Multigrid Cycle

protomucca · December 10, 2009, 10:45

Hi, what is the difference between Multigrid V-cycle, W-cycle and F-cycle? I'm studying 2D subsonic (Mach about 1,15 - 2,0) flow over an airfoil, and comparing different turbulence model.
I've saw that with F-Cycle the convergence is more smooth than with the other cycle type, it's normal?

tele · December 11, 2009, 04:24

Hi,

Philosophy:

The solution of the linear system is not calculated on the original grid (see below 1), but will be calculated on a coarse grid (in order to reduce its dimension...) and then the solution will pe prolonged on the original mesh.

Ex: V cycle:
Actual grid - delx - 1: o o or o o ...
Coarse grid - 2*delx - 2: o o o
Coarse grid - 4*delx - 3: o
...

Ex: W cycle:
Actual grid - delx - 1: o o o or o o ...
Coarse grid - 2*delx - 2: o o o o o
Coarse grid - 4*delx - 3: o o
....

F cycle means flexible cycle and the code will choose for you the best strategy.

The convergence depends on the structure of the linear system beeing solved. In your case it seems that the F cycle "behaves" better for your special setup.

December 10, 2009, 10:45	Multigrid Cycle	#1
protomucca New Member Daniele Join Date: Jul 2009 Posts: 2 Rep Power: 0	Hi, what is the difference between Multigrid V-cycle, W-cycle and F-cycle? I'm studying 2D subsonic (Mach about 1,15 - 2,0) flow over an airfoil, and comparing different turbulence model. I've saw that with F-Cycle the convergence is more smooth than with the other cycle type, it's normal?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
LES by Multigrid	mnabi	Main CFD Forum	8	July 11, 2009 20:21
multigrid v cycle for laplace equation	devalpandya	Main CFD Forum	0	April 22, 2009 04:02
multigrid	sureshkumar	Main CFD Forum	0	June 9, 2006 01:20
MULTIGRID at the wall	Jim	Main CFD Forum	0	February 19, 2003 12:22
Multigrid applied to k-e models	Paulo Zandonade	Main CFD Forum	9	May 24, 1999 08:10

December 11, 2009, 04:24		#2
tele New Member Ioan Join Date: Dec 2009 Posts: 5 Rep Power: 16	Hi, Philosophy: The solution of the linear system is not calculated on the original grid (see below 1), but will be calculated on a coarse grid (in order to reduce its dimension...) and then the solution will pe prolonged on the original mesh. Ex: V cycle: Actual grid - delx - 1: o o or o o ... Coarse grid - 2delx - 2: o o o Coarse grid - 4delx - 3: o ... Ex: W cycle: Actual grid - delx - 1: o o o or o o ... Coarse grid - 2delx - 2: o o o o o Coarse grid - 4delx - 3: o o .... F cycle means flexible cycle and the code will choose for you the best strategy. The convergence depends on the structure of the linear system beeing solved. In your case it seems that the F cycle "behaves" better for your special setup.