The end of unsteady flow

flow_CH · October 29, 2013, 11:43

Hi.
I am modeling an unsteady flow in fluent(implicit). With changing the time step, i can achieve to converging, but after converging the iteration is continue, so How can i understand the simulation is done and i can export the results?
Thanks

-mAx- · October 30, 2013, 00:42

by decreasing (I suppose) the time step size (dt), your solution converges faster FOR A TIME STEP (t). That means the solver switches on next time step (t+dt) untill time step t reaches the number of time steps you set (N)
In this example, dt = 5e-5 s
N = 2000
That means calculation will be over for t = N * dt = 2000 * 5e-5 = 0.1 s

flow_CH · October 30, 2013, 05:12

Thank you for reply
1- In my case, by decreasing the time step size (dt), the residuals are going up and more diverge .
2- I do not anything about the time of ending the flowing in my model, so which time i should stop the processing?

-mAx- · October 30, 2013, 05:26

1- weird
2- it depends on what you want to compute (your physics)

flow_CH · October 30, 2013, 05:41

Can i stop processing when the results do not change after more iterations?

-mAx- · October 30, 2013, 06:12

For sure
(you should ask: "Can i stop processing when the results do not change after more time steps")

flow_CH · October 30, 2013, 07:35

I am asking: Can i stop processing when the results do not change after more time steps ?

-mAx- · October 30, 2013, 07:44

Quote:

Originally Posted by -mAx-

For sure

I already answered, so yes you can

October 29, 2013, 11:43	The end of unsteady flow	#1
flow_CH Senior Member FHydro Join Date: Jan 2013 Posts: 197 Rep Power: 13	Hi. I am modeling an unsteady flow in fluent(implicit). With changing the time step, i can achieve to converging, but after converging the iteration is continue, so How can i understand the simulation is done and i can export the results? Thanks

October 30, 2013, 00:42		#2
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	by decreasing (I suppose) the time step size (dt), your solution converges faster FOR A TIME STEP (t). That means the solver switches on next time step (t+dt) untill time step t reaches the number of time steps you set (N) In this example, dt = 5e-5 s N = 2000 That means calculation will be over for t = N * dt = 2000 * 5e-5 = 0.1 s __________________ In memory of my friend Hervé: CFD engineer & freerider

October 30, 2013, 05:26		#4
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	1- weird 2- it depends on what you want to compute (your physics) __________________ In memory of my friend Hervé: CFD engineer & freerider

October 30, 2013, 06:12		#6
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	For sure (you should ask: "Can i stop processing when the results do not change after more *time steps*") __________________ In memory of my friend Hervé: CFD engineer & freerider

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Simulation of a single bubble with a VOF-method	Suzzn	CFX	21	January 29, 2018 00:58
No results for solid domain	Gary Holland	CFX	10	March 13, 2009 03:30
Concentric tube heat exchanger (Air-Water)	Young	CFX	5	October 6, 2008 23:17
error using combination of step function	xujjun	CFX	1	January 15, 2008 16:46

October 30, 2013, 05:12		#3
flow_CH Senior Member FHydro Join Date: Jan 2013 Posts: 197 Rep Power: 13	Thank you for reply 1- In my case, by decreasing the time step size (dt), the residuals are going up and more diverge . 2- I do not anything about the time of ending the flowing in my model, so which time i should stop the processing?

October 30, 2013, 05:41		#5
flow_CH Senior Member FHydro Join Date: Jan 2013 Posts: 197 Rep Power: 13	Can i stop processing when the results do not change after more iterations?

October 30, 2013, 07:35		#7
flow_CH Senior Member FHydro Join Date: Jan 2013 Posts: 197 Rep Power: 13	I am asking: Can i stop processing when the results do not change after more *time steps ?*