[Paraview] Fwd: VTK interpolation help

[Ruggiero Guida]

Thanks Ken.

The subdivision filter you suggest does help by making the edges a bit
blurred. Looking at the exact values though I have realized that it is a
more fundamental issue. I am looking into other options like
post-processing the contour with some kind of interpolation.

Ruggiero

On 29 September 2014 20:33, Moreland, Kenneth <kmorel at sandia.gov> wrote:

>   I couldn't see the files you posted to dropbox because they were
> deleted before I got a chance to see them. I also cannot even guess what
> Star-CCM is doing since I haven't seen that either.
>
>  One thing you can try is to apply the loop subdivision filter. That
> filter subdivides triangles in a polygonal mesh and applies some nonlinear
> smoothing to the data. To use this filter, first apply the "Extract
> Surface" filter to your data (to convert it to polygons), and then apply
> "Loop Subdivision."
>
>  -Ken
>
>   From: Ruggiero Guida <ruggiero.guida at gmail.com>
> Date: Saturday, September 27, 2014 at 5:41 AM
> To: Kenneth Moreland <kmorel at sandia.gov>
> Cc: "paraview at paraview.org" <paraview at paraview.org>
> Subject: [EXTERNAL] Re: [Paraview] Fwd: VTK interpolation help
>
>   Thanks Ken.
>
>  Got it. It makes sense. It seems that the only way to improve the
> quality is to have a finer grid and a more detailed calculation of the
> amount of radiation (the previous image included only the direct). New
> images are at this link
>
>  https://www.dropbox.com/s/z9m2nr7ybj5mqqp/finermesh.zip?dl=0
>
>  I was wondering whether there was a way to at least smooth the results
> (see other attached image). I have looked at the filters available, but
> there is no such thing. When I import these results onto StarCCM I
> definitely see a smoother interpolation.
>
>  Thanks
> Ruggiero
>
> On 26 September 2014 19:01, Moreland, Kenneth <kmorel at sandia.gov> wrote:
>
>>   Yes, ParaView will use the triangles to interpolate values between
>> points. The interpolation you are getting is exactly as I would expect
>> based on your data and topology. I'm not sure what you mean by "not
>> [interpolating] in the x,y directions". The triangles are in the x-y plane
>> and interpolation is happening in this x-y plane.
>>
>>  Moving into guesswork, I am imagining that you are expecting a smooth
>>  and fairly straight contour at some distance from the buildings. The
>> reason you don't see this is due to the representation in your data. I
>> notice that your data contains only two values for "RadiationOnPerson." The
>> value is either 0 or 205.7. Presumably there is some contour in space where
>> this field discontinuously transitions from 0 to 205.7, but there is no way
>> to accurately predict where that is from the data you are providing.
>> Instead, ParaView is just linearly interpolating between these values,
>> which is wrong but probably the best you can do with the data provided. The
>> problem is made even worse by the irregular sampling of the field.
>>
>>  -Ken
>>
>>   From: Ruggiero Guida <ruggiero.guida at gmail.com>
>> Date: Thursday, September 25, 2014 at 10:44 PM
>> To: "paraview at paraview.org" <paraview at paraview.org>
>> Subject: [EXTERNAL] [Paraview] Fwd: VTK interpolation help
>>
>>    Dear group,
>>
>>  I hope I will be able to find an answer to my problem here.
>>
>>  I am not sure I understand how the interpolation works. I have
>> generated a VTK file (attached) and I am trying to visualizing it in
>> Paraview. The data represent the solar radiation on the ground around
>> various buildings.
>>
>>  My algorithm calculates the radiation in each node of a constrained
>> delauney triangulation. The plan is to visualize this in paraview to have a
>> nice smooth fillled contour.
>>
>>  The image I get though, does not make sense to me; it seems that
>> paraview is interpolating only within the triangles and not in the x,y
>> directions. Am I missing something?
>>
>>  Thanks a lot for any help
>>  Ruggiero
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20141002/a99bda45/attachment-0001.html>