[jing113cn]

Hi,

I am doing a two phase flow simulation by VOF model as in the attachment. It is a transient simulation with two phases, air and water. The cavity is intially filled with air, and then the water enters into the cavity from the bottom.

My problem is, why the water is flowing along the walls in the simulation? In the experiments the water is straight. I am using a smooth wall boudnary condition with zero shear stress.

Thanks for any help!

Best,
Jane

[Saeid.doosty1984]

Hello how can a pours media moving rotation .

[Saeid.doosty1984]

Hello how can porous media rotation .

[michal.s]

Hello there,
I have quite similar project and facing the same issue.
Somebody can explain it?

[sfn]

you may try
- refine the mesh
- change timestep
- check settings of VOF (for example switch between sharp/dispersed interface)
- use different turbulence models

[michal.s]

My mesh is refined and good quality (structured and the domain isn't complicated)
Calculations are steady state, however the transient case shows the same behaviour,
I've already tried different settings
Turbulence model k-W SST
Don't know what could I change, with Eurelian model the shape of the water stream looks good.

Isn't it like that because of the VOF model calculates one velocity profile for both phases? BTW, mixture model shows the same behaviour as VOF

[usamaperwez]

i think so it depends on contact angle of water with wall.
Have you defined that?
I am also working on VOF and trying to refine my model.

[michal.s]

I am wondering which values should I type to be sure it's correct.
My simulation is the flow around a sphere (water jet cooling its surface).
Any suggestions?

[BlnPhoenix]

Originally Posted by michal.s

I am wondering which values should I type to be sure it's correct.
My simulation is the flow around a sphere (water jet cooling its surface).
Any suggestions?

Have you defined correct interfacial tension? Contact angle depends on the material of the solid. You can read it up the contact angle of water against different materials in the literature to be sure it's correct..

[michal.s]

In Models>VOF>Phase Interactions I have typed constant value of Surface Tension Coefficient as 0.072358 n/m and tried turning on option "wall adhesion". Continuum Surface Force selected (as default).

My domain is only the fluid area so actually the sphere was excluded and set as wall boundary condition (perhaps that is the reason?).
I've just set the roughness height as 0.03mm, roughness constant left default (0.5).
I can change the contact angle (in wall BC options) between the water and the air. I've tried default value of 90deg and also 50deg. Still the same result

[BlnPhoenix]

Can you maybe post some pics of your domain and mesh?
Also from the first user it looks like the wall is curved, also the geometry and the experimental picture don't look the same, so i supsect there was maybe an error in the geomerical setup altogether..

[michal.s]

The domain and mesh:
https://ibb.co/bXP2Gw
https://ibb.co/h5yjUG

Sphere diameter - 20mm
Mesh around the sphere prepared with the tool o-grid in ICEM

[BlnPhoenix]

Ok, very fine mesh, nice.
Can you show also picture of your problem with the water phase around the sphere? Do you expect the water splashing or something like that?

[michal.s]

I don't have my workstation with me, here is "quick" sketch of the issue.
Left side - what I expect, right - what is happening.
https://ibb.co/kLgXGw

Its high-pressure water jet so I expect the water to hit the surface and then flow in every possible direction

[BlnPhoenix]

Did you set the blue axis to symmetry?

Edit: From your sketch it look's like the flow attaches to the sphere, which would indicate low Reynolds number. You said its high pressure, but maybe there is something wrong with dimensions of the geometry.

[michal.s]

Yes, i've set the case as axisymmetric and the blue line is the axis BC.
Single phase flow looks fine, the problem is just that flow in VOF

[michal.s]

Hello again!
I've been working with my VOF case again. After changing turbulence model from k-w SST to k-e enhanced wall treatment the water started to flow in realistic way.
With RSM turbulence model the situation was the same as with k-w SST.

What could be the reason?
Wrong y+, sth with the mesh or maybe k-e EWT performs some unique features?

[usamaperwez]

What is your problem.
You are giving inlet to a fluid and it is striking sphere (artificial object)?

What is the courant number of your simulation?

[michal.s]

Yes, a stream of water flows from the inlet (mass flow/inlet BC with velocity around 1-3m/s - depends on the case being investigated).
It's axisymetric simulation (sphere and the water jet along the axis) with gravity turned on.
A stream should hit the sphere surface and change the flow direction (I have attached a pictures before).
I've left default value of courant number as 200, because the convergence was OK. I decided to use pressure based coupled solver.

With k-epsilon after switching the case from transient to steady state, the flow is again not realistic

My previous post:

Originally Posted by michal.s

I don't have my workstation with me, here is "quick" sketch of the issue.
Left side - what I expect, right - what is happening.
https://ibb.co/kLgXGw

Its high-pressure water jet so I expect the water to hit the surface and then flow in every possible direction

[usamaperwez]

Courant number should be less than 2 for VOF model to be successful.
Kindly try to solve it with 2 and you will see the change of results.