Learning Forum

[tatimaes]

Hello there, I am making a new post from my last question but it is more or less related.

I'm running VOF+Open channel submodel (free surface) / BC: Inlet/outlet: pressure in/out (I assigned values of free surface AND bottom levels in/out in BC setting) --- Top domain: pressure outlet --- wall: no-slip conditions

As you can see in the image, there are some arrows in the opposite direction of the flow at the inlet. I am not sure if there is reverse flow or actually, water is not entering the domain at this point? My solution is converging properly, but there is a mass imbalance shown in the fluxes report.

I already tried to extend the length upstream but it go worse. (Also I've been playing a lot with the mesh because of the limitations for student license).

I found the next comment somewhere, does it actually happen? if so how any recommendations on how to fix it?

[Rob]

Assuming flow goes right to left and gravity is into the page. Are the reverse flow vectors in the gas or liquid phase? VOF tends to be transient, so as well as checking the boundary flux check the mass of liquid in the domain. Assuming everything is converging well, in - out = accumulated mass .

[tatimaes]

They are shown in the iso-surface, here a plane view from the inlet: (for this one flow direction is from left to right)

[Rob]

OK. Other than the mesh looking a bit coarse the flow isn't too bad. Have a more careful look at what the air is doing, can you see any reason for the disturbed air flow?

[tatimaes]

Ok, so as in the image below these are vectors for the volume fraction of air. So water is not coming out, thanks for helping to check that! Now then no understanding of the mass imbalance.
I don't understand the disturbance of air. the only gravitational force I'm applying is y=-9.81. Then everything by default.
For being a transient simulation, there is the case that maybe checking from flux mass flow rate this is not accurate? I've been reading some posts, one of them says surface mass flow report can be tricky (?).
Which other way can I make the mass flow as a monitor to check this condition? or should I just visually identify the flow time in the mass flow plot when it shows a similar behavior trend and compute mass balance from there?
Thanks!

[Rob]

If you're using open channel boundaries you will have set something so flow is moving. Use the surface mass fluxes, but also the volume reports in Fluent to monitor the water mass: you want 2-3 reports to do this. Run the model on and see what happens.

[tatimaes]

Thank you so much. I will do that right now! to see

[tatimaes]

Hello there,
After I run my simulation again and decrease the residuals in one magnitude to help mass balance (it reduced the mass imbalance I've got previously) as is suggested in the user's guide, I got these graphs, Can I say I reached a "steady state" in my transient simulation? How can I judge it from the monitors, is it right to follow the trend from where the pattern seems to go quite similar?


Thank you so much!
Have a nice day

[Rob]

It's fairly steady. Your options are to use the data sampling, or just take a data set and decide it's good enough. Remember the real units are made from poured concrete so you're modelling something where the geometry tolerance might be 1" or so.

[tatimaes]

Thank you, Dear Rob,
Sorry I got lost about the units and follow part?

[Rob]

The real device that you're modelling. From the size it's a civil engineering project and if so, the "as built" dimensions compared to the "as drawn" tend to be a little further apart than (for example) the aircraft industry.