hello all; I'm writing a computer code simulating flow about an airfoil, in case of 2D, incompressible, laminar and viscous flow.Reynolds number equal 2x10**3. Governing equations were adimentionalised by setting U=u/ue and V=v/ue, .......(ue is the upstream velocity) My question is about the boundary condition at upstream, is it U=1, or U=Re*visc/C Re: Reynolds number Vis: Kinematic vicosity of air C: Corde

Thanx

Dear Bouencer,

The upstream BC should be U = ue/ue = 1. The kinematic viscosity would then become Vis/Re, after non-dimensionalisation.

Regards,

Ganesh

Yes,

Ganesh is right. Also, to add more :

u said:

" is it U=1, or U=Re*visc/C"

It should be u=Re*visc/C and not 'U'. Here 'u' has dimensions. But U doesnt have. If u calculate 'u' based on the above Re and divide by ue, u shud get '1' at the upstream.

regards,

Dominic

Thanx Ganesh, Please i've an other question: I'm using C-grid generation, so the upstream side in physical domain become the top side in computational domain, how to write the boundary condition U=1 (condition in physical domain in terms of cartesian coordinates) in computational domain where i must express U in terms of the new curvillinear coordinates ksi and eta

Hi,

If U' represents the solution in computational domain, and U in the physical domain then,

U' = [T] U, Where [T] = | d(x,y)/d(ksi,eta) |

T(1,1)=dx/dksi ; T(1,2)=dy/dksi ; T(2,1)=dx/deta; T(2,2)=dy/deta;

Here [T] is the Jacobian of the transformation..

Knowing U, you can compute U'...

regards, Dominic

thanx Dominic; If y're agree i'll send you the mathematical model in cartesian and curvillinear coordinates with descriptions, the finite volumes steps discretization and my code program in Fortran.

[imah]

hai all.. im also writting computer code using fortran.. i try to simulate airfoil in c-grid.. my problem is.. i not sure the boundary condition at double connected grid (after the trailing edge).. hope u all can help me.. thanks..