[LyngHoo]

Hi, everyone. I'm trying to use VoF model to simulate gas-liquid two-phase
flow in microchannels. firstly, I started with a 2-D simulation in a 500 micron
wide T-shape microchannel. Details are listed below,
# Segregated time dependent unsteady solver
# PRESTO! for pressure interpolation
# PISO for pressure-velocity coupling
# Second-Order Up-Wind scheme for momentum equation
# Geometry Reconstruction scheme for interface interpolation
# Implicit body force for body force formulation
# Courant Number 0.25 for volume fraction calculation
# d = 0.5 mm, gas and liquid inlet length is 3*d, respectively. Main channel is
15*d in length.
# U_L= U_G = 0.02 m/s
# surface tension of water is 0.072 N/m
# Contact angle 0
# Velocity inlets
# Pressure outlet ( 0 gauge pressure )
# No-slip wall condition

after that I tryed to reproduct one work of Qian et. al.( chem. eng. sci. 2006
) by simulating a ─┴───── shaped microchannel in 2-D domain. the channel
width, d, is aslo 0.5mm. And keep all the settings above unchanged, and
these seetings are uniform as Qian reported in his paper.
I can get the Taylor bubble flow pattern if the liquid goes in horizontally from
the left inlet, and gas geos into the channel vertically. But if I exchange the
two fluids, I got stratified flow pattern instead of Taylor flow. and refining grid
is not working. However, increasing surface tension by a factor of 1.5 works.
But according to Qian et. al.( chem. eng. sci. 2006 ), they got Taylor flow in
such ─┴───── shaped channel.
and I tried a Y-shaped channel with same configurations, and got stratified
flow. But when I incresing the surface tension of water (0.072 N/m) by a
factor of 1.2, Taylor flow can be obtained. So I wander why a lower serface
tension can get a correct flow pattern. In many literature, they can get
correct flow pattern with surface tension as low as approximately 0.02 N/m.
How could they done that. Did I do something wrong?

[zmh021810]

Well, I also saw that paper and recently I am going to do the same thing.

Well do you have the exact same setting comparing with Qian's paper? Because I know that the surface tension is very important to control the slung flow, you can only have slung flow when the surface tension is large enough.

Well, do you finish your simulaiton yet? I want to talk with you about this

field.for example, Qian's paper said both the inlets are fixed by uniform velocity distribution, do you also did it in the same way?

another problem is: do you use patch option in fluent?

Best wishes for you!

Justin