Re: "memory" problems in CG?

From: Mattias Lantz (Mattias.Lantz@mi.infn.it)
Date: Thu Jan 18 2007 - 12:32:26 CET

  • Next message: Silvia Pani: "Re: "memory" problems in CG?"

    Dear Silvia

    I just noted one thing in your geometry, you use integers for the black
    hole and the air:
    > RPP black -5e6 +5e6 -5e6 +5e6 -5e6 +5e6
    > RPP air -1e6 +1e6 -1e6 +1e6 -1e6 +1e6

    Try with
    > RPP black -5.e+06 +5.e+06 etc...

    and see what happens. Your strange effects may be due to this initial
    error in the geometry.

    Another thing, please send attached files as attachments instead
    of including them in the email itself.
    This makes it easier for people to help you, and input files written in
    fixed format do not face the risk of being modified by different email
    program.

    Best wishes,

    Mattias Lantz - INFN Milano

    On Thu, 18 Jan 2007, Silvia Pani wrote:

    > Dear all,
    > I am writing the geometry for a system for non-destructive testing
    > consisting of a total of 31 bodies and 16 regions.
    > I attach the geometry file below.
    > When I run the code with all bodies and regions it crashes, but it
    > doesn't crash if I remove ANY region out of a certain set (and
    > redefine the remaining regions, in particular the surrounding air,
    > accordingly), i.e. intcyl (a cylinder defined as the intersection-
    > subtraction of an infinite cylinder and two half-spaces), CCDhold,
    > extcase, case2, incase ("hollow boxes" defined as the subtraction of
    > a parallelepiped from another one).
    > The only common characteristic I have found within the set is that
    > the regions are obtained as a subtraction of bodies.
    > On the other hand the program still crashes if I remove one or more
    > "simpler" regions (i.e., those defined only as a sum of different
    > bodies).
    >
    > I guess it is not a problem of ill-defined geometry, otherwise it
    > wouldn't work with any of these regions removed (they are located in
    > different regions of space and they don't "interfere" with each
    > other). I am wondering whether there could be some sort of "memory"
    > problem, perhaps due to the fact that the more bodies I add the more
    > complex the "airound" region becomes.
    > Would that be possible? Any other suggestion?
    >
    > Many thanks
    >
    > Silvia
    >
    > ------------------------------------------------------------
    > 0 10 defining the NDT system
    > RPP black -5e6 +5e6 -5e6 +5e6 -5e6 +5e6
    > RPP air -1e6 +1e6 -1e6 +1e6 -1e6 +1e6
    > * ----------test object-------
    > XCC infrod 0.0 0.0 2.0
    > XCC infint 0.0 0.0 1.5
    > XCC infsp 0.0 0.0 0.2
    > * base surfaces of test object
    > YZP rodsurf1 -50.0
    > YZP rodsurf2 50.0
    > * base surfaces of internal cylinder
    > YZP isurf1 -25.0
    > YZP isurf2 25.0
    > * spheres for end hemispheres
    > SPH ball1 -50.0 0.0 0.0 2.0
    > SPH ball2 50.0 0.0 0.0 2.0
    > * ---------------------------------------------
    > * lead casing
    > RPP leadout -67.0 +67.0 -63.8 +60.0 -120.0 +80.0
    > RPP leadin -64.5 +64.5 -61.3 +57.5 -117.5 +77.5
    > * holes in lead casing
    > XCC holes 0.0 0.0 6.0
    > * -------------------------------------------------
    > * X-ray system holder
    > RPP Xhold1 -50.0 +50.0 -40.0 +40.0 +67.5 +77.5
    > RPP Xhold2 -30.0 +30.0 -20.0 +20.0 +47.5 +67.5
    > RPP Xhold3 -20.0 +20.0 -40.0 -5.0 +32.5 +46.5
    > RPP Xhold4 -40.0 +40.0 -20.0 +20.0 +46.5 +47.5
    > *---------------main detector----------------
    > * CCD camera
    > RPP CCD -1.5 1.5 -1.0 1.0 -10.21 -10.18
    > *CCD holder
    > RPP Chold1 -3.0 +3.0 -3.0 +3.0 -13.0 -9.0
    > RPP Chold2 -2.7 +2.7 -2.7 +2.7 -10.31 -9.1
    > * scintillator
    > RPP CsI -1.5 1.5 -1.0 1.0 -10.03 -10.0
    > * FOP
    > RPP fiber -1.5 1.5 -1.0 1.0 -10.18 -10.03
    > * Al substrate
    > RPP Alsub -1.5 1.5 -1.0 1.0 -10.31 -10.21
    > *-----------------line scan detector---
    > * Actual detector
    > RPP linesc -0.03 0.03 -30.2 -25.0 -10.21 -10.18
    > * GadOx scintillator
    > RPP gadox -0.03 0.03 -30.2 -25.0 -10.18 -10.17
    > * Al substrate
    > RPP Alsub1 -0.2 0.2 -30.2 -25.0 -10.31 -10.21
    > * plastic case
    > RPP plcase1 -3.2 3.2 -33.4 -21.8 -12.0 -9.0
    > RPP plcase2 -2.7 2.7 -32.9 -22.3 -10.31 -9.1
    > *----------------lead shielding----------------
    > RPP caseout -7.5 7.5 -40.0 -15.0 -15.0 -5.0
    > RPP casein -7.0 7.0 -39.5 -15.5 -14.5 -5.5
    > *-----------------------------------------------
    > END
    > * black hole (body 1)
    > black 5 +black -air
    > * Rod (body 2)
    > rod 5 |(+infrod -infint +rodsurf2 -rodsurf1)
    > |(+infsp -isurf2 +rodsurf2) |(+ball2 - rodsurf2) | (+ball1 +rodsurf1)
    > * Inner cylinder (body 3)
    > intcyl 5 +infint +isurf2 -isurf1
    > * Lead casing (body 4)
    > incase 5 +leadout - leadin - holes
    > * CCD (body 5)
    > detect 5 +CCD
    > * X-ray holders (body 6)
    > Xhold 5 |+Xhold1 |+Xhold2 |+Xhold3 |+Xhold4
    > * CCD holders (body 7)
    > CCDhold 5 +Chold1 -Chold2
    > * air around (body 8)
    > airound 5 +air - (|(+infrod -infint +rodsurf2 -rodsurf1)
    > |(+infsp -isurf2 +rodsurf2)|(+ball2 - rodsurf2) | (+ball1 +rodsurf1))
    > - (+infint +isurf2 -isurf1)
    > - CCD - (+leadout-leadin-holes)
    > - (|+Xhold1 |+Xhold2 |+Xhold3 |+Xhold4)
    > -(+Chold1 -Chold2)
    > - CsI -Alsub -fiber
    > -(+caseout -casein)
    > - linesc
    > -gadox -Alsub1
    > - (+plcase1 -plcase2)
    > * scintillator (body 9)
    > scint 5 +CsI
    > * FOP (body 10)
    > FOP 5 +fiber
    > * Al substrate (body 11)
    > substra 5 + Alsub
    > * lead case (body 12)
    > case2 5 +caseout -casein
    > * line scan detector (body 13)
    > scan 5 +linesc
    > * GadOx layer (body 14)
    > phosph 5 +gadox
    > * line scan substrate (body 15)
    > Alscan 5 +Alsub1
    > * detector case for line scan (body 16)
    > extcase 5 +plcase1 -plcase2
    > END
    > ------------------------------------------------------------
    > <><><><><><><><><><><><><><><><><><><><><><>
    > Dr Silvia Pani
    > University College London
    > Department of Medical Physics and Bioengineering
    > Malet Place Engineering Building
    > London WC1E 6BT
    > Tel +44 20 76790259
    > e-mail spani@medphys.ucl.ac.uk
    > <><><><><><><><><><><><><><><><><><><><><><>
    >
    >


  • Next message: Silvia Pani: "Re: "memory" problems in CG?"

    This archive was generated by hypermail 2.1.6 : Thu Jan 18 2007 - 12:54:05 CET