[omidiut]

Hi

I'm trying to simulate centrifugal compressor impeller. I try to use Peng-Robinson equation of state for mixture of 5 Gases indclud CH4, C2H6, C3H8, C4H10 & N2. I use MATERIALS-pengrob.ccl Library file.

for checking of solving procedure, I monitor Z (Compressibility factor) at inlet boundary as following Expression :

Total Pressure in Stn Frame / ( (Total Density in Stn Frame *(R/(mw)) )*Total Temperature in Stn Frame )

and Check the value with HYSYS.

unfortunately there is mismatch between HYSYS results and CFX, about 10% deviation. where as i reduce the Mixture Component (2 gas instead of 5), the deviation between results is less than 2%.

So how could I decrease deviation in calculation of Z at Inlet boundary? (CFX & HYSYS in comparison)

[evcelica]

Real gas equations of state like Peng Robinson will have some errors in material properties. RGP files are more accurate, but a lot more work to get set up in the proper format.
There is also the mixing model, looks like there are several.
First step is to use RGP files of all your individual fluids, and do the mixture model in CFX, if you know how to do the right mixture model?
One step further would be to make a single RGP file for the mixture, using a relevant mixture model, but this may be a much more difficult process to format the table manually.