temperature contour is unexpected

fluidflow · October 16, 2012, 12:46

hi all
i want to simulate fluid in a cavity so i solve it is cfx
but at cfx post when i want to see the temperature contour (about 95% )area of cavity is in avg temp but the boundary is the my temp that i specified at B.C it means that the temp changed very fast near the hot and cold wall that are B.C and i do not why i do not have distribution for temp?
PLEASE HELP ME...............

Possa · October 16, 2012, 14:39

Hello meysam,

Your post is a little confuse, so maybe I'm wrong here but I believe that what you are trying to say is that you have two adjacent walls with temperature boundary condition and these temperatures are not correct when you visualize in CFX Post. It varies on the border of the walls where it is conected one to another. Is that it?

If that is the case, first you have to be sure to visualize the variable with the hybrid option and not conservative to see the B.C.

Another thing is that CFX tries to correct your boundary condition to avoid discontinuity. So it interpolates the temperature values at the border of the walls to try to satisfy both B.C.

I hope that was your question,
Regards,
Possa.

ghorrocks · October 16, 2012, 19:24

This question is a FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

fluidflow · October 17, 2012, 04:33

thanks for your helpful answer
you are true but when i see the hybrid option and not conservative the B.C is correct(as you observe the line at the picture below) but my problem is chenging the temp very fast and reach to average temp near the wall both cold and hot wall(that are my B.C) and i think i must have distribution temp on all domain but i do not have
............

fluidflow · October 17, 2012, 04:38

thanks for your helpful link but i could not find my answer

ghorrocks · October 17, 2012, 06:58

Please note point 2 on the FAQ - has you simulation fully converged? You might simply need to run the simulation for a longer time.

Please post an image of what you are simulating and the CCL file.

fluidflow · October 17, 2012, 11:50

ok i think this problem will solve with this ........
turbulence and mass are convergence 10e-6 and energy imblance is really near zero but heat transfer not decrease under 1.6e-5 and only change near the number do you think is it the reason of problem?and do you have any way to solve it?
this is its pic:
http://www.keepandshare.com/doc/4837...y-jpg-52k?da=y
thanks a lot Glenn Horrocks

ghorrocks · October 17, 2012, 18:40

I cannot view your attachment. Just use the advanced features to attach it direct to the forum.

But your problem is likely to be that the thermal field takes much longer to establish than the fluid field. So the fluid variables are converged but the heat is not. The heat imbalance is the parameter to keep an eye on here. In these situations you can usually increase the time step size to achieve full convergence quicker.

This is all discussed in the FAQ

October 16, 2012, 12:46	temperature contour is unexpected	#1
fluidflow New Member meysam Join Date: Sep 2012 Posts: 5 Rep Power: 14	hi all i want to simulate fluid in a cavity so i solve it is cfx but at cfx post when i want to see the temperature contour (about 95% )area of cavity is in avg temp but the boundary is the my temp that i specified at B.C it means that the temp changed very fast near the hot and cold wall that are B.C and i do not why i do not have distribution for temp? PLEASE HELP ME...............

October 16, 2012, 19:24		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	This question is a FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F fluidflow likes this.

October 17, 2012, 04:33	hello possa	#4
fluidflow New Member meysam Join Date: Sep 2012 Posts: 5 Rep Power: 14	thanks for your helpful answer you are true but when i see the hybrid option and not conservative the B.C is correct(as you observe the line at the picture below) but my problem is chenging the temp very fast and reach to average temp near the wall both cold and hot wall(that are my B.C) and i think i must have distribution temp on all domain but i do not have ............

October 17, 2012, 04:38	hi ghorrocks	#5
fluidflow New Member meysam Join Date: Sep 2012 Posts: 5 Rep Power: 14	thanks for your helpful link but i could not find my answer

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Calculation of the Governing Equations	Mihail	CFX	7	September 7, 2014 07:27
Unexpected large decrease in Total temperature across the shock using sonicFoam	JLight	OpenFOAM	0	December 16, 2010 12:46
Unexpected large decrease in Total temperature across the shock using sonicFoam	JLight	OpenFOAM Running, Solving & CFD	0	December 15, 2010 06:47
Temperature contour as thermal condition in Workbench?	nuothe11th	ANSYS	0	April 23, 2010 13:11
temperature contour plot is discrete (why)?	susan_w_b	FLUENT	0	January 5, 2004 04:36

October 16, 2012, 14:39		#2
Possa New Member Talita Possamai Join Date: Sep 2012 Posts: 23 Rep Power: 14	Hello meysam, Your post is a little confuse, so maybe I'm wrong here but I believe that what you are trying to say is that you have two adjacent walls with temperature boundary condition and these temperatures are not correct when you visualize in CFX Post. It varies on the border of the walls where it is conected one to another. Is that it? If that is the case, first you have to be sure to visualize the variable with the hybrid option and not conservative to see the B.C. Another thing is that CFX tries to correct your boundary condition to avoid discontinuity. So it interpolates the temperature values at the border of the walls to try to satisfy both B.C. I hope that was your question, Regards, Possa.

October 17, 2012, 06:58		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Please note point 2 on the FAQ - has you simulation fully converged? You might simply need to run the simulation for a longer time. Please post an image of what you are simulating and the CCL file.

October 17, 2012, 11:50		#7
fluidflow New Member meysam Join Date: Sep 2012 Posts: 5 Rep Power: 14	ok i think this problem will solve with this ........ turbulence and mass are convergence 10e-6 and energy imblance is really near zero but heat transfer not decrease under 1.6e-5 and only change near the number do you think is it the reason of problem?and do you have any way to solve it? this is its pic: http://www.keepandshare.com/doc/4837...y-jpg-52k?da=y thanks a lot Glenn Horrocks

October 17, 2012, 18:40		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	I cannot view your attachment. Just use the advanced features to attach it direct to the forum. But your problem is likely to be that the thermal field takes much longer to establish than the fluid field. So the fluid variables are converged but the heat is not. The heat imbalance is the parameter to keep an eye on here. In these situations you can usually increase the time step size to achieve full convergence quicker. This is all discussed in the FAQ