Jon and other ODS and Modal Analysis Gurus

Is it possible for you to post an actual animation here in the forum and show us an example (like a broken machine foot) so us grasshoppers can start to get a better understanding?

Terry O

Here is an animation of one of our motors that I think is resonanting.

S31412MP290_30Hz_Movie_2.avi (1,046 KB, 190 downloads)

Hi Steve,
Would you check your database and verify:
84 and 60 are 180˚ apart in the Y direction
83 and 59 are 180˚ apart in the Y direction
If they aren't, you can edit them in the database to correct if this is an out-of-phase condition.

This message has been edited. Last edited by: Jon McFadden,

Hi Steve,
How can you tell that the system is resonating (or why do you have that suspicion)?

Because is moving (almost) completely in-phase?

thank you very much for the info

This message has been edited. Last edited by: felipe.figueroa,

Hi Felipe,
I was hoping Steve would jump in and answer your question.
Numbered points are locations where vibration measurements were made. Points in between these points are calculated to give an average movement between measured points.
Because points 59 and 60 are moving together and opposite to the other half of the motor, I suspect their DOFs are not oriented properly. A minus sign inserted in the block file would correct this problem.
There does appear to be a mode shape on the base at point 74.
Here is my interpretation of the animation:

This message has been edited. Last edited by: Jon McFadden,

Cool. I forgot this forum was here until a link was posted on the vib forum.

I will take a stab at answering felipe's question, although Steve and Jon would probably know better.

It looks like a classic case of rocking base resonance. Some things that lead to this conclusion:
1 - H much higher than V = directional.
2 - The base is moving a significant fraction of the amount that the machines are moving (50%?). That shouldn't be the case unless we have a resonant base.
3 - I am guessing the thing between the pump and motor is a gearbox??? (not sure). If that is the case, then the pump running speed is differernt, and the fact that it is vibrating at the motor speed suggests vibration transmitted through a resonant base.

I agree with Jon that points 59 and 60 look unnatural.

All,

Thanks for being patient.

Disclaimer - This was my first ODS/Modal job using ME'Scope and IOTech hardware/software, so there are a lot of potential opportunities for making mistakes. I am hoping to repeat this job and spend more time collecting data on the motor points. In addition to the structures shown on the model, the pump is mounted on a steel plate supported by I-Beams. I collected data on the plate, but there did not seem to be much response, so I didn't show it in the animation.

In order to collect data at points 59 and 60, I had to change the orientation of my accelerometer. The standard orientation that I used was:
+X / -Y / +Z
The flipped orientation was:
+X / +Y / -Z

I made the change to the orientation in the block file, so I think that the data is correct, but I admit that it looks confusing.

Why do I think it is a resonance? We have a highly directional (horizontal versus vertical) vibration at running speed. The FRF and Coherence data indicates a peak near running speed (30 Hz). Maybe Jon or another experienced person could provide some direction on confirming the resonance.

Thanks,
Steve

Pete - That's not a gearbox, it's the pump bearing housing/pedestal. I've attached a picture of the motor/pump.

quote:

There does appear to be a mode shape on the base at point 74.

More points could be added to the base to get better resolution.
Also, you could turn on the options "animate/contours/fill" and "animate/contours/values" to display the actual movement.
Since the main focus is on the motor end of this animation, you could turn off the sub-structures for the fan and bearing pedestal, then post a quad-view so we can get a better look.
I still think there is a bad entry in the data.
83 and 84 look OK, but 59 and 60 don't.
Check the block data for these points and make sure 83 and 59 are opposite in the "Y" direction, and 84 and 60 are opposite in the "Y" direction. I'm not sure why the "Z" direction had to be reversed.

This message has been edited. Last edited by: Jon McFadden,

Jon - Regarding the flipping of the Z-Axis. I used a tri-axial acclerometer mounted on a cube with 3 magnets as shown in the link below:

http://maintenanceforums.com/e...1082983/m/8911053104

When I mounted the accelerometer at points 59 & 60, I had to flip the orientation such that I got +X / +Y / -Z.

Steve

Steve,

I question the polarity (phase) of your measurement points on motor. It shows the motor case to be distorting, and I doudt this is occuring. Points 59 & 60 horizontal motion should be outward when base points 73 & 74 vertical motion is upward. We discussed triaxial accelerometer mounting issues in the previously mentioned Link. It is difficult to keep the accelerometer axis (data channels) and their polarity properly matched to the structures coordinate system (+X, +Y, +Z). I check mode shape plots (no fancy animations) to be sure that the Mode shape makes sense, otherwise additional data is needed to confirm anomolies. Does this make sense to you?

Walt

Hi Steve,
I would like to discuss this privately using an e-mail if you're willing.
Your animation shows horizontal radial being in the "Y" direction.
Your Vertical is the "Z" direction.
Your Axial is in the "X" direction.
This is why I have been asking for you to look in the "Y" axis data.
I don't know if you reviewed this thread, but here is a tidbit I learned the hard way:
Accelerometer Orientation

Walt - I agree with your concern and will attempt to recollect the data and pay close attention to sensor orientation. In regards to using a tri-axial as described in the other thread, I'm having a hard time seeing how to get data in 3 directions at some measurement points. How do you do it with a single axis transducer?

Another question: We are coming out of an outage, and the sister pump is now in service. Is it worthwhile to collect modal data with the other pump in service? My only other option is to do an ODS. The other caveat is that when the unit (Turbine-Generator) is online, I get a lot of background vibration and beating with the main generator.

Jon - Email sent.

Thanks,
Steve