[wlt_1985]

I did my model in unsteady case. I faced a big problem for long time. Each time further for next iteration, there was a sudden jump in residuals and it was going to reduce again repeatedly as attached. Please help me. Your opinion is appreciated.

[-mAx-]

It's normal: it is due to the change of time.

[wlt_1985]

Originally Posted by -mAx-

It's normal: it is due to the change of time.

Dear mAx,

Thanks for your help. Do you mean that sudden jump in residuals during change in time is correct for unsteady state? Do you mean that my model is acceptable with such trend of residuals? And is there any approach to reduce the sudden jump in residuals?

Thanks a lot.

[-mAx-]

Yes the jump is correct
I would investigate for having more convergence on continuity/x-y-z-velocity residuals.
Maybe by reducing your time step

[wlt_1985]

Originally Posted by -mAx-

Yes the jump is correct
I would investigate for having more convergence on continuity/x-y-z-velocity residuals.
Maybe by reducing your time step

Thanks mAx,

I had gone through the model by having smaller time step. The trend was similar to result of larger time step. The results only might differ with different under relaxation factor or courant number or explicit relaxation factor under coupled algorithm. I had gone through few forum and some thread replied that those value under 'under relaxation factor or courant number or explicit relaxation factor'should not been changed for model accuracy. Is it true?

Best regards

[-mAx-]

yes it will just give you faster or slower convergence

[wlt_1985]

Originally Posted by -mAx-

yes it will just give you faster or slower convergence

Thanks mAx. You definitely help me a lot

[enayath]

Hello,

What about this huge jump in continuity residual?
it is related to transient k-e model

Thank you,
Hooman

[-mAx-]

see my first reply --> "It's normal: it is due to the change of time. "