CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Mesh refinement (http://www.cfd-online.com/Forums/main/12224-mesh-refinement.html)

Beginner September 14, 2006 11:00

Mesh refinement
 
Hi everybody,

I've developed a small program to simulate incompressible bidimentional flows using finite differences.

I've read that refine mesh near boundaries gives better results but I don't know how to implement this technique.

The basic idea is to have the finest mesh near the boundary and increment it. To do this, I think that I need to define a function. But how and what else I need to define?

Hope that my question was clear so if you have any good course, paper, link or whatever else on the subject, thank you in advance.

Iain Barton September 22, 2006 09:21

Re: Mesh refinement
 
If you are dealling with Cartesian meshes (which assumably you must be since using finite differences), the usual rule is to apply an expoential growth of 1.1 to 1.4.

i.e.

first cell height = h second cell height = h * 1.1 nth cell height = h * 1.1^(n-1)

In very old CFD books you see they use functions, to find either the cell height or position. But this doesn't seem to be the current vogue in grid generation.

Iain

ps (ignoring turbulent flow simulations) you want to refine your grid close to your boundary because as you want the change in "properties" across a cell to remain reasonably constant in magnitude.



All times are GMT -4. The time now is 18:12.