There is a resonance on a centerhung fan (confirmed by Peak-Phase analysis)...
Stiffening the structure is not an option, nor is adding weight.
Can an dynamic absorber be used for this type of equipment?
Here is a quick sketch of the equipment.

Do you know which components are resonant? What is forcing the resonanance?

Just wondered. Dynamic absorbers can often help out, but they do fatigue, sometimes over a short period of time.

Have you tried just using a C-clamp and steel strap, just to find the sweet spot?

Not sure what component(s) is in resonance, but the flimsy floor is definitely not helping.

Haven't tried strapping anything down, the fan rotor is 20 feet in diameter and 10 feet wide... the only thing to strap it to would be the floor.

I just have never seen or heard of a dynamic absorber being used on this type of equipment, so wondering if it's an option.

Thanks.

You need to know exactly haow the structure is vibrating (mode shape) before even considering any type of modifications. Are we talking about a shaft critical (balance resonance), bearing pedestal, casing, foundation or what? If the fan rotor is 20-ft diameter, then that is a big fan. I would not consider C-clamps, cables or any ties to a "floor". Dynamic absorbers have been applied to large fans with bearing pedestal resonance. What forcing frequency does the resonance match?

Walt

quote:

Originally posted by Walt Strong:
Are we talking about a shaft critical (balance resonance), bearing pedestal, casing, foundation or what?

Walt,

A related question. How can one differentiate as to what component is resonating if, like in this case, all of them are rigidly connected (even rotor as part of the structure)? If, for example, foundation is at resonance then the bearing pedestal may vibrate at the same or even higher magnitude. Can the structure's mode shape be a guide here?

Dave

That's a really big fan. I would do like Walt said and use great caution in modifying it. I would also think that any dynamic absorber would have to be very large, given the high mass of any of the components.

Has anybody ever used a dynamic balancer? I think Balmac makes them and it tunes the balance during operation.

Assuming that imbalance is the exciting force, this might be a place for one.

Before trying a dynamic absorber I would balance it down to nothing and see if that makes it ok. If so then a dynamic balancer is probably a good investment.

Dave,

Both point impedance (force and response at same location) and mode shape can reveal what is important. Several tests can be used including coastdown, operating deflection shape (ODS), and impulse-response methods. On large fans, one needs to know the foundation type/design and soil properties. Bearing type/design is needed in case a stiffness or damping change is considered. The fan rotor can have a shaft critical (bending mode) or various wheel (blade, shoud, and backing plate) vibratin modes modes. If vibration frequency is at 1xSS, then a good trim balance (Steve's suggestion) and excellent shaft alignment are cheap fixes compared to structural modifications. If fan is new and under warranty, then by all means correct any resonant structure issues. I have worked on fans up to 8000-hp.

Walt

What is the frequency of the resonance? At or near running speed? Is it variable speed? Why use absorbers as the first option, has every other option been exhausted?

Has this thing all of sudden started vibrating or is this a long time history or is it a new system or what? Like Walt said, if it is a new system, correct the structure problem now.

Balancing the rotor, as others have mentioned, might help things unless the resonance is at some wild out of reasoning frequency far far away from being related to running speed. But something has to start the resonance vibration, again, what frequency is it?

There is not enough information in the post to suggest any solutions. For an absorber one would want to know the operating deflection shape and mode shape - can be approximated with hand methods.

Usually, one doesn't start at an absorber for the solution. The gear also spins at the same frequency as the motor (assuming this is of interest - not clear in post). Alignment and coupling issues should be checked in addition to balance prior to jumping to an absorber or multiple absorbers.

quote:

Originally posted by Walt Strong:
...The fan rotor can have a shaft critical (bending mode)

Walt,

This is one of the problems I meant: a rotor may be in a bending mode an analyst in real life has no way of knowing unless he can conduct a modal test on a free rotor or can model it.

Dave

Dave,

I have crawled inside many large fans for conducting an impact test. I have also impacted exposed shaft and coupling with some success. If you conduct impact tests on bearing housings, pedestles, and foundation combined with the other tests that I mentioned, then a rotor critical can be found without modeling. Modeling can be used as a tool for shaft/structural modifications provided that the model agrees with experimental test data. Remember that poor modeling may have caused the problem!

Walt

quote:

There is a resonance on a centerhung fan (confirmed by Peak-Phase analysis)...

Vibex, could you explain the process used to identify the resonance.

Once had a large centerhung fan with high 1x that did not respond well to balancing (by others). By phone, I suggested the bearing bolts and pedestal bolts be checked for tightness. They said this had already been done. When I arrived on site, I found the nuts were in fact tight, but several were bottomed out on the threads, leaving a slight gap between the base and the nut (found this with a bare finger run along the baseplate joints). They removed the nuts, added a heavy flatwasher, retightened the nuts, and the problem was solved.

Have you checked all the "simple" but less obvious possibilities? I find it unlikely that a fan this large would actually have a resonce problem.

Thanks for all the replies... I've been trying to follow what has been said, but I'm still new in this field.

Additional information:

- Resonance is at operating speed of fan (~710 RPM);
- Fan is on 6th floor of old building... the floor not very thick and is supported by random vertical beams;
- Balancing was performed on fan, reducing amplitudes from 0.46 in/sec (highest) to 0.18 in/sec (highest); 1x rpm remaining = 95% of amplitudes;
- Alignment was performed by client... so we do not have any specs, but no 2x or 3x showing in spectra;
- Took a Peak-Phase analysis during final balancing run (during coastdown), phase changed about 180o from 710 to 650 RPM;

Assuming a resonance is causing the vibration :
There really is nowhere to stiffen the structure. Another solution would be to stiffen the floor which the fan sits on, but that's not an option either (too complicated = $$$). Adding mass is out of the question. This is why a dynamic absorber was maybe an option. I haven't seen or heard of too many dynamic absorbers on centerhung fans though.

So what would be another option, assuming there was a resonance?

Would you suggest performing impact tests to locate the resonating component?

Client mentioned that their team has been having trouble balancing the fan ever since the fan blades have been repaired (repaired = welding plates to blades where pitting had gone through, all work was done inside the fan).

Thanks,

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

Which direction is this .18 in/sec (5 mils)? Vertical?

Does the floor seem to give the impression that the fan is out of balance more than it really is? People standing on the floor feel a very bad vibration?

How much weight did it take to bring the amplitude down from .46 to .18 in/sec? Which direction was it balanced in? Was it balanced in single or dual plane? What were the phase differences between the bearings in all directions before and after balancing?

OK, sounds like maybe you do have a resonance. Did you actually get a 180 phase shift, or did the phase plot just go "off the chart" at top or bottom (this happens when phase value passes through 360 degrees -- often mistaken for an abrupt phase change).

A simple, single channel impact test will verify this fairly easily. Best done with fan off. I'd use a large rubber mallet (not a deadblow hammer though).

Since this is in a building, it would not be unusual to mount the entire unit on isolators, assuming it has a skid-type frame. This would decouple the fan from the structure which is the most likely source of the resonance. This would have to be an "engineered" solution, but the actual implementation would not be that expensive. Also, stiffening the floor, in the correct location might not be as difficult as you'd think, but you don't want this to simply be a trial and error effort.

Many great suggestions here. But I would like more information on the fan and application itself. What process or air temp is it pushing? What type of bearings? (assuming babbit) oil type and temps? ambient air temp? Is this a new problem that crept up or has it always done this? Has the structure housing and the fan blade assembly been inspected for cracks or breaks that would have reduced the structural integrity and lowered the critical closer to run speed? Just some of many questions that might offer more insight into possible solutions. I'm a big believer in letting the data lead you to the problem. I get myslef into trouble by jumping to conclusions to often. More info would help me better understand your dilema.

quote:

Took a Peak-Phase analysis during final balancing run (during coastdown), phase changed about 180o from 710 to 650 RPM

This would suggest a very poorly damped resonance, since you have gone the full 180 degrees within 60 cpm. This could be a simple structure or piping resonance, but very unlikely a rotor resonance as others have stated.

You need to take coastdown data at numerous points along the structure (including piping to the bearing housing if that is where you are seeing the high vibration) to pinpoint the component that is being excited. Once the component in resonance is determined, then stiffen as appropriate. Without much information, I would suspect oil piping to the bearing housing.

2 cents for you. Is this your first set of data for the machine. If not has is always had this issue? Check ALL the simple things first. Check all foundation bolts on the machine for tightness. Perform a thorough inspection of the support structure under the machine looking for broken welds, missing bolts etc...
Gene Ford of Technical Associtates or Randy Fox of Entek/IRD maybe offer more insight on the design and installation of a dynamic absorber

quote:

Took a Peak-Phase analysis during final balancing run (during coastdown), phase changed about 180o from 710 to 650 RPM;

What did the amplitude do from 710 to 650? Has this been a problem for a while, or did it just show up recently?
If this is a new problem then I agree with looking for simple first. If you have a strobelight you can look over it and possibly find problems. You should set the strobelight slightly above or below the running speed to see the movement.