Two-phase flow in vertical pipe

Ladnam · May 9, 2011, 14:38

Have anyone been able to model two-phase flow (water-air) in a vertical pipe and reproduced the observed flow regimes (annular flow, slug flow etc.)? Or maybe have experience with problems related to this?

This is almost the same model as the Hibiki bubble column tutorial except it is upside down.

I have tried steady and unsteady simulation with different BC and IC with VOF and multiphase segregated flow, but the solution either diverges or will not converge to a steady or realistic solution.

abdul099 · May 10, 2011, 03:06

Are you sure, VOF is the right model?
You should have a mesh resolution small enough to have several cells in one bubble (I assume, there will be some bubbles of air, oil or whatever phase).
Therefore VOF is not suitable to simulate a flow with a very intensive mixture of the different phases.
And not to forget: A steady state VOF simulation is just crap.

When you want to run a case with very intensive mixture or you HAVE to do it steady state, think about a Eulerian / Eulerian multiphase. But be aware, this will NOT resolve single bubbles. And good luck when you try to prevent it from divergence...

Ladnam · May 10, 2011, 05:26

I actually got a steady VOF simulation to converge.
But it is not free falling water. I use a velocity inlet at one end with 30% water and a pressure outlet at the other end.

It converges to an annular flow which is realistic.

niriosrat · August 25, 2011, 04:40

Hi,

Did you model slug flow? it is because I am trying to model slug flow, but is not working, so I wonder which conditions did you used on the VOF?

Ladnam · August 29, 2011, 08:07

Yes I think I got some resonable results.
I got slug flow for Froude numbers larger than about 1.5.
I dind't do anything special with the VOF model.
I used a vertical pipe with air intake (pressure outlet) and velocity inlet and a pressure outlet at the bottom end.

niriosrat · August 30, 2011, 02:52

My vertical tube has a diameter of 2.53 cm and 4 m length, due to symemtry I am modelling just half of it.

In CCM+ I can not define a single inlet (I think) with air and water coming together with a specific mixture velocity (=0.5 m/s) and a volume fraction of 0.5-0.5. Therefore, I split the inlet in two velocity inlets (for water and air), where each inlet had a velocity of 0.5 m/s. I am using a time-step of 5e-4 sec and the bubbels develop at the inlet (red volume fraction) but after they just disappeared (green volume fraction) and I don't know why?

Ladnam · August 30, 2011, 07:29

You can use a mixture of two phases but it is not good to use with the VOF model I guess. You can try having two separate inlets for water and air. And maybe the outlet needs to be bigger to avoid backflow. E.g connecting a big box to the outlet. You can also try using the full model (no symmetry) to give the flow more freedom.

ping · September 3, 2011, 02:11

VOF is for modelling immiscible fluids - no mixing. Each cell in theory should have 100% of one fluid and none of the other(s). Transitions from one fluid to the next should ideally be across just one cell with a 'mixture' of the phases. To achieve this you need a fine mesh is some cases - eg to capture bubbles/drops as mentioned above. Otherwise you must use normal Eulerian multi-phase. In your contour plots as soon as you see zones of colours other than red or blue you are using VOF incorrectly. And so to try to model slug flow in steady mode is crazy - think about what slug flow looks like!

jignesh_thaker2007 · January 22, 2012, 03:28

I am also trying to solve the air-water two phase flow problem in horizontal pipe i also split this pipe in two part and define it to water and air and gave them different velocity. i also used the interDyMFoam solver. but my problem is it possible to get the bubbly flow regime? if possible then how?

Quote:

Originally Posted by ping

VOF is for modelling immiscible fluids - no mixing. Each cell in theory should have 100% of one fluid and none of the other(s). Transitions from one fluid to the next should ideally be across just one cell with a 'mixture' of the phases. To achieve this you need a fine mesh is some cases - eg to capture bubbles/drops as mentioned above. Otherwise you must use normal Eulerian multi-phase. In your contour plots as soon as you see zones of colours other than red or blue you are using VOF incorrectly. And so to try to model slug flow in steady mode is crazy - think about what slug flow looks like!

niriosrat · Yesterday, 08:26

Hi,

I have an horizontal pipe with ID of 5 cm and a length of 30 m. I defined and inlet (velocity inlet) and outlet (pressure outlet). I am not familiar with field functions and I don't have two-inlets in my geometry. So I wonder if someone knows how to write a field function for the inlet in Star CCM+ where I can set half of the cross-sectional area with air and the other half with water and put a fix velocity for each phase. E.g. the top part with air with a velocity of 0.15 m/s and the bottom with water and a velocity of 2 m/s.

ping · Yesterday, 09:07

the help has lots of detail and examples. and if you do the first vof training tutorial you will see how to do a field function which is very similar to what you require to change from air to water at a certain position. then create an equivalent one for the boundary normal velocity scalar. or a vector field function if you want to express the flow in components.

May 9, 2011, 14:38	Two-phase flow in vertical pipe	#1
Ladnam Member Join Date: May 2010 Posts: 30 Rep Power: 4	Have anyone been able to model two-phase flow (water-air) in a vertical pipe and reproduced the observed flow regimes (annular flow, slug flow etc.)? Or maybe have experience with problems related to this? This is almost the same model as the Hibiki bubble column tutorial except it is upside down. I have tried steady and unsteady simulation with different BC and IC with VOF and multiphase segregated flow, but the solution either diverges or will not converge to a steady or realistic solution.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
HOMOGENEOUS TWO PHASE FLOW IN A PIPE	beginner	FLUENT	2	July 27, 2007 12:17
How to adapt lower phase in 3 D pipe flow	K.Baker	FLUENT	2	July 21, 2007 15:19
compressible two phase flow in CFX4.4	youngan	CFX	0	July 1, 2003 23:32
Problem on boundry of two phase flow	youngan	CFX	0	June 30, 2003 02:32
one phase flow within a pipe	Liouskos Jiannis	Phoenics	2	October 17, 2001 05:35

May 10, 2011, 03:06		#2
abdul099 Senior Member Join Date: Oct 2009 Location: Germany Posts: 431 Rep Power: 7	Are you sure, VOF is the right model? You should have a mesh resolution small enough to have several cells in one bubble (I assume, there will be some bubbles of air, oil or whatever phase). Therefore VOF is not suitable to simulate a flow with a very intensive mixture of the different phases. And not to forget: A steady state VOF simulation is just crap. When you want to run a case with very intensive mixture or you HAVE to do it steady state, think about a Eulerian / Eulerian multiphase. But be aware, this will NOT resolve single bubbles. And good luck when you try to prevent it from divergence...

August 25, 2011, 04:40		#4
niriosrat New Member Nicolas Join Date: Jun 2009 Posts: 8 Rep Power: 4	Hi, Did you model slug flow? it is because I am trying to model slug flow, but is not working, so I wonder which conditions did you used on the VOF?

August 30, 2011, 07:29		#7
Ladnam Member Join Date: May 2010 Posts: 30 Rep Power: 4	You can use a mixture of two phases but it is not good to use with the VOF model I guess. You can try having two separate inlets for water and air. And maybe the outlet needs to be bigger to avoid backflow. E.g connecting a big box to the outlet. You can also try using the full model (no symmetry) to give the flow more freedom.

September 3, 2011, 02:11		#8
ping Member Ping Join Date: Mar 2009 Posts: 78 Rep Power: 5	VOF is for modelling immiscible fluids - no mixing. Each cell in theory should have 100% of one fluid and none of the other(s). Transitions from one fluid to the next should ideally be across just one cell with a 'mixture' of the phases. To achieve this you need a fine mesh is some cases - eg to capture bubbles/drops as mentioned above. Otherwise you must use normal Eulerian multi-phase. In your contour plots as soon as you see zones of colours other than red or blue you are using VOF incorrectly. And so to try to model slug flow in steady mode is crazy - think about what slug flow looks like!

Yesterday, 08:26		#10
niriosrat New Member Nicolas Join Date: Jun 2009 Posts: 8 Rep Power: 4	Hi, I have an horizontal pipe with ID of 5 cm and a length of 30 m. I defined and inlet (velocity inlet) and outlet (pressure outlet). I am not familiar with field functions and I don't have two-inlets in my geometry. So I wonder if someone knows how to write a field function for the inlet in Star CCM+ where I can set half of the cross-sectional area with air and the other half with water and put a fix velocity for each phase. E.g. the top part with air with a velocity of 0.15 m/s and the bottom with water and a velocity of 2 m/s.

Yesterday, 09:07		#11
ping Member Ping Join Date: Mar 2009 Posts: 78 Rep Power: 5	the help has lots of detail and examples. and if you do the first vof training tutorial you will see how to do a field function which is very similar to what you require to change from air to water at a certain position. then create an equivalent one for the boundary normal velocity scalar. or a vector field function if you want to express the flow in components.