|
|
This is the issue-tracking page for Vers. 99 of the NJOY Nuclear Data Processing System. Issues are entered by the developer (ryxm@lanl.gov) in response to postings to the NJOY list (njoy@nea.fr), items reported directly by email, observations by NJOY users at Los Alamos, problems noted by the developer, or new ideas from the developer. The summary table includes an issue number, issue status, and issue title. Click on the issue number to jump to a more detailed description of the issue and its resolution, if any. Issue status can be S (submitted), P (pending, meaning there are some comments posted on the problem), or R (resolved, meaning that there is an NJOY update that handles the issue). |
|
Number |
Status |
Title |
|---|---|---|
|
R |
Incorrect initialization in HEATR |
|
|
R |
Allow more Legendre terms in HEATR |
|
|
R |
Vertical segments in File 6 distributions |
|
|
R |
Incorrect law=61 distributions |
|
|
R |
More compact charged-particle energy-angle distributions |
|
|
R |
Possible infinite loop in PLOTR |
|
|
R |
MT=0 in ACER message |
|
|
R |
Interpolation for JENDL-3.2 U-235 nubar |
|
|
R |
Problem with redundant MT=10 |
|
|
R |
Error in MCNP/sabcol reported for nonthermal problem |
|
|
R |
Law=5 used in JEF U-233 fission spectrum |
|
|
R |
Need special skip6 routine in more places |
|
|
R |
Error in up11 for law=7 and newfor=0 |
|
|
R |
Missing initialization in acer/ptlegc |
|
|
R |
Reaction self-shielded cross section not printed correctly |
|
|
R |
Reich-Moore channel-spin format added |
|
|
R |
Allow more groups for photons |
|
|
R |
Improve sigfig truncation for multigroup matrix elements |
|
|
R |
Incorrect threshold for particle production cross sections in ACER |
|
|
R |
DTFR doesn't find correct material and temperature on PENDF file |
|
|
R |
Constant nlmax used before being initialized |
|
|
R |
Problem in DLAYXS with 69 groups |
|
|
R |
Unresolved with no resolved for JENDL materials |
|
|
R |
PDF same as CDF for law-7 charged-particle emission |
|
|
R |
Problems processing recent EFF-3.0 Be-9 evaluation |
|
|
S |
Make MT=875-890 available to MCNP |
|
|
R |
Problems with multigroup runs for 150 MeV data |
|
|
S |
ACER has problem opening some scratch files under Forte6 |
|
|
S |
Problem with 181-point tabulated angulur distributions in MF=4 |
|
|
R |
GROUPR error when running with 300 gamma groups |
|
|
R |
Failure for Pb-208 from ENDF/B-VI Release 6 |
|
|
R |
RECONR fails with 100 GeV photoatomic data |
|
|
R |
Charged-particle heating zero for some parts of Be-9 ACE file. |
|
|
R |
ACE interpolation-law printout |
|
|
R |
Excess timers in PURR listing. |
|
|
R |
Heating zero at first energy point in ACE file. |
|
|
R |
Allow for larger ACE files |
|
|
R |
Length of some ACE thermal files too large by one word |
|
|
R |
ACE parameter LANDH wrong for Be-9 |
|
|
R |
Problem with ACE law3/33 for isomeric targets |
|
|
R |
Error message for JENDL photonuclear evaluations |
|
|
R |
Two-body recoils for photonuclear evaluations |
|
|
R |
Tabulated fission spectra for photonuclear evaluations |
|
|
R |
Isotropic discrete recoils |
|
|
R |
Levels of the (n,2n) reaction in RECONR |
|
|
R |
Levels of the (n,2n) reaction in HEATR |
|
|
R |
Histogram interpolation for law=7 |
|
|
R |
Conversion of law=7 to law=1 |
|
|
R |
ACER fails for charged-particle set with tabulated angular distribution. |
|
|
R |
Levels of the (n,2n) reaction in ACER |
|
|
R |
Diffs in energy grids from BROADR |
|
|
R |
Bad vertical scale for 3-D plots. |
|
|
R |
Reactions with nonzero cross section at threshold |
|
|
R |
Missing threshold for charged-particle reaction |
|
|
R |
Common variable coeff not set |
|
|
R |
Photoatomic heating value given in eV-barns instead of MeV-barns |
|
|
R |
Problem with heating in probability tables |
|
|
R |
ACER doesn't include delayed neutron data as allowed for MCNP4c |
|
|
R |
Unitialized variables introduced in up57 |
|
|
R |
Undefined variables introduced in up63 |
|
|
R |
Fix radionuclide production for EAF2001 |
|
|
R |
Bad MT name string in GROUPR |
|
|
R |
Bad Kalbach a factor for photonuclear cases |
|
|
R |
Add photonuclear capability for MATXS |
|
|
R |
Confusion of pi versus pi |
|
|
R |
Temperature field in RECONR |
|
|
R |
MCNP Fluorescence Data |
|
|
R |
Delayed neutron consistency checks and plots |
|
|
R |
Cross sections or yields that are zero at all energies |
|
|
R |
8-group delayed neutron data in ACER |
|
|
R |
8-group delayed neutron data in GROUPR |
|
|
R |
Editing error in LWR EPRI weight function |
|
|
R |
Upscatter in getsed |
|
|
R |
Recent typographical errors |
|
|
R |
Increased energy resolution in PLOTR output |
|
|
R |
Plot differences or ratios |
|
|
R |
Tighten up tolerances for integral thinning in BROADR |
|
|
R |
Cross section grids in BROADR |
|
|
R |
Tighten up integral thinning tolerances in RECONR |
|
|
R |
Cross section grids in RECONR |
|
|
R |
RECONR printouts |
|
|
R |
Photoatomic edge discontinuities |
|
|
R |
Particle production thresholds |
|
|
R |
More energy points in ACE files |
|
|
R |
Plots with proton production present |
|
|
R |
ENDF-6 covariance processing in MODER |
|
|
R |
Photoatomic heating |
|
|
R |
Negative percent differences |
|
|
R |
Unresolved probability tables strange for ENDF66 |
|
|
R |
Photon production eV-barns |
|
|
R |
Coherent thermal scattering for aluminum, iron, lead |
|
|
R |
ACE files for incident charged particles |
|
|
R |
MATXS self shielding and group bounds |
|
|
R |
Increase arrays for some evaluations |
|
|
R |
PURR zeroes two flags |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
23 Jul 2000 |
|
By |
Margarete Mattes (mattes@ike.uni-stuttgart.de) |
|
Title |
Incorrect initialization in HEATR |
|
Details |
To process Ni-58 of JEFF-3.0, I had to make a small correction.... |
|
Comments |
The parameter ir was being initialized incorrectly. |
|
Impact |
Unknown beyond the case cited. |
|
Resolution |
Fixed in up15 (31 Jul 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
31 Jul 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Allow more Legendre terms in HEATR |
|
Details |
Saw a crash in heatr/h6ddx while testing the proposed JEFF-3T evaluation for Fe-56. |
|
Comments |
It turned out that this new evaluation used more Legendre terms than previous evaluations, so we increased the size of the arrays to handle up to 15 orders. |
|
Impact |
Only affects Fe-56 from the new JEFF-3T set. |
|
Resolution |
Fixed in up15 (31 Jul 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
21 Jul 2000 |
|
By |
Zhang Jingshang (zhangjs@iris.ciae.ac.cn) |
|
Title |
Vertical segments in File 6 distributions |
|
Details |
He observed divides by zero in HEATR for Zr-nat from CENDL. |
|
Comments |
When we looked at the evaluation, we found vertical segments in the distributions for some energies. These vertical segments were causing the divide by zero in the interpolation routine. We fixed the code to find such vertical segments, move the upper energy up slightly, and print a message to the user. Evaluators should be careful not to include such features in their files. |
|
Impact |
This only affects a few CENDL evaluations currently. |
|
Resolution |
Fixed in up15 for HEATR and up16 for GROUPR (21 Jul 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
31 Jul 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Incorrect law=61 distributions |
|
Details |
Originally noticed strange anisotropic distributions at threshold where the scattering cross section is zero. |
|
Comments |
It turned out that this was caused by an incorrect index in one array, and we originally thought that the patch would only be cosmetic in that it would change the shape of distributions with zero cross section. However, we now realize that it affects all the law=61 distributions. The corrected coding was tested against analytic calculations of the scattering probability at the forward and backward limits. |
|
Impact |
Affects all distributions that use law=61. This means MCNP4C libraries for such important ENDF/B-VI materials as Fe, Ni, Cr, Cu, and Pb. This is an important patch. |
|
Resolution |
Fixed in up17 (31 Jul 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
31 Jul 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
More compact charged-particle energy-angle distributions |
|
Details |
We want to use the compact law=4 format for isotropic charged-particle distributions and the more general law=61 for anisotropic cp distributions. This problem was noticed when processing a proposed evaluation for Fe-56 from JEFF-3. |
|
Comments |
The problem is that we cannot tell which is which until reading past the first few energies, so the patch is pretty complex! |
|
Impact |
Basically affects appearance and reduces the size of files. Numerical results when using MCNP should be the same either way. |
|
Resolution |
Fixed in up17 (31 Jul 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
3 Aug 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Possible infinite loop in PLOTR |
|
Details |
There is a possibility for an infinite loop. |
|
Comments |
Simple fix. |
|
Impact |
Probably not very significant. |
|
Resolution |
Fixed in up18 (3 Aug 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
3 Aug 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
MT=0 in ACER message |
|
Details |
The ACER message about energy-dependent yields gives MT=0 for the reaction type. |
|
Comments |
Just a small patch required. |
|
Impact |
This is a trivial cosmetic patch and will not affect any results. |
|
Resolution |
Fixed in up18 (3 Aug 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
3 Aug 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Interpolation for JENDL-3.2 U-235 nubar |
|
Details |
Problems seen in nubar. |
|
Comments |
Must set the nonlin flag. |
|
Impact |
Probably only this one material |
|
Resolution |
Fixed in up18 (3 Aug 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
3 Aug 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Problem with redundant MT=10 |
|
Details |
While studying the JEF-2.2 evaluation for lead to find the problem summarized above, we also noted that it used the reaction MT=10. |
|
Comments |
MT=10 was originally proposed to represent "continuum neutron emission," but it never really became a part of the ENDF scheme and NJOY never has handled it. Unfortunately, it is leading to a double counting of these reactions in the total cross section for ACER and affecting results from HEATR. |
|
Impact |
So far, it seems that only Pb from JEF is affected. |
|
Resolution |
Fixed in up18 and up19 (3 Aug 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
11 Aug 2000 |
|
By |
Waclaw Gudowski (wacek@neutron.kth.se) |
|
Title |
Error in MCNP/sabcol reported for nonthermal problem |
|
Details |
MCNP crashes with error in sabcol when using data from ENDF/B-VI Release 6. |
|
Comments |
There turned out to be two problems here. When processing angular distributions using ltt=3 with newfor=0 (MCNP4B compatibility), there was an extra tosend that caused the code to skip over the first reaction after the elastic MF=4. This lead to a bad tyr=0 value, which somehow causes MCNP to issue a confusing error message about sabcol, even when no S(alpha,beta) data are being used. Second, the code was finding the wrong value for the lct parameter (lab or cm) for sections of File 4 that were fully isotropic. This also can result in an incorrect value for the ACE tyr parameter, which leads to the strange message from sabcol. This problem was introduced when the NJOY97 coding was being converted to block structuring. |
|
Impact |
The first part of the problem affects runs using ENDF/B-VI Release 6 (which contains the 150-MeV extensions) with newfor=0. It can be avoided by using Release 5, which is identical to Release 6 for MCNP4B purposes. The second part is more serious, and it affects many evaluations for the minor materials where evaluators haven't gone to the trouble to construct realistic angular distributions. |
|
Resolution |
Fixed in up20 (16 Aug 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
11 Aug 2000 |
|
By |
Waclaw Gudowski (wacek@neutron.kth.se) |
|
Title |
Law=5 used in JEF U-233 fission spectrum |
|
Details |
There are some funny results for U-233.... |
|
Comments |
The unique feature of U-233 from JEF-2.2 is the use of law=5 to describe part of the secondary energy distribution for fission. This law is normally used only for delayed neutrons, and not for prompt neutron emission. Unfortunately, NJOY does not really process this representation (and never has). The numbers written into the file are just junk. MCNP does have a formal representation for law5 that uses equally probable bins. Equally probable bins are not very good for representing fission; the cummulative law4 is much better. The best solution here is to convert the law=5 part of MF=5/MT=18 to law=1 tabulated format. We are putting an error message in NJOY to alert users to this problem. |
|
Impact |
Only affects U-233 from JEF-2.2. |
|
Resolution |
Error message installed in up20 (16 Aug 2000), but that doesn't really solve the problem! |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
16 Aug 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Need special skip6 routine in more places |
|
Details |
The reading of MF6 is sometimes getting out of position. |
|
Comments |
There are some additional places where we need to use the special version of skip6 called skip6a. For law=7 sections, the tab2 is changed to a tab1 in order to pass on the integrated angular distribution. |
|
Impact |
This only affects processing of materials that use law=7 with newfor=1 (MCNP4C libraries). For ENDF/B-VI, this is Be-9. |
|
Resolution |
Fixed in up20 (16 Aug 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
16 Aug 00 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Error in up11 for law=7 and newfor=0 |
|
Details |
An error was made in the part of up11 that handles law=7. |
|
Comments |
This problem showed up when processing Be-9 from ENDF/B-VI. |
|
Impact |
Only affects ACER runs using newfor=0, that is files for MCBP4B, for evaluations that use law=7 for energy-angle distributions. For ENDF, this is only Be-9. |
|
Resolution |
Fixed in up20 (16 Sep 00). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
16 Sep 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Missing initialization in acer/ptlegc |
|
Details |
Missing initiazation for dco |
|
Comments |
Some platforms may have problems for incident charged particles in ACE files if this parameter is not initially set to zero. |
|
Impact |
Only affects ACE files for incident charged particles (e.g., the APT incident-proton 150-MeV data) on some platforms. |
|
Resolution |
Fixed in up20 (16 Sep 00). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
28 Sep 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Reaction self-shielded cross section not printed correctly |
|
Details |
The self-shielded cross sections for the reactions appear to be all the same! |
|
Comments |
The self-shielded cross sections are not being printed out correctly for the reactions, but the total is OK. The gendf file is OK, so libraries made with NJOY99 are OK. |
|
Impact |
Only affects the listings, not the data that goes on to production libraries. |
|
Resolution |
Fixed in up21 (28 Sep 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
10 Jun 1999 |
|
By |
Nancy Larson (nml@ornl.gov) |
|
Title |
Reich-Moore channel-spin format added |
|
Details |
There has been a problem in the ENDF Reich-Moore resonance format for a long time for materials with spin greater than zero in that it doesn't specify the widths for different channel spins explicitly. R. Q. Wright (ORNL) suggested using a minus sign on the J value as a flag to use the other channel spin, and this format modification was approved by CSEWG. Nancy Larson provided an NJOY patch to implement this option for NJOY97. John White provided test results from SAMMY to verify the operation of the patch in NJOY. |
|
Comment |
We made a few changes to Nancy's contribution to make it fit into the block structured coding used in NJOY99. We tested the results against the SAMMY numbers, and Red Cullen and Bob MacFarlane also did tests of NJOY against RECENT running at LLNL. |
|
Impact |
At present, this change does not impact any existing processing. Several evaluation under development will use it (e.g., Al-27 from ORNL). |
|
Resolution |
Fixed in up22 (28 Sep 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
28 Sep 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Allow more groups for photons |
|
Details |
LLNL requested that we allow for up to 400 groups for photoatomic photons and for photon production from particle reactions. |
|
Comments |
We needed to change some dimensions and sizes in gaminr, dtfr, and groupr. |
|
Impact |
No impact on existing processing. The change only has an impact if you want more than 150 gamma groups. |
|
Resolution |
Fixed in up23, up24, and up25 (28 Sep 2000 and 11 Oct 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
11 Oct 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Improve sigfig truncation for multigroup matrix elements |
|
Details |
There are too many differences in very small matrix elements when going from platform to platform. |
|
Comments |
The previous algorithm was only operating on the in-group element. We now work on all of them to try to make the results for elastic and discrete inelastic matrices the same on different machines. The basic idea is that these numbers are obtained by subtraction of numbers on the order of unity, so any results less than about 1e-7 are just random numbers and can be removed. |
|
Impact |
The affect of this change on real calculations is entirely negligible, but it does help to make it easier to do diffs between different platforms. |
|
Resolution |
Fixed in up25 (11 Oct 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
12 Oct 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Incorrect threshold for particle production cross sections in ACER |
|
Details |
The current coding sometimes gets the threshold for charged particle production off by one point. |
|
Comments |
It is necessary to supply a slight shift in the test for the threshold. |
|
Impact |
Very minor, but it looks better after being fixed. |
|
Resolution |
Fixed in up26 (12 Oct 2000) |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
27 Oct 2000 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
DTFR doesn't find correct material and temperature on PENDF file |
|
Details |
The PENDF cross sections used for the DTFR plots are not coming from the correct material and temperature. |
|
Comments |
A do loop was translated incorrectly in converting to block constructs without statement numbers. |
|
Impact |
Damages the DTFR plots, but not the multigroup tables. |
|
Resolution |
Fixed in up27 (27 Oct 2000) |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
6 Oct 2000 |
|
By |
Piet de Leege (p.f.a.deleege@IRA.TUDelft.NL) |
|
Title |
Constant nlmax used before being initialized |
|
Details |
In up15.13 (heatr) nlmax used before init, perhaps data statement missing. |
|
Comments |
The statement setting nlmax was inserted inside the "sw" conditional block instead of after both branches of the block. Therefore, it was only available to the short-word version of the code, but Piet was running at 64 bits on his alpha. |
|
Impact |
This only affects people using native 64-bit modes on their platform. |
|
Resolution |
Fixed in up30 (5 Nov 2000) |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
17 Oct 2000 |
|
By |
Cornelis Broeders (Cornelis.Broeders@iket.fzk.de) |
|
Title |
Problem in DLAYXS with 69 groups |
|
Details |
Found another bug in the routine ccccr.f. For larger group numbers (27 is OK, 69 is wrong) the output file dlayxs is wrong. After some debugging, the correction seems to be ... With this correction, my results are OK now. |
|
Comments |
His diagnosis was correct. The pointer in the e array for the moreio call was wrong, which caused the later parts of the data block to overwrite the first part, leading to incorrect results. |
|
Impact |
This only affects the generation of the CCCC delayed-neutron data file DLAYXS. |
|
Resolution |
Fixed in up29 (5 Nov 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
25 Oct 2000 |
|
By |
Andrej Trkov (A.Trkov@iaea.org) |
|
Ttitle |
Unresolved with no resolved for JENDL materials |
|
Details |
In our work on the WIMS-D library, we identified processing problems in JENDL-3.2 for the following nuclides: 44-Ru-103, 44-Ru-106, 51-Sb-125, 52-Te-127m, 61-Pm-148g, and 61-Pm-149. There may be more. These nuclides have: - Resonance range define entirely by unresolved resonance parameters, - No resolved resonance parameter ranges is defined, - For some nuclides the resonance range starts at relatively high energies (>300 eV for Ru-103). For these materials, the cross sections are zero below the resonance range. Below is a patch.... |
|
Comments |
We had noticed that here also, but just skipped the materials. The correction proposed by Andrej seems reasonable. |
|
Impact |
Only affects these JENDL fission-product evaluations (and maybe a few more not listed). There is no affect of this change for ENDF or JEF processing. |
|
Resolution |
Fixed in up32 (5 Nov 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
2 Nov 2000 |
|
By |
Chikara Konno (konno@fnshp.tokai.jaeri.go.jp) |
|
Title |
PDF same as CDF for law-7 charged-particle emission |
|
Details |
I think that acer.8863 and acer.8864 should be cut since PDF data become the same as CDF data in the angular distribution of the particle production data for law 7.... |
|
Comments |
The problem was an incorrect pointer for the statement that readjusted the normalization and adjusted the number of significant figures for the PDF in the routine for charged-particle emission in acer. |
|
Impact |
Only affects ACE files for evaluations using Law 7 for describing the energy-angle emission of secondary charged particles. The one case of this in ENDF/B-VI is alpha emission from the (n,2n)3alpha reaction in Be-9. |
|
Resolution |
Fixed in up28 (5 Nov 2000). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
15 Nov 2000 |
|
By |
Ulrich Fischer (ulrich.fischer@irs.fzk.de) |
|
Title |
Problems processing recent EFF-3.0 Be-9 evaluation |
|
Details |
We encountered some problems when tring to process the most recent Be-9 EFF-3.0 file (NMOD=3) with ACER/NJOY.14. This file is peculiar in the following sense: (a) In addition to MT=16, it has individual (n,2n) reaction channels stored in MT=875-890 (MF=3, 6, 33). (b) The file is very big (ASCII file size of 17 MB). With the following bug fixes, we succeeded in processing the file.... |
|
Comments |
Of the changes that Ulrich proposed, some had already been included in up20 in August before he made his report, but he also discovered that the pointer to the moreio call was not being incremented as new pages were brought in, which had the affect of losing some of the words further down in the record and clobbering the first part of the record that was read in earlier. The problem only occurs for very long records like those in the very detailed Be-9 evaluation from EFF-3.0. |
|
Impact |
This patch will only affect evaluations using Law7 for angle-energy distributions, and only those using very detailed tabulations. The only known example is this Be-9 evaluation from EFF-3.0. |
|
Resolution |
Fixed in up28 (5 Nov 2000). |
|
Number |
|
|---|---|
|
Status |
Submitted |
|
Submitted |
15 Nov 2000 |
|
By |
Ulrich Fischer (ulrich.fischer@irs.fzk.de) |
|
Title |
Make MT=875-890 available to MCNP |
|
Details |
The new Be-9 evaluation for EFF-3.0 uses MT=875-890 to represent different levels of the (n,2n) reaction. Even when considering only MF=3 data using the dosimetry xs option, we cannot persuade ACER to provide them on the ACE file (of course, they are redundant) to the MF=16 data. Furthermore, we would eventually like to use the MF=6 data for MT=875-890 instead of MT=16. |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
2 Oct 2000 |
|
By |
Harm Wienke (hwienke@sckcen.be) |
|
Title |
Problems with multigroup runs for 150 MeV data |
|
Details |
Harm first reported error messages from findex when running with 150 MeV data in GROUPR. He suggested a patch to fix that, and then he had some floating-point overflows and suggested a further change. |
|
Comments |
It turned out that there were two problems. An index was set to 1 in an array used by tab2io, which caused the data being read in to clobber the indexes used by the dynamic storage allocation scheme. This is what caused the findex errors. The other problem was related to the "over" constant, which allows the routine to extrapolate slightly above the highest energy point in the table. It is necessary to make sure that the code doesn't use "over" at the break between the 0-20 MeV data and the 20-150 MeV data that appears when the option LTT=3 is used. Otherwise, GROUPR finds an integration panel of zero width, which leads to the overflow. |
|
Impact |
This will cause trouble with multigroup processing for all the Los Alamos 150-MeV evaluations that were prepared in support of the APT project. |
|
Resolution |
Fixed in up31 (5 Nov 2000). |
|
Number |
|
|---|---|
|
Status |
Submitted |
|
Submitted |
23 Jan 01 |
|
By |
Jean-Christophe Sublet (jean-christophe.sublet@CEA.FR) |
|
Title |
ACER has problem opening some scratch files under Forte6 |
|
Details |
Using a Forte6 f95 version of NJOY99-24 or 32 for in07 and in14, ACER wants to open unit -13 or -10 through gamsum calling openz "open(nun,form=for,status=age)" but thinks that the unit (nun 13 or 10) is already connected !! |
|
Comments |
The Sun Forte6 f95 compiler seems to be very strict about opening a scratch file that is already open, although all other compilers used for NJOY so far were more forgiving. We will just have to be careful to close a unit used as a scratch file before reusing the unit for another purpose. |
|
Impact |
All ACER runs with codes compiled using Sun Forte6 f95. |
|
Resolution |
Fixed in up45 (8 Apr 01) and up46 (9 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Submitted |
|
Submitted |
23 Jan 01 |
|
By |
Jean-Christophe Sublet (jean-christophe.sublet@CEA.FR) |
|
Title |
Problem with 181-point tabulated angulur distributions in MF=4 |
|
Details |
I am trying to process a new pu-239 evaluation for deterministic code (GROUPR road) that does contain some tabulated angular distribution of secondary particles in file 4, mt 2, 51-55 with 260 to 288 energy points, each having a probability distribution f(mu,energy) given for 181 mu values included between -1 and +1. This seems to give no problem to the ACER road processing with an array size increase in ACER: <data nwmaxn/20000/ >data nwmaxn/200000/. However it fails or hangs when GROUPR is used to produces group constants. NJOY97-114 (f90 -g) fails with a core dump and a SIGFPE message. NJOY99-32 (f77 -g), goes into an infinite loop. |
|
Comments |
|
|
Impact |
|
|
Resolution |
|
|
Number |
|
|---|---|
|
Status |
Pending |
|
Submitted |
5 Feb 01 |
|
By |
Brad Sleaford (sleaford1@llnl.gov) |
|
Title |
GROUPR error when running with 300 gamma groups |
|
Details |
When running a problem with 300 gamma groups, GROUPR stops with the message "too many groups". |
|
Comments |
When we increased the number of gamma groups allowed in up25, we forgot to reset the maximum value that is used by the error check. Just change the value 150 to 400 in groupr.2000. |
|
Impact |
Only affects GROUPR jobs with more than 150 photon groups. |
|
Resolution |
Fixed in up35 (8 Feb 01) |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
8 Feb 01 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Failure for Pb-208 from ENDF/B-VI Release 6 |
|
Details |
This failure occurred while making an MCNP4C library for 150-MeV neutron data. |
|
Comments |
In up17, we checked for isotropic distributions in order to use a more compact presentation. The logic misses this one special case. |
|
Impact |
Only this one material is currently affected. |
|
Resolution |
Fixed in up36 (8 Feb 01) |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
8 Feb 01 |
|
By |
Red Cullen (cullen1@llnl.gov) |
|
Title |
RECONR fails with 100 GeV photoatomic data |
|
Details |
When trying to run NJOY using the new proposed ENDF version of epdl97, the code goes into an infinite loop in RECONR. |
|
Comments |
NJOY currently uses 1e10 eV as its idea of an infinite energy. Thus, when the energy in RECONR hit 10 GeV, the code started to fail. We are increasing "infinity" to 1e12 eV in the places that are envolved in photoatomic processing, but leaving 1e10 elsewhere for now. With this change, a sample element from the 100 GeV library runs OK using the pattern of test/in03, and all the other standard test problems run as before. |
|
Impact |
None with current libraries. This patch will be needed when ENDF/B-VI Release 8 comes out. |
|
Resolution |
Fixed in up37 (reconr), up38 (njoy), and up39(gaminr), all 8 Feb 01. |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
22 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
Charged-particle heating zero for some parts of Be-9 ACE file. |
|
Details |
The charged-particle heating for protons, deuterons, etc., is zero even above the reaction threshold, but sometimes gets to be nonzero at higher energies. Seen in the ACE file for Be-9 from ENDF/B-VI. |
|
Comments |
Due to a bad if clause, the contribution to heating from charged particles is not being included for MF=6, law 3 or 4. This shows up for reactions like (n,p) or (n,alpha) in Be-9. This if clause was removed, which allows the indentation level to be reduced by one notch. |
|
Impact |
This only affects heating in coupled neutron-proton calculations in MCNPX (the code doesn't currently handle the other charged particles). In addition, the error is extremely small because of the very small cross sections for emitting charged particles below 20 MeV. |
|
Resolution |
Fixed in up40 (23 Mar 01). |
|
Number |
114 |
|---|---|
|
Status |
Resolved |
|
Submitted |
22 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
ACE interpolation-law printout |
|
Details |
Some sections on the ACE listing are missing the string "nr = 0" that designates them as using linear-linear interpolation for all points. |
|
Comments |
Easy to fix. |
|
Impact |
Only affects the ACE listing, not any of the numerical results. |
|
Resolution |
Fixed in up41 (27 Mar 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
22 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
Excess timers in PURR listing. |
|
Details |
There are timers given for every ladder in the PURR output, and this make it hard to compare the results of successive runs using diff. There are so many hits that the real differences are hard to see and the diff listing is too long. |
|
Comments |
This is easy to fix, and we had already made this change in NJOY 2001. Note that when running these kinds of diffs between listings made on different computers, there will always be a large number of differences because the random number sequences will be different. This also will change with NJOY 2001. |
|
Impact |
Only affects the listing, not any of the numerical results. |
|
Resolution |
Fixed in up42 (27 Mar 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
22 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
Heating zero at first energy point in ACE file. |
|
Details |
The heating value is zero at first energy point in ACE file, and then it has reasonable values for higher points. |
|
Comments |
This actually comes from HEATR where the round-up value used to make sure that the energy is inside a table when extracting values was chosen back in the old days when we had to cope with short-word machines. Now that we use double precision for energies and can routinely handle 7 or more digits, the value of "rup" needs to be made closer to unity. People who look closely at PENDF files might notice that the first energy in MF=3/MT=301 is given as 1.000010-5 instead of the correct 1.000000-5. This is why ACER thinks the heating (and damage) is zero at the first energy point. |
|
Impact |
Very minor for realistic problems. |
|
Resolution |
Fixed in up43 (27 Mar 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
22 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
Allow for larger ACE files |
|
Details |
LANL/X-5 would like to be able to make ACE files with lengths up to 1.5 Mwords for some special files at lower temperatures. |
|
Comment |
It is necessary to increase the dimension of the main container array for ACE data, the number giving that size, and the integer field on the XSDIR cards that holds the file length. We hope that this doesn't adversly impact people using smaller machines. |
|
Impact |
No impact for runs that do not need the extra size. |
|
Resolution |
Fixed in up44 (29 Mar 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
13 Mar 2001 |
|
By |
Roberto Orsi (orsi@bologna.enea.it) |
|
Title |
Length of some ACE thermal files too large by one word |
|
Details |
The length of the data block for ACE thermal files is too large by one word in both the header of the file and the XSDIR line. Orsi was able to identify the line in NJOY 94, 97, and 99 that caused this problem. |
|
Comment |
This problem only occurs for materials containing incoherent elastic scattering. For endf, this is poly, h(zrh), and solid methane. In addition, it will only occur if the true length of the ACE data block is divisible by 4. this explains how we were able to get by for so many years without seeing the problem! |
|
Impact |
Impacts some thermal ACE files for hydrogenous solids only. |
|
Resolution |
Fixed in up47 (9 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
30 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
ACE parameter LANDH wrong for Be-9 |
|
Details |
The LANDH parameter should be zero (not -1) for isotropic subsections of MF=6 described using ACE law=3 in Be-9. |
|
Comments |
This occurs for the reactions (n,p0) through (n,a0) in Be-9 from ENDF/B-VI. |
|
Impact |
This error only affects MCNPX. The symptom is a cosine>1.0 message. |
|
Resolution |
Fixed in up48 (9 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
30 Mar 01 |
|
By |
Bob Little (rcl@lanl.gov) |
|
Title |
Problem with ACE law3/33 for isomeric targets |
|
Details |
In response to some problems that Waclaw Gudowski reported with isomeric fission products, Bob Little determined that the original specifications for the ACE-format law3/33 needed to be changed to use -Q instead of abs(Q). |
|
Comments |
This change allows some upscatter to take place when levels are de-excited in isomeric targets. |
|
Impact |
This change only impacts those few evaluations that describe isomeric targets. |
|
Resolution |
Fixed in up49 (9 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
9 Apr 01 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Error message for JENDL photonuclear evaluations |
|
Details |
ACER crashes with an I/O error for most of the JENDL photonuclear evaluations now available from the http://iaeand.iaea.or.at/photonuclear web site. |
|
Comments |
These evaluations use a non-conforming format where MF=6, MT=201-207 are used to represent particle production. We are adding a clearer error message for the users' convenience. These evaluations cannot be used in NJOY or MCNPX in their current form. |
|
Impact |
Affects the processing of all the JENDL photonuclear evaluations except for H-2 (which has different problems with NJOY). |
|
Resolution |
Fixed in up50 (9 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
12 Apr 01 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Two-body recoils for photonuclear evaluations |
|
Details |
ACER doesn't handle photonuclear evaluations that contain sections of File 6 describing two-body recoils (law=4). |
|
Comments |
We have added coding to handle this option. It has been tested using a modified form of the JENDL evaluation for gamma+D and a preliminary LANL evaluation for gamma+D. Full distributions are provided for both secondary neutron and secondary proton. |
|
Impact |
Currently only affects evaluations for gamma+D. |
|
Resolution |
Fixed in up51 (12 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
13 Apr 01 |
|
By |
Bob MacFarlane (ryxm@lanl.gov) |
|
Title |
Tabulated fission spectra for photonuclear evaluations |
|
Details |
ACER doesn't handle photonuclear evaluations that contain sections of File 5 using law=1 (tabulated secondary energy distribution). |
|
Comments |
Up until now, the code handled analytic fission spectra for photonuclear evaluations, and we have now added coding to handle tabulated fission spectra, as needed for some of the Russian photonuclear evaluations for the actinides, such as the plutonium isotopes. Note that due to other formatting problems, revised versions of the Russian evaluations are needed for processing by NJOY. These revised versions have been submitted to the IAEA photonuclear compendium. |
|
Impact |
Affects photonuclear processing for many of the Russian evaluations from the BOFOD library. |
|
Resolution |
Fixed in up52 (13 Apr 01). |
|
Number |
|
|---|---|
|
Status |
Resolved |
|
Submitted |
11 Jun 01 |
|
By |
Dieter Leichtle (leichtle@irs.fzk.de) |
|
Title |
Isotropic discrete recoils |