This one goes out to EP and anyone else who might have the answer.
I understand the mass of a tuned absorber should be calculated large enough to have as low as possible a deflection to increase the time to failure by fatigue of the beam. I also understand from EP's spreadsheet comment that a good estimate would be 3% of the reactive mass one is trying to control.
The question is this: If I am dealing with a rigid body mode on a fan (horizontal to vertical vibration is 10:1, and the running speed is 600 cpm below the calculated first bending mode of the fan shaft), is the 3% calculation applied to the fan bearing/pedestal weights (in this case approx. 1000 lbs.) or the entire fan/shaft weight (15,000 lbs.)
Your response(s) are greately appreciated.
Ron Brook

The 3% thumbrule was something I read on a forum but makes sense to me.

Here's my thoughts

The dynamic absorber removes the resonance at the tuned speed and creates two new resonances closely spaced on either side of it. The reason for wanting a large absorber mass (3% of main mass by one thumbule) is to keep those two new resonances from being too close to the center frequency.

A quick undamped calculation suggests:
3% mass moves those new resonant frequencies about 9% away from the center frequency.
1% mass moves those new resonant frequencies 5% away from center frequency.
0.5% mass moves those new resonant frequencies 3.5% away from center frequency.

The relevant main mass would be the modal mass treating the original system as SDOF. That's a little tricky if close to bending rotor critical as your case, but you said 600rpm away. If it's a 2-pole that is 16% away. Slower machine even further. I like the simple approach treat the rotor as rigid and the modal mass would include the rotor mass (somewhere close to 15000 pounds total).

Building a 3% mass for dynamic absorber I imagine would make life difficult. If you use less mass than the 3%, I would think it will just be a little more difficult to tune and a little more sensitive to changes, but would still work. If fatigue is a concern, make the main bar thicker to limit displacement. I would have to give some more thought to a systematic way to address fatigue. We would need to try to estimate displacement of the absorber ahead of time.

EP,
Thanks! I just did a calculation based on a 15000 lb fan/shaft assembly and the mass came out to 225 lbs/ per pedestal. That is easily doable with 12 inch x 12 inch x 2 inch plates, one on each side of the beam. The main intent of my question was what mass are we talking about? Is it the bearing/pedestal mass or the mass of the entire rotating assembly? I know that as soon as I mounted the tuned absorber I would know, wouldn't I? If the mass required was in reference to the entire rotating assembly and I calculated it based only on the mass of the bearing and pedestal, then the deflection of the mass would be excessive and lead to fatigue.
Right?

I don't see any downside to going with assuming the larger mass, other than higher cost to build it.

The difficulty with assuming the smaller main mass and going with smaller absorber mass could be more difficult to tune as discussed above. Also more concern for fatigue as you discussed. My statement "If fatigue is a concern, make the main bar thicker to limit displacement." should be revised to: "If fatigue is a concern, make the main bar thicker and longer". The should give less stress for a given mass qualitatively. How to calculate stresses would take some work.

That's my thoughts fwiw. We have only put these in on small vertical motor. Never tried on a large horizontal machine.

EP,
Thank you for your words of wisdom. We are still contemplating.
Ron

quote:

a rigid body mode on a fan

Talking about rigid and bending makes me think about rotor (dominated) modes. Are you talking about a rotor rigid mode?

Is the dynamic absorber your last resort and are you using it for a fix? Can you post a photo of your project?

Ron,
Sorry I'm not to sure what you mean buy the "failure of the beam". Are you talking about the frame or plinth that the motor and fan are mounted on or the shaft that the fan is attached to? Drawing or photo woould be great.
Hooch

Ron,

Assuming you are trying to decrease the response of a structural resonance and not a rotor critical, I would suggest you test the process with a smaller weight and all-thread rod. I often do this for a quick and dirty way to make sure the dynamic absorber will work before going to the time and expense of fabricating an absorber that will be a permanent fixture on the machine. The all-thread rod may fatigue and break in a short time (maybe even minutes to hours), but it will insure you are on the right track. If the vibration decreases, you know a larger absorber will do the trick. If it does not reduce the vibration, you will need to investigate the problem further. Use the root diameter of the threads to calculate the moment of inertia. Rough weight will work because you can adjust the weight with nuts on the all-thread to tune to the resonant frequency. Just a suggestion.

Regards,
John J

I too have used the all thread rod with success, we have had weights hang off it and we have also used it to add stiffness to a fan plinth/frame. One of the fan bases we filled with sand and tested the response. In the end a couple of months later they had some left over concrete and they filled it with that. That fan ran so smooth for the two more years that I was at that plant.

Hooch

That's a good idea to start with the small mass and easy construction.

I can see all-thread is easier to adjust.

But is the all-thread any easier to attach to the machine? I imagine if you wanted to take advantage of the threads, then you'd have to find some thick steel horizontal surface to drill and tap... and have access inside/underneath to catch shavings. That doesn't sound promising. Otherwise need to weld or use some kind of clamp, which would be at least as hard as attaching a bar, right?

This message has been edited. Last edited by: electricpete,

EPete,

You are right. The all-thread absorber has to be attached in a rigid manor, whether in a tapped hole or welded to a plate and the bolted or welded to the structure. The main point is, I can make an all-thread absorber in less than an hour and prove or disproove my theory. If you are absolutely sure that an absorber will work, then the temporary absorber is a waste of time. I have had a few situations where I wasn't sure that an absorber would work and used an all-thread absorber as a test. If the vibration decreased considerably, I fabricated an absorber that would last. If it didn't, I had to look further into the problem.

Regards,
John J

Thanks John. That makes sense.

Thanks to all,
The customer had someone in to trim balance the fan. Two shots dropped it into the dirt! My only concern is that if there really is a resonance issue here (my last thoughts are that it is a natural frequency in the fan bearing/pedestal/sole plate system, that as soon as there is a slight change in the balance, the vibration levels will be back at an unacceptable level.
Ron