|
[Sponsors] |
Why implicit solvers are stable at more cfl no |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
August 12, 2005, 16:33 |
Why implicit solvers are stable at more cfl no
|
#1 |
Guest
Posts: n/a
|
Hi, I just was wondering what makes implicit solvers more stable as compared with explicit ones in terms of CFL number.. Cheers endee
|
|
August 12, 2005, 20:10 |
Re: Why implicit solvers are stable at more cfl no
|
#2 |
Guest
Posts: n/a
|
This can be answered on so many levels, but it's such a basic question that almost any textbook on CFD will give you an answer. The CFL condition basically states that within any explicit time step, flow information is not allowed to propagate farther than from one grid point to the adjacent grid point. Given a grid and flow, this condition provides a maximum limit for the time step.
The exact stability condition will depend on the numerical scheme, and for linear equations you can find it by a von Neumann stability analysis as described in your textbook. The same stability analysis on implicit schemes will reveal that they are more stable (even unconditionally stable). You could argue that the reason is given by the implicit connection of all grid points in such schemes. In explicit schemes only the immediate neighbors (within the stencil) exchange information within each time step. Hence the CFL condition. With implicit schemes, the connectivity ideally reaches through the entire grid. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
implicit / explicit solvers | jk75 | FLOW-3D | 5 | March 10, 2017 15:32 |
Suggested unsteady, implicit solver stable with arbitrarily large time steps | djbungee | OpenFOAM Programming & Development | 45 | March 23, 2015 05:14 |
Implicit Reimann solvers | harish_0689@ | Main CFD Forum | 0 | September 18, 2011 14:09 |
CFL condition for implicit CD schemes | turb | Main CFD Forum | 4 | August 16, 2005 13:56 |
How to choose the CFL number with implicit scheme? | D.T. | Main CFD Forum | 2 | October 23, 2002 14:56 |