If I have a model of a open channel with, lets say, a square obstacle on the middle.

with the following conditions:

- obstacle nodes N = 21
- I have a grid resolution of 160x800
- Reynolds Re = 100: so vortex shedding occurs
- velocity = u_m = 0.1
- kinematic viscosity --> (21-1)*0.1/100 so I have a relaxation time by isolating the kinematic viscosity from a fixed reynolds and a fixed velocity.

Is it correct to make a mesh refinement of the following plan?

- I double the nodes on the x and y domain so that the grid resolution is now 320x1600
- therefore the nodes of the obstacle N = 41 (20*2+1)
- therefore to keep the Reynolds I increase the kinematic viscosity, now is (41-1)*0.1/100

and thatâ€™s it?

If it is not correct I would appreciate your appreciations to do it correctly.

Thank you in advanced.

puigar