Help needed

ethanhunt134 · December 12, 2011, 17:17

Hi,

suggestions needed for my current simulation.

I am currently running a heat transfer and flow model. There is mass flow inlet and heat sources and mesh is fine enough.

I can run either flow or thermal model by deleting the other regions, but cannot run the entire model. Solver stops saying AMG solver diverged or floating point exception. Couldn't get much help from support.

Can anybody advice me on this issue.

Thanks in advance.

willimanili · December 13, 2011, 11:38

Some more information about your simulation model would be helpful to identify potential causes for your problem.

Providing you are using segregated solvers, try to decrease the underrelaxing factors.

ethanhunt134 · December 13, 2011, 11:44

@willimanili thank you for the reply.

I am using segregated solvers and I have tried decreasing the under-relaxation factors but it doesn't change anything.

I have tried several things like, I added some some initial conditions obtained from cold flow model, fluid velocity, temperature. Unfortunately, nothing has changed.

Thanks

willimanili · December 14, 2011, 02:32

If you simulate steady state, try this one:
Freeze the energy solver since the solution for the flow is nearly converged. Than unfreeze the solver at really low underrelaxing factors (<0.1). You can increase them manually step by step to get an faster convergence.
If this also doesnt work check the mesh for mesh quality. Especially at the locations where the first pressure or temperature limit overstepping take place.

Zeppo · April 1, 2012, 04:55

In CCM+ 5.04 it's impossible to use segregated solver when you have at least 2 regions (fluid and solid) connected by an interface. Conjugate heat transfer requires coupled solver to be activated.

LuckyTran · April 23, 2012, 15:25

Quote:

Originally Posted by Zeppo

In CCM+ 5.04 it's impossible to use segregated solver when you have at least 2 regions (fluid and solid) connected by an interface. Conjugate heat transfer requires coupled solver to be activated.

You can still do this with segregated solver or coupled solver, the interface takes care of the coupling of the two problems.

You are mixing up terms.

December 12, 2011, 17:17	Help needed	#1
ethanhunt134 New Member Join Date: Dec 2011 Posts: 2 Rep Power: 0	Hi, suggestions needed for my current simulation. I am currently running a heat transfer and flow model. There is mass flow inlet and heat sources and mesh is fine enough. I can run either flow or thermal model by deleting the other regions, but cannot run the entire model. Solver stops saying AMG solver diverged or floating point exception. Couldn't get much help from support. Can anybody advice me on this issue. Thanks in advance.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
a CFD software is needed...	maxgzc	Main CFD Forum	3	September 27, 2011 01:51
Mesh generation software is needed	H.Dou	Main CFD Forum	12	May 4, 2011 15:20
attachDetach: Not all zones and patches needed in the definition have been found.	schmittp54	OpenFOAM Running, Solving & CFD	2	October 20, 2010 09:26
a dissertation in ETD needed very much	LiQiang	Main CFD Forum	0	March 11, 2005 03:09
CFD Software Engineer needed in Fort Worth, Texas	Ross	Main CFD Forum	0	November 19, 2001 14:48

December 13, 2011, 11:38		#2
willimanili Member Join Date: Feb 2011 Location: DE-PB Posts: 56 Rep Power: 3	Some more information about your simulation model would be helpful to identify potential causes for your problem. Providing you are using segregated solvers, try to decrease the underrelaxing factors.

December 13, 2011, 11:44		#3
ethanhunt134 New Member Join Date: Dec 2011 Posts: 2 Rep Power: 0	@willimanili thank you for the reply. I am using segregated solvers and I have tried decreasing the under-relaxation factors but it doesn't change anything. I have tried several things like, I added some some initial conditions obtained from cold flow model, fluid velocity, temperature. Unfortunately, nothing has changed. Thanks

December 14, 2011, 02:32		#4
willimanili Member Join Date: Feb 2011 Location: DE-PB Posts: 56 Rep Power: 3	If you simulate steady state, try this one: Freeze the energy solver since the solution for the flow is nearly converged. Than unfreeze the solver at really low underrelaxing factors (<0.1). You can increase them manually step by step to get an faster convergence. If this also doesnt work check the mesh for mesh quality. Especially at the locations where the first pressure or temperature limit overstepping take place.

April 1, 2012, 04:55		#5
Zeppo New Member Join Date: Dec 2009 Posts: 22 Rep Power: 4	In CCM+ 5.04 it's impossible to use segregated solver when you have at least 2 regions (fluid and solid) connected by an interface. Conjugate heat transfer requires coupled solver to be activated.