Page 1 of 1

Reading ACE data

Posted: Mon Sep 26, 2011 5:47 pm
by mathieu.hursin
Dear all,

I m trying to have SERPENT read an existing cross sections library (based on JEFF3.1 data w/ probability tables) in ACE format.
However, when I try to run SERPENT, I obtain the following error message:
=======================================
Reading data from ACE files:
Isotope 92235.31c (U-235)...
Isotope lwtr01.31t (H-1)...
Isotope 1001.31c (H-1)...
OK.

Reading energy arrays:
Isotope 92235.31c (U-235)...
Isotope lwtr01.31t (H-1)...
Isotope 1001.31c (H-1)...
OK.

- Main grid thinned from 77019 to 77019 points using tolerance 0.00E+00.

- 5839 important points added resulting in a total of 77019 points.

- Final grid size 76955 points (1.00E-11 < E < 20.0).

- Total 83 points in nubar grid.

- 4 energy groups in few-group structure.

Processing XS data:
Isotope 92235.31c (U-235)...

***** Mon Sep 26 16:30:37 2011 (seed = 1317047435)

Fatal error in function EnergyDistribution:

Unable to process E interpolation (mt 18)

Simulation aborted.
=====================================================

The xs data works fine on the same problem with MCNPX....

Did anybody have the same problem, if so, how did you solve it?

Thanks a lot,

Mathieu

Re: Reading ACE data

Posted: Mon Sep 26, 2011 8:41 pm
by Jaakko Leppänen
Mathieu,

This is a Serpent-related problem, not a problem in the data. The code does not support all features in the ACE data format, and some rarely encountered features may lead to errors like this (rather than using the data without the appropriate methodology).

I don't remember why I added this particular check, but you can try commenting out the if-statement on line 103 of energydistribution.c (version 1.1.16) and see what happens. If the calculation runs, there is a chance that the fission energy distribution of U-235 is handled incorrectly, but you may also get the correct result.

What data are you using? Some MCNP library distributed with MCNP5 or MCNPX?

Re: Reading ACE data

Posted: Tue Sep 27, 2011 3:09 pm
by mathieu.hursin
Jaakko,

when I comment out the line you mentioned, SERPENT runs...the results look ok but not great, the agreement with MCNPX is not so good (most likely some kind of input discrepancy...). I m investigating at the moment.

As for the data, it is the JEFF-3.1 library distributed by NEA: Ref. ZZ-MCJEFF-3.1NEA, http://www.nea.fr/abs/html/nea-1768.html

Thanks for your prompt reply.

Mathieu

Re: Reading ACE data

Posted: Fri Nov 13, 2020 1:05 pm
by augusto_vib
Hello,

I found this old post where there is an issue reading ACE data, and the error is the same that I'm just encountering. Currently, (for version 2.1.31) I am having having the following error:

*********************************************************
Fatal error in function ProcessEDistributions:

Unable to process E interpolation (94237.09c mt 18)
*********************************************************

The data is Pu-237 from TENDL-2019, and the ACE file has been processed with NJOY2016. On the other hand, MCNP6.2 is able to read and process this ACE file. Any idea why is this happening?

Thanks

Re: Reading ACE data

Posted: Fri Nov 13, 2020 7:27 pm
by Jaakko Leppänen
It's impossible to tell what the problem is based on the error message alone. I can take a closer look if the ACE file small enough to share?

Re: Reading ACE data

Posted: Fri Nov 13, 2020 8:28 pm
by augusto_vib
It seems the file is too large to be shared here (2.8 MB). But I hope is ok with you that I sent to you by email. Thanks
Jaakko Leppänen wrote:
Fri Nov 13, 2020 7:27 pm
It's impossible to tell what the problem is based on the error message alone. I can take a closer look if the ACE file small enough to share?

Re: Reading ACE data

Posted: Mon Nov 16, 2020 2:00 pm
by Jaakko Leppänen
It seems that there is something wrong with the nubar data. The six precursor groups are defined, but the decay constants are off (the first is 1.0 and the rest are zeros). The error comes from the energy distribution of delayed neutrons in the second group.

Does MCNP print anything about the delayed neutron data? At this point it is difficult to say whether the problem is in the data or the serpent routine that reads it.