[ong khi]

Hi,

The turbulence model that i am using is Realizable K-epsilon.

I know that for unsteady condition, the residuals should be below 10E-3. All my residuals look OK except for Tdr and Tke.

Does it affect the convergence of the solution? Or is this normal?


Thanks for the reply and help!

[fluid23]

Are these normalized residuals?

[ong khi]

Originally Posted by MBdonCFD

Are these normalized residuals?

How to determine whether the residuals are normalized? Is it we can determine it from the scale factor of X and Y? Since both the both the X and Y scale factor are 1.0 (as shown in the attached figure), means it is normalized?

Thanks in advance!

turbulence residuals.jpg

[harry123]

In StarCCM, they are mostly normalised I believe.

Sometimes I dont't get low residuals for my simulations, but the solution looks good enough. I usually depend more on the variation of a quantity at a fixed point for checking whether simulation has run for enough time, for example variation of a velocity component at a fixed probe point.

[lcarasik]

Originally Posted by ong khi

How to determine whether the residuals are normalized? Is it we can determine it from the scale factor of X and Y? Since both the both the X and Y scale factor are 1.0 (as shown in the attached figure), means it is normalized?

Thanks in advance!

Attachment 49298

Star-CCM+ automatically normalizes all residual quantities to the first ten iterations I believe (you should check the manual to make sure). Additionally, you can check/change how the residuals are normalized (or make them not normalized) under the Monitors section.

[ong khi]

Originally Posted by lcarasik

Star-CCM+ automatically normalizes all residual quantities to the first ten iterations I believe (you should check the manual to make sure). Additionally, you can check/change how the residuals are normalized (or make them not normalized) under the Monitors section.

If my residuals are all normalized. Can i just ignore Tke and Tdr and focusing on monitoring the stabilization of other variables (drag, lift,...)?