CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   About selecting Turbulence model in VOF model (http://www.cfd-online.com/Forums/fluent/100653-about-selecting-turbulence-model-vof-model.html)

mariam.sara April 27, 2012 14:37

About selecting Turbulence model in VOF model
 
Hi I am confused about selecting the type of flow as laminar or turbulence when I want solve two phase flow problem within VOF model should I estimate Reynolds number based on the mixture velocity (the sum of superficial velocities of phases) or estimate Reynolds for each phase separately based on the physical velocity of each phase? So which one is the correct one in determining the type of flow? knowing that my case is stratified flow?

LuckyTran April 28, 2012 00:16

Quote:

Originally Posted by mariam.sara (Post 357561)
Hi I am confused about selecting the type of flow as laminar or turbulence when I want solve two phase flow problem within VOF model should I estimate Reynolds number based on the mixture velocity (the sum of superficial velocities of phases) or estimate Reynolds for each phase separately based on the physical velocity of each phase? So which one is the correct one in determining the type of flow? knowing that my case is stratified flow?

If the flow is stratified, then it is mostly unmixed correct?

I do not foresee any issuse as to why you cannot calculate Reynolds number separately for each phase. But I definitely would not use the mixture velocity (that leads to a completely wrong interpretation of Reynolds number) if the flow is stratified.

Hopefully you end up with a case where both phases are both clearly laminar or both turbulent, since you can't choose separate models for each reason.

mariam.sara April 28, 2012 04:36

Hi Lucky thanks for the answer. In my case the oil volume fraction is 0.39 , oil superficial velocity is 0.03 m/s and its density is 828 kg/m^3 , viscosity is 0.3 N.m/s and water superficial velocity is 0.15 m/s , density 1000 kg/m^3 and viscosity 0.001 N.m/s the pipe diameter is 0.026 m and length 12 m. When I estimate Reynolds for oil phase it be just 4.3 while Reynolds for water phase is 4756 so is my case should be simulated as laminar or turbulent?

Mariam

LuckyTran April 28, 2012 13:49

Quote:

Originally Posted by mariam.sara (Post 357764)
Hi Lucky thanks for the answer. In my case the oil volume fraction is 0.39 , oil superficial velocity is 0.03 m/s and its density is 828 kg/m^3 , viscosity is 0.3 N.m/s and water superficial velocity is 0.15 m/s , density 1000 kg/m^3 and viscosity 0.001 N.m/s the pipe diameter is 0.026 m and length 12 m. When I estimate Reynolds for oil phase it be just 4.3 while Reynolds for water phase is 4756 so is my case should be simulated as laminar or turbulent?

Mariam

So is it straified flow or not? If the flow is straified then the volume fraction should be 0 or 1. Since you have a volume fraction of 0.39 your flow cannot be stratified, just a simple mixture. Is VOF even appropriate? Anyway, can you make a sketch of your domain? i.e. what are the boundary conditions? What would generate any turbulence?

Your units on viscosity are wrong.

If it is a well-mixed flow you use the proper weighting scheme to calculate a single Reynolds number.

If it is straified flow you actually have 3 Reynolds numbers, 1 for the oil in a pipe, 1 for the water in a pipe, and 1 for the interface between the oil and water. If you still have 1 laminar and 1 turbulent like you mentioned then there is difficulty now, as there is both a laminar and turbulent region. Unfortunately there are no models that can handle this well (and selecting both laminar and turbulent is not an option obviously).

LuckyTran April 28, 2012 13:50

Quote:

Originally Posted by mariam.sara (Post 357764)
Hi Lucky thanks for the answer. In my case the oil volume fraction is 0.39 , oil superficial velocity is 0.03 m/s and its density is 828 kg/m^3 , viscosity is 0.3 N.m/s and water superficial velocity is 0.15 m/s , density 1000 kg/m^3 and viscosity 0.001 N.m/s the pipe diameter is 0.026 m and length 12 m. When I estimate Reynolds for oil phase it be just 4.3 while Reynolds for water phase is 4756 so is my case should be simulated as laminar or turbulent?

Mariam

So is it straified flow or not? If the flow is straified then the volume fraction should be 0 or 1 depending on which region you are in. Since you have a volume fraction of 0.39 your flow cannot be stratified, just a simple mixture. Is VOF even appropriate? Can yo explain it more clearly?

Your units on viscosity are wrong.

If it is a well-mixed flow you use the proper weighting scheme to calculate a single Reynolds number.

If it is straified flow you actually have 3 Reynolds numbers, 1 for the oil in a pipe, 1 for the water in a pipe, and 1 for the interface between the oil and water. If you still have 1 laminar and 1 turbulent like you mentioned then there is difficulty now, as there is both a laminar and turbulent region. Unfortunately there are no models that can handle this well (and selecting both laminar and turbulent is not an option obviously).

mariam.sara April 28, 2012 15:42

Yes its a stratified flow case. The 0.39 is the oil-holdup at inlet and the oil level is 10.8 mm. The oil at the lower part of the pipe and water above it. I am trying to get wavy interface depending at the inlet velocities which i took it from a paper (I failed to attached it here coz its 1MB type me your email if you want see it) . The VOF model equations mentioned at the manual reference to the average velocities and properties for mixture hence I think selecting the type of flow is depend on the Reynolds number which it defined as :
Rem=(U(mix)*D*density(mix))/viscosity (mix)
where density(mix)=density(oil)*vof(oil)+density(water)* vof(water)
viscosity(mix)=viscosity(oil)*vof(oil)+viscosity(w ater)*vof(water)
U(mix) = U(oil)+U(water)
I am right?

Yes i type the viscosity units wrong its in Pa.s

LuckyTran April 28, 2012 16:26

I should've asked this from the start but: What are your inlet conditions? Is it turbulent inflow? If so then obviously it will not suddenly become laminar again. Your fluid at the inlet must come from some source, and is that source laminar or turbulent? Is this developing or fully-developed? If it is developing, where did the flow come from and was that flow already turbulent? These are important questions that will greatly assist deciding if the nature of the flow.

Quote:

Originally Posted by mariam.sara (Post 358215)
The VOF model equations mentioned at the manual reference to the average velocities and properties for mixture

The VOF model does do this. It solves a single set of equations for both fluids by taking the weighted properties of each fluid for each cell.

Quote:

Originally Posted by mariam.sara (Post 358215)
I think selecting the type of flow is depend on the Reynolds number which it defined as :
Rem=(U(mix)*D*density(mix))/viscosity (mix)

This interpretation of Reynolds number is very non-physical. Each fluid occupies a different portion of the pipe, and each therefore have different geometric length scales. Each fluid has different bulk velocities and therefore both have different velocity scales. Each fluid then also has different properties. Again, as I said before, there are at least 3 Reynolds numbers for this type of flow (1 for oil, 1 for water, 1 for the interface). You even said that it is a satrified flow, hence the two fluids never mix, hence they can never be treated as the same flow. It is two interacting flows.

Unfortunately, since a single set of equations is used for both zones, Fluent cannot solve a case where one fluid is laminar and the other is turbulent. You can try your luck with a transitional model. But to choose either laminar or turbulent is wrong, at least in my opinion.

mariam.sara April 29, 2012 16:21

Dear Lucky many thanks to the information you post. What is the transitional model ? is it another type of VOF model and can it solve stratified flow case? what it's capabilities?

LuckyTran April 29, 2012 16:23

Quote:

Originally Posted by mariam.sara (Post 358368)
Dear Lucky many thanks to the information you post. What is the transitional model ? is it another type of VOF model and can it solve stratified flow case? what it's capabilities?

It's a turbulence model for transitional flow cases, it's about as close as you can get to trying to do laminar & turbulent in the same simulation. None of the turbulence models are meant for VOF or stratified flow of different fluids so in general not really.


All times are GMT -4. The time now is 03:34.