Re: geometry precision errors

Hi Jaroslaw

Both prototype and dummy-region must fit exactly, including boundaries. An
RPP includes all its boundaries while a definition of an RPP with PLA
includes only those boundaries which are defined with PLA with positive
sign. I suggest that you correct this aspect and see if the message still
occurs.

GEOBEGIN/What(2) should not be used. It is reserved for program
development.

Greetings
Stefan

On Sun, 6 Apr 2008 j.zdebik@fz-juelich.de wrote:

> Dear FLUKA authors and users,
>
> I have got two questions related with problems with a precision in geometry.
>
> 1. Precision errors on boundary.
>
> I have defined a base cell using RPP declaration, but my lattice cells, being rotating of this last one, are decalred with PLA declaration,
> could I have some problems related with this?
>
>
> I am asking because I am trying to
> debug a geometry.
>
> When I hit my geometry with PHOTON BEAM (momentum = 10 GeV/c) I get the following error message:
>
> ---------------------
> "error messsage"
>
> NEXT SEEDS: 12AF7A 0 0 0 0 0 181CD 3039 0 0
> Geofar: Particle in region 2 (cell # 0) in position 2.499620401E+01 -1.121151881E+01 1.530000000E-03
> is now causing trouble, requesting a step of 1.000000000E+08 cm
> to direction -8.540679798E-02 -7.081037323E-01 -7.009242350E-01, error count: 0
> R2: 2.739540781E+01 R3: 2.739540785E+01 cm
> X*U (2D): 5.804072569E+00 X*U (3D): 5.803000155E+00 cm
> X*UOLD(2D): -9.155329813E+00 X*UOLD(3D): -9.154151914E+00 cm
> Kloop: 81377, Irsave: 591, Irsav2: 2, error code: -33 Nfrom: 5000
> old direction -5.274356424E-01 -3.593232260E-01 7.698691203E-01, lagain, lstnew, lsense, lsnsct F F F F
> Particle index 7 total energy 5.780450824E-03 GeV Nsurf16253
> Try again to establish the current region moving the particle of a 1.000000000E-06 long step
> We succeeded in saving the particle: current region is n. 2 (cell # 0)
> ---------------------
>
> This point is exactly on boundary, and I read in Fluka manual that this is related with "a rounding error near a boundary",
>
> but when I hit at the same point with MUON+ BEAM (momentum = 10 GeV/c) everything is correct and I saw normal energy deposits.
>
> I built my geometry with precision 15 digits after dot. Geometry debugger tels "no errors in geometry file" for this point.
> And I am able to make visualization all my geometry with FLAIR.
> In lattic.f I use a rotation subroutine with double precision.
>
>
> I know a similar problems were discussed before:
> http://www.fluka.org/web_archive/earchive/new-fluka-discuss/0045.html
>
>
> 2. What(2) geobegin card.
>
> I read too that it is possible to modify an accuracy of the geometry using WHAT(2) command in GEOBEGIN card.
> But I don't know how I can change this precision and if this is safe solution.
>
>
> Thanks for help,
>
> Best greetings,
> jaroslaw
>
>
>
> -------------------------------------------------------------------
> -------------------------------------------------------------------
> Forschungszentrum Juelich GmbH
> 52425 Juelich
>
> Sitz der Gesellschaft: Juelich
> Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
> Vorsitzende des Aufsichtsrats: MinDir'in Baerbel Brumme-Bothe
> Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender),
> Dr. Ulrich Krafft (stellv. Vorsitzender), Prof. Dr. Harald Bolt,
> Dr. Sebastian M. Schmidt
> -------------------------------------------------------------------
> -------------------------------------------------------------------
>
>

-- 
___________________________________
Stefan Roesler
CERN, SC/RP
CH-1211 Geneva 23
Switzerland
Phone:  +41-22-7679891
Fax:    +41-22-7669639
E-mail: Stefan.Roesler@cern.ch