Documentation on the ENDF60 Library

Released: 1994-10-01

The ENDF60 library was originally released as DLC-181 in October, 1994. To make it easier to obtain all of the X-Division currently supported data libraries for MCNP®, we have combined all of the available data libraries into one package, DLC-220, for release through RSICC as of Feb. 2003.

The ENDF60 library was processed with NJOY91 at room temperature (300 K), using flat weighting, and thinned such that most nuclides had no more than 400,000 words (or used approximately 3.2 megabytes). The ENDF60 library is made up of 124 nuclides and requires approximately 200 megabytes of disk space. The largest data table is for I-127 which requires 8.1 megabytes of disk space. Each of the data tables has the unique ZAID ending of .60c. Charged-particle production cross-sections in the MT=200 series are not provided. Two modifications have been made to the library since original release:

  • removal of data for 238Np
  • removal of the ground state of 242Am

Data Library Characteristics

There are a few common questions that have been asked. First, the upper energy limit for the data files of the ENDF60 library are 20 MeV with the following exceptions:

am 9524130.0
au 7919730.0
c 600032.0
ho 6716530.0
h 1001100.0
i 5312730.0
np 9323817.5
u 9223417.5

The table below describes the individual data tables in the ENDF60 library. The ZAID, evaluation group, evaluation type, release number, and photon production are given for each nuclide. The evaluation group is identified by the laboratory participants who performed the evaluation, and the evaluation type indicates whether the ENDF/B-VI evaluation is a new evaluation (N) or is a translation of the ENDF/B-V evaluation (T). New evaluations comprise 48% of the total in the ENDF60 library. If the data file is a new evaluation, the corresponding release number for the ENDF/B-VI release is specified, where release 6.0 may correspond to a new or translated evaluation. If the evaluation has been modified at Los Alamos, to add in photon production for instance, the revision number has been specified as LANL. Photon production availability is indicated in the last column. Determining whether an evaluation has enough improvements to be considered "New" is subjective. The file 1 information from each evaluation has been made available so that the user can determine the answer for their specific application.

The specification of a new or translated library is given with respect to the ENDF/B-V Release 0 evaluations on which most of the .50c data files are based. **

ZAIDSourceEvaluation TypeRelease #Photon
1001.60cLANLN6.1yes
1002.60cLANL,AWREN6yes
1003.60cLANLT6no
2003.60cLANLN6.1no
2004.60cLANLT6no
3006.60cLANLN6.1yes
3007.60cLANLN6yes
4009.60cLLNLN6yes
5010.60cLANLN6.1yes
5011.60cLANLN6yes
6000.60cORNLN6.1yes
7014.60cLANLNLANLyes
7015.60cLANLN6yes
8016.60cLANLN6yes
8017.60cBNLT6no
11023.60cORNLT6.1yes
12000.60cORNLT6yes
13027.60cLANLT6yes
14000.60cORNLT6yes
15031.60cLLNLT6yes
16000.60cBNLT6yes
16032.60cLLNLT6yes
17000.60cGGAT6yes
19000.60cGGAT6yes
20000.60cORNLN*6yes
21045.60cBNLN*6.2yes
22000.60cBRC,ANLT6yes
23000.60cANL,LLNL,+N6yes
24050.60cORNLN6.1yes
24052.60cORNLN6.1yes
24053.60cORNLN6.1yes
24054.60cORNLN6.1yes
25055.60cORNLN6yes
26054.60cORNLN6.1yes
26056.60cORNLN6.1yes
26057.60cORNLN6.1yes
26058.60cORNLN6.1yes
27059.60cANLN6.2yes
28058.60cORNLN6.1yes
28060.60cORNLN6.1yes
28061.60cORNLN6.1yes
28062.60cORNLN6.1yes
28064.60cORNLN6.1yes
29063.60cORNLN6.2yes
29065.60cORNLN6.2yes
31000.60cLLNL,LANLT6yes
39089.60cANL,LLNLN*6yes
40000.60cSAI,BNLT6.1no
41093.60cANL,LLLN6.1yes
42000.60cLLNL,HEDLT6yes
43099.60cHEDL,BAWT6no
47107.60cBNL,HEDLN*6no
47109.60cBNL,HEDLN*6no
49000.60cANLN6yes
53127.60cHEDL,RCNN*LANLyes
53129.60cHEDL,RCNT6no
55133.60cHEDL,BNL,+T6no
55134.60cORNL,HEDLN6no
55135.60cHEDLT6no
55136.60cHEDLT6no
55137.60cHEDLT6no
56138.60cORNL,HEDLT6yes
63151.60cLANLN6yes
63153.60cLANLN6yes
64152.60cBNLT6no
64154.60cBNLT6no
64155.60cBNLT6no
64156.60cBNLT6no
64157.60cBNLT6no
64158.60cBNLT6no
64160.60cBNLT6no
67165.60cLANLN6yes
72000.60cSAIT6no
73181.60cLLNLT6yes
73182.60cAIT6no
74182.60cLANL,ANL,+N*6yes
74183.60cLANL,ANL,+N*6yes
74184.60cLANL,ANL,+N*6yes
74186.60cLANL,ANL,+N*6yes
75185.60cORNL,LANLN6no
75187.60cORNL,LANLN6no
79197.60cLANLN6.1yes
82206.60cORNLN6yes
82207.60cORNLN6.1yes
82208.60cORNLN6yes
83209.60cANLN6yes
90230.60cHEDLT6no
90232.60cBNL,ANL,+T6yes
91231.60cHEDLT6no
92232.60cHEDLT6no
92233.60cLANL,ORNLT6yes
92234.60cBNL,GGAT6no
92235.60cORNL,LANLN6.2*yes
92236.60cHEDLN6no
92238.60cORNL,LANL,+N6.2yes
93237.60cLANLN6.1yes
93239.60cORNLN6no
94236.60cHEDL,SRLN6no
94237.60cHEDLN6no
94238.60cHEDL,AI,+T6no
94239.60cLANLN6.2yes
94240.60cORNLN6.2*yes
94241.60cORNLN6.1yes
94242.60cHEDL,SRL,+T6yes
94243.60cBNL,SRL,+T6.2*yes
94244.60cHEDL,SRLT6no
95241.60cCNDCNLANLyes
95243.60cORNL,HEDL,+N6yes
96241.60cHEDLT6yes
96242.60cHEDL,SRL,+T6yes
96243.60cHEDL,SRL,+T6yes
96244.60cHEDL,SRL,+T6yes
96245.60cSRL,LLNLT6.2yes
96246.60cBNL,SRL,+T6.2yes
96247.60cBNL,SRL,+T6.2yes
96248.60cHEDL,SRL,+T6yes
97249.60cCNDCNLANLno
98250.60cBNL,SRL,+T6.2yes
98251.60cBNL,SRL,+T6.2yes
98252.60cBNL,SRL,+T6.2*yes

* The changes in Release # indicate those LANL modifications that were accepted as part of ENDF/B-VI Release 2. An errata sheet for Table 1 of LA-12887 and LA-12891 is available.

** The data files np93238 and am95242 have been removed from distribution. While the data files accurately reflected the evaluations, we believe that the evaluations are too incomplete to be used for most transport calculations.

New Scattering Laws in ENDF/B-VI

Three new scattering laws of the ENDF/B-VI evaluations have been implemented in NJOY and MCNP®. This required new data formats and processing, transport physics, and next-event estimator sampling schemes. The three laws are the Kalbach-87 formalism (file 6, law=1, lang=2 : MCNP® law 44), correlated angle-energy scattering (file 6, law=7 : MCNP® law 67), and correlated angle-energy scattering phase-space law (file 6 law=6 : MCNP® law 66). In addition to the algorithms to sample these laws for neutron transport, new algorithms were developed for next-event estimators so that collisions using the new laws now contribute to point and ring detectors, as well as to the DXTRAN variance reduction method.

The Kalbach-87 formalism requires sampling the cosine between the incident and outgoing particle directions from a density function. This formalism was also used to approximate the new ORNL evaluations (F, Cr, Mn, Fe, Ni, Cu and Pb) which used file 6, law=1, lang=1. An approximation was required since MCNP® cannot as yet sample from a Legendre polynomial expansion. The major error in this approximation for the ORNL evaluations is that the tabulated outgoing laboratory energy is used in the formulation as if it were in the center-of-mass. This error is not substantial for the heavier elements, but was thought to be significant for the F evaluation. Therefore, F was approximated using file 6, law=7 instead.

The correlated angle-energy scattering law first samples the angle mu from 32-equiprobable cosine bins between the incident and outgoing particle directions with interpolation for a given incident energy E. With mu and E fixed, the outgoing energy, E', is sampled from tables interpolated on the mu and E tables. This law is used for the Be-9 and F evaluations only.

The correlated angle-energy scattering phase-space law first samples the outgoing particle direction cosine isotropically in the center-of-mass system. Then, given the number of bodies in phase space and the total mass ratio for all the bodies, the outgoing energy is sampled from a density function. This law is used for the H-2 evaluation only.

Data Library Testing and Errors

In the process of data testing, several problems were encountered and corrected. The O-16 data file was modified such that exponents less than -37 were increased to -30 to prevent MCNP® from crashing on HP workstations. The natural Cd file was not released due to a processing error by NJOY when photon production was added. As the Cd evaluation is simply a translation from ENDF/B-V, this was not considered critical to the library's release. Cf-249 was also modified to correct an evaluation problem concerning center-of-mass to lab energy conversion. Prior to this correction, neutrons could up-scatter above 20 MeV and cause MCNP® to crash.

All of the new ORNL evaluations (F, Cr, Mn, Fe, Ni, Cu, and Pb) use the Legendre expansion description for energy-angle distributions. Because MCNP® does not yet handle this scattering law, these were approximated with the Kalbach-87 formalism. The interpolation scheme for INT=12 in the Kalbach-87 formalism has been replaced with INT=2 for neutrons in the data library. For photons, INT=12 remains specified in the data library but is treated as INT=2 by MCNP®.

There are two files which have problems with TYR values (see Appendix F of the MCNP4A manual for discussion of these values). The first file is c6000, natural carbon. The (n,2n)p reaction, MT=41, should have a TYR value of 2 instead of the 1 contained in the data file. To correct, go to line number 1256 of the data file 'c6000' and replace the 1 in column 40 with a 2. The threshold for this reaction is En=29.7 MeV and hence not many applications will be affected by this error.

  • go to line number 13323 and change the '-' to ' ' in column 79

  • go to line number 13324 and change the '-' to ' ' in column 19 and 39

  • go to line number 13328 and change the '-' to ' ' in column 19

  • The space, ' ', is important as the TYR values must be in the proper column.

Further information on issues relating to the QA of ENDF60 can be found in LA-UR-98-533.

Benchmarking

There were five sets of MCNP® simulations originally performed for the ENDF60 library: (1) infinite medium simulations, (2) LLNL pulsed sphere benchmarks, (3) criticality simulations, (4) iron benchmarks, and (5) photon production assessment. For all tests, both the ENDF/B-VI and B-V libraries were used and compared with one another. Additionally, the recommended data library was used when it differed from the ENDF/B-V library. The results of each of these is briefly described below.

Infinite Medium Simulations

The infinite medium simulations used a 20-MeV neutron source in an infinite medium of the nuclide. This source was modified so that the reactions for the heaviest elements spanned the entire neutron energy range from 0-20 MeV. The neutron flux and heating, as well as the photon flux and heating, were tallied and coplots of the results were made for the different libraries. These simulations showed that the flux tallies were basically equivalent for the two libraries, with the obvious exception of when ENDF/B-VI has photon production while ENDF/B-V did not. NJOY has made improvements in calculating the neutron heating since the previous libraries were processed. NJOY now takes into account the recoil of the resulting nucleus as well as the outgoing particle's energy. Corrections to the Q-values of the B-V translated evaluations also improved the neutron heating results. This work was documented in LA-12887.

LLNL Pulsed Sphere Benchmarks

Simulations were performed for 28 neutron pulsed-sphere experiments orginally designed to benchmark several neutron transport codes at LLNL in the 1960's. The experiments used a pulsed-deuterium beam on a tritiated target centered within a spherical shell of material. The pulsed beam produced 14-MeV neutrons and the neutron flux spectrum was measred as a function fo time-of-flight. The spheres of material were 0.5 to 4.8 mean free paths thick. The simulatedresults were compared withexperimental values. Both the B-VI and B-V libraries gave similar results for O, Mg, Al, Ti, H2O, D2O, and concrete. The B-V library more closely matched the experimental value for Li-6 and C, while the B-VI library more closely matched the experimental value for Li-7, Be, N, F, Pb, CF2, and CH2. The special T-2 evaluation for Fe did marginally better than the B-VI library, both of which were better than the previous B-V library. This work is documented in LA-12885. **These benchmarks have been revised and expanded since this publication.

Criticality Simulations

The MCNP4 validation benchmarks, LA-12212 and LA-12415, which used the ENDF/B-V and recommended data libraries, were repeated using MCNP4A with all data libraries. The first set of benchmarks consisted of 9 critical assemblies used to benchmark the COG code, while the second set of benchmarks consisted of 25 simulation problems used to benchmark the KENO code. The results indicate very little difference between the libraries for these simulations. **These benchmarks have been revised and expanded, LA-13627.

Iron Benchmarks

Four sets of Iron benchmarks were performed; (1) the LLNL pulsed spheres, (2) ORNL Fusion Reactor Shielding benchmark, (3) a 76-cm diameter iron sphere benchmark from University of Illinois, and (4) the ORNL benchmark for neutron transport through iron using a fission reactor source. The special T-2 library did marginally better than the ENDF/B-VI data library for the experiments. In all cases, the ENDF/B-VI library did better than the B-V data library, both of which did better than the B-IV data library. This work is documented in LA-12884.

Photon Production Assessment

These simulations compared photon production for the data libraries using a thermal and 14-MeV isotropic neutron source in a thin shell of material. This comparison was made for natural elements, or isotopes for non-naturally occurring nuclides. The neutron and photon fluxes were tallied in 20-keV and 5-keV wide energy bins from 0 to 14 MeV. In general, photon production is represented by discrete-energy gamma-rays or by wide, histogram energy bins. For translated evaluations, photon production was equivalent. For the new evaluations, photon production generally changed substantially. There is often a combination of both discrete-energy and wide energy-bin photon production, particularly at higher energies. Some decrease in energy-bin width was also made for a number of nuclides for the ENDF60 library. All changes in photon production were verified by hand checking against the evaluation. This effort was documented in LA-13092.

The best data library to use will be dependent upon the specific application.

Downloads

ENDF60 can be downloaded either as a zip file or as a compressed tarball

  • endf60.tgz (sha512 1e065c4a9855a1af90021f4d6f0ac1f9f538a27fd612169b448d3930ea4c5af2cf4484c6706c5390164bfedabdd3aa3b9a47541fd1ca3560c715425ec20e2cc8)
  • endf60.zip (sha512 7d7034218165fbf33c2d413b2bf6300269f6dbc1feb71376a99c4863ca0f23ebf3a257dc2e344f8cdf08e5dbd44c7b18962d9b4d2e1da3351777b7d6c235d743)

If further information is required contact a member of the Data Team by e-mail at nucldata@lanl.gov