Dear Enrico
As Giuseppe pointed out, that messages comes from the geometry
interpreter.
It means that there is somewhere in your geometry description one (or
more) region defined using parentheses, such that when parentheses are
expanded there are more than 100000 resulting individual zones...
As surprising as it can be, even relatively simple (but convoluted)
expressions using parentheses can result in a very large number of
terms. Fluka can purge those which are obviously empty, but the
remaining ones can still be too many.
Increasing the dimension of that parameter would not help, a) there
could be many more terms than 100000, b) a region with so many terms
would be a CPU killer runtime.
There are two possibilities, 1) rework the culprit region using less
parentheses, 2) switch to runtime expansion of parentheses (don't worry,
runtime expansion does not generate > 100000 terms), which is a bit
slower than pre-expanded parentheses but it could be a reasonable way
out in your case.
Ciao
Alfredo
On 2024-10-29 15:55, Enrico Scomparin wrote:
> Dear experts,
>
> I am facing the following error immediately after starting my job
> (looks like the .inp is not entirely scanned, even if I do not get
> errors with flair)
>
> At line 105 of file rp2exp.f
> Fortran runtime error: Index '100001' of dimension 1 of array 'txnew'
> above upper bound of 100000
>
> Error termination. Backtrace:
> #0 0x73967ea23960 in ???
> #1 0x73967ea244d9 in ???
> #2 0x73967ea24ad6 in ???
> #3 0x57d10870f47e in rp2exp_
> at /home/alfredo/fluprogfor/comlatmvax/rp2exp.f:105
> #4 0x57d108700ff4 in flkcgi_
> at /home/alfredo/fluprogfor/comlatmvax/flkcgi.f:1602
> #5 0x57d10855b751 in geoinp_
> at /home/alfredo/fluprogfor/geolatmvax/geoinp.f:263
> #6 0x57d1083efe8a in flukam_
> at /home/alfredo/fluprogfor/mainmvax/flukam.f:3583
> #7 0x57d107fa5ad9 in fluka
> at /home/alfredo/fluprogfor/mainmvax/fluka.f:359
> #8 0x57d107fa5ad9 in main
> at /home/alfredo/fluprogfor/mainmvax/(USFLMD):23
>
> The file rp2exp.f and the array are not explicitly available in the
> code so I have no real hint on what to do. Any help would be greatly
> appreciated. Let me know if I should supply more info.
>
> Thanks a lot
>
> Enrico
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Thu Oct 31 2024 - 11:06:55 CET