Question about time step setting for different turbulence modeling

Anna Tian · November 16, 2012, 08:18

Hi,

May I ask shall I use different time step for different turbulence modeling? Why?

My simulation converges well with K-Omega model with time step = 10^(-4)s. Shall I use a larger or smaller time step for reynolds stress models?

atmcfd · November 19, 2012, 03:09

A very vague answer, but yes., and I am not sure of the time step you need for the RS model, without knowing the case.
Generally, It also depends upon the numerical scheme you are using for timestepping (implicit or explicit). Usually, if you are not sure, I would suggest running smaller time steps for any turbulence model initially.this is to be safe and make sure the simulation does not become unstable. Some turbulence models which focus a lot on explicitly "resolving" considerably small turbulent scales would definitely require small time steps, eg highly resolved LES , on a very fine grid. Likewise, flows in cases where you expect to see large transients would also need smaller time steps, since it would better resolve the transients you want to see. More importantly, the simulation would not diverge (unless of course, there are errors elsewhere, like B.Cs etc..)

November 16, 2012, 08:18	Question about time step setting for different turbulence modeling	#1
Anna Tian Senior Member Meimei Wang Join Date: Jul 2012 Posts: 494 Rep Power: 16	Hi, May I ask shall I use different time step for different turbulence modeling? Why? My simulation converges well with K-Omega model with time step = 10^(-4)s. Shall I use a larger or smaller time step for reynolds stress models? __________________ Best regards, Meimei

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
InterFoam negative alpha	karasa03	OpenFOAM	7	December 12, 2013 04:41
plot over time	fferroni	OpenFOAM Post-Processing	7	June 8, 2012 08:56
Problem with FloatingObject	Leech	OpenFOAM Running, Solving & CFD	10	March 29, 2012 16:24
directMapped problem	panda60	OpenFOAM Bugs	4	July 8, 2010 11:23
Warning 097-	AB	Siemens	6	November 15, 2004 05:41

November 19, 2012, 03:09		#2
atmcfd Senior Member ATM Join Date: May 2009 Location: United States Posts: 104 Rep Power: 17	A very vague answer, but yes., and I am not sure of the time step you need for the RS model, without knowing the case. Generally, It also depends upon the numerical scheme you are using for timestepping (implicit or explicit). Usually, if you are not sure, I would suggest running smaller time steps for any turbulence model initially.this is to be safe and make sure the simulation does not become unstable. Some turbulence models which focus a lot on explicitly "resolving" considerably small turbulent scales would definitely require small time steps, eg highly resolved LES , on a very fine grid. Likewise, flows in cases where you expect to see large transients would also need smaller time steps, since it would better resolve the transients you want to see. More importantly, the simulation would not diverge (unless of course, there are errors elsewhere, like B.Cs etc..)