supersonic flow in vane channel
 

Hi to all

Can anybody tell me what does mean the supersonic flow that I got in my axial compressor simulation? Is this effect a sign of something?
It's strange that it took place in stationary domain.
Velocity increases rapidly and then drops, static pressure drops and then rapidly increases, while total pressure increases smoothly as it should.
The solution convergence is good, grid is fine.
The worst (and the simplest) guess - wrong vane geometry, it won't be easy to check angles since I have only parasolid model and no draws.
But if not - may be there's any trick to abide this step? May be to change any air parameters for some time steps and then return to previous values?

Thank in advance

Is this a region where the flow goes supersonic then returns sonic again? If so it does not sound like a device running at high efficiency to me (if efficiency is what you are looking for).

Yes, it returns to subsonic flow within stationary domain. And yes again, efficiency and massflow values are relatively low. But anyway, flow pattern is not normal even for low mode, am I right?

I am no turbomachinery expert but if your design is doing that in flow regime where the device should be operating well - then I suggest the stator design is not very good. I suspect it is choking in the gaps between stator blades.

So it is not normal for well designed turbomachines, but for a poorly operating turbomachine it is an example of one type of problem you can face.

If your solution is right, the velocity contour obviously shows a shock wave in the stator.

Glenn and Turbo,
thanks for your answers. I think this compressor just suffers from choking at mass flow rate about 3/5 of expected value. I think here's problem with geometry indeed... But can you tell me what stage I should check firstly? The place where shock wave was generated, or upstream?

Reduce the stator vane count.

Oh, you mean to reduce number of vanes to increase the area between vanes that will lead to velocity decrease. I'll try it. But may be something is wrong with angles?