i need the best setup and setting for compressible flow test
 

أهلا شباب ...
.Note : all figures in this problem are inside this picture

http://imageshack.us/a/img703/9779/3344562121.jpg

I'm trying to calculate the Lift coefficinet of naca64012 ( with chord = 10 ) at different angels of attack and at difference speed (0.25 mach (incomp), and 0.7 mach (comp) 0
I used Structured mesh as shown in the figure A .

At 0.25 mach (incompressible flow ) test , i obtained Reasonable results of lift
coefficients at different .angles as shown in figure 2
Then i try to calculate at M=0.7 ( compressible flow ) by using the same mash
that used in the incompressible flow test ,
and the i put setup as that :
solver : figure 2
Viscous model : i select to ( K omega ) < figure 3
Materials : figure 4
But when i try that , i obtained results until angel of attack around 8 ,
they are nearly like that :
angle 0 CL 0
angle 2 CL 0.3
angle 4 CL 0.65
angle 6 CL 0.75
angle 8 CL

and i didnot obtained results at angles that more than , and there are comments appeared in The fluent console ( as shown in figure 5 )0

and the CL VS iterations divergence by high ratio as shown in this pic
http://imageshack.us/a/img22/5204/62015177.jpg

i want to know what are the problems ???
and what is the most suitable setup and setting for compressible flow test ?

Hi Blueaero,
I think you need to correct the initial spacing of your mesh to have an appropriate yplus value because you want to use the same mesh for different Mach numbers? (What about the Reynolds number, it will remain the same too?)

I think you also need to use the pressure-farfield boundary condition for the compressible flow.

You should also provide the proper reference area to calculate your lift and drag via Report Reference values.

As in your case its 2D, you use reference area= 10 x 1 = 10m^2.

My advice would be to start first with the Spalart-Allmaras turbulence model this should also give good result and this is also more robust due to its ease in implementation resulting in economical resources (this is just a comment for your information). Anyways their is no proper rule to choose specific turbulence model.

Best of luck and regards.

How i can correct the initial spacing of the mesh ???
is it bt using GAMBIT ???

( The Reynolds number remains same ^_^ )

If i can change it in Fluent , tell me how ? ^_^

Hi
is there a reason why you use the density-based solver? I always prefere the pressure-based one. You need a stable setup before you think about the turbulenz (btw. I recommend the k-w-SST). So my recommendation is: Pressure-based, coupled pv-coupling, maximum second order upwind advection, last squares cell based gradient, SST-model. I did calculations with this setup which worked perfektly on transsonic flows.
Oh one last thing: Fluent is a bad solver on meshes with an aspect ratio bigger than round about 100. If you want a finer layer-region, use CFX :D.
Greetings