Error: Isolated fluid regions

rushabh · October 11, 2011, 22:36

I was trying to simulate a heat exchanger. The solver returned an error saying there were '4 isolated fluid regions'. While troubleshooting I came across this thread (http://www.cfd-online.com/Forums/cfx...tml#post327563) where CycLone suggested opening the res.err file in CFX-Post and plotting a variable (or creating an isovolume) to identify isolated regions.

Could somebody guide me on that? What variable to plot and how? Or how exactly to create an isovolume?

Thanks,
Rushabh

ghorrocks · October 12, 2011, 07:47

This is basic post processing and is covered in the tutorials.

evcelica · October 19, 2011, 16:14

I'm guessing it immediately failed when you went to solve?
If so you have to specify the pressure somewhere in each fluid domain (even just a reference pressure) for any enclosed fluid.

ghorrocks · October 19, 2011, 18:35

The root problem is there are unconnected sections of mesh - hence isolated fluid volumes. Either the mesh is wrong or th CFX-Pre setup is wrong. But the question is a basic one about post-processing.

Having said that, I don't think CFD-Post will help here anyway. Check your mesh in your mesh generation.

rushabh · October 26, 2011, 23:54

I am getting the same error msg. I just dont know what I am doing wrong. I have created a vid of my meshing. Its a little over 4 mins. Here's the link: http://youtu.be/J5nG-y1rM4w

I've been breaking my head over this for a week now. I would really appreciate any help.

ghorrocks · October 27, 2011, 07:04

Are the internal circles a pipe system for the heat exchanger or are they some form of heating/cooling element?

rushabh · October 27, 2011, 12:01

Ideally there should be no pipe (holes) in the Shell and initially I was not planning on putting holes (internal circles) in the Shell but going by the tutorial-geometry for the HeatingCoil example, I did. In that geometry they have drilled a hole in the Annulus and put a solid pipe in the hole (which can be simulated as a wire or pipe) and thats what I did for my Heat Exchanger. I made a hole in the Shell and fitted tubes inside.

Here's few images of the HeatingCoil geometry.

Heating Coil assembly

Annulus hole

Copper wire/pipe in hole

ghorrocks · October 27, 2011, 17:51

Sure, but you did not answer my question. What are the holes you have in the geometry?

rushabh · October 27, 2011, 18:20

Yup, so my point was that the holes mean nothing to me and I know they shouldn't be there at first place but I had to include them in the geometry to select a surface during the 'Domain Interface' selection. If I didn't have the holes in the shell I would have had no surface to select for interface like I do in the following image.

In the pic the cold water zone is the Shell and the interface in shell region is the shell holes. The tubes carry hot water. Technically it is not a heating element but you could call it one therotically. So coming back to your question, the shell holes have no function as such and if I remove them and have one completely hollow shell I don't know what should I select for interface.

ghorrocks · October 27, 2011, 18:30

Quote:

The tubes carry hot water.

Thank you - that answered my question.

Do you want to model the flow in the tubes?

rushabh · October 27, 2011, 18:36

Yes sir but I would also like to see what's going on in the shell.

ghorrocks · October 27, 2011, 19:28

OK, now I understand the issue. You are always going to have isolated fluid volume in this model as the tube domains are not fluidically connected to the shell domain. You will have to use the expert parameter "check isolated regions = f" to turn off this check, as the error message says.

Mina_Shahi · November 20, 2012, 09:32

Quote:

Originally Posted by rushabh

Yup, so my point was that the holes mean nothing to me and I know they shouldn't be there at first place but I had to include them in the geometry to select a surface during the 'Domain Interface' selection. If I didn't have the holes in the shell I would have had no surface to select for interface like I do in the following image.

In the pic the cold water zone is the Shell and the interface in shell region is the shell holes. The tubes carry hot water. Technically it is not a heating element but you could call it one therotically. So coming back to your question, the shell holes have no function as such and if I remove them and have one completely hollow shell I don't know what should I select for interface.

Why did you define fluid-fluid interface? Shouldn't it be fluid-solid interface? since fluid is connecting to the shell both through the tube and out side that.

ghorrocks · November 20, 2012, 18:25

You would use a f-f interface if you did not want to explicitly model the tubes. The thermal resistance of the tubes can be added as a factor in the interface model. Doing this simplifies the model considerably compared to directly modellign the intermediate metal tubes.

October 11, 2011, 22:36	Error: Isolated fluid regions	#1
rushabh New Member Rushabh Join Date: Oct 2010 Posts: 8 Rep Power: 16	I was trying to simulate a heat exchanger. The solver returned an error saying there were '4 isolated fluid regions'. While troubleshooting I came across this thread (http://www.cfd-online.com/Forums/cfx...tml#post327563) where CycLone suggested opening the res.err file in CFX-Post and plotting a variable (or creating an isovolume) to identify isolated regions. Could somebody guide me on that? What variable to plot and how? Or how exactly to create an isovolume? Thanks, Rushabh

October 27, 2011, 07:04		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Are the internal circles a pipe system for the heat exchanger or are they some form of heating/cooling element? Mina_Shahi likes this.

October 27, 2011, 19:28		#12
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	OK, now I understand the issue. You are always going to have isolated fluid volume in this model as the tube domains are not fluidically connected to the shell domain. You will have to use the expert parameter "check isolated regions = f" to turn off this check, as the error message says. Mina_Shahi likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Questions of fluid pairs	fjalil	CFX	1	June 10, 2009 18:36
Fluid pairs	fjalil	Main CFD Forum	0	June 10, 2009 14:47
My Revised "Time Vs Energy" Article For Review	Abhi	Main CFD Forum	2	July 9, 2002 10:08
Terrible Mistake In Fluid Dynamics History	Abhi	Main CFD Forum	12	July 8, 2002 10:11
separating fluid regions	David C	FLUENT	1	March 7, 2001 11:52

October 12, 2011, 07:47		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	This is basic post processing and is covered in the tutorials.

October 19, 2011, 16:14		#3
evcelica Senior Member Erik Join Date: Feb 2011 Location: Earth (Land portion) Posts: 1,188 Rep Power: 23	I'm guessing it immediately failed when you went to solve? If so you have to specify the pressure somewhere in each fluid domain (even just a reference pressure) for any enclosed fluid.

October 19, 2011, 18:35		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	The root problem is there are unconnected sections of mesh - hence isolated fluid volumes. Either the mesh is wrong or th CFX-Pre setup is wrong. But the question is a basic one about post-processing. Having said that, I don't think CFD-Post will help here anyway. Check your mesh in your mesh generation.

October 26, 2011, 23:54		#5
rushabh New Member Rushabh Join Date: Oct 2010 Posts: 8 Rep Power: 16	I am getting the same error msg. I just dont know what I am doing wrong. I have created a vid of my meshing. Its a little over 4 mins. Here's the link: http://youtu.be/J5nG-y1rM4w I've been breaking my head over this for a week now. I would really appreciate any help.

October 27, 2011, 12:01		#7
rushabh New Member Rushabh Join Date: Oct 2010 Posts: 8 Rep Power: 16	Ideally there should be no pipe (holes) in the Shell and initially I was not planning on putting holes (internal circles) in the Shell but going by the tutorial-geometry for the HeatingCoil example, I did. In that geometry they have drilled a hole in the Annulus and put a solid pipe in the hole (which can be simulated as a wire or pipe) and thats what I did for my Heat Exchanger. I made a hole in the Shell and fitted tubes inside. Here's few images of the HeatingCoil geometry. Heating Coil assembly Annulus hole Copper wire/pipe in hole

October 27, 2011, 17:51		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Sure, but you did not answer my question. What are the holes you have in the geometry?

October 27, 2011, 18:20		#9
rushabh New Member Rushabh Join Date: Oct 2010 Posts: 8 Rep Power: 16	Yup, so my point was that the holes mean nothing to me and I know they shouldn't be there at first place but I had to include them in the geometry to select a surface during the 'Domain Interface' selection. If I didn't have the holes in the shell I would have had no surface to select for interface like I do in the following image. In the pic the cold water zone is the Shell and the interface in shell region is the shell holes. The tubes carry hot water. Technically it is not a heating element but you could call it one therotically. So coming back to your question, the shell holes have no function as such and if I remove them and have one completely hollow shell I don't know what should I select for interface.

October 27, 2011, 18:36		#11
rushabh New Member Rushabh Join Date: Oct 2010 Posts: 8 Rep Power: 16	Yes sir but I would also like to see what's going on in the shell.

November 20, 2012, 18:25		#14
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	You would use a f-f interface if you did not want to explicitly model the tubes. The thermal resistance of the tubes can be added as a factor in the interface model. Doing this simplifies the model considerably compared to directly modellign the intermediate metal tubes.