IGES from CATIA

Discussion in 'CATIA' started by Ian Reeves, Dec 14, 2004.

  1. Ian Reeves

    Ian Reeves Guest

    Hi,

    We use CATIA V4 and also have a Solidworks 2004 station to provide
    translations to our suppliers. On reading IGES files, I am noting alot of
    errors on the translated files. Does anyone know any methods to clean up
    files before placement in to solidworks.


    Cheers

    Ian
     
    Ian Reeves, Dec 14, 2004
    #1
  2. The way we convert is with STEP files - much better that IGES.

    WT
     
    Wayne Tiffany, Dec 14, 2004
    #2
  3. Ian Reeves

    Sporkman Guest

    Try using Rhino as a go-between. Import IGES from Catia, SaveAs IGES
    for SolidWorks.

    BTW, in trying to export IGES cavity and core surfaces to a mold maker
    for CAM tooling we've noticed that the imported geometry can be very
    squirrelly. Using Rhino seems to take out some extraneous data, leaving
    pure surfaces which the CAM software (Teksoft) imports without problems.
    The export options we used from SolidWorks (specifically for Teksoft)
    SHOULD HAVE left pure surfaces -- and perhaps it did -- but the data in
    the imported files were very problematic from multiple standpoints. The
    files seemed to contain duplicate sets of data causing duplicate tool
    path generation, and sometimes the "duplicate" tool paths weren't
    actually duplicates (different depths, and no the cavity and core
    surfaces were not offset) and sometimes they would bring the program to
    a halt.

    Substitute BIG for LARGE to reply directly.
    'Sporky'
     
    Sporkman, Dec 15, 2004
    #3
  4. Ian Reeves

    That70sTick Guest

    You can expect some problems with sheet metal type parts from CATIA.

    CATIA's modelling kernel has a higher tolerance for error when
    declaring whether entities are parallel or normal. I have had CATIA
    translations of part where supposedly parallel surfaces were off by
    3E-6 degrees or less. Not much, but enough to cause SW to say they are
    not parallel.

    The source of this error is in the CATIA file itself, not in the
    translation. I did have this verified by one intrepid CATIA operator
    who dug deep enough to find this.
     
    That70sTick, Dec 15, 2004
    #4
  5. Ian Reeves

    kmaren24 Guest

    CATIA

    The Catia IGES translator is based on IBM IGES Format (IIF). When you
    save a Catia model as an IGES file, it undergoes two translations:
    Catia=>IIF=>IGES.
    The IIF=>IGES step is done by the igesp program. This program has two
    parameters which affect the accuracy of the model and therefore
    increase the potential that SolidWorks will be able to form a solid out
    of the data. You can modify these values by editing the file
    igesinp.data. The 2 parameters which should be changed are:

    SIGFIG COEF n
    SIGFIG CORD n
    where n is the number of significant digits. The default values are 8
    and 6 respectively. The range of values are from 5 to 14. The
    recommended values are 14 and 14.

    COEF indicates the maximum number of significant digits in the IGES
    file for a real number that is a coefficient of an equation.
    CORD indicates the maximum number of significant digits in the IGES
    file for a real number that is a coordinate.
    By using the maximum values for these parameters, the size of the IGES
    file can become large, but the potential that SolidWorks will be able
    to form a solid out of the data is much greater.
     
    kmaren24, Dec 15, 2004
    #5
  6. Ian Reeves

    That70sTick Guest

    I've seen similar things happen in Unigrapihcs. A sheet metal part is
    unfolded, modified, refolded, and for some reason the faces are no
    longer parallel or the holes are off normal by a miniscule amount.

    In our position, we are not able to prompt the OEM's to fix their
    files. Most users also don't understand the variable tolerance thing
    enough to know how it can mess up a model.

    Just be aware that it can be an issue, and the source is not just
    "translation error".
     
    That70sTick, Dec 16, 2004
    #6
  7. Ian Reeves

    Cliff Guest

    Note that both UG & SW (and a few others) use the ParaSolid
    kernel and that it's the kernel that decides what a valid solid is
    (and is not).
    There must be kernel tolerances ....

    Sometimes I miss the old CADDS-III error "surfaces not
    contiguous" (IIRC).
     
    Cliff, Dec 16, 2004
    #7
  8. Ever hear of FormatWorks? It is designed to handle this.

    Lyle
     
    Lyle and Laurel Fischer, Dec 17, 2004
    #8
  9. Ian Reeves

    That70sTick Guest

    We had Capvidia (FormatWorks) do some repair for us. I was impressed.
    For us, it is more economical to employ this type of thing on a service
    basis (which Capvidia also provides).

    Don't underestimate the value of a clean translation and a good repair
    job. It saves a lot of time, and time = money.

    Lyle:
    I referred someone to you the other day. Can't remember his name,
    though.
     
    That70sTick, Dec 17, 2004
    #9
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.