Residuals

A subforum specific to the development of the OpenFoam-based workbenches ( Cfd https://github.com/qingfengxia/Cfd and CfdOF https://github.com/jaheyns/CfdOF )

Moderator: oliveroxtoby

Forum rules
and Helpful information for the FEM forum
IMPORTANT: Please click here and read this first, before posting on this CfdOF fourm
Post Reply
Sidemountyucatan
Posts: 114
Joined: Wed Apr 17, 2019 2:08 pm

Residuals

Post by Sidemountyucatan »

OS: Windows 10 (10.0)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.23074 (Git)
Build type: Release
Branch: master
Hash: 74b866ed3022fd1452c534dae31a721bcc668f2f
Python version: 3.6.8
Qt version: 5.12.1
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: German/Switzerland (de_CH)

Can somebody tell me the background why the residuals behave as to be seen in the attached picture ? All residuals make a smooth descend but just energy dissipation rate/frequency has data fluctuations ?

I have also attached the model. (.... maybe any hints for improvement ?)

Thx
Attachments
Air flow over one tank_flow characteristic.FCStd
(34.04 KiB) Downloaded 44 times
Residuals 07.png
Residuals 07.png (82.25 KiB) Viewed 1363 times
Top
thschrader
Veteran
Posts: 3157
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Residuals

Post by thschrader »

What is it? An empty wardrobe... :)
wardrobe.JPG
wardrobe.JPG (24.37 KiB) Viewed 1290 times
Top
Sidemountyucatan
Posts: 114
Joined: Wed Apr 17, 2019 2:08 pm

Re: Residuals

Post by Sidemountyucatan »

Hi, yes I know I could have shorten the air flow space. The topic is to see how flow behaves in the trough in the middle. So despite this little clumsy figure could you give me some hints for improvement ? .... If I shorten the air space do I have to define the new face on the very top as wall with slip ? Because a wall with no slip would be obviousely wrong I guess.
Thx
Top
thschrader
Veteran
Posts: 3157
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Residuals

Post by thschrader »

Sidemountyucatan wrote: ↑Thu Dec 17, 2020 9:07 pm ... If I shorten the air space do I have to define the new face on the very top as wall with slip ?
...
Thats what I did. The influence of the top wall is nearly zero, see velocity-arrows.
I reduced the height of the model and used a little different mesh.
pressure at inlet/outlet = 0 (why do you define a pressure at the outlet?).
But: at iteration 150 I get the (first?) "omega-bump" like you. Mmmh...
However, runtime to reach 200 iterations was nearly 2 hours, I stopped the solver.
But the flow-field looks like what I would expect. Velocity at bottom nearly zero, a dust-trap.

At v-inlet = 0,2 m/s and box-length 1,2 m I get a Re=16000 for air. Thats not much.
Maybe a laminar approach does the work...
tank_flow_TS.FCStd
(37.3 KiB) Downloaded 77 times
box.JPG
box.JPG (91.39 KiB) Viewed 1225 times
Top
Sidemountyucatan
Posts: 114
Joined: Wed Apr 17, 2019 2:08 pm

Re: Residuals

Post by Sidemountyucatan »

Great ! Thx. I put the pressure to certain value because pressure 0 seemed to me physicaly wrong. Doesn't it imply vacuum ? Or is it kind of referenced pressure ?
Top
thschrader
Veteran
Posts: 3157
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Residuals

Post by thschrader »

Sidemountyucatan wrote: ↑Sat Dec 19, 2020 6:50 pm Great ! Thx. I put the pressure to certain value because pressure 0 seemed to me physicaly wrong. Doesn't it imply vacuum ? Or is it kind of referenced pressure ?
Set all inlet/outlet to pressure =0.
Example: airfoil
You will get negative pressure at certain zones. But this is physically impossible.
Solution: it is the pressure difference to the outside world.
Top
Post Reply

Return to “CfdOF / CFD”