Vibration analysis of industrial machinery has been around for many decades, but gained prominence with the introduction and widespread use of the personal computer. Vibration Analysis refers to the process of measuring the vibration levels and frequencies of industrial machinery, and using that information to determine the “health” of the machine, and its components.
When an industrial machine (such as a fan or pump) is operated, it generates vibration. This vibration can be measured, using a device called an accelerometer. An accelerometer generates a voltage signal, proportional to the amount of vibration, as well as the frequency of vibration, or how many time per second or minutes the vibration takes place. This voltage signal from the accelerometer is fed into a data collector, which records this signal as either a time waveform (amplitude vs. time), as a Fast Fourier Transform (amplitude vs. frequency), or as both. This signal can then be analyzed by a trained vibration analyst, or by the use of a “smart” computer program algorithm. The analyzed data is then used to determine the “health” of the machine, and identify any impending problems in the machine, such as misalignment, unbalance, a bearing or lubrication problem, looseness, and more.
As an example, if we took a general industrial fan, removed one of the fan blades, and started the fan up, we could expect the fan to vibrate, due to an unbalanced fan wheel. This unbalance force would occur one time per revolution of the fan.
If we re-installed the fan blade, this vibration would be reduced.
In another example, if a bearing on this fan had a spall (a portion of the bearing race damaged, much like a “pot hole” on a highway), each time one of the bearing’s roller contacted this spall, it would generate a vibration.
If 3.2 rollers “hit” the spall per revolution, we should expect to see a vibration signal of 3.2 times the running speed of the fan.
The use of vibration analysis can determine problems caused due to improper installation, machining errors, insufficient lubrication, improper shaft or sheave alignment, loose bolting, bent shafts, and much more. It can, in most cases, detect these problems long before the damage can be seen by maintenance, and long before it damages other machine components.
The use of vibration analysis, condition monitoring, or predictive maintenance has made great strides increasing the usable life of machinery.
Related blog posts: