(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
Telecon Minutes
16 March 2000
2:00pm Eastern Time
The meeting was called to order at 2:00 Eastern time.
There were 18 attendees, listed below.
Bob Bush reviewed the minutes from the January meeting, noting that
they have been posted on the web. Having no comments, the minutes were
approved as posted. Bob also noted that following the January meeting, a
presentation was made to the AIAA CFD Committee on Standards proposing
to affiliate with that organization. The proposal was accepted, and we
are now a sub-committee of this organization.
- A round-table discussion of the current implementation and usage
status of CGNS in various organizations was held to familiarize the
Committee with current activities. The high points are summarized
here.
- Intelligent Light - Dave Edwards - IL is planning to release
a CGNS reader for their primary product in the July time
frame. During this effort, they will assess performance
and identify areas for improvement. One area is likely to
be incorporation of their 'Data Guide' technology to speed
access to large data sets. During the installation process
they also anticipate needing additional utilities. They hope
to provide these utilities to the CGNS community. This led
to some discussion of how to provide and support additional
utilities. This will need to be explored further. Steve
Legensky also offered to help with a web list of CGNS
contacts. See the viewgraph
presented by Dave Edwards.
- ADAPCO - Scott Ochs - ADAPCO is in the process of
transitioning to CGNS, and can currently read and write
grids and solutions (structured and unstructured). Their
next step is to utilize the mixed element capability, and
store time accurate data. The capability is already in beta
release, with a more complete release planned in a couple of
months. They expressed an interest in obtaining 'sample' codes
to help in the transition. This led to a discussion of the
possibility of creating a place on the web where sample code
can be posted. Scott (and Steve Feldman) also raised a problem
they were having with the cg_goto function on IBM
systems. Others on the line who had used IBM systems have not
experienced similar problems. They will continue to try and
identify the problem.
- AMTEC - Diane Poirier noted that Tecplot now has a CGNS
reader. Release date is unknown.
- Fluent - Barb Hutchings - Barb noted that Fluent has
downloaded the documentation, and assigned a programmer to
begin implementation. They anticipate a release of the ability
to read and write information in V5.5, which should occur in
the third quarter of 2000.
- NUMECA - Michel Delanaye - NUMECA is adopting CGNS for all
their products. They currently have a full implementation
in their grid generator, with the exception of the boundary
condition information. The structured code also uses it for
the solution (but again not coupling information yet. They
would like to include unsteady data as soon as the capability
is available. They would also like the ability to store
planes of solution data (they were referred to the Boeing
discussion to come). They anticipate a March release to their
user base. NUMECA has also been active convincing various
consortia to adopt CGNS as the data exchange standard. New
capabilities required include new element types (e.g.
hanging nodes) and potentially octree information. For
more details, see below.
- Lockheed - Steve Karman - Also mentioned the need for hanging
node information.
- Boeing - Todd Michal - Boeing has a structured grid
implementation of CGNS in a pre-release version of Wind. They
anticipate including unstructured grid capability, utilities
and post processing capability this year, and a release to
NPARC user community next year. Boeing is also bringing in
Overflow and CFL3D versions, which utilize CGNS. They have
made a proposal for storing boundary planes, which will be
distributed for discussion as an extension proposal.
- RocketDyne - Armen Darian - RocketDyne has converted a
rotating machinery post-processor to use CGNS.
- Vircinity - Andreas Wierse - We introduced Andreas Wierse
who gave an overview of Vircinity, a visualization and
post-processing capability with emphasis on virtual
reality. They are looking at CGNS, and should decide on a
development plan soon, and assign a CGNS contact point. They
also have contacts with the EU and will bring CGNS to their
attention.
- Extension Status (Diane Poirier)
- The unstructured grid element ordering extension has
been implemented and is on the web. Documentation is
underway. ADAPCO is utilizing already.
- The moving grid proposal is finished and ready for
implementation.
- The chemistry proposal is still stalled. Suggested to finalize
as is, and extend in the future if a more comprehensive
standard is required.
- Diane also brought up the need for a documentation focal
point. Anyone interested please contact Diane or Bob Bush.
- Steve Legensky offered to take on some of the testing on
multiple platforms to relieve Diane of this aspect of the
development.
- June Meeting - There was a brief discussion of the June
meeting, and it was decided to coordinate with the AIAA CFD CoS
with the CGNS team meeting first, and reporting out at the CoS
meeting. (Subsequent contact with Ray Cosner has identified
that our meeting night will be Tuesday, with the CoS meeting on
Wednesday night. Please plan accordingly!)
Attendees
| Bob Bush |
| UTRC
|
| Stuart Ochs |
| ADAPCO
|
| Steve Feldman |
| ADAPCO
|
| Dan Dominik |
| Boeing Rocketdyne
|
| Armen Darian |
| Boeing Rocketdyne
|
| Barb Hutching |
| Fluent Inc.
|
| Michel Delanaye |
| NUMECA
|
| Bill Thompson |
| ?
|
| David Kenwright |
| NASA Ames
|
| Steve Karman |
| Lockeed
|
| David Edwards |
| Intelligent Light
|
| Steve Legensky |
| Intelligent Light
|
| Mark Fisher |
| Boeing St-Louis
|
| Ray Cosner |
| Boeing St-Louis
|
| Todd Michael |
| Boeing St-Louis
|
| Mori Mani |
| Boeing St-Louis
|
| Andreas Wierse |
| Vircinity
|
| Diane Poirier |
| ICEM CFD Engineering
|
Intelligent Light's CGNS Effort
- Develop CGNS reader for Fieldview (Version 7, Sept 2000)
- Testing with CGNS files generated from various solvers
- Creation of matrix of CGNS data sets generated from from
various CFD solvers and grid generation tools
- Assessment of performance
- Identification of areas of improvement
- Integration with Fieldview Data Guide
- Develop Deskdrawer CGNS Utilities
- Utility for output of header information (tell user what's
in the file)
- Utility to determine if CGNS file can be used for
pre-processing, solver or post-processing
- Utility to convert CGNS to Fieldview Unstructured File Format
- Others (TBD)
Report from NUMECA
Subject: NUMECA input
Bob,
I thought you would be interested to get something written about what
I said during the last teleconference. So, NUMECA is adopting the CGNS
format as the only data transfer system between its different products,
IGG (multi-block structured grid generation), Euranus (multi-block
structured flow solver), CFView (postprocessing, visualization).
So far, the grid coordinates are written in the data base, the topology
is still stored in our own format, but we are progressively changing
to CGNS for the complete set of information, connectivity, boundary
conditions, etc. This should be completed fairly soon. The flow solver
outputs solutions in the CGNS database, and the database can be read by
CFView. For unsteady computations, solutions at different time steps are
stored for the moment in a different file. The CGNS will be part of the
next release of these products, i.e. end of March.
We have proposed the use of CGNS in two European research projects. The
CGNS format has been adopted in the first one, HPNURSA, to study
unsteady flows in turbomachinery. Partners in this project are TaskFlow,
SIMULOG which is writing a CGNS reader for Ensight, SEP, HYRDO-Vevay,
EPFL and Ecole Centrale de Lyon. In the second one, ESTEDI, we proposed
to test the CGNS format as a possible way to store data in a very large
data base which is intended for data posttreatment (data mining) for
large scientific computations (Climate modeling, Cosmology, Satellite
tracking, LES, etc).
I mentioned some problems regarding speed. Further investigation on
our side has highlighted some bad usage which has been corrected
since. I also mentioned our wish to be able to store quantities on
block boundaries. The Boeing proposal will most propably meet our
expectations. The last comment was about the ability to store our
hierarchal hexahedral unstructured grids. On this matter, I am going to
investigate different possibilities and probably contact Diane with some
propositions.
Hope this will help,
-Michel
Dr. Michel Delanaye
NUMECA International
Av. Franklin Roosevelt 5
B-1050 Brussels
Tel: +32 2 647.83.11 (main) ext 824
+32 2 642.28.24 (voice mail)
Fax: +32 2 647.93.98