Re: FLUKA: Too big material number
On Wed, 6 Jun 2001, Fan Lei wrote:
>
> >Concerning the geometry error messages, it is possible that they are
> >connected with the unusual size of your geometry. We are still
> >investigating them.
>
> Dear Albeto,
>
> I was also thinking alone this line. The dynamical range of my geometry may
> have excceded the limits in fluka?
No, several people have already run "Earth size" problems with FLUKA.
However, sometimes it is necessary to adjust a special parameter to ensure
that the tracking accuracy is preserved. This is a non-documented feature
and I know little about it. I hope some of my colleagues can tell you more
about it.
> I am doing a debug of the geometry with a gid of 1E3x1E3x1E3 following the
> advice
> from Paolo Sala, but it seems to take a very long time and I am still
> waiting for the execution
> to complete after more than four hours cpu time on a Pentium 500MH machine.
Since your geometry is spherically symmetric, you can speed things up
by scanning in only one dimension: 1E3x1x1. You should gain a factor 1E6
in speed.
> However I do have a couple of questions to ask:
> 1) To check whether my own source.f is being used or not I follwing the
> suggestion in the
> manual to have a print out statement
> WRITE(LUNOUT,*) ' users source.f being used '
> But i don't see the message is being printed out on screen. From simulation
> results I know
> that modified source.f HAS BEEN used. Where the message has gone and where
> is the value
> of LUNOUT is defined?
LUNOUT is not the screen, but the main output (atmos001.out). The value of
LUNOUT is currently = 11.
> 2) In using FLRNDM(XXX), is the variable XXX relevant? or is it really true
> that it can be
> any variale name.
XXX is irrelevant. You can use any floating point variable name,
and even nothing at all: FLRNDM( ). However,
it has happened in the past that some compiler was "optimizing"
by taking out of DO loops what it thought to be "a constant expression".
Such silly compilers probably don't exist anymore, but it is safe to
replace XXX by something that could have a different value at each call.
> How the radom number generator is controled?
The random number generator is initialized at the beginning of the run
based on the information contained in the ran... file. From then on, it
follows a deterministic sequence and cannot be controlled.
Alberto
--
Alberto Fassò
CERN-EP/AIP, CH-1211 Geneve 23 (Switzerland)
Phone: (41 22) 767 2398 Fax: (41 22) 767 9480 Alberto.Fasso@cern.ch
Partial thread listing: