Up Time Magazine - December 2008/January 2009 - (Page 57) Figure 3 - The peak close to 0.5x running speed was higher than the running speed peak, and there also appeared to be a harmonic at 1.5x running speed. Figure 2 – A multi-stage pump being lowered into position. often associated with rubs. The most likely place for a rub would be the impeller rubbing in the pump. Perhaps, there was some debris in the suction? The sub-synchronous frequency was 750 cpm, or .4x running speed. The upper frequency of interest was 2825 cpm, or 1.6x running speed. Neither was an exact ½ multiple. The possibility remained that this might be the cage frequency of the lower bearing in the motor. However, since most of the vibration was in the top of the motor, this theory was discounted. Another possibility was that the frequencies in question were non-synchronous vibration being excited by the pump or the motor. Researching the OEM manuals revealed the pump natural frequency between 500 cpm and 800 cpm. This could actually be a resonance. There were now three clues leading us to the conclusion that the problem was with the pump, possible resonance, possible rub, and possible fulcrum effect. The next day the vibration doubled again and Dispatch agreed www.uptimemagazine.com Figure 4 - Spectra of uncoupled, unloaded motor. Levels were 25% of the coupled and loaded reading, but contained a high 2x running speed peak. to allow the unit to come to half-load and remove the pump from service. We removed the coupling and ran the motor solo. The vibration was about 25% of the coupled readings, but these readings were still higher than what was expected of an uncoupled, unloaded motor. And, there was a very strong 2x running speed signal (see Figure 4). Because of the reduced overall vibration on the motor, the decision was made to pull the pump. The pump inspection revealed extensive damage. It was a complete wreck. Something had apparently gone through the pump and destroyed several impellers (see Figure 5). The first stage cast iron housing was broken off of the pump. All of the cutlass bearings were in need of replacement and one had seized to the journal and was spinning in the housing. Due to the machine work on the journals and bearing housings and the lead time to order impellers, we decided to send the motor in for a routine clean, dip, and bake. The motor inspection revealed no obvious Figure 5 – The completely trashed pump. 57 http://www.uptimemagazine.com
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.