Hi,
I am trying to output an EP file from my code. I have both solid elements and surface elements for my case. It is a 3D geometry, but only solved in 2D. Like a flow around a column in a channel.
Just wondering do I need to output all elements? Could I just output solid elements or surface elements for the animation?
Regards,
Cean
EP file need all elements?
-
- Site Admin
- Posts: 4826
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: EP file need all elements?
Hi Cean
You can mask the dofs so that they are only saved there where some field variable is active. For the .ep file defined in the Simulation section use
while for ResultOutputSolve in the Solver section say just
This should work with formats Vtu, Gmsh, and ElmerPost.
Just saving the bulk elements is not posisble, to my knowledge, but the boundary elements are usually not detrimental for performance.
-Peter
You can mask the dofs so that they are only saved there where some field variable is active. For the .ep file defined in the Simulation section use
Code: Select all
Post File Mask Variable = String varname
Code: Select all
Mask Variable = String varname
Just saving the bulk elements is not posisble, to my knowledge, but the boundary elements are usually not detrimental for performance.
-Peter
Re: EP file need all elements?
This is what I want to get.
But it is in VTK format. It only has value for inside solid elements for one time step.
I want to output in a single EP file which includes all time step data to animate the dam break process.
But it is in VTK format. It only has value for inside solid elements for one time step.
I want to output in a single EP file which includes all time step data to animate the dam break process.
- Attachments
-
- vtk.gif
- (44.89 KiB) Not downloaded yet
Re: EP file need all elements?
Hi Peter,
Sorry, I didn't understand what you said.
I have attached my EP file, it includes both the surface elements and inside solid elements. Because the values for
the surface element are all 0, it shows all BLUE. Though there are 3 solid elements that have value of 1, I can't
see through the surface. If I clip, It only shows a hollow space.
I tried to delete the surface elements, only keep the solid elements, but it shows nothing on the screen. "Color
Mesh" only shows the surface mesh. My value is at cell center. Maybe I need to get the value at vertex.
Regards,
Cean
Sorry, I didn't understand what you said.
I have attached my EP file, it includes both the surface elements and inside solid elements. Because the values for
the surface element are all 0, it shows all BLUE. Though there are 3 solid elements that have value of 1, I can't
see through the surface. If I clip, It only shows a hollow space.
I tried to delete the surface elements, only keep the solid elements, but it shows nothing on the screen. "Color
Mesh" only shows the surface mesh. My value is at cell center. Maybe I need to get the value at vertex.
Regards,
Cean
- Attachments
-
- dam.ep
- my ep file
- (5.69 KiB) Downloaded 386 times
-
- Site Admin
- Posts: 4826
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: EP file need all elements?
Hi Cean
Well, it seems that your problem is not to be selective on what you save but rather what you show. If your dof is zero at all boundaries you should use iso surfaces and clipping planes to access the internal features of your computation. Note also that in ElmerPost all dofs are nodal ones i.e. you do not have values defined element-wise. The nodal values are visualized using linear interpolation in between. This is inherent in ElmerPost and not likely to change. In Paraview, for example, you can have both point and cell values allowing discontinuities in the visalization. Is this also your wish? Most equations give continuous solutions (all solutions in standard nodal elements) so the alternative approach has not been that actively pursued.
-Peter
Well, it seems that your problem is not to be selective on what you save but rather what you show. If your dof is zero at all boundaries you should use iso surfaces and clipping planes to access the internal features of your computation. Note also that in ElmerPost all dofs are nodal ones i.e. you do not have values defined element-wise. The nodal values are visualized using linear interpolation in between. This is inherent in ElmerPost and not likely to change. In Paraview, for example, you can have both point and cell values allowing discontinuities in the visalization. Is this also your wish? Most equations give continuous solutions (all solutions in standard nodal elements) so the alternative approach has not been that actively pursued.
-Peter
Re: EP file need all elements?
Ok, I need to get the nodal value from cell centered value.
Thanks.
Thanks.
Re: EP file need all elements?
Hi Peter,
Is it possible for ElmerPost to assign the cell center value to every nodes of the cell? It's not precise, but good for a quick look.
Regards,
Cena
Is it possible for ElmerPost to assign the cell center value to every nodes of the cell? It's not precise, but good for a quick look.
Regards,
Cena
Re: EP file need all elements?
I assigned the cell center value to every n
odes of the cell in my code and finally got the right EP file.