Hello to everybody:

I have a doubt about Reynolds number. I know that is defined like: Re=(characteristic velocity * characteristic length)/(viscosity).

For my problem, I have:

-A cube(in 3-dimension) of (2*pi)*(2*pi)*(2*pi), so is it supposed that characteristic length is 2*pi?

-I know the initial values (t=0) for the 3 components of the velocity (u, v, w), what would be the characteristic velocity?

-What is the range of Reynolds number for working with steady flows, for working with nonsteady flows and for working with fully nonsteady flows?

Thank-you in advance. Best regards Isabel

The Reynolds number appears in the equations if you nondimensionalize the data of your problem. Usually, the problem data comes from a real world problem. You seem to have an artificial problem. If your data is not O(1), you should rescale it and use the rescaling to define your Reynolds number. If your data is O(1) you should choose a Reynolds number depending on what kind of flow you are interested.

Last, the Reynold's number has nothing to do with the flow being unsteady or steady, but with turbulence.

Your,

P. Birken

Hi,

What kind of flow is it? Is it a geometry with a wall? If there are no walls (isotropic turbulence) the characteristic length scale should be based on some characteristic length scale of the turbulence, not the size of the box. This length scale can be the integral length scale or the Taylor micro scale, see the text books on turbulence for the definitions. A typical velocity scale is the root mean square of the velocity fluctuations or some mean/bulk velocity.

Hope this helps,

Tom