Crash while processing USRYIELD results

Dear FLUKA experts,

I am having some weird issues with processing USRYIELD’s outputs. Whenever I finished a run and try to process all the fortran files, similar error would pop-up just like below:

Started 2020.07.17 09:44:53

- Processing: Nominal_21.yie
Cmd: /usr/local/fluka/bin/usysuw
<<< Nominal001_fort.21
<<< Nominal_21.yie

Type the input file:
DATE: 7/17/20, TIME: 9:44:33
At line 752 of file usysuw.f
Fortran runtime error: Index ‘2501’ of dimension 1 of array ‘accumg’ above upper bound of 2500

Error termination. Backtrace:
#0 0x7febc2e3b2ed in ???
#1 0x7febc2e3bed5 in ???
#2 0x7febc2e3c2a7 in ???
#3 0x7febc360432e in usysuw
at /root/rpmbuild/BUILD/fluka-4/src/tools/usysuw.f:752
#4 0x7febc3601152 in main
at /root/rpmbuild/BUILD/fluka-4/src/tools/usysuw.f:1407

e> ERROR: cannot access file ‘Nominal_21.yie’
Maybe the run is still going on?
Error processing: Nominal_21.yie
Ended: 2020.07.17 09:44:54

Possible errors: Nominal_21.yie
Correctly processed:

What I have tried is what Francesc Salvat Pujol suggested in Question on LET spectra. I delete all my files produced from the run along with flair files leave only my input file intact, then I restart the terminal and flair to load the input file. This does not solve the problem. What might solve the problem is that I found out if I make a new project with flair and then type up all the stuff I had with my old input file. However, this new project only process the results correctly once. Anytime I try to increase the number of bins or change the upper limit of the USRYIELD, aforementioned error appears.

The attached is all my files that associated with the error message above. I am running the latest version of FLUKA and flair in the WSL environment.

Nominal.out (938 Bytes) Nominal001.err (21.9 KB) Nominal001.out (106.5 KB) Nominal.flair (1.7 KB) Nominal.inp (1.4 KB)

Thanks in advance for the time and help.


Dear Martin,

currently the processing program usysuw has a hard limit on how many bins it can process, which is less than 2200.

USRYIELD scorings with less bins than that can be processed out of the box.

If you really need 4000 bin in a single USRYIELD scoring, you can download the .tgz package of FLUKA and modify the source of usysuw (src/tools/usysuw.f), by increasing the NMXEBN parameter at line 201 to a higher value, then running the make command as you would during the installation of FLUKA.


Dear @horvathd ,

Thank you so much for the help. I did not know this.