[DrZee]

I am trying to simulate a gear churning into oil in a gear box. I set up a simulation using Fluent sliding mesh. So I constructed an interface between the "inner fluid zone" and "outer fluid zone". The initial condition is that the gear is partly immersed in the oil. Then the gear rotates with a speed of 1000 rpm. The pitch circle speed is around 4.7 m/s and the minimum size of the cell is around 0.1 mm. Therefore I set up the time step to be around 1 E-5.

After a few time steps, the 2 phases look like the picture. The oil surface in the "inner fluid zone" seems to rotate with the region. I though while the inner region is rotating, the two-phase surface should be stable, especially for those far from the gear. This results look not physical in my opinion.

Capture.JPG

So, is this kind of result typical for two-phase sliding mesh simulation? Is it just an unstable transient solution that I should wait till the results stabilized?

P.S. this is not similar with what I got from dynamic mesh or Star-CCM+ overset mesh methods (which seems physical to me).

Thank you for your help!!

[Karl]

For me it seems you have a wrong setup for the rotating gear. Could you post your settings. Normally there is no problem using the Vof-model together with a sliding mesh.

Best Regards

Tip: Using the (rpsetvar 'patch/vof? #t) command in the text-console before patching the heavier phase in the gearbox will give you a smoother interface (right side on your picture).

[DrZee]

Originally Posted by Karl

For me it seems you have a wrong setup for the rotating gear. Could you post your settings. Normally there is no problem using the Vof-model together with a sliding mesh.

Best Regards

Tip: Using the (rpsetvar 'patch/vof? #t) command in the text-console before patching the heavier phase in the gearbox will give you a smoother interface (right side on your picture).

Dear Karl, here are the settings. I set up an inner rotating fluid region around the gear. And the gear is set to "stationary wall" relative to adjacent cell zone. Could this be the problem? Thanks! Also thanks a lot for your tip!

1.png
2.png
3.png
4.jpg
5.png

[DrZee]

Originally Posted by Karl

For me it seems you have a wrong setup for the rotating gear. Could you post your settings. Normally there is no problem using the Vof-model together with a sliding mesh.

Best Regards

Tip: Using the (rpsetvar 'patch/vof? #t) command in the text-console before patching the heavier phase in the gearbox will give you a smoother interface (right side on your picture).

Continue...
9.png

10.png

11.jpg

12.png

13.png

[Karl]

No, the settings for the gear wall are ok. Do you have a conformal or non-conformal interface zone ? Which solver settings are you using ?
You should include the "Implicit Body Force" option in your VoF-Settings.

Best regards,
Karl

[DrZee]

Originally Posted by Karl

No, the settings for the gear wall are ok. Do you have a conformal or non-conformal interface zone ? Which solver settings are you using ?
You should include the "Implicit Body Force" option in your VoF-Settings.

Best regards,
Karl

I built a conformal mesh initially. During the gear rotation, the mesh interface needs interpolation between two sides, so it should be a non-conformal interface.

Solvers I used are:
PISO for pressure-velocity coupling
Gradient: Least squares cell based
Pressure: PRESTO!
Momentum: Second order upwind
Volume Fraction: Geo-Reconstruct
TKE and TDR: Second order upwind
Transient: First Order Implicit

Is there anything inappropriate?

I guess my error may come from the fact that I did not choose "implicit body force" and did not change the VOF solver from default to "Solve vof every iteration". I am trying with these options selected. Thanks for pointing this out!

[DrZee]

Can anybody help? The problem still remains...

[SJSW]

Use VOF and implicit force .
Check "Sharp" in "Interface Modeling Type".
I think it will help the sharpness of the interface.