Hi all CFD experts!

I am currently carried out a project which is regarding modelling wingtip vortex.

The model is a naca0012 wing with aspect ratio of 5. free stream mach number is 0.5 Ma. The Re is about 6.5e^5.

I use SA as turbulence model. The lift coefficient and drag coefficient is get different from the experimental results from a wind tunnel test.

My concern is:-

will my cd and cl obtained influenced by modelling of tip vortex? if yes, what turbulence model should i used?

Hi Rooney,

You might wanna check out the wind tunnel conditions and see if the walls are affecting the flow. Also, is it wildly different from what you're getting? As for the turbulence models, S-A is sufficient for external flows, so, also check your boundary conditions and mesh quality near the boundaries.

http://www.cfd-online.com/Wiki/Spalart-Allmaras_model

regards, QX-2

Thanks QX-2

Thanks for the reply. The wind tunnel test is actually from a NACA report "characteristic of low-aspect ration wings at supercritical mach number" by John Stack and W.F.Lindsay.

my model was a box with a wing attached on a symmetry plane. The wall which wing attached to, the boundary was set to symmetry. freestream condition was set for the side,top,front,bottom faces.pressure outlet was set at farfield behind the wing.

I am using prism layer mesh of 15 layers,near wall thickness prism layer was 1e-06m.but i have an awkward TE meshing of the wing. prism layer mesh converge to a point at trailing edge, making poor meshing at the TE. Nothing can be done to improve the TE mesh even though prism layer parameter like convex angle, concave angle, boundary mach number is adjusted.

a)Do you know is there a way to avoid meshing at the TE to converge to a point?

b)I suspect mesh at wingtip to capture wingtip vortex give me poor results. Do you know which part of the wing should be refine?

c)Is this justify to used SA model to capture wingtip vortex?If i am not wrong, eddy viscosity model of turbulence model assume flow's strain is isotropic.is wingtip vortex anisotropic?should I used Reynolds Stress Transport model?

any suggestion?