[vavnoon]

Hello everyone,

I am trying to make a simulation of multiphase flow in a rotating cylinder using VOF and mesh motion.

The cylinder has 6 cm height and is filled with water to a height of 4 cm. Actually, there is a free surface.

I activated the VOF setting and marked level set and implicit body force.

I chose air and water for the primary and secondary phases, respectively.
I gave 0.072 to the surface tension coefficient for the interaction of the phases.

B.C.s are considered moving walls (of course no-slip) with zero relative velocity.

I have initialized the case with zero level set function and zero phase-two volume fraction.
Then I marked the lower two-thirds of the cylinder as a region.
So I patched it by selecting the region, phase-two, and volume fraction with value one.

Afterward, I plotted a contour of the volume fraction of the phase-two at a midplane initially.

It was expected to be one and zero in the lower and upper parts, respectively. But it was not so. Why?

Could anyone please guide me in this regard?
Thank you so much

[vinerm]

Level set is a distance function and needs to be define like volume fraction of one of the phases. If you initialize it with 0 everywhere, then only one phase would exist. It needs to be defined as a non-zero value at the interface, just like a dirac-delta function.

Disable LSM and then try with the initialization and patch.

[vavnoon]

Dear Vinerm,

I greatly appreciate your concern.

I replaced Cylinder with Hex in the categories of shapes in the region box. Then it worked well. I thought because the geometry is a cylinder, I have to mark a cylindrical region.

I tried it two times, with and without LSM, and I got the same result. Why?

[vinerm]

What kind of results do you get?

[vavnoon]

For example volume fraction contour.

[vinerm]

Let me rephrase the query

What kind of volume fraction do you get? And why do you think that the outcome is incorrect?