[CFDvasu11]

Hi

I am analyzing a centrifugal compressor with vaneless diffuser. My design conditons are 0.8 Kg/s, 3.2 pressure ratio and 70% isentropic efficiency. Upon running with multiple mesh densities (2.5 and 8 lakhs mesh elements) and turbulence models (SST and k-ephsilon) interface models (frozen rotor and mixing plane) I am getting an efficiency of 52% and pressure ratio of 2.3.

I am using total pressure of 1 bar, 273.15 K at inlet and average static pressure at outlet (which I vary to get my cahracteristic curve). I am using "air at 25 C" to get compressibility effect in my analysis.

What is your suggestion for my lower pressure ratio and efficiency? Also my pressure ratio keeps on increasing when I go near surge point where it should increase very slightly. Suggestions for this problem is also welcome.
I have attached the characteristic curve for reference.

[ghorrocks]

Please read the FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F

You need to do more mesh densities than just 2, and 2.5/8.0 is not a sufficiently large difference between the meshes. You want to double or halve the element edge length which should result in around 8x the node count in the finer mesh.

Also read the best practises guides in the CFX documentation. There are a few on turbomachinery, surge and stall so you should read these as background.