RE: [fluka-discuss]: Problem in geometry: something wrong with parentheses?

From: Vasilis Vlachoudis <Vasilis.Vlachoudis_at_cern.ch>
Date: Thu, 24 Oct 2013 16:09:12 +0000

This is a common problem with the use of parenthesis. It doesn't matter if they are nested or not.
e.g. an expression like +A -(subexp1) -(subexp2) -(subexp3)
during the expansion it will end up with all the combinations of all terms of A, subexp1-3. With a
few parenthesis the upper limit of 100k terms can be easily reached.
There is an internal optimization but works only on parallel planes with the axes and bounding boxes of objects.
When infinite objects are used or transformations then the optimization starts to fail.

You can graphically edit the problematic region with the GeometryEditor, directly into the expanded form (unions of zones).
The process is easy please look on the flair manual (zone-editing) as well the youtube examples.

Vasilis

________________________________________
From: owner-fluka-discuss_at_mi.infn.it [owner-fluka-discuss_at_mi.infn.it] on behalf of Giuseppe Battistoni [giuseppe.battistoni_at_mi.infn.it]
Sent: 24 October 2013 17:11
To: fluka-discuss_at_fluka.org
Subject: [fluka-discuss]: Problem in geometry: something wrong with parentheses?

Dear experts of combinatorial geometry:

I have a problem with the attached geometry (with relative input+flair).
It describes a drift chamber where the drifts cells (as defined by wires in reality) are
equivalent in space, from the simulation point of view,to tubes of exagonal cross section.

Each exagonal tube is built by intersecting PLA bodies.
There are 4 + 4 layers of such tube cells (4 per layer)
oriented in two orthogonal directions (U and V cells)
I also attach a picture (the Y-z projection where the U cells are clearly visible).

Flair return no error in geometry.
However at run time, or better, at initialization time, FLUKA crahes
in routine rpnorm:

At line 84 of file rpnorm.f
Fortran runtime error: Index '100001' of dimension 1 of array 'tx' above upper bound of 100000

Backtrace for this error:
   + function rpnorm_ (0x5FF599)
     at line 79 of file rpnorm.f
   + function flkcgi_ (0x5EDFA0)
     at line 1526 of file flkcgi.f
   + function geoinp_ (0x4E930C)
     at line 241 of file geoinp.f
   + function flukam_ (0x409299)
     at line 2958 of file flukam.f
   + function fluka (0x40782C)
     at line 290 of file fluka.f
   + /lib64/libc.so.6(__libc_start_main+0xed) [0x36f242169d]


I have a certain number of parentheses. but they are not nested!
Is the problem really due to them or to something else?

Thank you for any help

        Giuseppe
Received on Thu Oct 24 2013 - 19:04:26 CEST