water inject tube, huge different results by icoFoam and Fluent

cchuran · July 29, 2015, 17:52

Hi all, I am a beginner to use openfoam. A simple case: water injection into 2D tube, x*y=2*0.4. But the velocity distribution is huge different when use openFOAM and Fluent. Why?

The icoFoam and Fluent were both used to simulate the steady-state process. The BCs are:

openFOAM
left: fixed velocity 0.05 m/s, zeroGrident for p
right: zeroGrident for U, fixed 0 for p
upDown: fixed (0,0,0) for U, zeroGrident for p
frontback: empty

fluent
left: velocity-inlet 0.05 m/s, zeroGrident for p
right: pressure-outlet, fixed for pressure
upDown: wall, no-slip

Attached is the comparisons of the openFOAM and Fluent results. Thanks!

loook · July 30, 2015, 05:05

Hi c churan

what did you set for the turbulence in both cases? it seems that in openFoam you have laminar flow while in fluent the profile looks like turbulent.

cchuran · July 30, 2015, 13:21

Thanks for your reply. The fluent also uses laminar model (the same as openFOAM)

fabian_roesler · July 31, 2015, 06:27

Have a look into the transportProperties inside constant folder.
The kinematic viscosity of 0.01 is ten thousand times larger than the one of water 1.004e-6. This gives you a perfect parabolic profile even withing this large diameter tube.

Cheers

Fabian

hxaxtma · July 31, 2015, 08:30

It looks like that you are simulating different Reynoldsnumbers and in assumption that you set the velocity (0,05 m/s) and the viscosity (water=1.04e-06) for both Fluent and OF, there is one left, the scaling of your geometry!

Please, verify that your case in fluent AND in OpenFOAM is set up in meters.

If you want to simulate laminar pipe flow, the critical Reynoldsnumber should be

Re_crit=u*d/<2320

P.S.:
your OF simulation looks for laminar pipe flow right, cause the peak velocity of the parabolic profile should be around 2*u_in = 2*0,05=0,1m/s

hxaxtma · August 3, 2015, 08:05

you fixed your problem?

July 31, 2015, 06:27	kinematic viscosity of 0.01	#4
fabian_roesler Senior Member Fabian Roesler Join Date: Mar 2009 Location: Germany Posts: 213 Rep Power: 18	Have a look into the transportProperties inside constant folder. The kinematic viscosity of 0.01 is ten thousand times larger than the one of water 1.004e-6. This gives you a perfect parabolic profile even withing this large diameter tube. Cheers Fabian

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Curious results in icoFoam differ from fluent	hemph	OpenFOAM Running, Solving & CFD	6	December 16, 2005 14:27

July 30, 2015, 05:05		#2
loook New Member Lukasz Join Date: Apr 2012 Posts: 10 Rep Power: 14	Hi c churan what did you set for the turbulence in both cases? it seems that in openFoam you have laminar flow while in fluent the profile looks like turbulent.

July 30, 2015, 13:21		#3
cchuran New Member CA Join Date: Jul 2015 Posts: 2 Rep Power: 0	Thanks for your reply. The fluent also uses laminar model (the same as openFOAM)

July 31, 2015, 08:30		#5
hxaxtma Senior Member Join Date: Jan 2014 Posts: 179 Rep Power: 12	It looks like that you are simulating different Reynoldsnumbers and in assumption that you set the velocity (0,05 m/s) and the viscosity (water=1.04e-06) for both Fluent and OF, there is one left, the scaling of your geometry! Please, verify that your case in fluent AND in OpenFOAM is set up in meters. If you want to simulate laminar pipe flow, the critical Reynoldsnumber should be Re_crit=ud/<2320 P.S.: your OF simulation looks for laminar pipe flow right, cause the peak velocity of the parabolic profile should be around 2u_in = 2*0,05=0,1m/s

August 3, 2015, 08:05		#6
hxaxtma Senior Member Join Date: Jan 2014 Posts: 179 Rep Power: 12	you fixed your problem?