Vibration in Centrifugal Compressor?

Please do pick a unit you like. In the US typically mils pp are used. In countries that use SI units, then microns pp are used. I use both in my work. What country are you in?

A centrifugal compressor vibration should be low under normal circumstances. Perhaps 1 to 2 mils pp (25 to 50 micron pp). The higher the compressor speed the lower that vibration should be under normal conditions.

Permissible vibration is a function of speed and bearing clearance. A compressor running nominally 3000 to 5000 rpm should probably have alarms or shutdowns in the range of 3 to 4 mil pp (75 to 100 micron pp), assuming a bearing clearance of about 6 to 8 mil (150 to 200 micron).

A high speed machine like an air compressor that runs at 15000 to 25000 rpm would have much lower normal and permissible levels. 50% or less of the above.

The OEM operations and maintenance manual for the compressor should contain their recommended alarm and shutdown levels.

The full scale range on the 3300 is adjustable. You pull out the card, flip the right dip switches on the mother board, and insert a new scale paper into the display and your 4 mil pp full scale range is now 20 mil pp full scale.

This process is detailed in the 3300 users manual.

The max the system can measure is a function of the probe not the monitor. Standard prox probes have an 80 mil linear range, so really the most a standard probe can handle is 40 mil pp. Compressor vibration should never be more that about 10 mil pp even in malfunction conditions.

Standard scale factor on a Bently prox probe is 200 mv/mil, so yes you can measure the ac voltage off the signal and apply this scale factor and arrive at an overall vibration amplitude in mils pp. Do not us the RMS value, as the unit is peak to peak. A standard volt meter will generally give you an rms value, and if the vibration is at a single frequency (running speed ususally) you can convert that to pp by multiplying by 2X 1.414, but most vibration is not single frequency so that does not work so well.

Best bet is to use an ocilloscope and measure the peak to peak value right off the time wave form.

Hi,

I don't know what the "normal" vibration should be, that's over my head we get the vibration alarm values from our mechanical engineer.We use mils P to P as measuring units.To make changes on the 3300 system(alarms,ranges etc.) as mentioned by Steve S. the module has to be pull out from the rack a little ways to get to the dip switches, my memory is terrible so I have to refer to the user's manual almost every time there is a change to be made which is rare.

1. "got some answers and yours sounds good and more technically correct"
2. "if I follow the method answered on my previous question,which says I need to convert it to rms value first"

I agree with number 1, Steve S. has some impressive credentials I am just the nuts and bolts man

number 2: your voltmeter reads rms voltage so there is no need to convert to rms first my friend you already have rms voltage in your meter.Please read my answer again, in that example we knew how many mils and PP voltage and we were trying to find rms voltage

the output of the proximitor is PP but your meter is only capable of reading rms or half of the wave, to read PP you will need to use an oscilloscope,I personally have never used an oscilloscope to figure out vibration.

If you have 0.837 ac rms voltage,(half wave)NO conversion needed, divided by the constant 0.707 equals 1.1838 multiplied by 2 (to get PP or FULL wave) equals 2.3676 ac Peak to Peak voltage, now let's convert it to mils,(1 mil equals 0.2 volts), 2.3676 divided by 0.2 equals 11.838 mils peak to peak vibration.

When using the Tk3 rotate the indicator until the probe is in the center of the wobble plate and set the probe at 10 vdc then switch your meter to ac, as you move the probe out away from the center towards the edge of the plate the vibration increases, lock it in anywhere between the center and the edge take the ac rms value showing on your meter and use the method I just described above to figured out how many mils of vibration you have.

If there is a certain value of mils you want to know what the rms voltage will be use the method I describe in the answer I gave you a while back.

Thank you and I hope I didn't confused you even more......take care

HI,

reply number 5 was posted by me, I forgot to sign in .....Thank you