[openAC]

Hi,

I am performing some simple VOF, transient simulations in ANSYS Fluent v17.2.

The solution data obtained has fields such as: "Velocity magnitude", "Axial velocity", "Radial velocity" etc. (Note: mine is 2D axisymmetric case). My question is, what is this a velocity of? For eg., is the "axial velocity" same as the "axial velocity" of the mixture? If so, then what is its relationship with the individual phase velocities (e.g., gas velocity, liquid velocity) etc.?

As per the theory manual, "A single momentum equation is solved throughout the domain, and the resulting velocity field is shared among the phases."
Reference: https://www.sharcnet.ca/Software/Ans...c_vof_mom.html

After some more reading, it seems that the velocity in the VOF momentum equation could be a volume-averaged field. But I cannot find the exact definition of this field, without which there are too many doubts such as intrinsic or superficial velocity field.

I have to compare the liquid phase velocity and gas phase velocity against experimental data for liquid and gas phases, respectively.

Thanks in advance.

[LuckyTran]

Indeed there is only 1 velocity. When the VOF is 1 then the velocity corresponds to the velocity of phase-1 because there is no phase-2. And when the VOF is 0 then the velocity corresponds to phase-2 because there is no phase-1 present.

When you have 0<VOF<1, you cannot say what the velocities of individual phases are (they are the same in VOF-land). If you are not okay with this approach, don't use VOF.

[openAC]

Originally Posted by LuckyTran

Indeed there is only 1 velocity. When the VOF is 1 then the velocity corresponds to the velocity of phase-1 because there is no phase-2. And when the VOF is 0 then the velocity corresponds to phase-2 because there is no phase-1 present.

When you have 0<VOF<1, you cannot say what the velocities of individual phases are (they are the same in VOF-land). If you are not okay with this approach, don't use VOF.

Can you suggest any text supporting this?

You may be right but that sounds weird to me. For example, let's say VOF of Phase-2 is 0.99 in a region, then can I not say anything about the velocity of the second phase, but if it is 1.0 then we suddenly can.

Also, if we cannot say anything about the individual phase velocities what is the quantity "u" (mixture velocity or whatever that is) useful for?

Thanks for offering input.

[hamed.majeed]

Originally Posted by openAC

Can you suggest any text supporting this?

You may be right but that sounds weird to me. For example, let's say VOF of Phase-2 is 0.99 in a region, then can I not say anything about the velocity of the second phase, but if it is 1.0 then we suddenly can.

Also, if we cannot say anything about the individual phase velocities what is the quantity "u" (mixture velocity or whatever that is) useful for?

Thanks for offering input.

Vof method have velocity of individual phases, actually, its the velocity of the mixture. That is the biggest limitation of this model.
I would like to quote Fluent theory guide here: "One limitation of the shared-fields approximation is that in cases where large velocity differences exist between the phases, the accuracy of the velocities computed near the interface can be adversely affected."

See the momentum equation in the theory guide. It have velocity, v, as the velocity of the mixture.
http://www.afs.enea.it/project/neptu...th/node301.htm

[hamed.majeed]

Originally Posted by LuckyTran

Indeed there is only 1 velocity. When the VOF is 1 then the velocity corresponds to the velocity of phase-1 because there is no phase-2. And when the VOF is 0 then the velocity corresponds to phase-2 because there is no phase-1 present.

When you have 0<VOF<1, you cannot say what the velocities of individual phases are (they are the same in VOF-land). If you are not okay with this approach, don't use VOF.

VOF-land is funny terminology...like it

[openAC]

@hameed.majid

Thanks for the reply.

Also, does that mean that in VOF, the mixture velocity "v" can be written in some way as a function of the individual phase velocities, say v-phase-1 and v-phase-2, with some combination of their volume fractions, i.e., vol-frac-phase-1 and vol-frac-phase-2?

Best regards

[hamed.majeed]

No, the velocities of individual phases cannot be linked by volume fraction.
In VOF model there are no individual velocities.
You need to read a book that describe the physics of the method.


Please do verify the info I tell you on your own.

[openAC]

@hameed.majid

Thanks again for the response.

I have checked with a multiphase textbook: "Particles, Drops, and Bubbles" by Eric Loth. According to the theory, the "one-fluid velocity" or the "V" that is in the VOF momentum equation is related to the "phase velocity" as follows:
1. If the location has only phase-1 or phase-2, then the "V" obtained from the simulations belong to that phase. (this is obvious)
2. Within the "finite thickness interface" between the two phases (i.e., where phase volume fraction is neither 0 nor 1), the velocity obtained (V) is non-physical.

Both @LuckyTran and @Hameed.Majid have answered similarly. So thank you both.

One more thing, the text says that this "V" is not the same as the mixture velocity. This is why, I think, we should not think of the VOF "one-fluid" velocity as a combination of individual phase velocities.

Hope this discussion will help other people.

[hamed.majeed]

Originally Posted by openAC

@hameed.majid

Thanks again for the response.

I have checked with a multiphase textbook: "Particles, Drops, and Bubbles" by Eric Loth. According to the theory, the "one-fluid velocity" or the "V" that is in the VOF momentum equation is related to the "phase velocity" as follows:
1. If the location has only phase-1 or phase-2, then the "V" obtained from the simulations belong to that phase. (this is obvious)
2. Within the "finite thickness interface" between the two phases (i.e., where phase volume fraction is neither 0 nor 1), the velocity obtained (V) is non-physical.

Both @LuckyTran and @Hameed.Majid have answered similarly. So thank you both.

One more thing, the text says that this "V" is not the same as the mixture velocity. This is why, I think, we should not think of the VOF "one-fluid" velocity as a combination of individual phase velocities.

Hope this discussion will help other people.

I will look into the reference....that requires further study on my part.

[hamed.majeed]

So I ensured...the original vof model used a single velocity V in all the equations (continuity, momentum, and volume fr.)

Refer to the original work of Hirt & Nicholas, 1981, Volume of Fluid (VOF) method for dynamics of free surface boundaries.
They are using SOLA-VOF code...which have similar equations mentioned in their manual.

However, I looked into Openfoam...they seem to have made a modification and used two-fluid vof to come up with a better equation. Its the one currently used in interfoam. Look up a document which has all those details: Description and utilization of interfoam multiphase solver by Santiago Marquez.

Now the sad part. Ansys Fluent in the theory guide have mentioned the VOF equation....which has velocity of the phase 2....and its still confusing.
In conclusion, in VOF model the velocity is V..which is not in any relation to the volume fraction.
However, they do have another model...which has a mixture velocity V_m linked to volume fraction.

[Sachin Zanje]

Hi.
Could you please tell me then how FLUENT calculates velocity and pressure when alpha is between 0 and 1?
Is there any kind of wattage used for the same?

[Sachin Zanje]

I meant to say above with respect to VOF