rotor/stator interaction

July 7, 2000, 06:08

Hi colleagues,

I'm a beginer in turbomachinery flow analysis, and i 'm trying to analyze a flow field including one stage rotor/stator using a commercial CFD code(Fine/Turbo) assuming a steady flow. I selected a k-e model without wall function.

However, I found that my computational results deviates from an experiment at a wake region of the stator. Velocity profiles at the wake region does not agree with the experimental data.

My idea on one possible error source is that turbulent viscosities get reduced by an order of magnitude(from around 5 to 0.5 times miolecular viscosity ) when the flow crosses the rotor/stator boundary interface. The code employs the mixing plane method for the rotor/stator boundary, which averages pichwise variation of flow variables and puts them as inflow BC for a stator block.

Although it might be unavoidable in a computational viewpoint to average pitchwise disturbunces, the fact that values of turbulent viscosity are not continuous and jump down by one order of magnitute seems to me quite unnatural.

Also, I think the mixing plane method would always reduce turbulent viscosities and, therefore might cause similar problem with mine.

Would anybody answer me on this problem ?

Do I have to solve the flowfield with the code in unsteady mode without any pitchwise-averaging BCs to get consistent turbulent voscosity and accurate wake profile ?

Or does the magnitude of turbulent viscosity have little effect on wake velocity profiles, and I have another thing to be fixed ?

Thanks in advance.

Best Regards,

Hyoungjin.

July 7, 2000, 06:29

Rotor/Stator interaction is always transient. The only way to do such an analysis is a sliding mesh calculation.

July 7, 2000, 12:32

(1). I think, the first thing to do is to contact the support engineer of the software vendor. (2). The rotor/stator interaction simulation is an extremely complex problem. (3). Even for the single row simulation, there are flow separation, vortices, and complex secondary flow problems difficult enough to solve. (4). It is a research problem. And there is no simple answer to it.

July 7, 2000, 06:08	rotor/stator interaction	#1
Hyoungjin Kim Guest Posts: n/a	Hi colleagues, I'm a beginer in turbomachinery flow analysis, and i 'm trying to analyze a flow field including one stage rotor/stator using a commercial CFD code(Fine/Turbo) assuming a steady flow. I selected a k-e model without wall function. However, I found that my computational results deviates from an experiment at a wake region of the stator. Velocity profiles at the wake region does not agree with the experimental data. My idea on one possible error source is that turbulent viscosities get reduced by an order of magnitude(from around 5 to 0.5 times miolecular viscosity ) when the flow crosses the rotor/stator boundary interface. The code employs the mixing plane method for the rotor/stator boundary, which averages pichwise variation of flow variables and puts them as inflow BC for a stator block. Although it might be unavoidable in a computational viewpoint to average pitchwise disturbunces, the fact that values of turbulent viscosity are not continuous and jump down by one order of magnitute seems to me quite unnatural. Also, I think the mixing plane method would always reduce turbulent viscosities and, therefore might cause similar problem with mine. Would anybody answer me on this problem ? Do I have to solve the flowfield with the code in unsteady mode without any pitchwise-averaging BCs to get consistent turbulent voscosity and accurate wake profile ? Or does the magnitude of turbulent viscosity have little effect on wake velocity profiles, and I have another thing to be fixed ? Thanks in advance. Best Regards, Hyoungjin.

July 7, 2000, 06:29	Re: rotor/stator interaction	#2
Joern Beilke Guest Posts: n/a	Rotor/Stator interaction is always transient. The only way to do such an analysis is a sliding mesh calculation.

July 7, 2000, 12:32	Re: rotor/stator interaction	#3
John C. Chien Guest Posts: n/a	(1). I think, the first thing to do is to contact the support engineer of the software vendor. (2). The rotor/stator interaction simulation is an extremely complex problem. (3). Even for the single row simulation, there are flow separation, vortices, and complex secondary flow problems difficult enough to solve. (4). It is a research problem. And there is no simple answer to it.