Iteration mode

Old bugs, solved problems and other topics that are no longer relevant are moved here
Post Reply
emilf
Posts: 37
Joined: Mon Jun 21, 2010 3:22 pm
Security question 2: 0
Location: HZDR

Iteration mode

Post by emilf » Thu Aug 19, 2010 5:56 pm

Jaakko,

I tried to run a PWR assembly model in the B1 iteration mode. I got the following:

====================================================================
Starting the transport calculation cycle...

Sampling initial source...
OK.

Inactive cycle 1 / 25: k-eff = 1.11376 (DT thresh = 0.9000)
Inactive cycle 2 / 25: k-eff = 1.11677 (DT thresh = 0.9000)
Inactive cycle 3 / 25: k-eff = 1.11915 (DT thresh = 0.9000)
Inactive cycle 4 / 25: k-eff = 1.12349 (DT thresh = 0.9000)
Inactive cycle 5 / 25: k-eff = 1.12085 (DT thresh = 0.9000)
Inactive cycle 6 / 25: k-eff = 1.12095 (DT thresh = 0.9000)

***** Wed Aug 18 13:50:09 2010 (seed = 1282132193)

Fatal error in function NextCollision:

Total xs exceeds majorant (E = 1.225141E+00, f = 7.934123E-04)

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

I was wondering if you have any explanations for such a behavior? Note that without iterations everything runs smoothly.

User avatar
Jaakko Leppänen
Site Admin
Posts: 2167
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Iteration mode

Post by Jaakko Leppänen » Thu Aug 19, 2010 9:36 pm

The coding is probably fu***d up. I noticed the same problem while running some tests for the next update. I'll see if I can find the problem.

Have you tried running the calculation with an earlier version? Are you using probability table treatment for unresolved resonances? That may not work with the iteration modes.
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2167
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Iteration mode

Post by Jaakko Leppänen » Thu Aug 19, 2010 10:02 pm

I tried running my test case with different versions. Version 1.1.8 works, 1.1.9 doesn't.
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2167
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Iteration mode

Post by Jaakko Leppänen » Thu Aug 19, 2010 11:04 pm

The problem is in subroutine nextcollision.c, and it can be fixed by adding:

majorant = majorant + xs0;

after line 104 (where xs0 is added to totxs).

It's a stupid bug. I added a value check:

if (totxs > majorant)
{
if (totxs/majorant - 1.0 > 1E-6)
Die(FUNCTION_NAME, "Total xs exceeds majorant (E = %E, f = %E)", nn->E,
totxs/majorant - 1.0);
else
majorant = totxs;
}

in version 1.1.9 without realizing that in previous version:

if (totxs > majorant)
majorant = totxs;

actually adjusts the majorant, doing the same thing as the bug fix.
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2167
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Iteration mode

Post by Jaakko Leppänen » Thu Sep 09, 2010 10:03 pm

This problem was fixed in update 1.1.13.
- Jaakko

Post Reply