[Vesta]

Hey all, it's me again.. I am currently working on simulating the flow around a NACA0012 profile. My current problem is that I get very high deviations (about 30%) in the simulated drag coefficient compared to the one from the experiment (NACA Report 460). The lift coefficient on the other hand has a much better value with 1% deviation.

About the Set up:
- automated poly mesh
- Fluid temperature
- SST K Omega
- AoA ~ 11°

The experiments were performed at Re = 3.2E6 and my simulation has a Reynods number of 7.8E5. Is this the reason or does the comparison make sense because of the different Reynolds numbers? Do you have any suggestion what I should look at again?


Many thanks in advance!

[Chris2337]

Hey, if you're still having problems try the gamma transition model with a turbulence intensity of 0.001.

With the higher Reynolds number you should be getting a lower drag coefficient, but without the transition model it effectively models the whole boundary layer as turbulent thus overpredicting the drag.

[Vesta]

Thank you very much for your answer!

I tried it, but it didn't really change anything in my case. I then removed it to save computational effort. I tried it with a larger domain and looked again for experimental values that fit better in terms of the Reynolds number and are documented in a table. Now it is OK

[Aeronautics El. K.]

Originally Posted by Vesta

Hey all, it's me again.. I am currently working on simulating the flow around a NACA0012 profile. My current problem is that I get very high deviations (about 30%) in the simulated drag coefficient compared to the one from the experiment (NACA Report 460). The lift coefficient on the other hand has a much better value with 1% deviation.

About the Set up:
- automated poly mesh
- Fluid temperature
- SST K Omega
- AoA ~ 11°

The experiments were performed at Re = 3.2E6 and my simulation has a Reynods number of 7.8E5. Is this the reason or does the comparison make sense because of the different Reynolds numbers? Do you have any suggestion what I should look at again?


Many thanks in advance!

So you're comparing apples with oranges. Start with matching the Reynolds number first.
Secondly, the boundary layer in the wind tunnel test is possibly tripped. But good luck trying to match the onset of transition if you use a transition model. In my opinion you'll be better off with a fully turbulent boundary layer right from the start.
Also, as the name of the report suggests, the tests have been done in a variable density wind tunnel. So make sure you replicate the conditions.

That should get you going.