Learning Forum

[i.tamunodienye2]

Hi,

I am a bit worried about my observation regarding my simulation.

The mixture does not go around the first bend in large mass. This really worries me. See attached.

[Rob]

What is the inlet condition? The result looks relatively sensible, albeit I think you're losing liquid mass. 2d or 3d?

[i.tamunodienye2]

Thanks for your support always. See attached the inlet conditions.






I initialized with "Hybrid Initialization" method and pathed the entire domain as liquid


It is a 3-D simulation.
I have provided the inlet conditions as you requested.
Let me know if you need additional information.

Thanks

[i.tamunodienye2]

Others inlet conditions

[i.tamunodienye2]

And this as well.

[Rob]

VOF?
Check for mass in, mass out & mass in the domain. As the air is compressible are you just squashing that at present? You may want to have a look at a Baker Chart and see what regime you're in.

[i.tamunodienye2]

Yes, VOF

[Rob]

OK, and how's the cell aspect ratio in the domain?

[i.tamunodienye2]

Cell Aspect ratio in ICEM CFD
Min = 7.23
Max = 107
Cell Aspect ratio in Fluent

[Rob]

That'll not help - you really want an aspect ratio in the 2-3 range as the flow is changing along the length of the cells with VOF: ie the free surface is moving axially.

[i.tamunodienye2]

The Baker Chart is for horizontal flow regime case. My case is for a vertical flow regime at the first vertical pipe only. Also my pipe internal diameter is 0.21m

[i.tamunodienye2]

Do you mean aspect ratio should be reduced to 2 - 3 in the meshing process (ICEM CFD).
In essence i should increase the grid size?

[Rob]

Increase the cell count and decrease the cell size.

[i.tamunodienye2]

Ok i will do and let you know.
But the 2-3 aspect ratio value you mentioned, is it for minimum or maximum value. See my previous attachment

[Rob]

It's a general recommendation. Don't get too hung up on the near wall y+ as unless you're also resolving the axial flow the results won't be much use.

[i.tamunodienye2]

Cell count is 964700 hexa already in ICEM CFD. This twice the previous cell count
I did not border about improving the near wall y+. See the new aspect ratio.
Is it reasonable. Computational cost is another issue.
Thanks

[Rob]

The graph shows the range of cell values, what you need to look at is where these occur. If the solution is changing along the length of the cell then you need to have a low aspect ratio.

[i.tamunodienye2]

I have increase the cell counts and decreased the cell sizes, obtaining a lower aspect ratio.
The problem did not change as such.
The two vertical upward flow is the issue. The mixture do fall back too from the video i did.
Is there any solution to this?

Thanks for your help always