Residuals criteria

Kwiaci · March 25, 2012, 10:23

I want to ask what absolute criteria for residuals do you use in yours analyses? What value do you use?
For example - is 10e-12 a satisfactory convergence absolute criteria? Is there any theory providing the guidance in this manner?

murx · March 26, 2012, 05:04

The convergence criteria depends on your simulation and on your objective.

You should probably check how the convergence criterion affects the parameter you are interested in. Start with a weak convergence criterion, e.g. 10e-4 and have a look at the solution parameter of interest. Then do more simulations and lower the convergence criterion step by step. When your solution does not change significantly anymore, you have reached your satisfactory convergence criterion.

To my knowledge, 10e-12 for RMS or max residual is generally a pretty strict convergence criterion.

March 25, 2012, 10:23	Residuals criteria	#1
Kwiaci Member Robert Join Date: Mar 2011 Location: Warsaw Posts: 62 Rep Power: 15	I want to ask what absolute criteria for residuals do you use in yours analyses? What value do you use? For example - is 10e-12 a satisfactory convergence absolute criteria? Is there any theory providing the guidance in this manner?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Tutorials] Tutorial of how to plot residuals !	wolle1982	OpenFOAM Community Contributions	172	December 6, 2024 07:20
under-relaxation factors -> level of residuals	Zigainer	FLUENT	19	July 21, 2017 17:53
what to monitor besides residuals?	franzdrs	FLUENT	5	March 21, 2013 04:59
How can I set the converge criteria properly?	Alina	FLUENT	3	September 7, 2007 17:55
Convergence - scaled vs unscaled residuals	HS	FLUENT	1	November 7, 2005 06:45

March 26, 2012, 05:04		#2
murx Member Max Join Date: May 2011 Location: old europe Posts: 88 Rep Power: 15	The convergence criteria depends on your simulation and on your objective. You should probably check how the convergence criterion affects the parameter you are interested in. Start with a weak convergence criterion, e.g. 10e-4 and have a look at the solution parameter of interest. Then do more simulations and lower the convergence criterion step by step. When your solution does not change significantly anymore, you have reached your satisfactory convergence criterion. To my knowledge, 10e-12 for RMS or max residual is generally a pretty strict convergence criterion.