Transient simulation running better with larger timestep

Foxhunter · November 7, 2017, 07:08

I am simulating combustion with the premixed model, had the timestep on 0,001 and the residuals where barely converging, energy was around 1e-4, continuity 1e+1. Then I changed the timestep to 0,005 and now the residuals are converging more?! How does that work? I changed nothing else. Does Fluent simplify the calculations for larger timesteps or something?
Thanks!

litzj · November 8, 2017, 09:32

maybe some sort of turbulence motions are averaged?

this is what I guess

Foxhunter · November 9, 2017, 01:11

That could give a false indication of convergence right? The larger timesteps are not as trustworthy as smaller, harder to converge timesteps.

November 7, 2017, 07:08	Transient simulation running better with larger timestep	#1
Foxhunter New Member Join Date: Oct 2017 Posts: 26 Rep Power: 8	I am simulating combustion with the premixed model, had the timestep on 0,001 and the residuals where barely converging, energy was around 1e-4, continuity 1e+1. Then I changed the timestep to 0,005 and now the residuals are converging more?! How does that work? I changed nothing else. Does Fluent simplify the calculations for larger timesteps or something? Thanks!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Help running SST k-omega Stress-Blended Eddy Simulation	DA6righthand	FLUENT	0	July 15, 2017 10:44
Simulation FPEs - turbulence for transient and steady-state?	DaveR	OpenFOAM Running, Solving & CFD	5	March 5, 2017 15:06
Transient simulation : Static temperature and time averaged static temperature	saisanthoshm88	CFX	4	July 4, 2013 02:18
the problem of my transient simulation "Floating point exception: Overflow "	alloveyou	CFX	15	November 22, 2012 11:14
Resume Transient simulation	HMR	CFX	1	June 28, 2011 21:13

November 8, 2017, 09:32		#2
litzj Member Jaesan Yoon Join Date: Sep 2017 Posts: 36 Rep Power: 8	maybe some sort of turbulence motions are averaged? this is what I guess

November 9, 2017, 01:11		#3
Foxhunter New Member Join Date: Oct 2017 Posts: 26 Rep Power: 8	That could give a false indication of convergence right? The larger timesteps are not as trustworthy as smaller, harder to converge timesteps.