MCNP6.2 - ENDF/B reaction numbers for tally multiplier

  • #1
TL;DR Summary
How may I use ENDF/B reaction numbers?
Hi everyone,

I am trying to evaluate the spectral index of an nonelastic (n,n') reaction. For that I want to set up a tally multiplier on a cell (let's call it cell 10). The reaction is present in the ENDF/B library as MT=4 but I have not seen it in the table of the special reaction numbers which can be used in MCNP (table 5.18 of the manual LA-UR-22-30006, Code Version 6.3.0, . However, the manual says that either ENDF or special number reaction numbers could be used, but there is no example of how to set a reaction list using the ENDF library. do I set up the tally with this reaction? Using MT=4?

See my tentative example below:

F4:N 10
E4 1 20
FM4 (-1 100 MT=4) T
SD4 1

I set the energy with this energy binning because I'm interested in E>1 MeV and 100 is the virtual material I'm interested, so not used in defining any cell.

Thanks in advance

Answers and Replies

  • #2
Ah! Solved...I think. Example 6 (Lifetime Calculation/Reaction Rates) of the manual uses positive reaction numbers, which should be the ones corresponding to the reaction number of the ENDF/B
  • #3
Yes, and you can plot the reaction cross section in MCNP if you want to make sure it is what you think it is. Some of the more obscure ENDF reaction descriptions change between versions.
  • Like
Likes 19matthew89
  • #4
Ow! Thanks!

Indeed I wanted to check the cross section.
I know that in theory the MCNP plotter would allow you to plot both cross section and reaction rates but I have never managed to have it working.

I can run the MCPLOT (i.e. the command "MCNP6 Z" works) but if I try to launch

mcnp6 ixz i=inputfile

that does not work. I think it's not clear to me which input file I have to give as it says that "Cross section plots cannot be made from a runtape or MCTAL file" what other file shall be given?
  • #5
Try this, mcnp ixz inp=input
input needs to be a normal inp file which must describe materials with the isotope you want to plot the cross section of. i means process input file, x means process cross sections, z means mcplot them.

If mcnp runs the file without error when run normally you should instead get a mcplot prompt (and no run). I'm using a random HEU containing input file using the 69c library. Entering,
xs=92235.69c mt=4
Produces a plot on my system.

I do not know how to plot results.
  • Like
Likes 19matthew89
  • #6
Hi, thanks.

But unfortunately
mcnp ixz inp=input
does not work, despite the code running smoothly when launched for simulation. I am using the default library for cross section (so, not specifying any library and just defined my material without adding any .XXX notation). In fact, in the output file it specifies the library as ".80c".

The errors that pop up are:
"error: searched directories:
DATAPATH must be set by user
DATAPATH must be set by user
fatal error. data file xsdir_mcnp6.2 does not exist.
fatal error. cannot open file xsdir_mcnp6.2"

Any idea about it?
Last edited:
  • #7
It seems like you have not got the install correct. You need to set up paths so that MCNP can find the various libraries and datafiles.

The xsdir file is a listing of isotopes that exist in your library. If you get seriously "into" MCNP you can customize this file. For example, if you use NJOY to produce special purpose cross section libraries, you can add these to your xsdir file. The one that comes with MCNP seems to be the one it is asking for, and that seems not to be correctly hooked up.

The MCNP package comes with installation scripts. There should be some "readme" files in there with these scripts. Those should help you figure out how to install the various components.

On some operating systems, the usual thing is for there to be an install process that happens one time, and also a script you run each time you start a session to run MCNP. So the first part sets up the directories with the files and such that come with MCNP. And then when you start a login to run MCNP, you run a script to set up the paths for this login.

On Windows, you open a DOS prompt, then run a script to set paths (including DATAPATH), then run MCNP. You can set a DOS prompt icon to run the script then provide you a DOS prompt window. This is nice because you don't need to have system privs to install MCNP on Windows.

In LINUX you can add the script to your .login or some such, depending on your specific LINUX install.
  • Like
  • Informative
Likes Alex A and 19matthew89
  • #8
This has more the feeling of a corrupt input file causing mcnp to explode than an install issue for me. I really don't know 6 that well though. Do check what is being suggested, but if you can share your input file could you rename it to add .txt and attach it to a post please?
  • #9
Thanks both.
Unfortunately, I'm also working on the cluster of my institute so I didn't actually install anything and cannot really modify anything there.

Here a minimal working example (I manage to plot the geometry so I hope it'll be enough to check if it's possible plotting the cross section).

I also found the ACE files of the library, but I guess I need to write a script in Python to read them and plot the cross sections based on those files.

Thanks in advance


  • Minimal_example.txt
    2.1 KB · Views: 1
  • #10
I only have 5.1.60 and 6.1 to hand, neither is capable of generating the error you get and both just plot the xs fine. If you have an older copy on the cluster it might be worth trying that.

If you can, check if you have a DATAPATH variable, and what xsdir files are present. Maybe you are tripping a 6.2 only feature and it is an install issue like @Grelbr42 thinks but I am at a total loss as to why it only happens for an xs plot.
  • #11

I had indeed to set the DATAPATH environment.
Moreover, I also discovered that cross section plotter (at least in my version) has a conflict with SSR card in case you read a source collected via SSW. In the minimal working example I sent above I replaced the SSR originally present in my code with the SDEF and, once set the DATAPATH, it indeed worked and plotted the cross section.
However, if I tried to re-insert back the SSR card, it gave me problems again. So...mind that, in case.

Suggested for: MCNP6.2 - ENDF/B reaction numbers for tally multiplier