We have a high vibration problem on a fan. The fan is horizontally mounted, direct driven, 75 HP motor, with two propeller type fans, one fan mounted on the OB and one fan mounted on the IB side of the motor. There are 9 blades on the impellers. The nominal fan speed is 1780 RPM. The motor is supported by tie rods mounted 90 degrees apart.
The current vibration levels are at 0.8 ipsp dominant at 15.33Hz. The previous vibration levels were at 0.3 ipsp dominant at fan speed. This fan is inaccessible during plant operation. Vibration data at other locations (on the tie rods) are very similar.
This is the first time seeing this signature on these fans. The data was taken multiple times to confirm. The acceleration spectra is basically made up of 9X and harmonics of 9X with the apparent 1/2X sidebands. I am suspecting looseness for starters although it does not met most of the criteria. Bearing FTF and BS frequencies don't line up. I appreciate your help.

CRDM_Fan_2A.pdf (75 Kb, 76 downloads)

I would consider a subharmonic resonance excited by flow noise or looseness as a possible cause. You might try a bump check when you can get to the fan. You might also try taking a higher resolution spectra to better define the frequency. Maybe the rods are loose.

Michael Titone

Thanks for the reply. We suspect looseness and plan on bump testing. We had a crew go in last night and use hand tools to check all fastners and visually inspect for structural fatigue. "Nothing significant was found". We are considering re-inspection.

A second inspection was performed last night. The report back was that some of the supporting tie rods were "slight loose". I guess enough was seen to decide to go further so tonight the same crew will perform inspections for the third time. I have added some additional data. If more data is needed please let me know. Again, thanks.

CRDM_Fan.pdf (79 Kb, 36 downloads)

You might check that the motor and fan are still centered in the housing. If the tie rods are tighten unevenly they can distort the housing. It will produce blade pass frequencies.

I will pass this on to the NS crew. Still can't figure out the subharmonic signature. I have never had to deal with this condition before. I wonder if ODS would help. Thanks for the reply.

It seems like you may have flow induced excitation as Michael suggested. I've seen simialr events caused by improper damper setting/positon on FD Fans. I would suggest you inspect correct open/close position and condition, if a damper/diverter is present.

Are the blades fixed pitch, or adjustable?

The blades are fixed pitch. There are turning vanes that are also fixed. There is a dampener on the suction side. We will confirm its position on next run. The attached image is from a different fan but same size and type.

I hope my image went out OK. If I messed up I would appreciate someone correcting my mistake if possible and please accept my appology. Thanks

quote:

The acceleration spectra is basically made up of 9X and harmonics of 9X with the apparent 1/2X sidebands.

Is the SBs realy 1/2x or are they the 15.33 Hz frequency?

I noticed the data is 16 days old. Have any more current data?

I see you are running CSI software. Can you extract out the machine into a small database and email it?

Why is the motor mounted by "tierods"? Is it some kind of special application? I can't visualize how it is made. Got picture of the whole thing? Just a S.W.A.G., I would say it has something to do with this "tierod" setup.

What resolution is your data? Seems like it is low judging from the 0 - 500 Hz expanded spectrum. Can you get a low Fmax with more lines, say, 500 Hz 3200 lines?

You image went up ok except it is a little too large. Be nice if you could shrink it to a smaller size before uploading it.

Your correct. The side bands are actually closer to 15.3 Hz. The fan can't be operated under current conditions but we will soon be collecting additional data. Good catch. Sorry about the image.

I am working on the seperate DB. It's too large to send but I am working on it. We have a 8 MB firewall. The person that collected the original data took a 400Hz high res spectrum but accidentally had a Peak-Vue filter on. When we remonitor additional data will be obtained.
The motor and fan are mounted on tie rods inside a circular housing. The tie rods mount to a collar that is bolted around the motor and then bolted to the side on the circular housing.

quote:

The tie rods mount to a collar that is bolted around the motor and then bolted to the side on the circular housing.

WOW!
Some configutation, huh?

I have a 10 meg limit on receiving. The one machine should be less tha 2 meg, huh? Wonder what happened?

Forgot the email address.
ralph_stewart@alertanalytical.com

Ralph, I sent you a split zipped file. I hope it works. I cain't explain why it's so large.
Maybe the third fan inspection tonight will find something. Thanks

I was just looking at your data and realized I've seen something very similar. We had an Iso-Phase Bus cooling fan that suddenly started vibrating very similar to your signature. The discharge air flow had pulsations severe enough to break rivets on the discharge ducting. Finally traced it to a broken damper that was restricting the fan suction. I think someone already mentioned flow induced excitation, which is what this looks like. Sort of a rotating stall condition.
I would look at anything that could be blocking the suction.

Thanks Ed for taking the time. I think your reply along with the others is convincing enough to do a inspection.

Just to expand on the possibility of damper issues. If the position setting is automated, please verify that the damper's physical position correlates to what the system (PC) calls for. Due to some maintenance activities the calibration may have shifted/drifted significantly without anyone realizing that this has occured. And so when the system calls for 80% Open, it may be reversed or completely closed.

Cheers...Rajan Muthukrishnan

The fan is in a Nuclear Power Station. I will contact the System Engineer about the calibration. I believe we perform test on those dampeners. Good information and thanks for the reply.

RajSha, I reviewed our procedure for inspecting the suction dampener. The following step is one of many:
Visually verify that the damper moves from the OPEN to FULLY CLOSED position with no evidence of binding or need for lubrication, adjustment
or tightening of linkage.

I am waiting on the group that performed the inspections to give me an up-date. Thanks.