|March 18, 2003, 16:33||
Re: Turbulence Modeling
Well, turbulent flows are characterized by a wide range of scales (both spatial and temporal). With the actual computer power, the idea of solving for all the scales (no modeling - DNS approach) is only affordable for simple geometries with a limited range a scales involved.
Therefore, for engineering purpose, and/or in case of complex flows, one has to separate the scales so that you don't have to solve for all of them. With this idea,some scales are going to be resolved (usually the larger scales) and some others will be modeled. That is the idea of turbulence modeling!!
Now as far as the modeling is concern, several approach are possible: RANS, LES, Galerkine-POD ... It is up to you to decide which approach is the most appropriate in terms of cost, reliability and accuracy!!
For more insight on: Why do we need a turbulence model?, just grab any book on turbulence modeling (like "Turbulence modeling for CFD" by Wilcox, 1993).
I hope this helps. Sincerely,
|Thread||Thread Starter||Forum||Replies||Last Post|
|Turbulence postprocessing||Mohsin||FLUENT||2||October 3, 2016 14:18|
|A small confusion regarding terminology of turbulence modeling||Chander||CFX||6||September 28, 2011 20:05|
|Discussion: Reason of Turbulence!!||Wen Long||Main CFD Forum||3||May 15, 2009 09:52|
|May Focus Area: Turbulence Modeling||Jonas Larsson||CFD-Wiki||0||May 2, 2006 11:48|