Implemented ccx's CENTRIF label from *DLOAD card
Moderator: bernd
Forum rules
and Helpful information for the FEM forum
and Helpful information for the FEM forum
Implemented ccx's CENTRIF label from *DLOAD card
Just implemented centrigual body loads.
Concerning CENTRIF label from ccx's *DLOAD card.
Here we have a not very glamorous example.
Two rotating steel rings at 100 rev per sec.
one is made of steel density 8.0 kg/dm^3, the other is made of Scotty's steel
(same properties as usual but only half density)
So we have a analytic result to compare ccx' results against.
Concerning CENTRIF label from ccx's *DLOAD card.
Here we have a not very glamorous example.
Two rotating steel rings at 100 rev per sec.
one is made of steel density 8.0 kg/dm^3, the other is made of Scotty's steel
(same properties as usual but only half density)
So we have a analytic result to compare ccx' results against.
Re: Implemented ccx's CENTRIF label from *DLOAD card
- download the file from this post
- purge results
- run it
- I get same results
BUT
- clear the mesh
- remesh it without group meshing (which is default FreeCAD pref)
- ccx gives an error ...
- does the mesh have to have group data for the centrif constraint to run proper?
- purge results
- run it
- I get same results
BUT
- clear the mesh
- remesh it without group meshing (which is default FreeCAD pref)
- ccx gives an error ...
- does the mesh have to have group data for the centrif constraint to run proper?
Re: Implemented ccx's CENTRIF label from *DLOAD card
BTW: I set up the example in Python for a unit test. See https://github.com/berndhahnebach/FreeC ... 39e41966bd But it only runs with your Mesh and of course your object names (to match the group names)
If the mesh has no groups ccx gives an error.
Does the example work for you without mesh groups?
If the mesh has no groups ccx gives an error.
Does the example work for you without mesh groups?
Re: Implemented ccx's CENTRIF label from *DLOAD card
It's only implemented for group meshing.
Because it's greyed out and checked, I thought that's the way to go.
Should be faster than sorting out nodes by python, isn't it?
Because it's greyed out and checked, I thought that's the way to go.
Should be faster than sorting out nodes by python, isn't it?
Re: Implemented ccx's CENTRIF label from *DLOAD card
it is much much faster than sorting out. That was the reason I did implemented this. But has problems in some circumstances. Thus standard setting is off. Because of the problems I only recommend this for power users like you. But we should make work analysis with this setting to off. All unit tests uses meshes without groups.
The problem is not the use of the groups but the meshing. The problems have happened on the meshing with groups and than the analysis did not run at all.
Re: Implemented ccx's CENTRIF label from *DLOAD card
https://github.com/berndhahnebach/FreeC ... femcentrif
How about this? Should work with and without group data. Would you give it a try?
cheers bernd
How about this? Should work with and without group data. Would you give it a try?
cheers bernd
Re: Implemented ccx's CENTRIF label from *DLOAD card
Once more I did some heavy history rewriting and squashing. We do have a clean history for this PR ATM. Thus very good documentation for adding mre constraints.
If you would give your ok I would go for a merge.
If you would give your ok I would go for a merge.
Re: Implemented ccx's CENTRIF label from *DLOAD card
Added a unit test. The unit test is before all the gui tools and code. The unit test does not need all the gui tools and code.
cheers bernd
cheers bernd