In my job we have just recently come across with the phenomena of Rotating Stall which was the source of shaft vibrations and pulsating flow from the discharge of one of the compressor. The vendor of the compressor was however not able to provide a CFD calculation of the flow in impellers and diffusors. My question basically is does any of you have experience in modelling Rotating Stall in centrifugal compressors, if so, can you send me some developed CFD results showing these vortices. (in JPG file please)

Many thanks, Paul

Hi, Paul.

Is the rotating stall located in the impeller or the diffuser ? If diffuser, is it a vaned or vaneless diffuser ? The last one is important because it is quite different mechanisms controlling the stall in those cases. I have been simulating and studied rotating stall in axial diffuser pumps and I can give you a lot of information about that if necessary, especially the influence on vibrations and shaft forces. You can really find out a lot of information about the stall propagation and strength by looking at the shaft vibrations.

Erik

Hi Erik,

Thanks for your reply on Rotating Stall. The rotating stall that we experience is most likely occuring in the diffusor of the last two wheels of a 7 wheel (2 side load) centrifugal compressor. The diffusors are vaneless. We have vibration measurements done and find basically subsynchronous vibration at 25-29% running speed. When this frequency coincides with the critical speed the rotor is excited and high vibrations up to 150 microns have been measured. Appreciate your further response on this issue

Regards, Paul

Hi again, Paul.

I haven't seen too many papers about CFD and vaneless diffuser stall, but there are several CFD papers studying vaned diffuser rotating stall. The typical low propagation frequency of the diffuser stall (vaned or vaneless) makes this types of simulations quite time consuming, especially if you want to cover at least one period of the stall propagation (which often is the case). I also think that vaneless stall is more difficult to simulate because it most probably needs a very fine near wall grid in the radial vaneless area to describe the main controlling mechanisms of the stall. Anyway, I can reccomend a paper which can be useful for you and that is:

"Flow control of rotating stall in a radial vanelss diffuser" by Tsurusak et al., ASME FEDSM99-7199.

It describes active control of the stall and could be helpful. Your measurements of vibrations is very interesting. We have been carrying out similar measuremnts for centrifugal pumps and have published some papers about rotating stall and shaft vibration measurements, especially coupling between blade passing frequencies and rotating stall frequencies, incl. CFD simulations describing the shaft forces as well.

Hope this helps.

Erik