NJOY2016.76

This update fixes the following issues:

• Increase an array size to properly process Pt covariances in ENDF/B-VIII.1.

NJOY2016.75

This update fixes the following issues:

• Corrected a typo in the gateff subroutine in thermr (changed 1220 into 1200, see issue #76)

In addition, this update made added the following features and changes:

• A small change to the NXS array for continuous energy and photonuclear ACE files (iopt = 1 and 5): in both cases, the isomeric state S, atom number Z and mass number A are now stored in NXS(9), NXS(10) and NXS(11) respectively.
• A new input option for acer was introduced for setting the zaid identifier of the continuous energy and photonuclear ACE files (iopt = 1 and 5):
  • use za in the zaid regardless of the metastable state of the nuclide (default, this is how NJOY has always worked)
  • use the metastable zaid rules (MCNP 6.3 or lower) in which we use za for ground state nuclides and use za + 300 + s * 100 for metastable nuclides (Am242 and Am242m are exceptions to this rule, for these we use 95642 and 95242 respectively)

NJOY2016.74

This update fixes the following issues:

• Correct the Euler-Mascheroni constant (this has an influence on the Coulomb wave functions and therefore influences LRF=7 evaluations that have charged particle channels)

NJOY2016.73

This update fixes the following issues:

• Fix an issue in ACER for thermal scattering leading to energy values being out of order when plotting the coherent elastic scattering cross section (this issue only affects plots, the thermal scattering ACE files do not change).
• Increased allocation of an array in LEAPR to accommodate ENDF/B-VIII.1 thermal scattering evaluations and added a check to avoid an infinite loop when using a very fine beta grid. In addition, LEAPR will now warn the user about potential excessive calculation times and print out progression in the phonon expansion sum when the phonon expansion order is large.
• Added logic to MODER to read background R-matrix element information from LRF=7 resonance parameter data.
• Updated RECONR to use background R-matrix element information from LRF=7 and added test 81 using ENDF/B-VIII.1 Sr88.
• Fixing a few thing related to intel compiler warnings and errors.

NJOY2016.72

This update fixes the following issues:

• Fixed an issue in GROUPR related to an error coming up in production matrix calculations. Depending on when a user asks for a production matrix associated to a reaction, it is possible that the reference frame of the previous reaction is used instead (caused by erronously defining an already declared global variable as local with a "save" attribute). In some circumstances, this causes NJOY2016 to error out (with a message related to unsupported reference frames). No test results had to be updated due to this change.
• Fixed issues in acer to properly print already existing dosimetry and photoatomic ace files when running a stand-alone acer iopt=7 job.
• The meaning of legord and ifissp in the ERRORR input file has been repurposed when mfcov=34. The values now represent the L,L1 values of the MF34 sub-subsection to be calculated (instead of the first one). By default, the L=1,L1=1 sub-subsection will be calculated which in almost all cases will correspond to the first sub-subsection in the MF34 data (as a result, the default behaviour of NJOY2016 will not change).
• Increased allocation of arrays to accommodate ENDF/B-VIII.1 evaluations.
• Fixed a typo in the name for MT195 in ACER.

NJOY2016.71

This update adds the new MF7 MT451 (thermal scattering general information) ENDF format to MODER so that this module will be able to interpret the new MF7 section. No other capability in NJOY2016 currently uses the information in this section.

This update also resolves an issue encountered when processing some JENDL5 evaluation that use LAW=7 (which ACER converts into LAW=1) due to errors in the temporary files. See issue #293 for more information.

NJOY2016.70

This update fixes a number of minor issues:

• Fixed an issue in HEATR when reading evaluations with large multiplicity tables in MF6.
• Fixed an issue in HEATR when calculating the average outgoing energy from a distribution that uses multiple interpolation ranges in TAB1 records (test 79 was added to detect this issue in the future). Mainly nuclides using MF5 instead of MF6 are impacted by this change (e.g. Sn119 and Sn122 from ENDF/B-VIII.0).
• Fixed an issue in HEATR where the photon recoil needed to be multiplied by the photon multiplicity to obtain the photon recoil per interaction.
• Fixed a crash in THERMR when asking for S(a,b) processing (iinc=2) while no ENDF tape is given (nendf=0).
• Multiple ERRORR calls can now be made in the same input file without crashing. This is of interest to users that wish to process MF34 and MF35 (where ERRORR needs to be called for each sub-subsection and incident energy group). The issue was related to arrays being allocated but not unallocated in the previous ERRORR run in NJOY's Sammy routines (evaluations using MF2 LRF=7 had this issue).
• Fixed an issue in ACER where the number of photons given in the ENDF file was larger than the hardcoded limit. The new limit is now adaptive.
• Fixed an issue in ACER where NaN values were produced in the postscript file for the checking plots.

A few compiler warnings have been resolved as well (unused variables). For source files that were corrected in this way, the remaining warnings relate to equality comparisons for real values, unused dummy arguments in subroutines and potential 0 indices into arrays (in all cases, if statements prevented this from happening).

NJOY2016.69

This update fixes a number of minor issues:

• PURR now writes Bondarenko data obtained from the probability tables to MF2 MT152 instead of the Bondarenko data obtained from the direct sampled cross sections (for very low dilutions, the Bondarenko data obtained using these two methods does not align, with the direct sampled data leading to extremely low P1 values). When comparing with the Bondarenko data at low dilutions obtained with UNRESR, the Bondarenko data obtained from the probability table directly seems to be the best.
• MF6 LAW=2 represents discrete two body scattering in which only angular distribution data is given (knowing that the outgoing energy of the secondary particle can be determined through kinematics when the angle is known). When calculating heating numbers based on LAW=2, ACER assumes that the yield of the secondary particle is 1, which is correct in all cases except when MT5 is used as a lumped reaction. Heating numbers in ACER for photonuclear files using LAW=2 in an MT5 entry are now correctly multiplied by the yield. A warning message is printed out whenever this situation is detected. Test 78 was added as part of this correction.
• Previously, ERRORR would segfault for LRF=7 resonance evaluations when MF33 was present without MF32. A check for this situation now avoids this.
• Fixed an issue in GROUPR when reading some of the FENDL3.2 evaluations.

A few compiler warnings have been resolved as well (unused variables). For source files that were corrected in this way, the remaining warnings relate to equality comparisons for real values, unused dummy arguments in subroutines and potential 0 indices into arrays (in all cases, if statements prevented this from happening).

NJOY2016.68

This update fixes a number of minor issues:

• there was an indexing error in the calculation of inelastic thermal scattering mubar in ACER for IFENG=2 ACE files. Test 74 was added to track this issue.
• there was a division by zero that caused a segmentation fault in Debug mode, also for IFENG=2 ACE files.
• removed an erroneously added FEND record in the PENDF file when no MF12 data gets added to the PENDF file (related to issue #250).
• correction in ACER to fix issue #250. Previously it could happen that ACER picked the wrong temperature from a multi-temperature PENDF file when the materials in the PENDF only contained linearised MF3 data (no MF10, MF12 or MF23 data was added by RECONR).
• fixed an indexing problem in CCCCR for delayed neutron data and added test 76.
• there was a use of wrong memory addresses in CCCR not accounting for GROUPR's energy structure offset.

NJOY2016.67

This update resolves a number of issues encountered when processing the newly released JENDL5 nuclear data library and TENDL-2021 library for evaluations with sub-actinide fission, and provides fixes for a few minor issues.

Due to an internal compiler error, gcc-11 will not be supported. Users should consider moving to gcc-12 or any version from gcc-7 to gcc-10.

The following minor changes and fixes are included in this version:

• several corrections to the ERRORR source code to remove compiler warning messages (these were made while looking into issue #211 for gcc-11 compilation)
• increased the nxcmax array size parameter in ACER to resolve issue #228
• increased the nthr array size parameter in PURR to resolve an issue in JENDL5 Mn55 (issue #233)
• fixed an issue for processing JENDL5 U235 in ACER (issue #237)
• correction in HEATR to properly retrieve the fission energy release component's Q value at 0 eV (issue #209 - there currently is no evaluation impacted by this change)
• corrected ACER following issue #188. Evaluations using LAW=7 in MF6 data that are added to the DLWH block are impacted by this change (i.e. for secondary particle types that are not the incident particle type). For ENDF/B-VIII.0, only the secondary alphas from MT16 in Be9 is impacted by the change.
• properly set the za in the case of sub-actinide fission (MF10 MT18 data) when using MF10 automated processing in GROUPR (issue #241)
• exclude fission from a>4 particle production matrices in GROUPR
• when using the ign option in GROUPR for predefined group structures, some of the newer options did not return the correct structure, this has been fixed
• an array size for reading in ENDF data was increased to accommodate a new evaluation

A large number of compiler warnings have now been resolved (unused variables, large static arrays converted into allocatable arrays, etc.). There are still a lot of warnings but their number is heavily reduced. For source files that were corrected in this way, the remaining warnings relate to equality comparisons for real values, unused dummy arguments in subroutines and potential 0 indices into arrays (in all cases, if statements prevented this from happening).