Skip navigation links
(CGNS Documentation Home Page) (Steering Committee Charter) (Overview and Entry-Level Document) (A User's Guide to CGNS) (Standard Interface Data Structures) (SIDS-to-ADF File Mapping Manual) (SIDS-to-HDF File Mapping Manual) (Mid-Level Library) (ADF User's Guide) (CGNS Tools and Utilities)

CGNS Steering Committee

Meeting Minutes

24 June 2003
7:00 PM Eastern Daylight Time
Orlando, Florida (in conjunction with AIAA meeting)

  1. This general informational meeting and discussion was called to order at 7:05 PM. There were 8 attendees, listed below.

  2. This was not an official steering committee meeting, so minutes from the last telecon were not reviewed or approved.

  3. Current status of CGNS and outstanding issues (HTML, PDF) - was presented by Rumsey

  4. Discussion of current issues:
    1. HDF-5
      1. AEDC is actively evaluating and working with Poinot's HDF-5 prototype (summer intern Brian Pittman).
      2. Poinot has also expressed an interest in bringing it to completion at some time in the future.
      3. Question raised: what will happen to existing CGNS files that use ADF? Will the MLL automatically switch them over? Or will an explicit translation step be necessary? Will HDF-5 implementation be a new Version (3.0)? Will the MLL be able to read CGNSLibraryVersion_t node if it is an HDF file (i.e., how will the MLL software detect that the file is HDF-5 as opposed to ADF)?
    2. UserDefined extension - no issues.
    3. BCDataSet extension - will this proposal work for unstructured datasets also?
    4. CGNSVersion
      1. The CGNSLibraryVersion automatically gets altered (updated) by MLL when it opens an older file (Q: is this true when going from V1 to V2?).
      2. Should there be a CGNSVersion number (under base) that NEVER changes? This would indicate what version was used to write the particular base - kind of a way to document the dataset. But this gets confusing if V2 is ADF and V3 is HDF-5. Do we really want to keep a file with part ADF and part HDF-5? This does not seem like a good idea.
    5. GridLocation=FaceCenter in ZoneGridConnectivity.
      1. This issue is for unstructured grids that connect multiple zones.
      2. Seems reasonable to allow this, but will this work for non-point-matched connectivity?
    6. General issues:
      1. Confusion among attendees as to if/how CGNS allows face-based info for unstructured to be stored (Partial answer: CGNS easily allows for defining face elements, e.g. for use to define BCs at faces rather than at vertices. See the "Single-Zone Unstructured Grid" section in the User's Guide to CGNS.)
      2. We probably need to review and update the Charter; some of it is outdated.

Attendees

   Dan Dominik    Boeing
Bill Jones NASA Langley
Steve Karman UT Chattanooga
Mori Mani Boeing
Chris Nelson AEDC
Christopher Reed LM Aero - Ft. Worth
Chris Rumsey NASA Langley
Tom Shih Michigan State