Everyone, What is mesh free methods? How is it possible to do CFD analysis without mesh generation first? Any idea? Regards, Joe

i actually am thinking one thing , if we do have meshless methods, why it is that finite volume solvers are so popular????

The term mesh-free is used in two contexts: for methods which are of the boundary-element type where you dont need an interior mesh and for methods which are of particle/lagrangian type. Recently there is considerable activity in the second type of methods. They are becoming very popular in structural mechanics, see the review paper by Belytschko et al. (available online). This reference and more links can be found here.

Also another interesting paper Do finite volume methods really need a mesh ?

my reason for thinking in this direction was this, for some time i am trying to construct an algorithm by which we shall eb able to handle _almost meshless_ cases. by _almost meshless_ i mean that , user provides surface mesh and interior point cloud.

i will start writing code about it in mid of next month (i am already writting a unstructured grid solver that uses multigrid acceleration ... i aspect it to finish by start of next month).

i belive that one of the reason for using finite volume methods is multigrid acceleration ..that it converges very fast, i guess meshless methods are not so fast on convergence ,

someone with experience with meshless type of method can clear this thing.

by the way i looked at the paper u mentioned , my idea is very much similar but not exactly the same. its nice to see that some people think alike.