OOFem

About the development of the FEM module/workbench.

Moderator: bernd

User avatar
PrzemoF
Posts: 2627
Joined: Fri Jul 25, 2014 4:52 pm
Contact:

Re: OOFem

Postby PrzemoF » Mon Oct 08, 2018 10:10 pm

HarryvL wrote:
Mon Oct 08, 2018 8:33 pm
bernd wrote:
Mon Oct 08, 2018 8:20 pm
Ahh but your master branch is wrong. Your master branch should have exact the same commit id than FreeCAD master
I updated my fork today as per the following advice: https://github.com/KirstieJane/STEMMRol ... he-browser
Interesting, I didn't know that it's possible! I normally use alias gcm (1) and "git push" with configured ssh key, so I don't need to type login/password.

(1) defined as: alias gcm='git checkout master && git pull'
User avatar
bernd
Posts: 8233
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: OOFem

Postby bernd » Tue Oct 09, 2018 11:05 am

for result reading you could directly read vtk format in a vtk pipeline if oofem does support vtk. The idea is to move to vtk pipeline anyway in FEM. Elmer does it too this way. It has no standard result object anymore.
User avatar
HarryvL
Posts: 1042
Joined: Sat Jan 06, 2018 7:38 pm

Re: OOFem

Postby HarryvL » Tue Oct 09, 2018 11:45 am

HarryvL wrote:
Mon Oct 08, 2018 8:27 pm
Bernd, I am ploughing through the solver framework. In writer.py class FemInputWriterZ88 refers to self.femmesh.Nodes. Can you give me a hint where this attribute gets defined / initiated. Thanks. Harry
Thanks Bernd on the VTK. Could you help me on the above query so that I can understand the structure of element and node numbering in the FC mesh?
User avatar
Kunda1
Posts: 5219
Joined: Thu Jan 05, 2017 9:03 pm

Re: OOFem

Postby Kunda1 » Tue Oct 09, 2018 11:57 am

HarryvL wrote:
Tue Oct 09, 2018 11:45 am
HarryvL wrote:
Mon Oct 08, 2018 8:27 pm
Bernd, I am ploughing through the solver framework. In writer.py class FemInputWriterZ88 refers to self.femmesh.Nodes. Can you give me a hint where this attribute gets defined / initiated. Thanks. Harry
Thanks Bernd on the VTK. Could you help me on the above query so that I can understand the structure of element and node numbering in the FC mesh?
@HareyvL here's a tip: if you are reading through the code on GitHub you can click on a line number or click shift and hold and click again on a greater number to highlight several lines. Then copy the url and paste. Clicking on it will take the user directly to that line or block of code
Want to contribute back to FC? Checkout:
#lowhangingfruit | Use the Source, Luke. | How to Help FreeCAD | How to report FC bugs and features
User avatar
HarryvL
Posts: 1042
Joined: Sat Jan 06, 2018 7:38 pm

Re: OOFem

Postby HarryvL » Tue Oct 09, 2018 12:53 pm

User avatar
HarryvL
Posts: 1042
Joined: Sat Jan 06, 2018 7:38 pm

Re: OOFem

Postby HarryvL » Tue Oct 09, 2018 1:19 pm

bernd wrote:
Tue Oct 09, 2018 11:05 am
for result reading you could directly read vtk format in a vtk pipeline if oofem does support vtk. The idea is to move to vtk pipeline anyway in FEM. Elmer does it too this way. It has no standard result object anymore.
Interesting. In that case, could we perhaps consider running Paraview as a sub-process from within FC? It saves us from building and maintaining the rich feature set that Paraview has.
ickby
Posts: 2914
Joined: Wed Oct 05, 2011 7:36 am

Re: OOFem

Postby ickby » Tue Oct 09, 2018 1:25 pm

HarryvL wrote:
Tue Oct 09, 2018 1:19 pm
Interesting. In that case, could we perhaps consider running Paraview as a sub-process from within FC? It saves us from building and maintaining the rich feature set that Paraview has.
Salome does it like this, and frankly, it sucks :) Back than I looked at this possibility, and decided against it, as I wanted to have everything in a single 3D view, model, mesh and result. This makes working just so much nicer and more useful... And not to talk about the annoyance of having to different gui styles, tree views etc.

And all feature of paraview are not needed, as we easily can export VTK files from freecad if someone needs the additional processing power.
User avatar
bernd
Posts: 8233
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: OOFem

Postby bernd » Tue Oct 09, 2018 2:27 pm

ickby wrote:
Tue Oct 09, 2018 1:25 pm
as we easily can export VTK files from freecad if someone needs the additional processing power.
But we can only export to vtk file format a FreeCAD FEM result ATM. We can not export a vtk pipeline to vtk even this would be not difficult to implement, because it just need to be writen. BTW, stefan would be something very easy to tackle down for you ?!
User avatar
bernd
Posts: 8233
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: OOFem

Postby bernd » Tue Oct 09, 2018 2:36 pm

HarryvL wrote:
Tue Oct 09, 2018 12:53 pm
Thanks @Kunda1

https://github.com/HarryvL/FreeCAD/blob ... ter.py#L67

Code: Select all

self.femnodes_mesh = self.femmesh.Nodes
It is similar to the following:

- start FreeCAD
- open FEM 3D from startWB
- rund following command

Code: Select all

App.ActiveDocument.Box_Mesh.FemMesh.Nodes
Means the .Nodes is a standard FreeCAD FemMesh attribute. We gone save it in the class anyway because if you call App.ActiveDocument.Box_Mesh.FemMesh.Nodes the nodes gets calculated. In the contrary in the code line we put them as class attribute which read them into memory at instanziation. Thus if we need the nodes 10 times we have them instead of calculating them ten times if they would not be a class attribute of the writer. Hope you got the point.

bernd
User avatar
bernd
Posts: 8233
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: OOFem

Postby bernd » Tue Oct 09, 2018 2:37 pm

ickby wrote:
Tue Oct 09, 2018 1:25 pm
Salome does it like this, and frankly, it sucks :) Back than I looked at this possibility, and decided against it, as I wanted to have everything in a single 3D view, model, mesh and result. This makes working just so much nicer and more useful... And not to talk about the annoyance of having to different gui styles, tree views etc.
I with you ickby, I like the FreeCAD way more than the salomone way in this regard.