Vibration Analysis

There are piezoelectric sensors available with built in amplifier, you can Analyze output waveform on a CRO.

I would be thankful if someone have some good literature for vibration analysis

Have you tried vibration analysis.

When you go for slow speed machines, try for the vibration spectrum on amplitude (mm) or at the worst velocity (mm/s).

What are the features you are trying to monitor will largely define what exactly you should attempt.

For our machines usually an FFT suffice to monitor the gears and bearings. However the time series analysis some times provide much more information than the FFT (which averages over a period).

The vibration analysis is a huge subject and going through web may not be enough. Better way is to get hold of consultants/experts which you can hire from universities or other places (eg FAG, SKF etc have their condition monitoring groups who will do the coondition monitoring for you with their equipment and give you the results and they conduct regular classes too for this (of course on chargeable basis)

Thanks,

We have purchased ProofTechnik make instrument and measuring Shock pulse measurement.But it is required to normalize the readings taken by the instrument.

For normalizing ,we need to provide shaft dia. and RPM-On this basis software calculates dBi value.But we have to provide dBa value also.

Can anyone guide how dBa value should be feed to get normalized readings.?

Many accelerometer chips are designed to specifically work at these low frequencies. You can even get these in three axes orientation that permit use in inertial navigation systems. Before you or anybody else here freaks at some of the prices you'll find for some of these devices, no budget was specified. Similarly no resolution or anticipated energy levels was specified either.

If you choose an FFT analyzer (which includes DSP processing, and usually programmability), be sure to use synchronous sampling. Low RPM machines frequently change RPM due to load variations, and synchronous sampling will assure that harmonics are always sampled at integral multiples of the RPM. This will make level measurements relevent for each harmonic.

See the attached URL for more on synchronous sampling:

http://www.wavelet.org/tutorial/synch.htm