Page 1 of 1

Looking for Serpent cross section directories

Posted: Mon Dec 14, 2020 7:40 am
by Joshua Guertler
I recently tried to run a molten salt input I found on Serpent 2.0 beta and received the following error:

Code: Select all

Unable to locate cross section directory file "endfb71_edep.xsdata"

Environment variable SERPENT_DATA is set to "/home/software/serpent/endfb70"
I'd imagine that this is due to how I don't believe that I have a directory by the name of endfb71_edep.xsdata in any of my Serpent folders. Upon further examination, I noticed that the original txt file I was using as my input had the following files used as the libraries:

Code: Select all

% --- Libraries:

set acelib "endfb71_edep.xsdata"
set declib "sss_endfb7.dec"
set nfylib "sss_endfb7.nfy"
Likewise, I was curious if these directories are something that exist as openly available files that I'd be able to get from RSICC or if these are directories that only exist in the computer of the original creator. Thank you.

Re: Looking for Serpent cross section directories

Posted: Mon Dec 14, 2020 3:29 pm
by Ana Jambrina
Serpent 1.1.17 comes with a set of processed XS-data (see: http://montecarlo.vtt.fi/links.htm), however, other nuclear data - processed or raw (to be processed through software such NJOY) - is available through other projects (e.g. MCNP - SLA through RSICC) or databases (JEFF , ENDF, etc.).
The name given to the nuclear libraries is free, up to the user. Aside the name given to the XS-library files,
- Environment variables (see: setting_up_data_libraries) or absolute paths to the cross-section library files are needed (check XS-card syntax: e.g. ‘acelib’ - set acelib).
- ‘xsdata’ should point the absolute path where the ACE files are, for each of the nuclides included (to be set up when converting ‘xsdir’ -> ‘xsdata’ through ‘xsconvert.pl’).

Note: ‘endfb71_edep’ corresponds with the cross-section library distributed by Serpent developer team (upon request) based on ENDF/B-VII.1 including the necessary data to use the new energy deposition modes (ACE files include: KERMA coefficients, MT458 data, additional unresolved resonance range probability tables). For further info, see: viewtopic.php?f=15&t=3187.

Re: Looking for Serpent cross section directories

Posted: Mon Mar 08, 2021 5:35 pm
by mathieu.hursin
Dear Ana,

I would like to have access to the ENDF/B-VII.1 library (endfb71_edep). How should I proceed?

Let me know,

Thank you,

Mathieu

Re: Looking for Serpent cross section directories

Posted: Tue Mar 09, 2021 11:06 am
by Ana Jambrina
Download it from the link (still active) provided in the post: viewtopic.php?f=15&t=3187&start=20#p11639.

Re: Looking for Serpent cross section directories

Posted: Tue Jun 01, 2021 7:05 pm
by mathieu.hursin
Hello,

could you give a bit of information on how you have produced the thermal scattering libraries (njoy version, possibly an input). NJOY evolves quickly these days and it seems that some significant differences can be obtained.

FYI, generating my own Sab lib based on the ENDF/B-VII.1 endf files, I obtained ~ 100pcm difference for a pincell model compared to your library, while using NJOY-2012 patch 82....

Thank you,

Mathieu

Re: Looking for Serpent cross section directories

Posted: Wed Jun 02, 2021 10:01 am
by Riku Tuominen
If you are referring to the xs-data with the additional energy deposition data, I can provide the following information. Since the thermal scattering data did not need any modification for energy deposition calculations, I took the thermal scattering files from nowadays quite old beta cross section library for Serpent 2. These files were created by a colleague who is no longer working with us so I do not have specific information on the processing. To the best of my knowledge NJOY 99.364 was used. Nowadays, I would recommend using NJOY2016 or NJOY21 since they are the currently maintained versions.

Re: Looking for Serpent cross section directories

Posted: Wed Jun 02, 2021 1:54 pm
by mathieu.hursin
Dear Riku,
I m talking about the production of the thermal scattering part of the file lwtr_EDF71.ace in the package (endfb71_edep.tar.gz). I m not after the additional energy deposition data.
I went back in the Serpent-v1.1.7 documentation and found that indeed NJOY99.364 was used then....I just wanted confirmation that the same NJOY version / approach was used to produce the tsl (lwtr_EDF71.ace) of the ENDF/B-VII.1 library (is that what you call "quite old beta cross section library for Serpent 2").

Using NJOY2016 and NJOY2021 instead of NJOY99 might lead to fairly different results using the same initial ENDF information....

Thank you,

Mathieu

Re: Looking for Serpent cross section directories

Posted: Wed Jun 02, 2021 2:09 pm
by Riku Tuominen
The thermal scattering data has not been taken from Serpent-v1.1.7 cross section data but instead from Serpent 2 "beta" cross section data that we used internally a while back. As I said in the previous post, I do not have specific information about the NJOY options used. You can of course create your own thermal scattering ace files with NJOY and use them.