interTrackFoam timestep

virginie_e · April 3, 2009, 06:28

Hello,

I am trying to simulate the flow of a viscous liquid (viscosity 100) over a slope, and I am having problems with the tolerated timestep.

Indeed, the more viscous my liquid is, the smaller the timestep should be for my simulation to run. I managed to make it bigger by choosing a CrankNicholson ddt scheme and by implementing a PISO loop rather than a SIMPLE one but it is still very low.

I am quite astonished because in Fluent, I can get far bigger timestep for the same kind of simulation (but without the moving meshes).

Could someone explain to me why it is so? Or could give me any advice on how I could get a bigger timestep for my simulation. Thank you.

Virginie

sega · April 3, 2009, 07:30

Hi!

What is your setting for maxCo in controlDict?

I experienced far better solution times with interDyMFoam when using the GAMG solver. I don't know if this is applicable to interTrackFoam...

deepsterblue · April 3, 2009, 11:42

Can you quantify the timeStep? What order of magnitude is it?

Simulations involving interface tracking use explicit surface updates, even if the bulk fluid is solved implicitly. Convection and diffusion is solved implicitly in most solvers that I've seen in FOAM.

virginie_e · April 6, 2009, 05:20

Hello,

I did not set a maxCo because it is not originally taken into account in interTrackFoam (and at the beginning I did not want to change anything in the solver). So the maximum time step I could set was 2.10-5 (for a viscosity of 100) and my Co values are about 0.001. By reading the user guide, I thought that I could hace Co values which could reach about 0.2...

And if I set a greater timestep than 2.10-5, I get a floating point exception very quickly.

I tried to change the viscosity of my fluid, and typically, with a viscosity of 10, I could reach a timestep of 2.10-4... I am astonished, I would have thought that I would have got a lower time step.

Thank you for your help anyway.

Virginie

virginie_e · April 6, 2009, 06:02

Sorry I had forgotten some information:

the order of magnitude is: 0.05 m.s-1
the typical size of the edges of my mesh is of order 0.005 m
and my total simulation domain is over a 1m length

hope it will help

Virginie

April 3, 2009, 06:28	interTrackFoam timestep	#1
virginie_e Member Virginie Ehrlacher Join Date: Mar 2009 Posts: 52 Rep Power: 17	Hello, I am trying to simulate the flow of a viscous liquid (viscosity 100) over a slope, and I am having problems with the tolerated timestep. Indeed, the more viscous my liquid is, the smaller the timestep should be for my simulation to run. I managed to make it bigger by choosing a CrankNicholson ddt scheme and by implementing a PISO loop rather than a SIMPLE one but it is still very low. I am quite astonished because in Fluent, I can get far bigger timestep for the same kind of simulation (but without the moving meshes). Could someone explain to me why it is so? Or could give me any advice on how I could get a bigger timestep for my simulation. Thank you. Virginie

April 3, 2009, 07:30		#2
sega Senior Member Sebastian Gatzka Join Date: Mar 2009 Location: Frankfurt, Germany Posts: 729 Rep Power: 20	Hi! What is your setting for maxCo in controlDict? I experienced far better solution times with interDyMFoam when using the GAMG solver. I don't know if this is applicable to interTrackFoam... __________________ Schrödingers wife: "What did you do to the cat? It's half dead!"

April 3, 2009, 11:42		#3
deepsterblue Senior Member Sandeep Menon Join Date: Mar 2009 Location: Amherst, MA Posts: 403 Rep Power: 25	Can you quantify the timeStep? What order of magnitude is it? Simulations involving interface tracking use explicit surface updates, even if the bulk fluid is solved implicitly. Convection and diffusion is solved implicitly in most solvers that I've seen in FOAM. __________________ Sandeep Menon University of Massachusetts Amherst https://github.com/smenon

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Timestep in Free Surface Wigley Hull	Carlos AndrÃ©s	CFX	8	April 29, 2007 17:38
Timestep	Joe	CFX	4	August 23, 2006 07:51
Use of Timestep in obtaining solution.	hagupta	CFX	7	February 28, 2006 14:14
Timestep selection	Jindra Kosprdova	CFX	9	April 28, 2005 07:41
timestep in transinet simulations	caty	CFX	1	April 18, 2005 20:54

April 6, 2009, 05:20		#4
virginie_e Member Virginie Ehrlacher Join Date: Mar 2009 Posts: 52 Rep Power: 17	Hello, I did not set a maxCo because it is not originally taken into account in interTrackFoam (and at the beginning I did not want to change anything in the solver). So the maximum time step I could set was 2.10-5 (for a viscosity of 100) and my Co values are about 0.001. By reading the user guide, I thought that I could hace Co values which could reach about 0.2... And if I set a greater timestep than 2.10-5, I get a floating point exception very quickly. I tried to change the viscosity of my fluid, and typically, with a viscosity of 10, I could reach a timestep of 2.10-4... I am astonished, I would have thought that I would have got a lower time step. Thank you for your help anyway. Virginie

April 6, 2009, 06:02		#5
virginie_e Member Virginie Ehrlacher Join Date: Mar 2009 Posts: 52 Rep Power: 17	Sorry I had forgotten some information: the order of magnitude is: 0.05 m.s-1 the typical size of the edges of my mesh is of order 0.005 m and my total simulation domain is over a 1m length hope it will help Virginie