[Paraview] Experimental Data Interpolate/Extrapolate to Custom Volume Mesh

Thu Dec 1 14:40:24 EST 2016

If I understand you correctly, you were able to take each experimental plane and sample it in a the CFD data on the same plane? Now you would like to stack your experimental planes together and interpolate it onto the CFD grid?

I haven't done this, so I'm not sure if it's going to work, but can't you do a little preprocessing and then do the steps in reverse from before? What I mean is:

1. Take all your experimental planes (u, v, w, x, y) at a given z and concatenate the files together, adding the z coordinate at the end, to get (u, v, w, x, y, z)

2. Load this into Paraview and use the TableToPoints filter

3. Load your CFD grid

4. Select your TableToPoints filter and do:

    a. ResampleWithDataset -- select the TableToPoints as the input and the CFD grid as the Source

This should then give you the data points from the experimental data interpolated onto the CFD grid.

Does that seem like what you are trying to do?

Tim

________________________________
From: treem22 at gmail.com <treem22 at gmail.com> on behalf of Mike Tree <treem22 at gatech.edu>
Sent: Thursday, December 1, 2016 2:30 PM
To: Gallagher, Timothy P
Cc: paraview at paraview.org
Subject: Re: [Paraview] Experimental Data Interpolate/Extrapolate to Custom Volume Mesh

Tim,

I successfully used my experimental data planes to sample my CFD data for comparison, but now I've got another procedure to tackle.

I am looking to interpolate between my experimental planes and fill in the volumetric velocity data. I've spent significant time working with the GaussianResample filter, but it seems to eliminate some of my data. Specifically, GaussianResample can only interpolate scalars, so I have to decompose my velocity vector into its components. This seems to work fine, except that when the GaussianResample filter is applied the negative values of these components are eliminated. Then, when I re-calculate my velocity vector as a sum of the component it is obviously wrong because none of the components are permitted to be negative.

Are there any other filters in existence that I can use for interpolation?

Any help will be most appreciated.

-Mike Tree

On Tue, Nov 8, 2016 at 9:42 AM, Mike Tree <treem22 at gatech.edu<mailto:treem22 at gatech.edu>> wrote:
Tim,

Seems pretty straight forward. Thanks! I'll let you know if I have any issues.

-MT

On Mon, Nov 7, 2016 at 12:00 PM, Gallagher, Timothy P <tim.gallagher at gatech.edu<mailto:tim.gallagher at gatech.edu>> wrote:

Mike,

Your state file doesn't work without the data files that go along with it.

That said, here is how you can do it:

1. Load your .dat file and use the TableToPoints filter to create points from the data table -- select your x, y, z appropriately

2. Load your CFD data file

3. Select your CFD data file and then do:

    a. ResampleWithDataset -- select the CFD data as the Input and the TableToPoints as the Source

4. This will give you the data interpolated to the points in the table, if you want a plane from this data, apply the Delaunay2D filter and select Best-Fitting Plane (if the points are not X-Y points only).

That should give you the data from your CFD simulation on the same plane as your PIV data.

I see you are at GT -- let me know if you are still unclear on these steps and would like to talk in person.

Tim

________________________________
From: ParaView <paraview-bounces at paraview.org<mailto:paraview-bounces at paraview.org>> on behalf of Mike Tree <treem22 at gatech.edu<mailto:treem22 at gatech.edu>>
Sent: Tuesday, November 17, 2015 5:32 PM
To: paraview at paraview.org<mailto:paraview at paraview.org>
Subject: [Paraview] Experimental Data Interpolate/Extrapolate to Custom Volume Mesh

I have three planes of experimental PIV data with velocity point data in .dat format loaded in Paraview. I am hoping to use these planes to interpolate/extrapolate velocity values onto a custom volumetric mesh (i.e. from a CFD simulation) to compare experimental results with simulation results in hopes of validating the simulation. I have attached my Paraview state file. Any help will be much appreciated.

Thanks,
-MT

--
Mike Tree, PhD Student
Cardiovascular Fluid Mechanics Laboratory
Georgia Institute of Technology
Atlanta, GA
treem22 at gatech.edu<mailto:treem22 at gatech.edu>
678-249-0922<tel:678-249-0922>
[https://mailfoogae.appspot.com/t?sender=adHJlZW0yMkBnbWFpbC5jb20%3D&type=zerocontent&guid=82092feb-f151-4b75-8a88-41da2f2b684b]?

--
Mike Tree, PhD Candidate
Cardiovascular Fluid Mechanics Laboratory
Georgia Institute of Technology
Atlanta, GA
treem22 at gatech.edu<mailto:treem22 at gatech.edu>
678-249-0922<tel:(678)%20249-0922>

--
Mike Tree, PhD Candidate
Cardiovascular Fluid Mechanics Laboratory
Georgia Institute of Technology
Atlanta, GA
treem22 at gatech.edu<mailto:treem22 at gatech.edu>
678-249-0922
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20161201/d76b00b4/attachment.html>