[DDG]

Hi everybody,

I'm analysing a transonic rotor at ISA conditions (rotational speed w=20260 rpm), I've used the Total P Inlet - Mass flow outlet boundary conditions. Firstly I used the Auto-timescale option, which did give me some sort of convergence but the Turbulence residuals were oscillating a lot. I used then a Physical timescale (time factor= 4*E-5) but although the residuals were reaching a plateau the oscillation was even more accentuated. I then used the local timescale, in this case, the solution is damping but still not converging.

On the other hand, the residuals are quite high (only the mass residual lower than 1*10-4).

I'm open to suggestions

P.S. The mesh I used has >250000 nodes, y+=1.08

[ghorrocks]

FAQ: https://www.cfd-online.com/Wiki/Ansy...gence_criteria

[DDG]

Originally Posted by ghorrocks

FAQ: https://www.cfd-online.com/Wiki/Ansy...gence_criteria

Hi ghorrocks, thanks for your reply,

Actually, I have already read this page and I managed to reduce the oscillation in the residuals. I used the High Velocity Option and a Local Timescale setting. Although my efforts there is still a smaller oscillation in the residuals, while in my output parameters, the pressure ratio and the efficiency, the oscillation is really feeble but is still there.

In the CFX documentation I have read that

"If the period of oscillation of the residual plot changes by changing the time step size, then the phenomenon is most likely a numerical effect."

Is it possible to eliminate this numerical effect (if it is the problem I'm facing) , and if not how is it possible to validate the model? May I use an average of these output parameters?

[ghorrocks]

The FAQ discusses this - try a larger time step and see if it fixes it. And the FAQ also mentions that have a look to see if the result is sufficiently converged despite the lack of residual convergence.