I have been trying to run a simple case of a flow over a cylinder. Though, I could get going with everything, I have not been able to fix the Reynolds number (Re=U*L/nu) (of my choice) for which I need to analyze the flow. A little investigation revealed that the solver was taking the Global length (L) as (domain size)^(1/3) where domain size=Lx*Ly*Lz, the product of domain length in all the three directions, though I was able to fix U=1.0 and nu=required value for air at standard conditions, I was not able to fix, L (global length). Physically, L is the length scale and for flow over a cylinder it is typically the diameter of the cylinder (say L=1, since my cylinder was of that diameter). But, instead it takes a value as explained above. Can anybody help me to fix this problem or tell me how to go about this.

thanks, ravi.

Hi ravi, I did the same thing to fix the reynolds number in Backward facing Step and i found the results with good agreement.You should make sure that the Re is set ; e.g in my case Re = Density*Umax*InletHeight/Viscosity i fixed the Re by varying the viscosity as it has very little influence.And all the other things are fixed(Constant). Solver computes it on the basis of Global Length. Hope this will work.

Manu

Hi Ravi,

The Reynolds number calculated by the solver has no influence on your solution. It is simply an approximate Reynolds number produced as a reference. Don't modify your domain size to set this Reynolds number to anything. Rather, you should calculate your own Reynolds number based on the velocity, length scales and material properties that are relevant to your analysis.

Regards, Robin

Hi Robin/Manu,

Thanks for the responses. This should help me to some extent.

regards, ravi.

Hi Robin,

Do you have an email ID to which I can write?. Let me know.

thanks, ravi.