unsteady simulation

hamid1 · February 4, 2011, 09:43

Hello guys

for unsteady simulation should one first have a converged solution in steady and then shift to unsteady or thats is not necessarily compulsory?

Amir · February 4, 2011, 10:55

Hi hamid,
it's not necessary to initialize unsteady problem with converged steady one.
but in some special cases like small perturbations and ... it helps you to have fewer iterations per time steps.

hamid1 · February 4, 2011, 11:41

thanks Amir,
but if I dont have a steady converged initialized solution does it mean that I cant rely on my unsteady simulation at the first few steps of time duration?

Amir · February 4, 2011, 13:47

your answer in unsteady simulation is reliable if and only if the residuals after each time step satisfy your convergence criteria. but as I said before , in some special cases if you have a good initialization data field , the convergence criteria may takes place in fewer iteration per time step . (this tip doesn't change the generality of the first sentence)

itsuetian@hotmail.com · February 4, 2011, 16:37

For unsteady simulation it is not mandatory to have converged solution in steady and then shift to unsteady simulation.You can start with unsteady simulation and can rely on that.I have worked on the unsteady simulation of flow arround cylinders and it gives reliable results.

February 4, 2011, 09:43	unsteady simulation	#1
hamid1 Senior Member hamid Join Date: Nov 2010 Posts: 185 Rep Power: 16	Hello guys for unsteady simulation should one first have a converged solution in steady and then shift to unsteady or thats is not necessarily compulsory?

February 4, 2011, 13:47		#4
Amir Senior Member Amir Join Date: May 2009 Location: Montreal, QC Posts: 735 Blog Entries: 1 Rep Power: 23	your answer in unsteady simulation is reliable if and only if the residuals after each time step satisfy your convergence criteria. but as I said before , in some special cases if you have a good initialization data field , the convergence criteria may takes place in fewer iteration per time step . (this tip doesn't change the generality of the first sentence) Last edited by Amir; February 5, 2011 at 09:04.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Adding heat source to chtMultiRegionFoam	maddalena	OpenFOAM Programming & Development	61	February 17, 2018 09:33
Unsteady vs Steady simulation - any difference?	Sally	Main CFD Forum	16	December 16, 2005 10:15
About unsteady flow simulation	Ji	Fidelity CFD	3	June 10, 2003 11:04
Procedure to run unsteady simulation?	STN	Main CFD Forum	2	February 16, 2002 05:37
1D simulation of unsteady gas flows in pipelines	Andrew Chernousov	Main CFD Forum	1	September 5, 1998 00:42

February 4, 2011, 10:55		#2
Amir Senior Member Amir Join Date: May 2009 Location: Montreal, QC Posts: 735 Blog Entries: 1 Rep Power: 23	Hi hamid, it's not necessary to initialize unsteady problem with converged steady one. but in some special cases like small perturbations and ... it helps you to have fewer iterations per time steps.

February 4, 2011, 11:41		#3
hamid1 Senior Member hamid Join Date: Nov 2010 Posts: 185 Rep Power: 16	thanks Amir, but if I dont have a steady converged initialized solution does it mean that I cant rely on my unsteady simulation at the first few steps of time duration?

February 4, 2011, 16:37		#5
itsuetian@hotmail.com New Member Farhan Join Date: Dec 2010 Posts: 4 Rep Power: 15	For unsteady simulation it is not mandatory to have converged solution in steady and then shift to unsteady simulation.You can start with unsteady simulation and can rely on that.I have worked on the unsteady simulation of flow arround cylinders and it gives reliable results.