centrifugal compressor 1D vs. CFD
 

Simulating centrifugal compressor with vaneless diffuser i usually get good agreement with 1d calculation at impeller exit. However CFD shows much bigger static pressure growth along vaneless diffuser. I performed my calculation using SA turbulence model. I repeated computation using diffrent y wall, but tendency remains the same. Vaneless diffuser seems to be easy element to predict, so I believe 1d calculation are correct. Is it nature of mentioned turbulence model, that it can't predict accurate static pressure along static components ?

"Is it nature of mentioned turbulence model, that it can't predict accurate static pressure along static components ?"

Definitely not, the SA-Model is a well-proven turbulence model in many turbomachinery applications from my experience.

May I ask how much bigger the static pressure at the diffuser exit is (comparing 1D and CFD)? How do you obtain your 1D results, maybe from a preliminary design tool?

Maybe you want to share your boundary conditions (inlet / outlet) with us?

Secondly you could compare the SA-model with SST or EARSM model computations.

I am looking forward to hearing from you.
Holger

Hi;
One point that should be considered in turbomachinery design cycles like what you are doing now is the different natures of 1D and 3D design tools. 3D solver will never replicate results of a 1D code even being a perfect one. So, some difference is always seen between the two and I believe that even running a LES turbulence model will not fill this gap.
What is the difference level between the two?
To my experiences if the difference is large, As "DarylMusashi" mentioned, changing the turbulence model will not be helpful. In such a case it seems you need to recheck your 1D code or boundary conditions of your 3D model.
If the difference is not too much, changing some other settings like mesh sizes or turbulence model can always help lower the difference.

Regards