|July 2, 2011, 05:52||
convergence criterium without normalization
Join Date: May 2011
Posts: 6Rep Power: 6
im working on an free jet with STARCCM+ and try to get an converged solution.
In the Dokumentation of CD-Adapco they propose an criterium of 10^-3 for all Residuals.
For normalized Residuals i am nearly not able to keep this condition.
When i turning the normalizationoption off, it is no Problem to get Residuals under 10^-6.
What do u think is the better way to judge the Residuals. With normalization or without? And what do u think is the absolute value the residuals should fall under, when turning the normalization off?
Thanks for every help ;-) And if u dont understand because of my bad English skills ask and i will try to formulate the parts new.
|July 10, 2011, 05:38||
Join Date: Oct 2009
Posts: 637Rep Power: 13
I think, you should judge convergence by looking on some engineering values, not just residuals. For example, look on a mass flow through a boundary, lift or drag or whatever coefficients, or the values you are interested in.
|Thread||Thread Starter||Forum||Replies||Last Post|
|convergence criterium for unsteady 3D VOF||jack1980||FLUENT||2||February 28, 2010 10:04|
|Convergence of CFX field in FSI analysis||nasdak||CFX||2||June 29, 2009 01:17|
|the criterium of convergence for the SIMPLE algo||Lam||Main CFD Forum||1||July 25, 2007 09:08|
|Defect correction and convergence||ganesh||Main CFD Forum||4||June 30, 2006 14:20|
|HELP!Convergence Criterium||Teresa||FLUENT||4||July 27, 2005 11:53|