I need some ideas or references on how to treat sharp corners like the ones we come across in lid driven cavity flow or flow over a flatplate.

Increase mesh density in those regions.

Try & maintain fairly evenly-proportioned elements - ie. tending towards square (if 4 sided), or isosceles (if 3 sided).

Make sure your mesh is well blended back into regions of lower mesh density.

Most importantly - time your dynamics around these regions.

See if it is possible to use higher-order elements in those corners, or over the whole mesh. P2 is good.

desA

I do not use unstructured grid or a finite volume solver.I use simple finite differences on a structured grid.I tried to maintain a the spacing pretty much constant in both directions close to the leading edge and it seems to work.

Thanks

using the finite difference there really isnt much you can do except, as diaw said, make your spacing very small. But, that will lead to a lot of computations. The trick is the boundary conditions at the corner. Also, you should get Ghia's paper and compare results. Gupta also has a good paper on it. That will, more or less, let you know if you are on the right track. Depending on your Re number you will get recirculation zones in the corners.

It's been a while since I last used FD, but, I would wonder if you could think of using a higher order stencil in these corner regions? Perhaps you could run up your system matrix as usual & then bring in the high-order stencil?

desA