|
[Sponsors] |
November 15, 2006, 04:27 |
AMR or Octree
|
#1 |
Guest
Posts: n/a
|
Hello,
I want to implement a code that use immersed boundary method to solve navier-stokes equations. My question is about mesh refinement and particularly I try to compare the number of cells that are needed in an octree mesh and in an AMR mesh. Any advice is welcome. Thanks |
|
November 16, 2006, 01:00 |
Re: AMR or Octree
|
#2 |
Guest
Posts: n/a
|
altouth AMR (Adaptive Mesh Refinement) is general concept and in fact its mesh refinement based geometry gradient (curvature) is equivalent with octree but in some literature AMR is called for hierarchical structured grid refinement (only difference with single level is due to coupling), but octree is called for locally refined grid, so implementation of AMR is simpler but its number of grids is very larger than octree especialy for complex geometry and large scale problems.
One waekness of octree is smoothing due to essential interpolations while in AMR final solution is on the finest grid and we don't have this problem |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[ICEM] Octree surface mesh smoothing | siw | ANSYS Meshing & Geometry | 11 | August 9, 2012 03:13 |
Octree Vs Advancing Front tetra meshing techniques | Sam | CFX | 4 | February 18, 2008 09:54 |
About AMR Library | C.-G. CHEN | Main CFD Forum | 0 | October 31, 2007 01:34 |
Octree-block data structure for AMR | xyuan | Main CFD Forum | 0 | July 3, 2007 18:36 |
AMR with Incompressible Flow | Shahriar | Main CFD Forum | 7 | March 7, 2003 09:53 |