Question about accelerating the convergence by changing the timestep

Anna Tian · September 15, 2012, 18:55

Hi,

We often change the time step in the middle of the simulation running to accelerate the convergence. For example, according to the FAQ of this forum, we can set the time step to local time step with factor of 5 at the beginning. Then when the RMS does not decrease anymore, we can change the time step to let it continue to decrease.

But how do we know what time step should I change to?

ghorrocks · September 16, 2012, 08:50

The CFX documentation describes setting the time step - a good starting point is the fluid residence time in the domain. But note this is only a guide - if the simulation is converging well a larger time step is possible, if convergnce is difficult a smalle rtiem step will be required.

September 15, 2012, 18:55	Question about accelerating the convergence by changing the timestep	#1
Anna Tian Senior Member Meimei Wang Join Date: Jul 2012 Posts: 494 Rep Power: 16	Hi, We often change the time step in the middle of the simulation running to accelerate the convergence. For example, according to the FAQ of this forum, we can set the time step to local time step with factor of 5 at the beginning. Then when the RMS does not decrease anymore, we can change the time step to let it continue to decrease. But how do we know what time step should I change to? __________________ Best regards, Meimei

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
problem with Min/max rho	tH3f0rC3	OpenFOAM	8	July 31, 2019 10:48
Convergence of CFX field in FSI analysis	nasdak	CFX	2	June 29, 2009 02:17
Convergence criterion	Moose	Main CFD Forum	5	June 9, 2005 21:39
timestep in transinet simulations	caty	CFX	1	April 18, 2005 20:54
A question about NS convergence!	zqnwpu	Main CFD Forum	1	September 10, 2003 09:00

September 16, 2012, 08:50		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	The CFX documentation describes setting the time step - a good starting point is the fluid residence time in the domain. But note this is only a guide - if the simulation is converging well a larger time step is possible, if convergnce is difficult a smalle rtiem step will be required.