the step size used in Gridgen
Hi there, I've got a confusing problem with the spacing size when using Gridgen.
As we all know, based on the grid generated by Gridgen, we could discretise NS equations on a rectangular domain(computational domain). The question is how to determine the step size since Gridgen doesn't provide any information on this. I've read many articles, it seems like the choice of stepsize in the computational domain is arbitrary. For convenience, many people choose 1.0 as the stepsize. Is it true? Thanks a lot. 
Re: the step size used in Gridgen
The grid size in computational domain will not affect the solution, rather, it is the metric coefficient that will determine the solution.

Re: the step size used in Gridgen
Thanks versi. Since the grid size plays no role in computation, what about the Courant number,which involves step size,i.e. dx?

Re: the step size used in Gridgen
hi i really want to know about the grid generation please let me inform. thankig you papiya roy

Re: the step size used in Gridgen
In curvilinear coordinates
dt = CFL/ ( U_contra + V_contra + a sqrt ( (\xi_x)^2+ (\xi_y)^2 +(\eta_x)^2+(\eta_y)^2) ), so again you see that only metric coefficents are concerned. Dx is already taken into account in metric coefficient. 
All times are GMT 4. The time now is 03:21. 