OECD Nuclear Energy Agency / L'Agence pour l'énergie nucléaire OECD-OCDE

 Home > Data Bank > Computer program services

 

Links to NJOY Web pages and Sites
 

NJOY:  Data Processing System of Evaluated Nuclear Data Files in ENDF format.

Recent Unofficial Updates at the NEA Data Bank


 Official Updates (LANL)  
     Archived Unofficial Updates and comments at the NEA Data Bank
Through interaction with Dimitri Rochman and Go Chiba, the following patches have been added:

  • upnea047 (Go Chiba) fix processing of fission width uncertainties in Reich-Moore compact format representation and uncertainties in MLBW compact format representation.
  • upnea048 (Go Chiba) include the contribution from the uncertainty of the scattering radius. Note: this is a trial patch because the format for storing the scattering radius uncertainty has not yet been officially approved.
  • upnea049 (A. Trkov) trivial corrections to keep the compilers happy.

Note by Andrej Trkov: By patching NJOY to conform with the ENDF-6 formats, the JENDL-3.3 fission spectrum covariance files become unprocessable. To circumvent this drawback, the chmf35 code is provided, which is designed to check and correct MF35 covariance files. It includes the following features:

  • The ENDF-6 code for the JENDL libraries is NLIB=6. The JENDL-3.3 library is released without the NLIB value set. CHMF35 corrects for this deficiency.
  • JENDL evaluators omitted discrete level photon data for metastable states. Physically this is a feasible representation since the photons are emitted with a delay and the ENDF-6 manual is not very specific on this issue. Practically, it causes processing problems with NJOY. An interim solution is to insert the missing level data, define full transition to the ground state (TP=1), but set the conditional transition to zero (GP=0). CHMF35 inserts the missing levels as described above.
  • JENDL-3.3 files contain fission spectra covariances for fission spectrum probability distributions while ENDF-6 format requires covariances for the normalised bin probabilities. The covariances are multiplied by the bin width to conform with the ENDF-6 format specification.
  • In the evaluation process the JENDL-3.3 evaluators adopted the criterion for the row-sum of fission spectrum covariance matrices of 1E-5. This criterion was found inadequate and was recently redefined. The fission spectra covariances were renormalised according to the revised prescription; the corrected files satisfy the revised zero-sum rule.
  • Note: the impact of unnormalised covariances matrices can be several hundred pcm. The Japanese approach was to take normalisation into account in the sensitivity coefficients, which was shown to be equivalent, but correcting the covariance matrices is safer in general. The CHMF35 source code can be downloaded.User instructions are embedded in the comments in the code. (30 March 2009)
  • Patch upnea045 fixes a problem in upnea041, pointed out by Skip Kahler and implemented by A. Trkov on 8 March 2009.
  • Patch upnea044, prepared by A. Trkov on 15 February 2009 increases the reserved fieldlength max from 500 to 800 in covr.
  • Patch upnea043, prepared by A. Trkov on 14 February 2009 fixes an inconsistency introduced in errorr in upnea041. The mffis was not properly initialised and the logic then skipped processing of MF33. upnea043 takes care of this deficiency.
  • Patch upnea042, prepared by A. Trkov on 6 February 2009 concerns the groupr module and refines upnea029 to decrease the tolerance by a factor 1000, and abolish it if processing mt=18 spectrum. In fact, truncation of the minimum cross section causes problems in the case of the fission spectrum when processing covariance matrices.
  • Patch upnea041, prepared by A. Trkov in January 2009 concerns the errorr module and
    • Fixes the selection of fission spectrum at higher energies
    • Adds an option to define the average energy for MF35
    • Improves diagnostic messages
  • Patch upnea040 solves a problem with thermal scattering matricies in GROUPR. There were small discrepancies between the sums thermal scatter matrices and the group cross-sections. This update renormalises the interpolated spectra (produced during integration) to sum to unity.
  • Patch upnea038 fixes a warning message when asymmetric covariance matrices are encountered in covr. Patch upnea039 fixes processing of covariances in evaluated data files containing MF32
  • Patches upnea036 and upnea037 fix a problem in patches upnea033
  • and upnea034 processing Th-232 resonance covariance files from ENDF/B-VII. The patches have no impact on existing processed libraries.
  • The patches upnea033 to upnea035 affect the processing of new evaluations with resonance covariances in compact format representation.
  • The patches upnea025 to upnea028 apply only to dosimetry library processing (such as IRDF-2002 or activation data in dosimetry ACE format). Processing of the data in ENDF File-10 is affected. Patch upnea028 increases the reserved array size to process U-238 and Gd-nat from IRDF-2002. Patches P262 to P264 were in the meantime withdrawn.

The processed IRDF-2002 library in dosimetry ACE format for the MCNP family of codes is available from www-nds.iaea.org/irdf2002/