Re: [Paraview] Good file format for point data on a surface over 60, 000 timesteps

[Tyson Whitehead]

On 20/11/17 12:49 PM, Berk Geveci wrote:

What kind of visualizations are they looking to do? It may be a good idea to 
treat time differently here given that it is the most dense dimensions. I am 
thinking of some sort of tabular format that can then be used with filtering to 
extract a time step. Potentially, even a database may work depending on what 
the end visualization product will be.


My understanding is they want to start by making a movie of the recorded 
pressure values interpolated (as some color ramp) across the structure surface 
and as this is what they are currently doing with their MATLAB code.

I ultimately expect they may come up with something else once they get the data 
in though and are able to play around with it (enabling playing around with it 
being a big part of why they are wanting to get it into ParaView).

While the local expert, I've only really go very very basic experience so far 
(all I've really done is work through the various tutorials and teach them in 
workshops), so I'm interested in any advice.

Compared to the tutorials, the strange thing about this data is the large 
number of time steps, that it is spatially irregular and sparse, and that it 
doesn't fully define the underlying object geometry (i.e., the geometry of the 
surface extends out past the last sample points due to each measurement point 
requiring drilling holes for measurement taps in the physical object).

Currently I'm think, if we can come up with a decent way of importing the data 
as a point set, of

https://blog.kitware.com/dataset-resampling-filters/

1. import the sample data as a point data set
2. use the point volume interpolate (or resample to image?) to create a mesh 
from it
3. use the re-sample with dataset to put that onto the actual physical geometery
4. save animation to get the desired movie

Is there a better way to do this?  Should I try and create some sort of 
irregular grid out of the sample data to avoid step 2?

Thanks!  -Tyson


___
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

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] Good file format for point data on a surface over 60, 000 timesteps

[Berk Geveci]

What kind of visualizations are they looking to do? It may be a good idea
to treat time differently here given that it is the most dense dimensions.
I am thinking of some sort of tabular format that can then be used with
filtering to extract a time step. Potentially, even a database may work
depending on what the end visualization product will be.

On Fri, Nov 17, 2017 at 6:44 PM, Tyson Whitehead 
wrote:

> Hi,
>
> I'm trying to help one of our users (HPC center) use ParaView to work
> with their wind tunnel data.  They have a large number of time samples
> (60,000) from a series of surfaces with from a small number of sample
> points (64).
>
> The sample points are irregular but structured as they are laid out to
> be more dense in areas that should see more differences (i.e., at best
> you could say it was a non-uniform grid with several of the points not
> used).
>
> What file format should this data be put in to import into ParaView?
>
> The VTK legacy formats and XML description formats seem painful to use
> with this number of timesteps and I believe (please correct if wrong)
> that the NetCDF wants a regular grid.
>
> Thanks!  -Tyson
> ___
> 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
>
> Search the list archives at: http://markmail.org/search/?q=ParaView
>
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview
>
___
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

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

Re: [Paraview] Good file format for point data on a surface over 60, 000 timesteps

[Samuel Key]

Tyson,

You won't get a common answer, but here is my $00.02. The EnSight format 
has a nice feature that every variable has its own file. So, each 
'sample point' will have its own file. The file names can be as you 
want, numeric or alphabetic. The EnSight format has a geometry file, but 
if you want to use single vertex cells it won't become very large.  If 
you want to be a little more graphical, you could use 64 4-node quads 
with their respective coordinate positions. Additionally, you could add 
a relatively coarse mesh of the structure. (ParaView has a Filter that 
will take a point set (your 64 single-point cells) and construct a 
triangular mesh to provide some geometry.


The EnSight format uses a meta-file (aka *.case file) that references 
all of the files in the datum set, example attached. It will be some 
work the first time around, but well worth it in my opinion.


(I use this file-format myself for million-element 3-D meshes for 
transient dynamic simulations. I would not leave home without it.)


Samuel Key
FMA Development
1005 39th Ave NE
Great Falls, MT 59404



On 11/17/2017 11:44 AM, Tyson Whitehead wrote:

Hi,

I'm trying to help one of our users (HPC center) use ParaView to work
with their wind tunnel data.  They have a large number of time samples
(60,000) from a series of surfaces with from a small number of sample
points (64).

The sample points are irregular but structured as they are laid out to
be more dense in areas that should see more differences (i.e., at best
you could say it was a non-uniform grid with several of the points not
used).

What file format should this data be put in to import into ParaView?

The VTK legacy formats and XML description formats seem painful to use
with this number of timesteps and I believe (please correct if wrong)
that the NetCDF wants a regular grid.

Thanks!  -Tyson
___
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

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview



FORMAT
type: ensight gold
GEOMETRY
model:   1  1  fmaego.results.geomchange_coords_only
VARIABLE
vector per node: 1  1  Displacement  fmaego.results.ndis
vector per node: 1  1  Velocity  fmaego.results.nvel
vector per node: 1  1  Acceleration  fmaego.results.nacc
scalar per element:  1  1  Material  fmaego.results.emat
scalar per element:  1  1  Material-Statefmaego.results.esta
scalar per element:  1  1  Pressure  fmaego.results.eprs
scalar per element:  1  1  Bulk-Strain   fmaego.results.elnv
scalar per element:  1  1  Strain-Energy-Density fmaego.results.esed
scalar per element:  1  1  Effective-Stress  fmaego.results.edev
vector per element:  1  1  Stress*Velfmaego.results.estv
tensor symm per element: 1  1  Stressfmaego.results.estr
TIME
time set:1
number of steps:   101
time values:  1.412425480266E-08  2.400133725957E-04
  4.800116413781E-04  7.200087675218E-04  9.600085065557E-04
  1.200013506904E-03  1.440004395631E-03  1.680012453687E-03
  1.920003546411E-03  2.160008706146E-03  2.42268722E-03
  2.64825978E-03  2.880008516105E-03  3.120007260160E-03
  3.360004355506E-03  3.600011259104E-03  3.840002535435E-03
  4.080002209989E-03  4.320008836543E-03  4.560001245923E-03
  4.800013108817E-03  5.040001157019E-03  5.280014080032E-03
  5.520013787104E-03  5.760010120856E-03  6.06744585E-03
  6.240010362247E-03  6.480011393759E-03  6.720005078932E-03
  6.960009814293E-03  7.23845679E-03  7.440002044729E-03
  7.680008165642E-03  7.920003445528E-03  8.160001173762E-03
  8.47812238E-03  8.64624252E-03  8.88733655E-03
  9.120008566836E-03  9.360009135538E-03  9.65315679E-03
  9.840004635407E-03  1.008001339743E-02  1.032000754625E-02
  1.056001376543E-02  1.080001393196E-02  1.104001032624E-02
  1.12801058E-02  1.152000634945E-02  1.176000815323E-02
  1.20831238E-02  1.224000912002E-02  1.248000571827E-02
  1.272000588352E-02  1.296001289990E-02  1.32451785E-02
  1.344000690212E-02  1.368000696938E-02  1.392000271408E-02
  1.416000989430E-02  1.44070735E-02  1.464000160285E-02
  1.488000797845E-02  1.512000435611E-02  1.536000661064E-02
  1.56307197E-02  1.584000938572E-02  1.608000643769E-02
  1.632000546980E-02