Foundation Mass/Looseness

[nnjunger]

I have a rather large fan (250 HP, belt driven, 600 rpm, 9'2" fan wheel diameter). We had some vibration on the fan prior to our shutdown, we opened and inspected the shaft bearings. Upon inspection, we found bearing looseness, so we replaced both bearings. After reassembly, we are still getting high vibration readings. Our field mechanics drove wedges between the fan foundation and the building slab, which lowered the level. I believe that this just increased the mass of the foundation to "dampen" the vibration of the fan. We will attempt to rebalance the fan on our next shutdown, but I would like some info/literature to try to calculate what the foundation mass should be for this fan. Any help, suggestions, or other guidance that you could give me would be greatly appreciated. Thanks.

 

[MachineryWatch]

You left out a lot of information that we would need to help you. What was the frequency of the vibration? Do you have spectral data? Is the fan mounted on "isolators"? Did the wedges "short out" the isolators (if isolators were installed)?

If there were no isolators and the wedges resulted in a significant reduction in amplitude, it may simply mean that you have looseness between the fan foundation and the building slab. This looseness should be corrected. This can be accomplished with grouting, but first correct other problems revealed in a carefull inspection, like anchor bolts pulled loose from the concrete, cracked or stripped bolts and nuts, cracked welds in the structure, erroded concrete (voids) under the fan foundation, etc.

Looseness also often shifts the machine's "natural frequencies" to a point where they coincide with one of the machine's forcing frequencies (RPM harmonics or belt frequency harmonics, or blade pass frequency harmonics). This creates a resonant condition which actually amplifies the vibration. Insertion of the wedges may have stiffened the structure enough to shift the natural frequency higher to reduce or eliminate the effect of the resonance.

You are right that a more solid connection to the building slab would increase the effective mass. Increasing the effective mass would result in a reduction of energy. Remember, F=MA, so increasing mass reduces acceleration (vibration). This is not a bad thing in terms of improving the life of the machine, but may not be correcting the core problem.

You really need to do a complete diagnostic investigation using all the vibration analysis tools that are available (time domain, frequency domain, phase) before you attempt to make corrections, or you are likely to waste a lot of effort on solutions that at best will treat the symptom, and at worst will make matters worse.

Skip Hartman

http://www.machinerywatch.com

 

[nnjunger]

The fan foundation is supposed to be a seperate mass, there is an expansion joint shown per original construction. The motion was 1X, shown large amplitude in all directions. No isolators were installed that I have detected in either field observations nor per original construction prints. The natural resonance is not believed to be the problems since the phenomenon is new. Phase also indicate a machine looseness. Initial plant assumptions were balance related problems, and while this may be partially to blame, if the machine does not have proper footing/anchoring, then we cannot correct this with balance.

I do know that bearings were installed and torqued to proper specifications, the fan wheel was not removed from the shaft, the wheel is clean, visual inspection of the exposed fan foundation shows no evidence of cracking.

 

[MachineryWatch]

Here are a list of things that are usual sources of 1X RPM vibration.
1)Unbalance
2)Misalignment (not just couplings, also sheaves)
3)Eccentricity
4)Resonance
5)Looseness
6)Bent shaft

There are many kinds of looseness. As stated before some of these looseness conditions lead to resonance. Just because it is a new vibration does not mean it is not a resonance problem. Most resonance issues (and eccentric rotating components) produce highly directional vibration, but not always. I worked on a problem similar to what you are describing one time, where vibration was noticeably higher in all directions at 1X RPM on a belt driven fan for the paint booth at a car manufacturing plant. The fan wheels cracked, which caused the high vibration. They cracked because the fan wheel was resonant at 1X RPM, but high vibration did not develop until the wheel cracked. All fan wheels in all five fans cracked within hours of each other. They had been running for several years before this happened, with relatively low vibration levels.

I would inspect the fan wheel for cracks, inspect the fit of the fan wheel hub on the shaft, inspect the fit of the sheaves on the shafts, check sheave alignment, check motor-to-foundation looseness.

Skip Hartman

http://www.machinerywatch.com

 

[nnjunger]

I appreciate your comments. I can tell you from the remainder of the PM that was performed on this unit during the bearing changeout that the following conditions were inspected. Fan wheel to hub fit was inspected, sheave fit to shaft (motor and fan), belt alignment was checked on reassembly with a laser, motor to foundation was checked, all mounting bolts secure, as are the anchors for the pillow block bearings.

I appreciate that resonance could be a possiblity. I am more inclined to believe that it is a balance problem being compounded by a marginally adequate fan base, but I am having trouble getting a definitive handle on how "massive" your foundations should be. I know that this is a consideration for design of new facilities, but this is the first time I have really run into the problem.

Also, since this fan is mounted on a variable frequency drive, if the problem is resonance, wouldn't that go away if we adjusted the turning speed of the fan? We have tried that, and maintain the vibration level (only very slight changes in overall measurements).

Thanks again for your help.

 

[Tmoose]

There are rules of thumb about inertia block type foundations being at least 3X equipment weight. Of far greater importance is block width vs cg height (wide and low) and soil condition. A heavy concrete canoe (or too narrow foundation)would roll like a harbor seal pup.

I'd start by feeling for looseness as differential motion between frame and concrete at each mounting bolt location. I'd use my vibration analyzer to quantify the results. Frame, block and bolt should all be moving in-phase and with equal amplitude.

 

[ccw]

By process of elimination, it seems that the fan rotor is severely compromised somehow, ie. a rotor imbalance. You could continue to subdue all vibration transmission to the outside world from the rotor, but would only be masking the real problem if there is rotor imbalance. The next time the system is down, look for assymetric dirt build-up around the rotor vanes.