Continuity convergence problem

ankur gupta · December 13, 2010, 16:54

Hello,

I am using a 2-phase, euler model to model laminar mixing flow across static mixers.
I am using a uniform tet mesh. Using 3-D,double precision

My continuity does not seem to converge. Residual is around 1.0e-03.

I have tried -:
1) increasing mesh density
2) increasing downstream length
3) changing the under-relaxation factor, it helps only a little.

When I use two similar fluids (of same viscosity) the convergence is still better, around 2.0e-05. But when the fluids are of extremely different viscosity (ratio 1:1000), the convergence is 1.0e-03

Can somebody please help me. I really need help with this

Thanks

ComputerGuy · December 13, 2010, 19:36

2e-5 seems reasonable, although often the best way to determine convergence is to look at the flow variables (mass flow, for instance) and see the total imbalance in the system.

If the mass in is approximately equal to the mass out, the pressures are behaving as you expect, etc., you might consider the solution converged, whether or not it reaches some arbitrarily low value for residuals.

Also, how are you refining the mesh? If you're refining in Fluent and you have a number of highly skewed cells, Fluent won't really decrease the skewness. It's better to return to design modeler or gambit and simply improve the mesh that way.

ankur gupta · December 13, 2010, 20:23

hello,

thanks for your reply..

2.0e-5 is coming when the viscosities of the fluids are same..

and 1.0e-3 is coming when the viscosities are very different..(1000 times)

i am refining skewness from fluent..

but do you think creating a dense mesh near the high gradient region will help ??

December 13, 2010, 16:54	Continuity convergence problem	#1
ankur gupta New Member Ankur gupta Join Date: Aug 2010 Posts: 5 Rep Power: 16	Hello, I am using a 2-phase, euler model to model laminar mixing flow across static mixers. I am using a uniform tet mesh. Using 3-D,double precision My continuity does not seem to converge. Residual is around 1.0e-03. I have tried -: 1) increasing mesh density 2) increasing downstream length 3) changing the under-relaxation factor, it helps only a little. When I use two similar fluids (of same viscosity) the convergence is still better, around 2.0e-05. But when the fluids are of extremely different viscosity (ratio 1:1000), the convergence is 1.0e-03 Can somebody please help me. I really need help with this Thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Convergence of Continuity	Dodi	FLUENT	0	February 28, 2009 17:03
Convergence Problem	AAJ	FLUENT	2	July 3, 2008 14:31
Convergence problem for P1 & Energy	HP	FLUENT	5	May 21, 2005 16:01
Non Convergence of 3D Heat transfer cfd problem	Balraj	Main CFD Forum	3	December 9, 2004 01:24
High Continuity Residual for 3D problem	Gearoid Lydon	FLUENT	5	July 27, 2000 09:09

December 13, 2010, 19:36		#2
ComputerGuy Senior Member Real Name :) Join Date: Jan 2010 Location: United States Posts: 192 Rep Power: 16	2e-5 seems reasonable, although often the best way to determine convergence is to look at the flow variables (mass flow, for instance) and see the total imbalance in the system. If the mass in is approximately equal to the mass out, the pressures are behaving as you expect, etc., you might consider the solution converged, whether or not it reaches some arbitrarily low value for residuals. Also, how are you refining the mesh? If you're refining in Fluent and you have a number of highly skewed cells, Fluent won't really decrease the skewness. It's better to return to design modeler or gambit and simply improve the mesh that way.

December 13, 2010, 20:23		#3
ankur gupta New Member Ankur gupta Join Date: Aug 2010 Posts: 5 Rep Power: 16	hello, thanks for your reply.. 2.0e-5 is coming when the viscosities of the fluids are same.. and 1.0e-3 is coming when the viscosities are very different..(1000 times) i am refining skewness from fluent.. but do you think creating a dense mesh near the high gradient region will help ??