search

njoy / NJOY2016

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

NMIX card problem in ACER for thermal data #260

Closed luohao0925 closed 1 year ago

luohao0925 commented 1 year ago

I have encountered a problem with the nmix card in ACER module with for the thermal data. The definition of nmix in NJOY2016 manual is number of atom types in mixed moderator and two examples are provided (2 for beo and c6h6). However, when i processed the thermal data for beo and set the nmix to 2, the inelastic scatter xs in my ace file is half of the corresponding data in ace file released by LANL(https://nucleardata.lanl.gov/lib/ENDF80SaB2.tgz).

According to the defination of nmix and also the two examples, it seems that nmix should also be 2 for HinH2O and 3 for 'C5O2H8'. I tried and got the same results. The value of reaction xs was 1/2 and 1/3 of the ace file data provided by LANL. i wonder if misunderstood or there is more precise description of the nmix card in ACER module.

I'm attaching the constrast figure for BeO and HinH2O. Feel free to contact me if additional information is required.

Be in BeO thermal scatter law xs

H in H2O thermal scatter law xs

whaeck commented 1 year ago

This is indeed a confusing case, and I admit that I have had the same question as you 15 years ago when I started processing thermal scattering data.

The short answer is as follows: nmix represents the number of atom types in a mixed moderator, i.e. the scattering from the entire molecule is represented in one evaluation. Examples would be BeO and benzene as you mention. Evaluations like H in H2O, H in Lucite or Be in BeO are not mixed as they relate to a single principal scatterer type (in the above examples, respectively H, H and Be). As such, for these evaluations nmix = 1.

With modern evaluations, you should rarely have to change the value of nmix to anything else but 1. Mixed moderators were used in the past but even then they were relatively rare.

Now, nmix should not be confused with natom in THERMR. The value to be used for natom is actually given in the evaluation - as you are probably aware.

luohao0925 commented 1 year ago

Ok, it seems to be a description problem just left in the user manual. Based on my tests results, the nmix should be 2 only for benzene in the modern thermal scattering sublibrary. And as you mention, some keywords like natom, emax, and even the thermal scattering temperatures temp are actually given in the evaluation. However, mti/mte ( #24) and nmix mentioned in this issue (#260) can not be extracted directly from evaluation. For a fresh evaluator, better guidance in the manual seems to be warranted. Thanks for your clarification @whaeck!

whaeck commented 1 year ago

You're welcome ;-)

mti and mte are essentially user choices for internal use in NJOY (so that ACER and other modules can find the data). I always pick 221 and 222 for these. There are some restrictions on these number but if you stick to these numbers you should be fine. Some older modules might indeed need specific values (there is a table with MT numbers for different thermal scattering materials in the THERMR manual) but I do know that is not the case for ACER. I suspect (don't quote me on this) that these differing MT numbers are artifacts from older ENDF versions.

There is indeed a lot of work to be done, both in NJOY2016 and its manual as well as the modernisation of NJOY as a whole. We'll get there eventually ;-)

whaeck commented 1 year ago

As a side note: with the mixed mode elastic being added the the ENDF format for ENDF/B-VIII.1, NJOY will now use 3 MT numbers internally - but as a user you only need to specify mti and mte (if mixed mode is used for elastic, coherent elastic will be stored in MT=mte and incoherent elastic in MT=mte+1).