Reliability Engineering Snapshot TM

The Illustrated Newsletter For Failure Analysis, Non Destructive Evaluation, and Predictive Maintenance

Vibration

November 1999

Subtle changes in a vibration waveform can help pinpoint problems. What follows is the method in which that question was answered.

This agitator gearbox was taken out for overhaul. The machine was shut down before any serious damage was done to the gears; the entire gear set is in good shape and is salvageable. The operators had noticed a high vibration that was shaking the tank and reported it. The question to answer was whether it was the motor or the gearbox. Some vibration readings were taken and it was determined that not only was the vibration coming from the gearbox, what's more, the vibration was coming from a 14 tooth gear. Field inspection of the unit revealed that there was a 14-tooth gear and it was getting beat up by another gear that was loose. The cost to make this call in terms of a mechanic's wages was $95, and represents the cost to gather vibration data 11 separate times over the last four years.

The value in long term monitoring is that one gets to see how a piece of equipment vibrates on any normal given day. There are always subtle changes in the waveform depending upon loading conditions, but when the shape of the waveform changes dramatically, then there is a problem. The shape of that waveform helps to pinpoint exactly what it is that has caused the change. The graph shownabove contains five of the 11 waveforms of the gearbox that were collected over the last four years, beginning with Oct. '95 on the top, and ending with Nov 8, '99 on the bottom. Between Oct. '95 and Mar. '98, the waveform shape looks about the same. However, the waveform for Nov. '99 is dramatically different. What has caused the change?

The graph to the left shows a close up of the waveform for Nov. '99. It represents three complete revolutions of the gearbox high speed input shaft (left to right), and each revolution is highlighted by a vertical black line. The waveform shows the actual change in direction of the vibration.

Now imagine that the centerline of whatever is causing the vibration is located at the number zero shown in the middle of the left side. Imagine that the centerline follows the exact pattern of the blue line as it moves from left to right in three revolutions. If you count how many times the centerline actually crosses the zero line and then comes back to its starting point (i.e. zero), you will come up with approximately 14 times that it does this during each revolution. The easiest way to determine this is to just count the number of clean sharp peaks on one side of the zero. If what you're seeing is really happening then it should show up on another type of graph known as a frequency spectrum (shown below). If it doesn't show up there, then you are imagining things and need to go home.

Agitator Gearbox Vibration Frequency Spectrum

In the graph to the right the vibration is shown in a different format. Instead of showing vibration in the form of amplitude and direction versus time, as in the waveform above, the vibration is shown as amplitude versus the frequencies that make up that vibration. The spectrum has some distinct peaks and are labeled as to how many times faster they are vibrating then the rotating speed of the high speed input shaft. As can be seen, there is a very pronounced peak at 14 times running speed. The only thing that could cause this would be a 14-tooth gear. The severity of the problem shows up as multiples, or "harmonics" of that "14 times" peak. These harmonics are labeled at 28, 42, and 56 times running speed.

If the overhaul of this gearbox is good, then the harmonics at 28, 42, and 56 will disappear and the frequencies close on either side of the 14 X frequency will disappear (graph to the right). The sidebands around the 14 X frequency are highlighted by square boxes. These sidebands represent harmonics of the high speed shaft speed (1 X) as it bounces around.