Recent Advances in the ENDF/B Library

Transcription

1 Recent Advances in the ENDF/B Library P. Obložinský National Nuclear Data Center, Brookhaven National Laboratory, Upton, NY 11973, USA Abstract. We describe recent advances in the US evaluated nuclear-reaction data file maintained by CSWEG. Since 2001, CSWEG has been developing a new version of the library, ENDF/B-VII. Altogether 231 new and improved evaluations have been submitted, including 51 neutron, 20 charged-particle, and 160 photonuclear evaluations. Neutron files were developed primarily by LANL in the fast region, in particular for actinides, followed by ORNL evaluations in the neutron-resonance region, and BNL evaluations for fission products. The current focus is on the data validation process. INTRODUCTION The US-evaluated nuclear-reaction data file is maintained by the Cross Section Evaluation Working Group, CSEWG [1]. Technical work is organized by four CSEWG committees: Measurements, chair D. Smith, ANL, Evaluations, chair M. Chadwick, LANL, Formats and Processing, chair M. Greene, ORNL, Validation, chair R. McKnight, ANL. In 2001, CSEWG decided to develop a new version of the ENDF/B library, ENDF/B-VII. It is scheduled for release in December The present paper mentions the current status of the library and describes the extensive ENDF/B-VII effort. The paper attempts to convey the excitement of new advances in the database as grounded in recent experiments and model calculations, highlighting new evaluations and data testing. We mention users of the database and applications that are motivating the work. ENDF/B-VI RELEASE 8 The current version of the library, ENDF/B-VI, release 8, was issued in October Since then, the library has remained frozen. It includes neutron cross-section evaluations for 328 materials (313 isotopic and 15 elemental). All files can be easily accessed from a new Web interface developed by the NNDC in collaboration with the IAEA Nuclear Data Section [2]. This service, launched in April 2004, offers many new features including extensive graphical comparison with experimental data and other evaluated libraries. The request form for the ENDF retrieval can be seen in Fig. 1. FIGURE 1. Request form for advanced ENDF retrieval from the new Web interface [2], DEVELOPMENT OF ENDF/B-VII Background Funding supporting the nuclear-data improvements in ENDF/B-VII comes from the DOE (NNSA, Office of Science, NE). A major practical motivation is to consolidate numerous releases of the ENDF/B-VI library issued in , make use of the favorable situation at LANL, ORNL, and BNL for evaluation work, and utilize the still-available expertise of several key nuclear data gurus before their retirement. New features of ENDF/B- VII: New version of neutron-standard cross sections, Energies up to 150 MeV for selected materials, Improved n-evaluations of actinides, Improved n-evaluations for criticality safety, Improved n-evaluations of fission products, Photonuclear library, Charged-particle evaluations for astrophysics, and Evaluations relevant to RIA, ADS, and Gen-IV. 165

2 Some of these features are, at least partially, in response to the needs of such applications as criticality safety, non-proliferationtechnology using photon beams, Rare Ion-beam Accelerators (RIA), Accelerator Driven Systems (ADS), and the next generation of reactors (Gen-IV). In 2002, CSEWG decided that ENDF/B-VII would be issued in the current ENDF-6 format. The reason behind this important decision was a lack of funding for extensive modifications of the processing codes. Thus, there will be no ENDF-7 format, though various options including free format and the XML approach were discussed. Milestones in the ENDF/B-VII development: 2001: Definition of the project, 2002: Decision on format, evaluations, 2003: Evaluations and validation, 2004: Validation, standards, 2005: Standards, validation, final review, and 2005: Release (December). Several contributions related to ENDF/B-VII library are given in the present conference. Among them are invited talks on cross-section standards [3], uranium evaluations [4], neutron resonance evaluations [5], and data validation [6]. New Evaluations Altogether, 231 new or improved evaluations were submitted for inclusion into ENDF/B-VII. The list that does not include standards is as follows: Neutrons: 51 materials (14 actinides, 10 materials up to 150 MeV, etc.) 232;233;234;235;236;237;238;239;240;241 U, 237 Np, 239;241 Pu, 241 Am (LANL, mostly up to 30 MeV, with 232;233;234;235;238 U and 241 Pu resonances by ORNL) 27 Al, 28 Si, 196;198;199;200;201;202;204 Hg, 208 Pb (LANL, up to 150 MeV) 19 F, 35;37 Cl (ORNL, resonance region) 95 Mo, 99 Tc, 101 Ru, 103 Rh, 105 Pd, 109 Ag, 131 Xe, 133 Cs, 141 Pr, 143;145 Nd, 147;149;150;151;152 Sm, 153 Eu, 155;157 Gd, 160;161;162;163;164 Dy (KAERI- BNL collaboration) Photonuclear: 160 materials (LANL, largely based on the IAEA photonuclear project) Charged particles: 20 materials (all LANL) Protons: 3 H, 6;7 Li, 196;198;199;200;201;202;204 Hg Deuterons: 2;3 H, 3 He, 6;7 Li, Tritons: 3 H, 3 He, 6 Li, and 3-He: 3 He, 6 Li Neutron files were produced primarily by LANL, in particular for actinides. This is followed by ORNL evaluations in the neutron-resonance region motivated by criticality safety applications, and BNL-led work on fission products. Standards The new version of neutron cross-section standards is a particularly challenging part of the ENDF/B-VII development. An IAEA CRP project, with considerable CSEWG participation, is underway to produce new evaluations and to resolve an old issue of unrealistically low uncertainties from R-matrix fits. New evaluations will be prepared for 8 standard reactions 1 H(n,n), 3 He(n,p), 6 Li(n,t), 10 B(n,α), 10 B(n,α 1 ), 197 Au(n,γ), 235 U(n,f), and 238 U(n,f). In addition, 238 U(n,γ) and 239 Pu(n,f) will be evaluated because of strongly coupled databases with standards. The CRP has made considerable progress; cross sections should be finalized in 2004, followed by covariances in See V. Pronyaev et al. [3] for more details. Actinides Neutron cross-section evaluations for actinides are the primary focus of LANL evaluation activity. LANL has impressive expertise in evaluating these elements using the well-known nuclear reaction model code GNASH. This capability is considerably strengthened by close collaboration with the nuclear data measurement program at LANSCE, as well as with LANL integral data testing. LANSCE measurements are made with several instruments, in particular GEANIE (n,xγ), FIGARO (n,xn+γ), and NZ-spectrometer (n,ch.p.) in the fastneutron region, and by DANCE (n,γ) in the 1 ev - 20 kev region [7]. Of particular importance are critical uranium evaluations, especially for 233 U, 235 U, and 238 U. LANL used extensive sets of measurements for fission, (n,xn), and capture, and these were performed iteratively with important feedback from integral data testing (critical assemblies in the fast and thermal region, and 14-MeV pulsed spheres). See Phil Young et al. [4] for more details. An example of the improvement achieved for 238 U at 14-MeV incident neutrons is shown in Fig. 2. The new evaluation gives an excellent description (direct reactions) of the double-differential neutron spectrum just below the elastic peak. Another example illustrates recent experimental results obtained by the GEANIE spectrometer for the 239 Pu(n,2n) reaction. This important measurement provided the basis for the improved evaluation shown in 166

3 Cross Section (b/sr MeV) U+n Neutron Emission E n =14.05 MeV, θ=90 0 Baba, 1989 Los Alamos, 2004 ENDF/B VI JEF 3.0 JENDL Emission Neutron Energy (MeV) FIGURE 2. Recent improvement in the 238 U(n,xn) evaluation by Los Alamos supported by the integral pulsed-sphere benchmark of Livermore. σ (mb) Pu(n,2n) Reaction GEANIE Frehaut Lougheed Other Data LANL ENDF/B VI ENDF/B V.2 JENDL E n (MeV) FIGURE 3. Improved 239 Pu(n,2n) evaluation performed by Los Alamos, driven by the recent GEANIE measurement. Fig. 3. Note that both ENDF/B-VI and JENDL-3.3 are much higher than the new LANL evaluation. Modern treatment of direct reactions at high energies was incorporated into an entirely new evaluation of 233 U+n. As shown in Fig. 4, fission cross sections differ considerably from ENDF/B-VI and JENDL Am(n,γ) was considerably improved by providing a new isomeric (branching) ratio of 242g Am to 242tot Am. New values by T. Kawano et al. (this conference) based on careful model calculations and fit to integral data can be seen in Fig. 5. Resonances Evaluations in both the resolved and unresolved neutron-resonance region, in general motivated by nuclear data needs for criticality safety, were performed Cross Section (b) U + n Fission Cross Section ENDF/B VI JENDL 3.3 GNASH Calculation LANL, 2003 Meadows, 1974 Meadows, 1988 Carlson, 1978 Kanda, 1986 Behrens, Neutron Energy (MeV) FIGURE 4. Fission cross sections for 233 U+n by Los Alamos, part of an entirely new evaluation with modern treatment of direct reactions. Branching Ratio (Ground/Total) Shinohara (1997) Wisshak (1982) Mughabghab (1984) 0.5 Dovbenko (1961) LANL Chemistry, C-INC ENDF/B-VI JENDL-3.3 Present Incident Neutron Energy [ev] FIGURE 5. New evaluation of the branching ratio for 242g Am to 242tot Am in the 241 Am(n,γ) reaction (Kawano, LANL). by ORNL. A key technical ingredient behind this work is the unique capability at ORNL to perform highresolution measurements with ORELA. Another key factor is the powerful R-matrix code SAMMY [8], maintained by N. Larson (this conference), to analyze these data. Among the new features of SAMMY are the capability to accurately account for experimental resolution and generation of resonance-covariance data in both the resolved and unresolved energy regions. This latter capability, termed the retroactive method, is of particular interest. So far, the method was used to get covariance 167

4 FIGURE 6. Total and fission cross sections for 233 U+n in the resonance region measured and fitted by Oak Ridge (data by Guber et al., NSE 2000). estimates for 7 isotopes of Gd and 2 isotopes of Re. It remains open, though, whether these results will make it into ENDF/B-VII. Of particular importance are the new evaluations for 233 Uand 238 U in the resolved resonance region as well as for 235 U in the unresolved resonance region. These were already included in the new LANL files. Recent cross-section measurements with ORELA and fits by SAMMY are illustrated in two examples. These examples, shown below, demonstrate the importance of high-resolution ORELA measurements along with the power of the code SAMMY to fit these data using Reich- Moore resonance formalism. For more details see Luiz Leal [5]. 233 U total and fission cross sections in the resonance region can be seen in Fig. 6. Comparison of transmission data on nat Cl with SAMMY calculations using resonance parameters is shown in Fig. 7. At BNL, Mughabghab is preparing a comprehensive Atlas of Neutron Resonances [9]. This new BNL-325 contains evaluations in the thermal, resolved, as well as unresolved resonance regions. Some of these results are already incorporated in new fission-product evaluations. Fission Products Neutron cross-section evaluations for 24 fission products were performed by BNL in collaboration with KAERI, South Korea. An important ingredient behind this work is the recent re-evaluation by Mughabghab in the low-energy region [9]. The second key ingredient is the BNL capability in fast-neutron evaluations, based on the code EMPIRE [10]. As an example of the fission-product evaluation, we show in Fig. 8 a part of neutron capture for 157 Gd. FIGURE 7. Transmission for nat Cl+n in the resonance region measured at ORELA compared with SAMMY (data by Guber et al., PRC 2002). Cross Section (barns) Empire ENDF/B-VI Wisshak 1988 Beer 1986 Voignier 1984 Nakajima 1981 Joly 1974 Shorin 1995 Wisshak 157 Gd(n,γ) Incident Energy (MeV) FIGURE 8. Neutron-capture cross sections for 157 Gd evaluated by BNL (EMPIRE) illustrate improvement over ENDF/B- VI.8. A evaluation performed by EMPIRE is compared with ENDF/B-VI.8, which already contains a new low-energy evaluationuptothe1 st excited level (54.9 kev). In addition, BNL led an international project (WPEC Subgroup 21) to review and assess neutron cross-section evaluations for all fission products. Final recommendations are summarized in Table 1. Photonuclear Library Evaluations for 160 materials were supplied by LANL. These are mostly taken from the IAEA Coordinated Research Project on Photonuclear Data completed in 1999 [12]. In general, these evaluations cover incident photon energies up to 150 MeV, though in several important instances (actinides) evaluations go up to 20 MeV only. 168

5 TABLE 1. Summary of recommendations for the 218 best fission-product evaluations [11]. Some of them will be considered for ENDF/B-VII. Library or Full Resonance Fast Data Source File Region Region ENDF/B-VI Pre-ENDF/B-VII 27-2 JEFF JENDL CENDL BROND New BNL EMPIRE Total No of materials Cross Section (mb) Caldwell(1980) Varlamov(1987) Zhuchko(1978) BOFOD 235 U(γ, F) E γ (MeV) FIGURE 10. Management flowchart for preliminary ENDF/B-VII including 3 checking codes, ENDVER package, 2 plotting codes, 5 evaluated libraries, and CSISRS data. Data Testing Data testing consists of two phases, data verification and data validation. Primary responsibility for data verification, phase 1 testing, is on the NNDC. Phase 2 testing is much more involved, it includes complex benchmarking and validation done by several other laboratories (LANL, ANL, KAPL, ORNL, and Bechtel Bettis). FIGURE 9. Photofission data for 235 U, taken from the IAEA photonuclear project [12]. We note that 160 materials represent selected evaluations done by LANL using GNASH, by KAERI using LANL methodology, plus actinides done by Obninsk (BOFOD library). An example of a photonuclear evaluation is shown in Fig. 9 comparing photofission on 235 U with experimental data [12]. Charged-Particle Reactions All 20 new evaluations were produced by LANL. They include proton evaluations up to 150 MeV (LA150 library [13]) for 7 isotopes of Hg. These evaluations are based on the GNASH code [14], validated against available experimental data. The remaining 13 evaluations are for light nuclei at low energies, of interest to nuclear astrophysics. They are due to G. Hale, based on his R-matrix approach incorporated in the code EDA. Phase 1: Data Verification To facilitate data verification (checking and plotting), the NNDC has set up a preliminary ENDF/B-VII Web page, available to CSEWG members only. A related flowchart, shown in Fig. 10, covers runs by checking codes, comparison with other evaluations, comparison with CSISRS experimental data, and plotting. All data files, checking results, and an extensive set of plots are readily available to reviewers. This procedure will be also used for final ENDF/B-VII assembly, including all files that will be transferred from ENDF/B-VI.8 into ENDF/B-VII. Phase 2: Data Validation Quantitative data validation by integral benchmarks represents an important component in the development of the ENDF/B-VII library. Below, we will focus on recent results from Los Alamos on both fast- and thermaldata validation, see R. MacFarlane [6] for more details. Fast criticals. LANL calculations for fast critical assemblies were performed by the Monte Carlo code MCNP5 using both the new ENDF/B-VII cross sections and the previous cross sections from ENDF/B-VI.8. Both CSEWG benchmark models and the newer models 169

6 TABLE 2. MCNP results for fast assemblies showing values of k ef f. Assembly Experiment B-VII B-VI.8 Godiva (10) (19) (19) Jezebel (20) (18) (18) Flattop (10) (21) (19) Flattop-Pu (14) (23) (22) Jezebel (10) (18) (18) Flattop (14) (22) (23) Bigten (30) (16) (17) from the International Handbook of Evaluated Criticality Safety Benchmark Experiments (ICSBEP 2003) were used. CSEWG benchmarks included: Godiva (bare sphere of 235 U), Jezebel (bare sphere of 239 Pu), Flattop-25 (sphere of 235 U reflected by U), Flattop-Pu (sphere of 239 Pu reflected by U), Jezebel-23 (bare sphere of 233 U), Flattop-23 (sphere of 233 U reflected by U), Bigten (interleaved plates of 235 U and U reflected by U). Results for critical assemblies are summarized in Table 2. In general, preliminary ENDF/B-VII shows more reactivity and better k ef f compared to the current ENDF/B-VI. In particular, both Godiva and Jezebel show good k ef f due to improved fission and nu-bar. A dramatic improvement was achieved for 233 U. The reflector bias for 235 Uand 239 Pu is largely reduced (Flattop- 25 and Flattop-Pu versus unreflected k ef f ), though k ef f for Flattops still remain too large. A dramatic improvement is seen in Bigten, mostly resulting from new inelastic data in 238 U. The Godiva spectral index, C/E ratio for 238 U/ 235 U fission of 0.96, indicates that the spectrum is too soft. Thermal lattices. Thermal-data testing at LANL focused on high-enriched and low-enriched uranium-water systems. Calculations were done with MCNP5. For HEU lattices, ENDF/B-VI.8 was considered to be very good, and ENDF/B-VII shows a similar performance. For water-moderated LEU lattices, underprediction of k ef f was substantial. This longstanding problem, thought to result from 238 U, is also addressed by an important WPEC activity [15]. Current results for LEU lattices show that for the LCT006 series the new ENDF/B-VII gives an average value of compared to for ENDF/B-VI.8. This is a considerable improvement supported by the LCT001 series. Half of this improvement is from highenergy data (inelastic scattering) in the LANL evaluation of 238 U. The other half comes from new resonance parameters and lower thermal-capture evaluations by ORNL. On the other hand, the LCT002 series indicates a possible enrichment bias. CONCLUSIONS In 2005, CSEWG effort will concentrate on the data validation. If CSEWG keeps its current momentum, ENDF/B-VII will be released as planned - in December ACKNOWLEDGMENTS This presentation would not be possible without the collective effort of CSEWG. The present work was performed under the auspices of the U.S.D.O.E. Contract No. DEA-AC02-98CH REFERENCES 1. Cross Section Evaluation Working Group, NNDC web page (2004), URL 2. ENDF Retrieval, NNDC web page (2004), URL 3. Pronyaev, V., Status of International Neutron Cross Section Standards File, in Proceedings ND2004, Young, P. G., Systematic Analysis of Uranium Isotopes, in International Conference on Nuclear Data for Science and Technology 2004, AIP Conference Proceedings (Melville, New York, 2005). 5. Leal, L., Recent Cross Section Evaluations in Resonance Region at Oak Ridge, ND2004, op. cit. 6. MacFarlane, R., Testing new actinide cross sections for ENDF/B-VII, ND2004, op. cit.; see also Reports LA-UR and 3053, May Haight, R., Nuclear Data Experiments at LANL, in CSEWG Minutes, 2003, URL csewg. 8. Larson, N. M., Updated user s guide for SAMMY: Multivelel R-matrix fits to neutron data using Bayes equation, Tech. Rep. TM-9179/R4, ORNL, Mughabghab, S., Atlas of Neutron Resonances,Academic Press, (assumed to be published in 2005), manuscript under preparation. 10. Herman, M., et al., Nuclear reaction model code EMPIRE-2.19 (2004), to be released by the end of Obložinský, P., ed., Minutes of the Workshop on Assessment of Fission Product Evaluations, April 19-23, 2004, BNL, 2004, URL Chadwick, M., Obložinský, P., et al., Handbook on photonuclear data for applications: Cross-sections and spectra, Tech. Rep. TECDOC-1178, IAEA, Vienna, Chadwick, M. B., et al., Nucl. Sci. Eng. 131, 293 (1999). 14. Young, P. G., Arthur, E. D., and Chadwick, M. B., Comprehensive nuclear model calculations: Theory and use of the GNASH code, in Proc. IAEA Workshop Trieste 96 (World Scientific, 1998), pp Courcelle, A., Nuclear Data for Improved LEU-LWR Reactivity Prediction, ND2004, op. cit. 170