[ajinkya1]

While running a simulation on NACA 0012 airfoil for steady state solution for a compressible flow problem with mach no. of 0.1 and angle of attack of 18, 20. Using JST and Runge Kutta explicit, running it for 20000 reynolds no., direct problem, CFL no. of 1.0, CFL ramp: 1.0,50, 2.0, MG level -0, spatial order flow- 2nd order limiter, I am not able to get a steady state solution even after 3 lacs iterations. Cl value is not converging. I have put convergence criterion as Residual, Residual reduction as 10, min val = -10. Can anyone suggest the probable reason why i am not getting a steady state solution.

[hlk]

Originally Posted by ajinkya1

While running a simulation on NACA 0012 airfoil for steady state solution for a compressible flow problem with mach no. of 0.1 and angle of attack of 18, 20. Using JST and Runge Kutta explicit, running it for 20000 reynolds no., direct problem, CFL no. of 1.0, CFL ramp: 1.0,50, 2.0, MG level -0, spatial order flow- 2nd order limiter, I am not able to get a steady state solution even after 3 lacs iterations. Cl value is not converging. I have put convergence criterion as Residual, Residual reduction as 10, min val = -10. Can anyone suggest the probable reason why i am not getting a steady state solution.

Thank you for your interest in SU2, and for providing me with the opportunity to learn that 1 lac = 1e5

Are you running viscous or inviscid? Please check the case most similar to yours in the TestCases folder (probably under either the euler or rans folders) to make sure that the boundary conditions and freestream values are set correctly.
Note that if the physical case is naturally unsteady (as might be the case at a high angle of attack which may experience separation) that you may get oscillations in the force coefficients when you try to solve it as a steady problem.

[Sachchit]

Hi Ajinkya,

NACA 0012 at Angle of Attack as high as 18 or 20 will have flow separation. This will give rise to wake region behind the airfoil which is unsteady in nature. In this case the steady state solver residuals, become periodic and rather than going down and down they oscillate about a value. So the criteria of -10 will never be reached.

To verify this first check conservative variable residuals oscillating and also check Cd and Cl plots oscillating. Magnitude of oscillation should be small. To capture the unsteady nature. Restart the solution with the solution generated for steady case and use unsteady global time steppping. Then you can capture the unsteady phenomena.

It is also helpful to use Cauchy convergence criteria to terminate the solution after oscillations are becoming periodic.