[star-user]

Hi Dear's
I have been working on simulation of centrifugal compressor in star ccm software applying steady-state approach.
The issue that I have faced is related to the divergence of the residuals.
I would be grateful if you can share any of your experience in this field (the minimum quality of the mesh, the initial and boundary condition and other important factors are our main concerns)
The boundary condition for inlet is 90 kP and 38 C and static pressore outlet is 185 kPa.
polyhedral mesh have been used and the rotational speed of the rotor is 150000 rpm.

[SESP]

Hi Ali.

the information is a bit too vague to be able to help you. In principle, this case will run in STAR-CCM+, we are running similar cases as routine simulation jobs.

Maybe a few thoughts that can help you:
- check that the boundary conditions are defined correctly
- use ideal gas in case of air
- try the coupled solver, depending on the problem with CCA and initial ramping
- segregated solver (I wouldn't recommend) with much larger under-relaxation only
- mesh should be small enough to resolve the tip gap, make sure you dont get any bad quality cells there
- if you are unsure about the grid, add the "post mesh optimization" and give it a few iterations here

Good luck,
Sebastian

[star-user]

Originally Posted by SESP

Hi Ali.

the information is a bit too vague to be able to help you. In principle, this case will run in STAR-CCM+, we are running similar cases as routine simulation jobs.

Maybe a few thoughts that can help you:
- check that the boundary conditions are defined correctly
- use ideal gas in case of air
- try the coupled solver, depending on the problem with CCA and initial ramping
- segregated solver (I wouldn't recommend) with much larger under-relaxation only
- mesh should be small enough to resolve the tip gap, make sure you dont get any bad quality cells there
- if you are unsure about the grid, add the "post mesh optimization" and give it a few iterations here

Good luck,
Sebastian

thank you so much

Please explain more about ''CCA and initial ramping'' and initail pressore

[SESP]

Originally Posted by star-user

thank you so much

Please explain more about ''CCA and initial ramping'' and initail pressore

CCA = Continuity Convergence Accelerator. It is worth a try, I have mixed experience with it: from cutting down simulation times by a factor of 15 to zero effect...
You can also try to combine it with Grid Sequencing as Expert Initialization. Akk above are settings for the Coupled Implicit solver.

If you face stability problems, try adding increased dissipation in the beginning or activate enhanced stability treatment, but make sure to remove it later.

For initial pressure just take your inlet pressure to avoid backflow.

Hope this helps!

[star-user]

Originally Posted by SESP

CCA = Continuity Convergence Accelerator. It is worth a try, I have mixed experience with it: from cutting down simulation times by a factor of 15 to zero effect...
You can also try to combine it with Grid Sequencing as Expert Initialization. Akk above are settings for the Coupled Implicit solver.

If you face stability problems, try adding increased dissipation in the beginning or activate enhanced stability treatment, but make sure to remove it later.

For initial pressure just take your inlet pressure to avoid backflow.

Hope this helps!

Thank you very much for sharing your experience