solution won't converge

sieginc. · March 10, 2012, 03:10

Can anyone tell me why these results are oscillating as such? The plots are surface averages of the temperature. I have have air blowing on a block with heat flux on each surface, and am looking at the temperatures. The rear side will not converge as you can see from the pictures. It moves from side to side.

Obviously the temperature is affected by the air flow, so I took a velocity cut and sure enough there are regions of air that keep changing back and forth in certain directions. I have a volumetric control around this entire region and the mesh is pretty fine. Does anyone know what else I can do to fix this problem?

abdul099 · March 12, 2012, 03:13

Ever thought about the physics behind your problem?
For me it looks like the flow is not steady, which is fine. You will have separations behind most blunt bodies, causing vortices to separate. Therefore the velocity (and hence the temperature) behind your body CAN NOT be uniformly.
Use any search engine and look for "Karman vortex street", there are plenty of explanations.

sieginc. · March 13, 2012, 17:51

So I should switched from "steady" to "unsteady" then? Some of the shrouds I use do not cause this oscillating result. I guess it depends on the characteristics of the vortices that form.

abdul099 · March 15, 2012, 03:20

Either you ignore the oscillations or you should switch to unsteady. But even that will not make them disappear, it will just catch your time dependent physics better.
Depending on what you want to achieve, it's maybe the easiest to average your values over some oscillations.

March 10, 2012, 03:10	solution won't converge	#1
sieginc. Member adam Join Date: Oct 2011 Posts: 52 Rep Power: 14	Can anyone tell me why these results are oscillating as such? The plots are surface averages of the temperature. I have have air blowing on a block with heat flux on each surface, and am looking at the temperatures. The rear side will not converge as you can see from the pictures. It moves from side to side. Obviously the temperature is affected by the air flow, so I took a velocity cut and sure enough there are regions of air that keep changing back and forth in certain directions. I have a volumetric control around this entire region and the mesh is pretty fine. Does anyone know what else I can do to fix this problem?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Naca 0012 (compressible and inviscid) flow convergence problem	bipulsaha	FLUENT	1	July 6, 2011 07:51
simple algorithm, converge of solution	bbasal	Main CFD Forum	0	February 13, 2010 02:58
Solution Does Not Converge	Rif	Main CFD Forum	3	March 7, 2008 05:33
My Solution wont Converge!!!	Chris	FLUENT	6	April 16, 2005 18:57
what to do when solution doesn't converge?	Jenny	CFX	5	January 28, 2004 04:00

March 12, 2012, 03:13		#2
abdul099 Senior Member Join Date: Oct 2009 Location: Germany Posts: 636 Rep Power: 21	Ever thought about the physics behind your problem? For me it looks like the flow is not steady, which is fine. You will have separations behind most blunt bodies, causing vortices to separate. Therefore the velocity (and hence the temperature) behind your body CAN NOT be uniformly. Use any search engine and look for "Karman vortex street", there are plenty of explanations.

March 13, 2012, 17:51		#3
sieginc. Member adam Join Date: Oct 2011 Posts: 52 Rep Power: 14	So I should switched from "steady" to "unsteady" then? Some of the shrouds I use do not cause this oscillating result. I guess it depends on the characteristics of the vortices that form.

March 15, 2012, 03:20		#4
abdul099 Senior Member Join Date: Oct 2009 Location: Germany Posts: 636 Rep Power: 21	Either you ignore the oscillations or you should switch to unsteady. But even that will not make them disappear, it will just catch your time dependent physics better. Depending on what you want to achieve, it's maybe the easiest to average your values over some oscillations.