We have a new fan here with a direct drive system on it which has been all over the board since we've had it. I attached a trend on the 4HVEL point which is our bearing against the housing. We have never balanced this fan, yet the vibration levels go up and down at will. Other than process I haven't a clue what is causing it...

We have two of these identical in design. the other is even worse...This is pushing hot air about 300-350 degrees...I talked about these fan a little over a year ago. But I wanted to reopen this case after I had trended it awhile...Virtually no vertical or axial vibration at all.1 to .2 at best...No sign of blade pass either.

Maybe some of you have had a similar experience. Just curious...

Thanks

Rod,

What does the vibration specrum show for each of the four high vibration events? If dominant requency was all at 1xSS, then I would suspect there is a natural frequency near 1xSS. Does the vertical vibrations have the same events, but at a lower amplitude? I think you need to extract some more useful information out of the data.

Walt

Rod,
Does this fan have any dampers controlling the air flow? If so which side of the fan? What type of bearings?

Don

Just adding to what Don said about dampers......If you do have some dampers, get into the habit of recording there settings at the time of vibe readings. You could be in for a big surprise and have the odd mystery solved.

Sure do. I have all the data at the high points and low...Vertical is not a problem nor axial. I seems that right after you hit the off button all vibration stops after a second or two. Makes me wish they would have indtalled VFD drives on these. At least I could have slowed them down a bit and played with the speeds, vs vibration levels...

quote:

Originally posted by Walt Strong:
Rod,

What does the vibration specrum show for each of the four high vibration events? If dominant requency was all at 1xSS, then I would suspect there is a natural frequency near 1xSS. Does the vertical vibrations have the same events, but at a lower amplitude? I think you need to extract some more useful information out of the data.

Walt

No, the only way to control the flow is one valve on the outlet side. There is nothing on the inlet side. I was questioning that when these fans were installed. Seems to me that it would have been better to have some sort of control over the inlet flow of air, just like we do on our Hoffman blowers. None of these spectrums shows any sign of blade pass. They are all basically clean velocity spectrums. Posted here is the worse point of the lot 4HVEL at 1.29ips

quote:

Originally posted by Don Jones:
Rod,
Does this fan have any dampers controlling the air flow? If so which side of the fan? What type of bearings?

Don

No dampers...These have Dodge Imperial bearings on them. We have two of these fans built the same time by the same company. Look like twins...

quote:

Originally posted by Don Jones:
Rod,
Does this fan have any dampers controlling the air flow? If so which side of the fan? What type of bearings?

Don

Hi Walt,

I have yet to delve deeper into the natural frequency of shaft speed. Is it a common problem with inline drive systems on overhung fans? And how can I check this while it's running?

Is there a way to determine shaft natural freq. by a formula based on shaft length, diameter, and metal qualities?

quote:

Originally posted by Walt Strong:
Rod,

What does the vibration specrum show for each of the four high vibration events? If dominant requency was all at 1xSS, then I would suspect there is a natural frequency near 1xSS. Does the vertical vibrations have the same events, but at a lower amplitude? I think you need to extract some more useful information out of the data.

Walt

I did the best I could gathering this data. I used my tach light instead of strobe, for more accuracy.

I really don't see any drastic phase changes on the fan itself, as far as phase shifting. The motor vertical does however have about a 100- 140 degree difference across the coupling. I'm not seeing that 90 degree phase difference from the Hor & Vert positions which indicate balance, yet I have a huge 1X vibe spectrum...I couldn't get to all the areas for the rest of the axial readings. But looking at the spectrums I see no evidence of a bent shaft, at the middle, or near the coupling.

I really don't do a lot of phase work, so if you see something here I'm not seeing, let me know. Makes me wish I had that modal software...

Thanks

Amplitude > 1.3 ips 0-p is certainly severe.
Physically where is this point of measurement? (Fan I/B brg?)

I wonder if the air is clean and if cleanliness of the fan was checked.


If I were you i would collect the following data and analyze:
1. 1x amp and 1x phase trend over the enough time to capture high and low amplitude.
2. Overall and 1x coast down Bode Plot
3. All trendable operational parameters like air pressures and temperatures

Yes, the blades are extremely clean. All this fan pushes is hot air and some moisture and steam. This goes into a bagfilter.

It would be tough to do a run up and coast down. This is a critical fan for production.

quote:

Originally posted by Shurafa:
Amplitude > 1.3 ips 0-p is certainly severe.
Physically where is this point of measurement? (Fan I/B brg?)

I wonder if the air is clean and if cleanliness of the fan was checked.


If I were you i would collect the following data and analyze:
1. 1x amp and 1x phase trend over the enough time to capture high and low amplitude.
2. Overall and 1x coast down Bode Plot
3. All trendable operational parameters like air pressures and temperatures

Forgot to mention that I took these readings on the bearing housings. Worse vibration at #4 by the fan blade.

quote:

Originally posted by Shurafa:
Amplitude > 1.3 ips 0-p is certainly severe.
Physically where is this point of measurement? (Fan I/B brg?)

I wonder if the air is clean and if cleanliness of the fan was checked.


If I were you i would collect the following data and analyze:
1. 1x amp and 1x phase trend over the enough time to capture high and low amplitude.
2. Overall and 1x coast down Bode Plot
3. All trendable operational parameters like air pressures and temperatures

Thinking out loud, I can't help but wonder if this is a design flaw with the shaft & wheel...Both of these fans have the same problem, only one is worse than the other.

I'm going to go measure the shaft length vs wheel and shaft diameters. Also the CL distance of the bearings, and distance between the 4 bearing and overhung blade.

Is there a formula one can use to determine what is acceptable?

In my humble opinion I think most of the problem is caused by too heavy a wheel overhung on too long & too small a diameter of shaft. Couple this with the heat factor, you have the ingredients for a sloppy rotating wheel...

Shaft length 45 inches, shaft diameter 2.937in. See for yourself on the wheel.. I may grab this thing and go weigh it. It has a solid stainless steel 5/8 inch plate on one side, and you can see the other...The wheel is about 12-14 inches away from the bearing housing.

Am I barking up the wrong tree or could this be the problem? Don't know what else I can add for you.

Any fan designers in this group?

This may be why our tech had trouble balancing it too. The microlog kept telling him to add weigh to the same spot after three different trim runs... The wobble factor...


***********************

quote:

Originally posted by Rockin' Rod:
In my humble opinion I think most of the problem is caused by too heavy a wheel overhung on too long & too small a diameter of shaft. Couple this with the heat factor, you have the ingredients for a sloppy rotating wheel...

Shaft length 45 inches, shaft diameter 2.937in. See for yourself on the wheel.. I may grab this thing and go weigh it. It has a solid stainless steel 5/8 inch plate on one side, and you can see the other...The wheel is about 12-14 inches away from the bearing housing.

Am I barking up the wrong tree or could this be the problem? Don't know what else I can add for you.

Any fan designers in this group?

Rod,

Are these fans physically close enough that you could have transmitted vibration and a beat? If so, can you take very high resolution data to see if there are 2 peaks? If not, I'm with Walt on the excited natural frequency and/or poor design. Any way to stiffen horizontally or add mass to the base?

Yes Bill, they are only about 5 feet apart. One fan is running really good, the other you can see here. Both running the same speed give or take a few.

You can see from my previous post I think the same thing. But I need to prove it. I just got back from taking two readings axially just behind the wheel on the inlet flange. One side has a 2.2ips the other has .43ips. Looks like it's moving back and forth. Also need to find a way to prove Nat Freq. of SS...

quote:

Originally posted by Bill Kilbey:
Rod,

Are these fans physically close enough that you could have transmitted vibration and a beat? If so, can you take very high resolution data to see if there are 2 peaks? If not, I'm with Walt on the excited natural frequency and/or poor design. Any way to stiffen horizontally or add mass to the base?

"I may grab this thing and go weigh it."

That is to say we have a spare wheel I can weigh, (:>

Can you collect a long time waveform of 2 minutes or so? Any way to monitor the phase for 2 or 3 minutes? If there is a beat, you should see amplitudes and phases drift.

How about a bump test for the natural frequency?

I probably could do that. Probably not the bump test though, but I can do extended real time monitoring. I've been using my magnetic tach to hit that tape on the shaft, and I can stand right there and watch. Guess I can increase the resolution of the readings as well.