Test request - 1D flow

About the development of the FEM module/workbench.

Moderator: bernd

KgoaOT
Posts: 66
Joined: Mon Jul 18, 2016 9:08 am

Re: Test request - 1D flow

Post by KgoaOT »

makkemal wrote:Nice Ofenste it is working well in Windows and Linux :D
You only need a few more fluid section types to support all of them.

I suggest reproducing the following exmaple
http://feacluster.com/CalculiX/ccx_2.11 ... ode13.html
and adding it to the unit tests
An example file similar to the one in http://feacluster.com/CalculiX/ccx_2.11 ... _deck=pipe
Resevoir.fcstd
(18.66 KiB) Downloaded 41 times
The pressure results I am getting are fairly close to the example but the mass flow rate is off. Perhaps the is a unit conversion error?
Assistance will be very much appreciated in resolving the error.

Couldn't yet rebase onto Bernds development branch as I was getting too many conflicts.

The branch can be found in: https://github.com/stoflap/FreeCAD/tree/FluidSection
Resevoir.fcstd
(18.66 KiB) Downloaded 41 times
Attachments
Resevoir.PNG
Resevoir.PNG (240.38 KiB) Viewed 2065 times
KgoaOT
Posts: 66
Joined: Mon Jul 18, 2016 9:08 am

Re: Test request - 1D flow

Post by KgoaOT »

So I just created 2 different files with the same geometry and conditions but applied in different unit systems.

With this one I get similar results to the Calculix manual example in http://feacluster.com/CalculiX/ccx_2.10 ... _deck=pipe
ResevoirExampleCalculixManualUnitSystem.inp.txt
(4.28 KiB) Downloaded 38 times
When I take same input file and convert it to FreeCAD CalculiX system, the pressure is similar to that in the manual but themass flow rate is different.
ResevoirExampleFreeCADUUnitSystem.inp.txt
(4.41 KiB) Downloaded 38 times
I just can't figure out why I am getting this error.
ulrich1a
Veteran
Posts: 1957
Joined: Sun Jul 07, 2013 12:08 pm

Re: Test request - 1D flow

Post by ulrich1a »

KgoaOT wrote:I just can't figure out why I am getting this error.
Could it be, that the Manning coefficient also needs a conversion?
The ccx-manual says, it has the unit time/length^(1/3). By dividing the Manning coefficient in the FreeCAD-version with 10, I am getting the same result with the Calculix version.

In the posted file of the Calculix-version, node 25 was not converted right. But that did not influence the result.

Ulrich
KgoaOT
Posts: 66
Joined: Mon Jul 18, 2016 9:08 am

Re: Test request - 1D flow

Post by KgoaOT »

ulrich1a wrote:
KgoaOT wrote:I just can't figure out why I am getting this error.
Could it be, that the Manning coefficient also needs a conversion?
The ccx-manual says, it has the unit time/length^(1/3). By dividing the Manning coefficient in the FreeCAD-version with 10, I am getting the same result with the Calculix version.

In the posted file of the Calculix-version, node 25 was not converted right. But that did not influence the result.

Ulrich
Thanks @Ulrich. For some reason I assumed the manning coefficient was just a factor.

The latest branch can be found here, rebased onto the latest master:
https://github.com/stoflap/FreeCAD/tree/FluidSection

The resevoir example file can be found here:
Resevoir.fcstd
(19.01 KiB) Downloaded 37 times
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Test request - 1D flow

Post by bernd »

User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Test request - 1D flow

Post by bernd »

Have some question in this regard. In FEM we have the following object types:
- analysis container object
- constraints objects
- materials objects
- meshes objects
- results objects
- solver objects

and three more objects:
- shellthickness
- beamsection
- fluidsection (soon)
- How would you guys call the object type for them?
KgoaOT
Posts: 66
Joined: Mon Jul 18, 2016 9:08 am

Re: Test request - 1D flow

Post by KgoaOT »

bernd wrote:Have some question in this regard. In FEM we have the following object types:
- analysis container object
- constraints objects
- materials objects
- meshes objects
- results objects
- solver objects

and three more objects:
- shellthickness
- beamsection
- fluidsection (soon)
- How would you guys call the object type for them?
Perhaps a mesh section type?? Its long but atleast it describes the object
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Test request - 1D flow

Post by bernd »

rebased the first commit of your branch on up to date master. Got rid of the line endings problem and made some changes to adapt it to the new module ObjectsFEM.py. By accident I rebased it in your commit so may use it as it is.

branch:
https://github.com/berndhahnebach/FreeC ... fem1dfluid

commit
https://github.com/berndhahnebach/FreeC ... it/1437134
KgoaOT
Posts: 66
Joined: Mon Jul 18, 2016 9:08 am

Re: Test request - 1D flow

Post by KgoaOT »

Thanks Bernd.

I have just rebased on the branch and then on the latest master.

The branch can be found here: https://github.com/stoflap/FreeCAD/tree/FluidSection
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Test request - 1D flow

Post by bernd »

KgoaOT wrote:Thanks Bernd.

I have just rebased on the branch and then on the latest master.

The branch can be found here: https://github.com/stoflap/FreeCAD/tree/FluidSection
Great! I found something ...
https://github.com/stoflap/FreeCAD/comm ... 9cc0f0L847

would you post your example files udated with this branch
Post Reply