Re: problem in running for more than 1000 histories

From: Giuseppe Battistoni (Giuseppe.Battistoni@mi.infn.it)
Date: Wed Nov 28 2007 - 17:36:05 CET

  • Next message: pksahani@cat.ernet.in: "Re: problem in running for more than 1000 histories"

    Hello!
    If you look at the first line of your arm1001.log you see that the run
    has been stopped because of TOO MANY GEOMETRY ERRORS.

    I recommend to you to study the use of geometry debugger (activated
    by GEOEND) in order to learn how to find the bad definitions.
    In your case the problem arises because you defined two concentric
    cylinders having diffent raidus, but exactly the same height!
    Morover your beam was shot exactly grazing to the base at X=0!

    I attach here a corrected input file where I arbitrarly gave a thickness
    of 1 cm to the two extremes of the cylindric container.
    I had also to move the X position of your beam injection in order to shoot
    at the middle of the cylinder.
    However, how Sebastien Wurth properly warned,
    I am not sure that this is the geometry you need...

    Another recommendation: it can be important to learn to avoid RCC.
    Although it seems more complicated, we try to recommend people to define
    closed cylinders by using infite cylinder bodies (XCC, YCC, ZCC) cut and
    delimited by infinite planes (XYP, YZP, XZP).
    This is much more advantageous from the point of view of computing time.
    Of course you must define regions with an accurate boolean logic.

    Please give a look at http://www.fluka.org/course/Houston-07/Lectures/
    where you can find the slides of the lectures given at one of the last
    FLUKA courses. There you can find the Geometry lectures and learn more
    and better.

    Understanding how to manage these geometry issues is the major attention
    that is required to FLUKA users.

            Good luck
                    Giuseppe Battistoni

    On Wed, 28 Nov 2007 pksahani@cat.ernet.in wrote:

    > Dear fluka users,
    > I am a new Fluka user.I m trying to score energy depostion with the
    > USRBIN command for a simple geometry of a Aluminium cylinder filled with
    > nitrogen by 1.25 Mev photon.When i m trying to run this program for
    > histories 10 or 100 , i am getting results.But when i ran it for
    > histories 1000,10000 or more, it didn't run and some problem in Seed
    > generation occured (as observed from output file).but i m not able to
    > point out the fault properly.
    > So please help me to find the solution.I m attaching my input file and
    > its output files.
    >
    > With regards
    > Prasanta Kumar Sahani
    > Scientific officer
    > Raja Ramanna Centre for Advanced Technology
    > Indore
    > INDIA
    >




  • Next message: pksahani@cat.ernet.in: "Re: problem in running for more than 1000 histories"

    This archive was generated by hypermail 2.1.6 : Wed Nov 28 2007 - 18:08:07 CET