|
[Sponsors] |
May 21, 2012, 14:00 |
Resetting all fields and mesh motion
|
#1 |
New Member
Pal Schmitt
Join Date: Aug 2010
Location: Belfast
Posts: 21
Rep Power: 16 |
Dear all,
I am attempting to write some implicit solver for a body moved by a fluid. I would like to move the body boundary, calculate the fluid, obtain accelerations and then reset to where I started and repeat as often as I need before advancing the timestep. The .prevIter(); seems to store all fields all right, so before entering into my implicit loop I am able to save all my fields like this. Inside the implicit loop I use something like Code:
U=U.prevIter(); I can run a case without moving the mesh and it seems to repeat every timestep a couple of times which shows i am OK, I suppose. Moving meshes run but diverge, so I must be resetting something badly. The same case as an explicit method works, so with my additional iterations it should only take longer... I use the same for cellDisplacement and pointDisplacement as above. But how do i reset the mesh motion best? Is something like this correct? Code:
mesh.movePoints(mesh.oldPoints()); Any comments are much appreciated, Cheers, Pal |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Converting Starccm+ mesh | Ladnam | OpenFOAM | 0 | September 14, 2011 07:30 |
Problem of reconstructing pointVectorFields with parallel mesh motion (1.5.x) | LESlie | OpenFOAM | 0 | November 6, 2009 07:07 |
PostChannel | maka | OpenFOAM Post-Processing | 5 | July 22, 2009 10:15 |
Fixed mesh method for rigid body motion | philippose | OpenFOAM | 1 | January 12, 2009 05:57 |
Problems with mesh motion in parallel | thomas | OpenFOAM Running, Solving & CFD | 3 | July 4, 2007 03:48 |