[newbie384]

Hi

I am simulating the rotor-stator interaction in centrifugal compressor with transient simulation. I created a probe at the trailing edge of the rotor and I monitored the pressure fluctuation history throughout the simulation.

Below are the things I observed the time history of pressure (see attached picture)

a) Since I have 9 stators and I set the number of time step to be 10 for one stator blade passage, I expect to get a peak in pressure every 10 time steps and a total of 9 peaks in pressure for every rotor revolution. However, the monitoring history in CFX Solver shows that there is a peak in pressure in about every 200 time steps.

b) Also, since I do not know how long should be the total time, I just set it to 20 rotor revolutions and I was hoping to stop it when I see some reasonable periodic fluctuation. However, as the time proceeds, the peak of the pressure get smaller and it seems that the pressure is going to converge to a steady single steady value.

Below are my questions.

1) Is the phenomenon described in a) above if the number of time step is too little? or it is more likely caused by other factors? I know that the number of time step I set might be too rough and I need to try a finer time step. I just want to make sure that I do not think in a wrong direction.

2) As for b), I wonder why the pressure seems to converge to a steady solution. This is a transient simulation and the pressure should fluctuate in time no matter how long is the total time. If I am wrong, is there other possible reason for this phenomenon?

Thank you in advance.

[brunoc]

By default, CFX will designate a mesh node to the 3D coordinates you specified in Pre, and will then monitor the value on that node through the simulation. This means the results you're seeing are from a node fixed at the TE of the rotor blade, and not static to the 3D position you've given it. So maybe the effects of the rotor blade passing through a stator blade just isn't that strong on your geometry, or maybe they're just small relative to the large oscillations that appeared when you started the simulation. Try to change the limits shown by your monitor point in Solver Manager, or set the monitor to a fixed position in Pre (I think this is available from 14.5 on).