CFD Online Logo CFD Online URL
Home > Forums > Main CFD Forum

local time stepping

Register Blogs Members List Search Today's Posts Mark Forums Read

LinkBack Thread Tools Display Modes
Old   February 28, 2006, 12:42
Default local time stepping
Posts: n/a

I would like to ask something about local time stepping and I hope somebody can help me.

Would you say that this method is applied to a solver to increase stability - so a steady problem can be treated as "unsteady" with a time step which is dependent on the CFL number?

  Reply With Quote

Old   February 28, 2006, 16:03
Default Re: local time stepping
Posts: n/a
No. Local time stepping is used as a solution acceleration method - it can occasionally lead to robustness problems as the solution is no longer entirely 'physical', except in the limit t -> infinity (i.e. the 'steady' solution).
  Reply With Quote

Old   March 1, 2006, 01:04
Default Re: local time stepping
Posts: n/a
you actually lose the temporal accuracy if you use the local time stepping. It is mainly used when we have a grid clustered on one side. So for large cells the CFL does not demand the small time step that a small cell would need. Hence local time stepping is used to accelarate the steady state convergence. It actually means that the different cells move in different time regimes if at all it makes any sense !!!. No physics here. A considerable lot of people are not comfortable with the local time stepping due to this aspect. You can use Dual time stepping to regain the temporal accuracy and use local time stepping for unsteady problems.

  Reply With Quote

Old   March 1, 2006, 03:06
Default Re: local time stepping
Posts: n/a
Sorry, but now I am completely mixed up.

Could you just try to give me another short explanation (or a free reference)?
  Reply With Quote

Old   March 3, 2006, 05:26
Default Re: local time stepping
Posts: n/a

Whatever the informations which we are applying interms boundary conditions(Physics of the problem)... cant move more than one grid space per unit time. So we have to give optimum timestep sothat the information shouldnt jump more than one grid space per unit time...

This is the funda of CFL number... Next Local and Global timestep..

Local timestep:: In each cell the available information moves with local velocity.. so Local timestep= (Size of the cell/ Local speed)

Global timestep is calculated based on max speed of information otherwise min timestep required :: Global timestep= Min. all( size of cell/ corres.infor speed)

I hope it will help

Regards Mohammed

  Reply With Quote


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On

Similar Threads
Thread Thread Starter Forum Replies Last Post
Dynamic moving mesh Pei-Ying Hsieh (Hsieh) OpenFOAM Running, Solving & CFD 64 June 7, 2012 10:04
alphaEqn.H in twoPhaseEulerFoam cheng1988sjtu OpenFOAM Bugs 14 September 18, 2011 13:46
Error while running rhoPisoFoam.. nileshjrane OpenFOAM Running, Solving & CFD 8 August 26, 2010 12:50
BlockMeshmergePatchPairs hjasak OpenFOAM Native Meshers: blockMesh 11 August 15, 2008 07:36
Info on local time stepping Frank Muldoon Main CFD Forum 1 October 9, 1998 07:44

All times are GMT -4. The time now is 06:28.