[lcbarreto]

Hi guys,

I am simulating the flow over an aircraft. I have already done my mesh convergence study and I found a optimum mesh for my case. Right now I am developing my drag polar so I am varying my angle of attack (AoA) from -10 to 20 in order to find the behavior of my CL, CD and CM.

I have tried 2 different approaches:
1) I ran my simulations, close to the stall, manually and I have tested initializing them with both standard and Hybrid methods and I am getting a stall angle of 11 deg. Above 11 I get a transient flow which is expected for stalled stages.

2) I have parametrized my AoA in order to use my table of design points and automatize my simulations. For the initialization I am using the data from the previous solution so when I ran for 11 degrees I initialize with the data obtained on my simulation with 10 deg. The simulations runs very similar to the method 1 for the linear part of the CL vs Alpha curve, but I am getting a higher stall angle (approximately 14.5 deg) and also a higher CL max And I get a convergence for CL, Cd CM, continuity, momentum and nut very qick with a factor of 10e-5.

I know that the initial guess is very important for a numerical methods, however I am not sure if it would make that difference or if there is something wrong with my simulations. Has anyone experienced this kind of behavior before?

Thanks in advance

[litzj]

I cannot memorize the title of specific paper, but load previous aoa data is appropriate approach for post stall aoa...

Maybe there is a aiaa paper dealing that issue.

vortex or separated flow is well developed when load previous aoa result

[lcbarreto]

Originally Posted by litzj

I cannot memorize the title of specific paper, but load previous aoa data is appropriate approach for post stall aoa...

Maybe there is a aiaa paper dealing that issue.

vortex or separated flow is well developed when load previous aoa result

Hi Jaesan, thanks for the reply.

So you're saying that the approach of loading previous aoa result is a better first guess for the simulation right? I just didn't think would make that big difference.

[litzj]

developing vortex or separated region on the airfoil is sensitive to initial status.

At high AoA, data from previous AoA is not only helping right solution, also helping for stability and convergence of the solution

[lcbarreto]

Originally Posted by litzj

developing vortex or separated region on the airfoil is sensitive to initial status.

At high AoA, data from previous AoA is not only helping right solution, also helping for stability and convergence of the solution

Thank you very much for the help!