Struggle processing TENDL nuclear data for DT fusion activation

I am doing activation of steels under DT neutrons irradiations. I have been using the basic ENDF VIII.1 and chain_endfb81_sfr.xml from Cross sections - OpenMC.

After reading this

FNS chains (n,gamma) is only the fourth largest group behind (n,na), (n,3He) and (n,2n), so SFR and PWR leave out most of what an activation calculation needs

on openmc_data, I wanted to move on to cross sections and depletion chain that are more suited for fusion irradiation, including MT 9/10. I aimed at the most recent data (TENDL 2025 + JEFF3.3 cross sections and TENDL 2025 FNS).

The issue I have is that my computer is struggling to do the NJOY nuclear-data processing (> 30 hours for the conversion) and I was wondering if anyone could send me the nuclear data (cross section and depletion chain) already processed in the .xml format. I was also wondering why these were not directly available on Cross sections - OpenMC. I would be okay to share mine there if I manage to convert them.

Thank you all for your help,

I recommend that you produce a TENDL chain using Shimwell’s tools
https://github.com/openmc-data-storage/openmc_data
https://github.com/openmc-data-storage/openmc_data/tree/main/src/openmc_data/depletion

I have TENDL2021 chains here you can take a look at:
https://github.com/yrrepy/Chains_GENDF-Isomeric/tree/master/TENDL2021

with and without my special processing for isomeric branching.
By MT 9/10, I presume you mean MF 9/10.

Wow this is super helpful thanks !

I was trying to use Shimwell’s tools but it takes 30 hours (using 10 CPU).

I’ll look into your data instead. And yes I meant MF 9/10, my bad!

@yrrepy Thank you again for your depletion chains. I still have to generate the cross sections myself right ?
I am also wondering why it is not made available directly on OpenMC’s website the same way it is done for SFR and PWR ? Is it an ongoing project ?

No, you don’t have to generate the cross sections.

OpenMC uses the transport cross-sections for activation. Either as direct MC Estimator or collapsed with mg-flux.

You point it to the standard h5 XS and to a chain.
The chain knows how to link reactions, nuclides, etc.

SFR and PWR chains are chains that have fixed isomeric branching yields at a given single, fixed neutron flux spectrum.
Shimwell has FNS chains with a fusion spectrum.

I’d suggest that you get the standard workflow working first.

Okay but If I use TENDL chain I need TENDL cross section (h5 + cross_sections.xml) to match it right? I cannot use TENDL chain with ENDF VIII.1 cross sections (the one I currently use)?

And what do you mean by standard workflow here ?

Just get the deplete workflow working with the OpenMC vanilla data and chains, eg: ENDFB8.1
Once it’s working then iterate the nuclear data.

Cross-sections:
https://anl.box.com/shared/static/6qr7jezzihkj9p9esl5jn19qgpujyjyz.xz
Chain:
https://anl.box.com/shared/static/n0pkqe66uotskoljr93szvjyvtvycgze.xml

If dead set on TENDL:
Cross-sections:
https://tendl.imperial.ac.uk/tendl_2019/tar_files/tendl19_hdf5.tar.bz2
Chain:
https://github.com/openmc-data-storage/openmc_data/blob/main/src/openmc_data/depletion/chain_tendl_2019_endf80_fns.xml

You can mix chains with different cross-sections. The code will auto filter out nuclides that are on the chain but not on the h5 cross-sections.

Oh I see. I have been using basic OpenMC data for a long time and it works perfectly well.

However the point was to improve the accuracy of my simulation by using data (depletion +XS) more suited for steel activation under fusion irradiation (including MF9/10).

By the way, once I manage to get TENDL running do you think I should trim it to keep only the nuclides I care about ? Especially if I intend to do some expensive R2S/parametric sweep ?

Thank you again for your guidance

I checked the data I have, the effect of isomeric branching is something like 1% on the activity of steel (SS304, SS316).
You’re going to get a greater effect by switching between JEFF40, ENDFB81, TENDL
https://nds.iaea.org/media/docs/fusion/UKAEA-CCFE-RE-20-004.pdf

Filtering nuclides: just make sure to chain reduce your chain, level=3
https://docs.openmc.org/en/stable/_modules/openmc/deplete/chain.html#Chain.reduce

Hi all,

I just wanted to mention there are a few more chains that were recently added and can just be downloaded. As noted earlier it takes hours to reproduce so it is nice to have the ability to reproduce but also to be able to download them directly is a nice shortcut.

These FNS TENDL ones might be the best fusion option we have today. That is until energy dependent branching ratios arrive of course. In the mean times these branching ratios are done for the DT fusion spectrum in the FNS machine.

Filtering nuclides: just make sure to chain reduce your chain, level=3

About nuclide filtering, I always wondered what number I should choose. For example, if I have an important nuclide that gets created after 4 fast decays reactions, I would miss it right ? So when selecting 3 we hope that most of the long lived nuclides are generated after less than 3 reactions. Conversely, when selecting 3 I would still process thousands of unnecessary reactions for long term activation. Is there a workflow around it ? Should I manually select each reaction I am interested in ?

I’d counsel starting high and slowly reducing the level, and watch for discrepancies (Bq, dose, etc).
Pref. with a not overly complex but still representative model.

This is however making me reflect that it might be useful to have a level depth for reaction and a separate level depth for decay.