Fully implicit scheme

johndeas · June 24, 2008, 12:14

Maybe a dumb question but... How is it that Fluent is capable of fully-implicit time discretisation while OpenFOAM is not ?
Or, if put another way, what would be required to have fully-implicit time discretisation in OF ?

eugene · June 24, 2008, 12:39

What do you mean by "fully implicit" time discretization, and why do you think there is a difference between Fluent and OpenFOAM in this regard?

billy · June 24, 2008, 12:46

You would need to solve the equations of mass, momentum and energy simultaneously instead of segregated.

johndeas · June 24, 2008, 12:49

For example, you won't have limitation on your Courant Number in Fluent.

This question arised as the Courant number can be a limitating factor and force you to have timesteps smaller than the one required by the eddy dissipation time.

eugene · June 24, 2008, 13:11

Most of the time schemes in OpenFOAM are "fully implicit". There are probably other things that Fluent does to run stable at very high CFL numbers.

On the other hand I regularly run LES with maximum CFL numbers of 2-3 without problems. It depends on the solver and your limiters I guess.

hjasak · June 24, 2008, 13:50

Sorry Eugene - time to read a book. The guy is talking about a block solution of the compressible flow system, which is "more implicit" than the segregated solution you are used to.

This is coming with my new block matrix implementation.

Hrv

liu · June 24, 2008, 13:59

Hrv,
"This is coming with my new block matrix implementation"

When? I have been looking forward for this for a long time.

hjasak · June 24, 2008, 14:37

As you may have guessed, it is all about funding: fully financed projects get priority, deadlines etc, meaning that interesting pieces of work get delayed.

It has been delayed for an inexcusable number of times, but it will make you happy to know that some people are already running block-coupled solutions.

Apologies for the delay,

Hrv

billy · June 24, 2008, 14:46

We should thank you for developing such a great software!

liu · June 24, 2008, 18:23

Great! Hrv

I should also thank you for sharing.

pbohorquez · June 25, 2008, 08:40

Dear Hrv

I wondered if the community can enjoy such piece of work via the development version? This would be thankful news (among others).

Best wishes
Patricio

June 24, 2008, 12:14	Maybe a dumb question but... H	#1
johndeas Senior Member John Deas Join Date: Mar 2009 Posts: 160 Rep Power: 17	Maybe a dumb question but... How is it that Fluent is capable of fully-implicit time discretisation while OpenFOAM is not ? Or, if put another way, what would be required to have fully-implicit time discretisation in OF ? Kummi likes this.

June 24, 2008, 12:39	What do you mean by "fully imp	#2
eugene Senior Member Eugene de Villiers Join Date: Mar 2009 Posts: 725 Rep Power: 21	What do you mean by "fully implicit" time discretization, and why do you think there is a difference between Fluent and OpenFOAM in this regard?

June 24, 2008, 12:46	You would need to solve the eq	#3
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	You would need to solve the equations of mass, momentum and energy simultaneously instead of segregated.

June 24, 2008, 12:49	For example, you won't have li	#4
johndeas Senior Member John Deas Join Date: Mar 2009 Posts: 160 Rep Power: 17	For example, you won't have limitation on your Courant Number in Fluent. This question arised as the Courant number can be a limitating factor and force you to have timesteps smaller than the one required by the eddy dissipation time.

June 24, 2008, 13:11	Most of the time schemes in Op	#5
eugene Senior Member Eugene de Villiers Join Date: Mar 2009 Posts: 725 Rep Power: 21	Most of the time schemes in OpenFOAM are "fully implicit". There are probably other things that Fluent does to run stable at very high CFL numbers. On the other hand I regularly run LES with maximum CFL numbers of 2-3 without problems. It depends on the solver and your limiters I guess.

June 24, 2008, 13:50	Sorry Eugene - time to read a	#6
hjasak Senior Member Hrvoje Jasak Join Date: Mar 2009 Location: London, England Posts: 1,905 Rep Power: 33	Sorry Eugene - time to read a book. The guy is talking about a block solution of the compressible flow system, which is "more implicit" than the segregated solution you are used to. This is coming with my new block matrix implementation. Hrv Kummi likes this. __________________ Hrvoje Jasak Providing commercial FOAM/OpenFOAM and CFD Consulting: http://wikki.co.uk

June 24, 2008, 13:59	Hrv, "This is coming with my	#7
liu Senior Member Xiaofeng Liu Join Date: Mar 2009 Location: State College, PA, USA Posts: 118 Rep Power: 17	Hrv, "This is coming with my new block matrix implementation" When? I have been looking forward for this for a long time. __________________ Xiaofeng Liu, Ph.D., P.E., Assistant Professor Department of Civil and Environmental Engineering Penn State University 223B Sackett Building University Park, PA 16802 Web: http://water.engr.psu.edu/liu/

June 24, 2008, 14:37	As you may have guessed, it is	#8
hjasak Senior Member Hrvoje Jasak Join Date: Mar 2009 Location: London, England Posts: 1,905 Rep Power: 33	As you may have guessed, it is all about funding: fully financed projects get priority, deadlines etc, meaning that interesting pieces of work get delayed. It has been delayed for an inexcusable number of times, but it will make you happy to know that some people are already running block-coupled solutions. Apologies for the delay, Hrv Kummi likes this. __________________ Hrvoje Jasak Providing commercial FOAM/OpenFOAM and CFD Consulting: http://wikki.co.uk

June 24, 2008, 14:46	We should thank you for develo	#9
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	We should thank you for developing such a great software!

June 24, 2008, 18:23	Great! Hrv I should also th	#10
liu Senior Member Xiaofeng Liu Join Date: Mar 2009 Location: State College, PA, USA Posts: 118 Rep Power: 17	Great! Hrv I should also thank you for sharing. __________________ Xiaofeng Liu, Ph.D., P.E., Assistant Professor Department of Civil and Environmental Engineering Penn State University 223B Sackett Building University Park, PA 16802 Web: http://water.engr.psu.edu/liu/

June 25, 2008, 08:40	Dear Hrv I wondered if the	#11
pbohorquez Member Patricio Bohorquez Join Date: Mar 2009 Location: Jaén, Spain Posts: 95 Rep Power: 17	Dear Hrv I wondered if the community can enjoy such piece of work via the development version? This would be thankful news (among others). Best wishes Patricio

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Is it feasible to apply implicit scheme in LES	EagleGao	Main CFD Forum	2	October 24, 2011 14:04
implicit scheme for time step	vivian	Main CFD Forum	4	June 23, 2006 03:47
time_step, implicit scheme, number of iterations	Isa	Main CFD Forum	3	May 11, 2006 12:27
How to choose the CFL number with implicit scheme?	D.T.	Main CFD Forum	2	October 23, 2002 13:56
Implicit Scheme for 1-D nozzle flow	Yogesh Talekar	Main CFD Forum	1	June 8, 1999 09:49