Issue with domain interface

shgho · October 21, 2013, 11:59

Hello everyone!
I am simulating heat transfer in a solid domain coupled with fluid flow in an adjacent domain. The interface between the two is GGI type with conservative interface flux. In CFX setup, I prescribed a liquid inflow temperature of 25 C and a fixed temperature of 25 C for the solid wall.
But in the results, the solid wall temperature has for some reason been violated near the interface. I hope the attached image can make things clearer. This image is from results of a converged mesh which makes me unsure as to whether this is a meshing issue. What am I missing here?
Thanks in advance.

ghorrocks · October 21, 2013, 17:32

Can you describe what you are modelling? What is the fluid - is it compressible? turbulent?

shgho · October 21, 2013, 23:41

I am doing conjugate heat transfer with a solid around which there is a liquid (water, as available in CFX materials library) flowing in the axial direction. The cold end of the solid is at 25 C and the hot end is at 65 C. Heat will be conducted in the solid from 65 C to 25 C. The water is flowing from the cold side of the solid to the hot side. The idea is for water to pick up heat along the length of the solid and redeposit it at the hot (65 C) side. The fluid is incompressible and the flow is laminar.

ghorrocks · October 21, 2013, 23:43

Can you show a more complete picture of the temperature field?

shgho · October 22, 2013, 00:24

Here is the temperature plot.
The middle domain is solid and both the left and right domains are water.

ghorrocks · October 22, 2013, 05:14

Yes, and what does your top image show? Where is that? You really need to explain what you are doing if we are to have any hope of helping you.

shgho · October 22, 2013, 05:37

Here is the top image you wanted. I have zoomed in near the interface so that the variation can be observed properly.
Is there some other info you need? I have already stated the boundary conditions and the nature of the interface.

ghorrocks · October 22, 2013, 06:59

This is probably just the boundary getting a bit confused with hybrid/conservative values. If that is the case then it is just a post-processing artefact and is not important.

shgho · October 22, 2013, 07:29

Thanks for all your help.
In the plots that I sent you, the values were all hybrid. In that case, shouldn't the real boundary condition values have been displayed?

ghorrocks · October 22, 2013, 07:34

Yes, but on which boundary? The interface or the wall? The wall is defined as one temperature and the interface will be at another - so how does the post-processor render that? That is why I suggest it is just a post-processing issue.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Wind turbine simulation	Saturn	CFX	58	July 3, 2020 01:13
Domain Imbalance	HMR	CFX	5	October 10, 2016 05:57
GGI and 1:1 Domain Interface connections	isabel	CFX	3	April 2, 2007 05:37
Replace periodic by inlet-outlet pair	lego	CFX	3	November 5, 2002 20:09

October 21, 2013, 17:32		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Can you describe what you are modelling? What is the fluid - is it compressible? turbulent?

October 21, 2013, 23:41		#3
shgho New Member Join Date: Oct 2013 Posts: 5 Rep Power: 12	I am doing conjugate heat transfer with a solid around which there is a liquid (water, as available in CFX materials library) flowing in the axial direction. The cold end of the solid is at 25 C and the hot end is at 65 C. Heat will be conducted in the solid from 65 C to 25 C. The water is flowing from the cold side of the solid to the hot side. The idea is for water to pick up heat along the length of the solid and redeposit it at the hot (65 C) side. The fluid is incompressible and the flow is laminar.

October 21, 2013, 23:43		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Can you show a more complete picture of the temperature field?

October 22, 2013, 05:14		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Yes, and what does your top image show? Where is that? You really need to explain what you are doing if we are to have any hope of helping you.

October 22, 2013, 06:59		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	This is probably just the boundary getting a bit confused with hybrid/conservative values. If that is the case then it is just a post-processing artefact and is not important.

October 22, 2013, 07:29		#9
shgho New Member Join Date: Oct 2013 Posts: 5 Rep Power: 12	Thanks for all your help. In the plots that I sent you, the values were all hybrid. In that case, shouldn't the real boundary condition values have been displayed?

October 22, 2013, 07:34		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Yes, but on which boundary? The interface or the wall? The wall is defined as one temperature and the interface will be at another - so how does the post-processor render that? That is why I suggest it is just a post-processing issue.