pisoFoam basics...

marupio · August 12, 2012, 21:00

Hello Foamers,

How do we know what time step or mesh density to use with pisoFoam? I know this should be introductory CFD basics, but it escapes me.

The pisoFoam tutorial, for instance... it is currently set to deltaT = 0.005. If I re-run the simulation with the time step 1/10 the size (deltaT = 0.0005), the results are different. The pressure field differs by 15% and phi differs by 8%.

Any insights would be much appreciated.

-Dave

akidess · August 13, 2012, 02:51

I don't think there's a simple answer to that. There are a priori error estimates (e.g. http://www.ams.org/journals/mcom/199...18-96-00701-6/), but their use is limited. In most cases, people have a posteriori error estimates, such as rerunning with a smaller timestep or higher mesh resolution, and then estimating the error e.g. with GCI or Richardson extrapolation. Jasak's thesis also might be a good read.

marupio · August 13, 2012, 18:32

Thanks. So it is just plain old trial and error, then. Man, CFD is littered with it.

Tarak · August 16, 2012, 20:52

Hii,

Do you have any idea about what the Mean/Max Courant Number (hence specifying the time step accordingly) should be for getting meaningful results with pisoFoam?

marupio · August 16, 2012, 20:56

If this is regarding the problem I was facing, the max was very small... around 10e-6. If this is a general question, then no, that's partly what I wanted to know.

-Dave

Tarak · August 17, 2012, 12:41

That was just a general question I was curious about.

August 12, 2012, 21:00	pisoFoam basics...	#1
marupio Senior Member David Gaden Join Date: Apr 2009 Location: Winnipeg, Canada Posts: 437 Rep Power: 21	Hello Foamers, How do we know what time step or mesh density to use with pisoFoam? I know this should be introductory CFD basics, but it escapes me. The pisoFoam tutorial, for instance... it is currently set to deltaT = 0.005. If I re-run the simulation with the time step 1/10 the size (deltaT = 0.0005), the results are different. The pressure field differs by 15% and phi differs by 8%. Any insights would be much appreciated. -Dave __________________ ~~~ Follow me on twitter @DavidGaden

August 13, 2012, 02:51		#2
akidess Senior Member Anton Kidess Join Date: May 2009 Location: Germany Posts: 1,377 Rep Power: 29	I don't think there's a simple answer to that. There are a priori error estimates (e.g. http://www.ams.org/journals/mcom/199...18-96-00701-6/), but their use is limited. In most cases, people have a posteriori error estimates, such as rerunning with a smaller timestep or higher mesh resolution, and then estimating the error e.g. with GCI or Richardson extrapolation. Jasak's thesis also might be a good read. marupio likes this. __________________ On twitter @akidTwit Spend as much time formulating your questions as you expect people to spend on their answer.

August 13, 2012, 18:32		#3
marupio Senior Member David Gaden Join Date: Apr 2009 Location: Winnipeg, Canada Posts: 437 Rep Power: 21	Thanks. So it is just plain old trial and error, then. Man, CFD is littered with it. __________________ ~~~ Follow me on twitter @DavidGaden

August 16, 2012, 20:56		#5
marupio Senior Member David Gaden Join Date: Apr 2009 Location: Winnipeg, Canada Posts: 437 Rep Power: 21	If this is regarding the problem I was facing, the max was very small... around 10e-6. If this is a general question, then no, that's partly what I wanted to know. -Dave __________________ ~~~ Follow me on twitter @DavidGaden

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

August 16, 2012, 20:52		#4
Tarak Senior Member Tarak Join Date: Aug 2010 Location: State College, PA Posts: 111 Rep Power: 15	Hii, Do you have any idea about what the Mean/Max Courant Number (hence specifying the time step accordingly) should be for getting meaningful results with pisoFoam?

August 17, 2012, 12:41		#6
Tarak Senior Member Tarak Join Date: Aug 2010 Location: State College, PA Posts: 111 Rep Power: 15	That was just a general question I was curious about.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Differences in solution method for pisoFoam and buoyantBoussinesqPisoFoam	mchurchf	OpenFOAM	7	August 6, 2023 09:12
pisoFoam fails with locDynOneEqEddy	gfilip	OpenFOAM Running, Solving & CFD	2	March 17, 2014 11:32
pisoFoam : error floating point	Dbutant	OpenFOAM Running, Solving & CFD	10	October 9, 2012 09:45
Add temperature to pisoFoam	danvica	OpenFOAM Programming & Development	11	September 5, 2012 05:30
pisoFoam compiling error with OF 1.7.1 on MAC OSX	Greg Givogue	OpenFOAM Programming & Development	3	March 4, 2011 17:18