Gas-liquid pipe flow

iloc86 · June 17, 2015, 16:00

Hi guys, im new in CFX and im trying to simulate a gas liquid multiphase flow but im getting weird results, i just want to be able to identify some horizontal pipe flow patterns (wave, stratified, etc) but i reallt dont know very well how to set the bc and multiphase models, i have read some tutorial and i have done this.

i set a simple horizontal pipe 3 m lenght and 0.20 m diameter

at inlet i set 0.5 of gas and 0.5 of water
1 m/s both

a simple outlet with av pressure 0 atm

im starting with laminar flow, both continuos phases, inhomogeneuos. fluid pair with mixing model, and drag coefficient.

i just want to know what best settings i have to usea to model stratified flow

as i run the model i get some waves but at some time i have "picks" in convergence and i see like suddenly "explosions" in the flow. i know i have alotof things wrong.

In simple words what a have to do to set a stable model.

ghorrocks · June 17, 2015, 20:13

For free surface flows - they are quite sensitive to time step size. I would recommend you use adaptive time stepping so it can automatically find a good time step.

Use adaptive time stepping, homing in on 3-5 coeff loops per iteration. Make sure the max and min time step size are wide enough that you do not hit them, and give it a sensible starting time step size.

iloc86 · June 18, 2015, 10:33

thanks glenn ill try that today.

JuPa · June 18, 2015, 11:13

Curious question for the knowledgeable:

Can you run this problem in steady state, with a small physical time step (e.g. t = 0.001 s) for 10000 iterations?

This will give you a pseudo simulation time of 0.001 * 10000 = 10 s.

Then use the steady state results as the initial guess for a transient run, but setting the initial start time as 10 s?

Thanks

June 17, 2015, 16:00	Gas-liquid pipe flow	#1
iloc86 New Member ivan Join Date: Dec 2014 Posts: 14 Rep Power: 11	Hi guys, im new in CFX and im trying to simulate a gas liquid multiphase flow but im getting weird results, i just want to be able to identify some horizontal pipe flow patterns (wave, stratified, etc) but i reallt dont know very well how to set the bc and multiphase models, i have read some tutorial and i have done this. i set a simple horizontal pipe 3 m lenght and 0.20 m diameter at inlet i set 0.5 of gas and 0.5 of water 1 m/s both a simple outlet with av pressure 0 atm im starting with laminar flow, both continuos phases, inhomogeneuos. fluid pair with mixing model, and drag coefficient. i just want to know what best settings i have to usea to model stratified flow as i run the model i get some waves but at some time i have "picks" in convergence and i see like suddenly "explosions" in the flow. i know i have alotof things wrong. In simple words what a have to do to set a stable model.

June 18, 2015, 10:33	thanks	#3
iloc86 New Member ivan Join Date: Dec 2014 Posts: 14 Rep Power: 11	thanks glenn ill try that today.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
sonicFoam - pressure driven pipe: flow continuity violation and waveTransmissive BC	Endel	OpenFOAM Running, Solving & CFD	3	September 11, 2014 17:29
Imposing gas and liquid mass flow rate BC for hyperbolic equations	omid1361	Main CFD Forum	0	October 17, 2012 03:07
Gas solid flow in a pipe with 90 bend	pandaba	FLUENT	0	September 14, 2010 02:43
What is the difference between liquid reactive flow and gas reactive flow?	James	Main CFD Forum	6	May 15, 2009 13:14
Two-Phase Buoyant Flow Issue	Miguel Baritto	CFX	4	August 31, 2006 13:02

June 17, 2015, 20:13		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	For free surface flows - they are quite sensitive to time step size. I would recommend you use adaptive time stepping so it can automatically find a good time step. Use adaptive time stepping, homing in on 3-5 coeff loops per iteration. Make sure the max and min time step size are wide enough that you do not hit them, and give it a sensible starting time step size.

June 18, 2015, 11:13		#4
JuPa Senior Member Mr CFD Join Date: Jun 2012 Location: Britain Posts: 361 Rep Power: 15	Curious question for the knowledgeable: Can you run this problem in steady state, with a small physical time step (e.g. t = 0.001 s) for 10000 iterations? This will give you a pseudo simulation time of 0.001 * 10000 = 10 s. Then use the steady state results as the initial guess for a transient run, but setting the initial start time as 10 s? Thanks