[Paraview] Best format for time dependent datasets (polygonal mesh) with constant topology

[Samuel Key]

Manuel,

A easy data format for your application would be the EnSight binary format:

1) The mesh topology/geometry is stored in a seperate file. (I have seen 
remarks that it can be a one-time write, however, I write topology and 
coordinates every time step but "direct" the reader to only read the 
coordinates -- my mesh distorts with time during execution.)

2) It provides for scalar, vector, and 2nd-order tensor data at each 
note and in each cell.

3) It stores each point/cell variable vs time in separate files.

4) The association between variable name and the data (in its file) 
takes place in a *case file, see the attached example. (A simple trick: 
*.case files are ASCII text and easily edited. After two runs, one can 
write a third *.case file (assuming the geometry files and time stepping 
is identical) and reference two files with two, new variable names like 
Velocity-Reference and Velocity-Trial-Algorithm#1 and their respective 
data files. A delta-Vel difference can be done with ParaView's 
Calculator filter.)

5) File Format: http://www2.ensight.com/90_manuals/90-UserManual.pdf 
Chapter 11

If you want to see a Fortran-90 routine that writes EnSight, let me know.

Samuel

Manuel González wrote:
> I am planning to use Paraview to post-process results from my in-house
> developed simulation software. 
>
> After reading the “Paraview Guide”, the “File Formats for VTK” and googling
> data formats, I still find very difficult to select the data format (among
> those supported by Paraview) which is more appropriate to my needs:
>
> - Type of dataset: polygonal mesh. 
> - Time dependent. Topology is constant over time, but geometry (point
> coordinates) and attributes are not. 
> - Dataset size: 1e5 points, 1e5 lines, 1e5 2D polygons, 2-4 attributes per
> line and polygon, 100-500 time steps.
>
> Of course, I would like to avoid writing the same (constant) topology for
> each time step. 
> And it would be great if the topology could be stored in a separate file
> from the rest of the dataset (I often run several simulations for the same
> topology).
>
> I would appreciate any help to select the right data format for this kind of
> dataset.
>
> Best regards,
> Manuel
>
> *********************************************************
> Manuel J. Gonzalez Castro
> Laboratorio de Ingenieria Mecanica
> University of A Coruña
> Spain
> *********************************************************
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView
>
> Follow this link to subscribe/unsubscribe:
> http://www.paraview.org/mailman/listinfo/paraview
>
>   
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: fmaego.data.case
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100205/b7f7219a/attachment.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: samuelkey.vcf
Type: text/x-vcard
Size: 282 bytes
Desc: not available
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100205/b7f7219a/attachment.vcf>