search

njoy / NJOY2016

Nuclear data processing with legacy NJOY
https://www.njoy21.io/NJOY2016
Other
95 stars 82 forks source link

ptleg2 error on ACER #247

Closed chubsme closed 2 years ago

chubsme commented 2 years ago

Hello.

I'm new to NJOY and while processing multiple materials from ENDF/B-VII.1 for multiple temperatures I encountered this error:

---message from ptleg2---negative probs found 60 for mt= 2 e= 2.000E+07

---message from ptleg2---negative probs found 270 for mt= 2 e= 2.200E+07

It is the same error from #164, but I encountered it for multiple materials such as B-10, U-238, Pu-238, etc.. From #164 it was related to the evaluation data and got resolved with errata. However I cannot check the link since it's giving me: image

Here is the sample input file for U-235 (given as example from the NJOY training held last 2021): moder 20 -21 reconr -21 -22 / 'A PENDF for U-238 from ENDF/B-VII.1' / 9237 / 0.001 / 0 / broadr -21 -22 -30 / 9237 5 0 1 / 0.001 / 293.6 600 900 1200 2500 / 0 / heatr -21 -30 -40 45 / 9237 7 / 302 303 304 318 402 443 444 / purr -21 -40 -50 / 9237 5 1 20 64 1 0 / 293.6 600 900 1200 2500 / 1E10 / 0 / gaspr -21 -50 -60 acer -21 -60 0 70 75 / Card1 1 1 1 .10 0 / Card2 'U-238 @ 293.6K from ENDF/B-VII.1' / Card3 9237 293.6 / Card5 / Card6---all defaults / Card7---all defaults acer 0 70 81 91 92 / Card1 7 1 / Card2 'U-238 @ 293.6K from ENDF/B-VII.1' / Card3 viewr 81 86 . . . stop

whaeck commented 2 years ago

The message you are getting from ACER is actually not an error: ACER is just telling you that the evaluated data leads to negative probability values when converting a Legendre polynomial expansion to a probability function. It can potentially indicate that the evaluation should use more coefficients in the expansion (a Legendre expansion has to be cut at a specific order and that cut can lead to negative probabilities). While it is something you need to know, it is not a showstopper - hence this being a message and not an error. In many cases you can also use the manual to look up these messages and errors and see what they mean.

Issue #164 was specifically for ENDF/B-VIII.0 B10 causing an error while writing out the xss array to an ACE file (in that case, NJOY2016 signals an unknown LAW in the xss array which can be traced back to the evaluation, and which has been fixed in an errata file). This is unrelated to the ptleg2 messages you are getting. However, I find it strange that you cannot access the errata page so I checked the link you provided. It should be the following: https://www.nndc.bnl.gov/endf-b8.0/errata.html (there should be a '-' instead of a '/' between 'endf' and 'b8.0'). I'll check issue #164 to see if I need to edit the link there.

whaeck commented 2 years ago

So the link in issue 164 was bad, I edited it.

chubsme commented 2 years ago

Thank you @whaeck! Thought it was the same since the problem was later mentioned in #164 and can't find the message from the manual.