I am still quite new to grid generation. I have been reading a few posts on this forum which states that if the grid stretching ratio is bigger than around 1.2, there 'll be errors.

However, must this practise be followed even for region such as far upstream/downstream ie places where the velocity/pressure is more or less constant? Can the ratio be higher e.g. 1.5 or even 2.0?

By the way, I am using my own 2D FVM fractional step structured solver.

1.2 is a rule of thumb, and not a hard and fast rule. Ideally, in regions of strong gradients even 1.2 might be a little high, while in regions of small gradients 1.5 may be perfectly acceptable. Of course, what you'll find is that even sticking to 1.2 will give very large cells after you get any distance away from the body.

I'm using too my own 2D FVM to study incompressible flow about airfoil. I've used SIMPLER algorithm and second order upwind scheem for convective terms. although i've used verry small relaxation parameters (2*10**-2) my code program converge to 10**-4 and diverge again. I don't know if the problem is in the formulation of boundary conditions in numerical domain. If y're agree i'm ready to send more details and discuss further ideas.

I just touch grid generation, and I hope to know the function of logarithm grid. Could you please give me some help? LXT