What difference between results obtained with different meshes is acceptable to say that no additional refinement is needed?

I would say 5%, but it can depend on a lot of issues!!

Have you got the Ercoftac domumentation yet? It gives guidelines to a lot of these basic questions you're asking

allan

Hi Allan,

No, don't have the Ercoftac documentation. How can I get it?

Thanks for your response,

Julie

Have you ever heard of Richardson extrapolation? This gives you as well a mean to check for the mesh independent solution. See also Ferziger & Peric, Computational Methods for Fluid Dynamics, Springer, 2001

How exactly do I perform the check for independent solution using Richardson extrapolation (for example, applying the material given here: http://www.math.ubc.ca/~burggraf/math256/richardson.pdf)?

go to their website:

http://www.ercoftac.org

click on the link ercoftac publications then best practice guidelines

allan