[Paraview] Native OpenFOAM reader for ParaView 5.3.0 has some weird visualization issues.

Wed Apr 19 19:01:26 EDT 2017

Hi Tom and Utkarsh,

I finally had time to take a closer look.
There is a corresponding merge-request https://gitlab.kitware.com/vtk/vtk/merge_requests/2716

Cheers,
/mark

________________________________________
From: Tom Fahner <tom.fahner at gmail.com>
Sent: Monday, April 3, 2017 3:24:19 PM
To: Utkarsh Ayachit
Cc: Mark Olesen; ParaView
Subject: Re: [Paraview] Native OpenFOAM reader for ParaView 5.3.0 has some weird visualization issues.

Hi Utkarsh,

I finally had some time to look into this issue a bit further. I have a testcase for you of an Ahmed body that also shows the weird missing of the data.

The case with some intermediate result (approx 400 MB) can be found here:
https://we.tl/ZZwy4ZGL1K

The case without that result (approx 125 MB) can be found here:
https://we.tl/3dSqgyd4zG

Hope this helps in finding the issue. Please let me know if you need more information.

Best Regards,
Tom

2017-03-16 13:25 GMT+01:00 Tom Fahner <tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>>:
Hi,

Let me check if that is possible, I do not think I can share this particular one. I will come back to you.

Tom

2017-03-16 13:12 GMT+01:00 Utkarsh Ayachit <utkarsh.ayachit at kitware.com<mailto:utkarsh.ayachit at kitware.com>>:
Tom,

Do you have a dataset to reproduce this issue? There were changes in
the reader in this past release, so it's conceivable that it
introduced a bug.

Utkarsh

On Thu, Mar 16, 2017 at 8:08 AM, Mark Olesen <Mark.Olesen at esi-group.com<mailto:Mark.Olesen at esi-group.com>> wrote:
> Hi Tom,
>
> So if foamToVTK works, it sounds like a reader bug. Maybe somewhere where the prism order is getting swapped - or something else where the reader's cell modeller is getting into trouble.
> It may not work, but if you try to export from paraview to VTK format (ie, use the OpenFOAM reader for reading). Then with some effort it might be able to identify which cells are different. Presumably the cell ordering will be preserved in both cases, but still not 100% trivial. Either way, we need to figure out which cell types are causing the issue, or an off-by-one of whatever.
>
> /mark
> ________________________________________
> From: Tom Fahner <tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>>
> Sent: Thursday, March 16, 2017 12:47:05 PM
> To: Mark Olesen
> Cc: ParaView
> Subject: Re: [Paraview] Native OpenFOAM reader for ParaView 5.3.0 has some weird visualization issues.
>
> Hi Mark,
>
> I have no polyhedral cells (hybrid tetrahedron/prism mesh), so I guess changing "decompose polyhedral on/off" does not make sense. In fact the smaller mesh that I tested with before did show the correct visualization and that mesh had polyhedral/hexagonal cells. It did not matter whether I used decompose polyhedral on/off. I also have experienced the slicing issues from time to time with various paraview versions with polyhedral meshes, but the effect was typically much less severe than in the current situation.
>
> Using foamToVTK (with and without the -poly option) does indeed work and provides the correct visualization. So yes this can be a workaround for now, but I than rather go back to 5.1.2.
>
> Regards,
> Tom
>
> 2017-03-16 11:44 GMT+01:00 Mark Olesen <Mark.Olesen at esi-group.com<mailto:Mark.Olesen at esi-group.com><mailto:Mark.Olesen at esi-group.com<mailto:Mark.Olesen at esi-group.com>>>:
> Hi Tom,
>
> Sorry, no answers - only questions/ideas:
>
> Is decompose polyhedral on or off? How do things look using the other setting?
> If the number of cells corresponds to checkMesh - then you either have no polyhedra, or decompose polyhedral is off.
> If you use foamToVTK (with and without -poly) and then read the VTK file back in, how does that compare?
> Using foamToEnsight could be yet another check, but without the option to decompose polyhedra.
>
> In most paraview versions I've had issues with slicing through polyhedron (eg, from the motorBike tutorial).
> The symptoms are either holes in the visual (like you have) or crashing the program. I haven't checked if this is better/worse in 5.3.0.
>
> Cheers,
> /mark
> ________________________________________
> From: ParaView <paraview-bounces at paraview.org<mailto:paraview-bounces at paraview.org><mailto:paraview-bounces at paraview.org<mailto:paraview-bounces at paraview.org>>> on behalf of Tom Fahner <tom.fahner at gmail.com<mailto:tom.fahner at gmail.com><mailto:tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>>>
> Sent: Thursday, March 16, 2017 9:39:49 AM
> To: ParaView
> Subject: [Paraview] Native OpenFOAM reader for ParaView 5.3.0 has some weird visualization issues.
>
> Dear all,
>
> Please find attached two images that show my OpenFOAM model in ParaView 5.3.0. It was the standard Binary Installers package downloaded from the website for 64 bit linux (OpenSUSE Leap 42.1).
>
> The correct number of cells is reported in the information tab (when compared to OpenFOAM's checkMesh result). The patches are
> all rendered nicely, but it looks like some cells are missing in the domain/volume.
>
> Is there some setting that I need to apply?
>
> Please note that the mesh has about 25.8 million cells, for smaller meshes this was not a problem. A colleague had similar issues.
> ParaView version 5.1.2 did not have this problem, ParaView 5.2 was not able to open the OpenFOAM files at all.
>
> Best regards,
> Tom
>
> --
> T.C. Fahner
> e: tom.fahner at gmail.com<mailto:tom.fahner at gmail.com><mailto:tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>><mailto:tom.fahner at gmail.com<mailto:tom.fahner at gmail.com><mailto:tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>>>
>
>
>
>
> --
> T.C. Fahner
> e: tom.fahner at gmail.com<mailto:tom.fahner at gmail.com><mailto:tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>>
> t: +31-6-52642814<tel:%2B31-6-52642814>
> a: Groene Woud 48
>     4834 BC Delft
>     Netherlands
> _______________________________________________
> Powered by www.kitware.com<http://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

--
T.C. Fahner
e: tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>
t: +31-6-52642814<tel:+31%206%2052642814>
a: Groene Woud 48
    4834 BC Delft
    Netherlands

--
T.C. Fahner
e: tom.fahner at gmail.com<mailto:tom.fahner at gmail.com>
t: +31-6-52642814
a: Groene Woud 48
    4834 BC Delft
    Netherlands