Learning Forum

[scabo]

Hi

I am trying to simulate air-water flow in a straight pipe with periodic left and right boundaries with Volume of Fluid. After making the boundaries periodic, there is no option to specify mass flow rate, only pressure gradient option is there. I want to specify the mass flow rate at the inlet. How will I initialise the problem? Will it be through normal Initialisation? I am slightly confused-any help is greatly appreciated..thanks!

AB

[Rob]

Have a look in the documentation - there's a section on limitations. 

[scabo]

Hi documentation means user guide right which is available online?

 

[scabo]

 

https://www.#######/project/neptunius/docs/fluent/html/ug/node252.htm

this one??

 

[Rob]

Not the third party and very out of date one, no. If you click on Help in the solver you'll get the current set. Or  Public Ansys Help tends to work equally well. 

[scabo]

This is the one https://ansyshelp.ansys.com/public/account/secured?returnurl=//Views/Secured/corp/v242/en/flu_ug/flu_ug_sec_periodic.html?q=periodic%20boundary%20conditions%20fluent

Here it is written multiphase flow can be modeled, that means i can use VOF models also right?

[Rob]

Correct, but read that line carefully relative to your original question. 

[scabo]

Yes i can only use a pressure gradient with VoF. Then will i just hybrid initialise? Because in standard initialisation we have to provide velocities which are not known on the periodic face..

[Rob]

No, you need to use standard. Then patch the liquid. Do NOT patch a non 1 or 0 volume fraction of phase: read the VOF model approach to see why. 

[scabo]

Hi, What is meant by your line: Do NOT patch a non 1 or 0 volume fraction of phase. Does that mean i should only patch a surface with volume fraction of either 1 or 0?

[Rob]

Cell zone, but yes VOF needs to be 1 or 0 as it's an interface tracking model. Eulerian/Mixture are fine with values between zero and one depending on other factors. 

[scabo]

Hi, I have checked, after standard initialisation i have applied patch and selected volume fraction>air=0 on the periodic face. It is correct right?

[scabo]

Also, i have left the standard initialise panel as it is without giving any X, Y velocities-just initialised and applied the patch..i hope it is fine?

[Rob]

You need to patch into the cell (volume) zone. 

[scabo]

I created a region through cell register and then patched water into it after initialisation. Is that correct?

[Rob]

Yes, assuming the patched region gives you the correct liquid volume in the domain. 

[scabo]

Thanks-I will solve and see what happens

[scabo]

But one thing is there: i have specified pressure gradient on the periodic faces and then created a 3D region near the periodic face and patched it with water. So the pressure gradient is not acting on the water right? but instead it is acting on the inlet face. but still i am getting movement of water in the results.

[Rob]

I'd patch to a set depth rather than from a boundary. Pressure drop is for the domain, it's not phase related. 

[scabo]

What is meant by patch to a set depth? did you mean just patch anywhere in the pipe with a certain depth? I have attached my patched region as a picture. The patch is near the periodic face and is a 3D box type. Is it fine?

[Rob]

It'll work, but take time to find a peridic flow level. I'd generally patch so the liquid is in about the right place - ie start as near to the expected solution as possible. 

[scabo]

okay thanks

[scabo]

I was just asking after standard initialisation i do not need to provide any X,Y velocities right? just patch and apply a pressure gradient on the periodic face..

[Rob]

You don't. However, the better the initialisation data the more stable the solution. For a transient that also means a faster overall solution as it takes less time to reach equilibrium. 

[scabo]

Hi

I successfully applied Periodic BC in Volume of Fluid in Fluent, but the the pressure gradient is driving the air phase to a veryy high speed and therefore there are some waves on the water surface. Is there any way to get rid of this; like applying the pressure gradient only to the water phase and will this type of fix actually work? My initial water patch is just a flat bed of water of some depth inside the length of the cylinder shown above in previous replies.

thanks 

[Rob]

Doesn't look like it. However, if there are no waves why do you need to model the free surface? 

[scabo]

I want a stable free surface because i wanted to visualise the secondary currents in the water region in an air-water system. Secondary currents are cross-section phenomenon which are correctly visualised when we have a stable free surface. When i modeled a full pipe of very long length with VoF modeling with normal velocity inlet of water i am getting correct results but needed to save computational time so planned to use a periodic BC in VoF.

[scabo]

Can i use some UDF to apply the pressure gradient only in the water region? That may solve the problem? or something like that

[Rob]

Not easily. The periodic condition is old, and was made to allow repeated sections to be modelled cheaply. It works for VOF simply because nothing causes it to break. 

[scabo]

thanks-"but what do you mean by nothing causes it to break"

[Rob]

It means just that. VOF works with periodics because it doesn't break any of the solver routines. There aren't any extra functions for VOF in periodic flows. 

[scabo]

Okay

[scabo]

So you are saying that there are some limitations of using Periodic BC in VoF models?

[Rob]

Only what's covered in the manual. If you want a different pressure drop for each phase that implies a different flow rate, and is that then periodic? 

[scabo]

Yes because air is resting above water in a cylindrical pipe half filled with water initially Please refer to the attached picture. SO if different pressure gradient is applied in air and water section one lies top of another so it will remain periodic in the streamwise direction. Is that right?

[Rob]

Not necessarily as I suspect it'll try and set up a standing wave. Have a look at the Baker Chart for flow regimes. 

[scabo]

Yes it is creating a wave because air vel is higher. Can i give the pressure gradient in proportion to the volume fraction somehow? 

[Rob]

I don't think so.