[omidiut]

Hi all

I'm trying to simulate a 5 stage centrifugal compressor. i simulate it by Ideal flow mixture and have not any error. i try to simulate it by real gas fix component mixture. the mixture components are: C2H6 - C3H8- C4H12
C3H8 has maximum mass fraction about 0.98.

i use several equation of state ping-robinson, redKw, but i can't give any accurate result, for example the pressure ratio of compressor be 50 !!
for most investigation i try to simulate only one impeller but this errors occurs.
Now what is the reason of this big difference ? Is there any consideration for simulation of real gas flows?

Best regards.

[ghorrocks]

The more complex material model is much harder to converge - this is pretty obvious. So you will need to use smaller timesteps and possible double precision to get it to converge. Also mesh quality is very important - the better the quality the easier it will be.

[omidiut]

Thank you Dear Ghorrocks for your reply

Is there any expert parameter setting to improve convergence procedure?
it seems to be real gas has very expensive computational lost.

at the first iteration there is notice about high mach number about 80.1, and then overflow error.. its only occurs when real gas modeled. have you any suggestion?

best regards.

[ghorrocks]

Is there any expert parameter setting to improve convergence procedure?

No..... If there was, don't you think it would be set by default anyway?

Read the section on obtaining convergence - Small time steps, double precision, a good initial condition, high quality mesh.