i'm an aeronautical undergraduate doin my final year project entitled 'structural studies and analysis of micro turbine (centrifugal compressor)'...Before i proceed with my structural analysis , i need to perform CFD on the centrifugal compressor ...my model a flow passage between a pressure blade and suction blade together with a splitter...I used FLUENT 6.1 to perform my CFD ..my 1st model where i used incompressible flow managed to converged but then i hv some difficulties in converging for the compressible model...i used mass flow inlet and pressure outlet as the BC...i really need help..my mass flow rate is 0.8kg/s , RPM =70000 and pressure ratio is 4

Hi Siva,

if you start with the converged incompressible solution, you might try solving the compressible flow transient (first order...with default under-relaxations) with about 3-4 iterations/time step. This thing is really spinning so you need a small time step. Perhaps something like 1e-4[s] (or smaller) to get it going.

Let us know how you make out.

Regards,

Bak_Flow

A good starting point is to log on studentfluent.com, search the tutorials lib for tutorial 9. Cheers and Good Luck

hi Bak_Flow!

thanx for ur info...let me explain my simulation procedure ...my impeller was set as stationary wall in absolute frame while my fluid is rotating at 70 000 rpm in absolute frame(boundary conditions)...i used segregated solver for my incompressible flow simulation..my viscous model is k-epsilon(standard)..after getting the results my this simulation , i tried using couple solver(unsteady)and follow the procedure as suggested by you for my compressible simulation..end of the simulation, 'Error: Floating point error: invalid number' appeared...at certain period before the error mesage was displayed there was this message too(divergence detected - temporarily reducing Courant number to 0.5)...can u explain to me this error messages?and how can i prevent this from happening?

Hi,

the set-up issues are not clear to me. There are tutorials which show how to set up rotating domain problems. It is best to follow one of them.

Since you had already solved incompressible using the segregated pressure based solver, I was suggesting that you use this solver transient with a few iterations per time step to advance the solution to steady state.

Best of Luck,

Bak_Flow

hi,

May i know what are the setup issues that u would like to know ?i referred to one of tutorial given in the Fluent manual regarding single rotating frame...are there any other tutorial that i can refer to?