. M. Larson, D. Wiarda, K. Guber and G. Arbanas , “Neutron resonance parameters of 55Mn from Reich-Moore analysis of recent experimental neutron transmission and capture cross sections,” Proceeding of International Conference on the Physics of Reactors ‘Nuclear Power: A sustainable Resource’, Interlaken, Switzeland,14-19 September, Vol. 2, pp. 473-474 (2008). 305.7 [7] A. Milocco, A. Trkov and R. Capote, “Nuclear data evaluation of 55Mn by the Empire code with emphasis on the capture cross section,” Nuclear Engineering and Design, 241, pp.1071-1077 (2011). [8] I. Kodeli, A. Trkov, A. Milocco and L. Leal, “Selection of Activation Foils for Thermal Flux Measurements in FNG-HCLL Benchmark (TPR Validation),” EFF Meeting, Issyles-Moulinaux, France, 2 December, (2010). [9] D. Kent Parsons, ANISN/PC MANUAL, EG&G Idaho Ink, Idaho Falls, Idaho, US (1988). [10] M. Herman, A. Trkov, ENDF-& Formats Manual – data formats and procedures for the Evaluated Nuclear Data File ENDF/B-VI and ENDF/B-VII, Brookhaven National Laboratory, Upton, NY, USA, (2009). [11] 11. R. E. MacFarlane, A. C. Kahler, “Methods for Processing ENDF/B-VII with NJOY,” Nuclear Data Sheets, 111, pp.2739-2890 (2010). [12] R. E. MacFarlane, TRANSX 2: A Code for Interfacing MATXS Cross-Section Libraries to Nuclear Transport Codes, Los Alamos National Laboratory, Los Alamos, New Mexico, US (1993). [13] D. B. Pelowitz, MCNPX User’s Manual, Version 2.6, Los Alamos National Laboratory, Los Alamos, New Mexico, US (2007). [14] A. Milocco, A. Trkov, I. A. Kodeli, “The OKTAVIAN TOF experiments in SINBAD: Evaluation of the experimental uncertainties,” Annals of Nuclear Energy, 37, pp. 443449 (2010). [15] “Fusion Evaluated Nuclear Data Library (Dec 2004)” http://wwwnds.iaea.org/fendl21/index.html