Was it loose or were the operating forces greater than the design limits? The fan rotor on a very large boiler fan had been replaced (point No. 1, below). Along with the new fan rotor came new pillow block bearing inserts. Somehow water had infiltrated into the new bearings, but not before having run for a day. The oil contamination problem was corrected and the equipment was placed on-line. Immediately, there was an increase in vibration (point No. 2). The first thing that came to everyone's mind was that the new bearings were ruined. It was figured that the fan rotor had "rolled" (i.e. physically displaced) some babbit and was now loose because of the excessive clearance in the bearings. It was tempting to just replace the bearings, logic would seem to indicate that path forward. However the vibration signatures told a different story.

The story went like this. The high fan vibration was actually due to fan rotor unbalance. The new fan bearings were still good, but they were getting beat up by the fan unbalance. The unbalance was bad enough to make the bearings feel, by hand, as if they were loose. This idea was contrary to doctrine. Fan rotors were always balanced by the repair shop, and they had never ever ever ever had a balance problem in the past. The vibration analyzer, it was felt, must be wrong. But a well known vibration expert by the name of Art Crawford once said that the "little black box never lies." Well, he was right, because after balancing the fan, the fan problem went away. In the midst of all this analysis it was determined that the motor was ready to fail soon. So much attention was given to the fan that nobody ever considered that the motor might be bad. The motor was replaced and the vibration levels on both pieces of equipment reached a six year low.

Vibration Trend - Inboard Fan Bearing

Overall vibration level trend shows how changing the fan rotor (point No. 1) improved the vibration for only a moment. Within 3 days the vibration level was back up to its alarm limits (point No. 2).

The effect of changing the motor can be seen by the vibration level drop from point No. 2 to point No. 3.

The effect of balancing the fan rotor in the field can be seen in the vibration level drop from point No. 3 to point No. 4. The fan bearings were still good and did not need to be replaced.

Vibration Trend - Inboard Motor Bearing

This is the trend over the last several years (i.e. 2300 days). The vibration level jumped significantly at point No. 2.

The effect of changing the motor can be seen by the vibration level drop from point No. 2 to point No. 3.

The effect of balancing the fan rotor in the field can be seen in the vibration level drop from point No. 3 to point No. 4.

Point No. 1 Motor Vibration - Inboard Bearing

After Replacing Fan Rotor and Before Replacing Motor

The fan rotor was replaced at point No. 1 on the trend chart. The fan bearing vibration level had alarmed (point No. 2 both trend charts) and vibration readings on the motor indicated a bad bearing. It was felt that the vibration on the inboard fan bearing was okay. The signature on the fan bearing did not show any rubbing or other indication that babbit had been "rolled" (i.e. physically moved), so it was felt that the motor had priority and required replacement as soon as possible before any action was taken on the fan.

The spikes in the mountain of vibration in the graph to the left, are nonsynchronous harmonics, and are indicative of a bad bearing.

Point No. 1 Fan Vibration -Inboard Bearing

After Replacing Fan Rotor and Before Replacing Motor

The fan bearing shows very little broadband frequency generation. Rubbing and metal to metal contact are usually broadband in nature. For instance, the small mountain of vibration at 60 to 80 orders is rubbing. If that rubbing was really severe, then the mountain might extend out to 100 to 140 orders (note: 60 orders = 60 x running speed).

Point No. 3 Motor Vibration - Inboard Bearing After Replacing Motor

The vibration level drops significantly. Notice the obvious disappearance of the mountain of vibration in the 80x to 120x order range. This reading was taken before balancing the fan.

Point No. 4 Motor Vibration - Inboard Bearing After Balancing Fan

The vibration level of the motor doesn't drop much. However, sideband frequencies disappear. The vibration signature is cleaner. This shows the subtle affects of vibration carryover and how it effects the position of the motor rotor within the stator. Any deviation of the rotor from its magnetic center, both radially and axially, will show up as sidebands around the rotor bar and stator slot frequencies.

Point No. 2 Fan Vibration - Inboard Bearing

After Fan Rotor Change and Before Motor Change

This is what the vibration looked like before changing the motor and balancing the fan. There is a dominating 1x frequency (far left of graph) that indicates fan unbalance. The waveform (not shown here) also shows a clean 1x sinusoidal wave. The multitude of spikes are all harmonics of 1x frequency. This indicates looseness. Was it really looseness, or was it good bearings trying to restrain a severe unbalance?

The vibration level did not change significantly when the motor was changed and still showed the multiple harmonics.

Point No. 4 Inboard Fan Bearing

After Balancing Fan

After balancing the fan, the vibration level dropped from 3.6 mils down to 0.4 mils, a factor of 9. Note the significant drop in the 1x frequency (spike far left). The majority of the harmonics of 1x are also gone. The harmonics indicated looseness, but in this case it wasn't really looseness. It was a condition of overload.

Anything will behave as if it were loose if one applies enough force to it.