Learning Forum

[janetj]

I am simulating Fx 63-137 airfoil and it is a 2-D model. However, my solution is not getting converged. In the boundary conditions for inlet, I used components and 1xcosx6 (For x component) and 1xsinx6 for y component. Goes without saying this is for 6 degree angle of attack. My analysis is 2-D steady state and flow is inviscid. Chord length is 1m.

I tried to find the solution at 1e-3 convergence and I got it. While I tried 1e-6, the convergence is not happening even after 20,000 iterations. 

And as I am simulating individually for different degree alphas, my convergence at 1e-3 was going okay from 0 to 8 degree angle of attack and these are all individual workbench projects for different alphas. Just as I started with 10 degree alpha, I wasn't able to converge the solution even at 1e-3 despite the 10,000 iterations. Why would that be?

Then I read somewhere I should try transient state simulation. I tried that and my solution was converged. May I get a good understanding of how transient state is able to converge my solution and how exactly does the time stepping is performed? 

Also, I changed the Courant number and for Implicit formulation to 0.4, 1, 2 and 3 but again solution did not converge in steady state.

I guess I should stick to steady state but how to get the solution converged for those alphas that are beyond 8 degree? Please help me in convergence as this is very important for my project.

Besides, just need a confirmation if I am doing it all right. Once the solution is converged and I obtain the Cl and Cd, say for 6 degree alpha, I enter the following values

For Cd: x = 1 cos 6 and y = 1 sin 6.

For Cl, I enter, x = 1 sin -6, y = 1 cos -6.

And thus I get the Cl and Cd at 6 degree alpha. Velocity is obviously 1m/s here. Am I doing it right?

[Rob]

I answered in the other thread, but given the problem you had creating a new thread will let you off. 

Check for flow separation and stall condition on that airfoil. You may find that as you near separation the flow field is slightly transient, so the steady solver struggles to converge the solution.