10 cpm sensor recommendation

[Tmoose]

I need to measure a large slurry tank's wall motion down to 14 cpm or so. MILS/thousandths of an inch are the preferred units.

Old IRD DataPac 1500 says it will go to 10 cpm non integrated, or 21 cpm integrated. The attached image shows measurements made with a garden variety 100 mv/g accelerometer, double integrated (?) to displacement. I made a test rig and was happy with the amplitude and frequency accuracy. I will repeat when I can figure out how to spin the rig down at 5 rpm or so. I'm expecting I will need at least a lower frequency sensor. And maybe borrow an old HP 3561 that claims to be able to analyze signals way below 1 Hz.

Any recommendations for a sensor?

thanks

Dan T

 

[Tmoose]

Should have said the attached image was an actual measurement on the vessel, and the HP 3561 specs say it handles signals below 1 cpm
(0.000125 Hz to 100 kHz)

Dan T

 

[DLite30]

Take a look a PCB's website...they have several seismic accelerometers that span your frequency range.

My first though was to use a prox probe or an lvdt if you're looking for discplacement, but I'm guessing you're needing to reach several points on the tank wall where there's no place to mount them.

 

[MachineryWatch]

Dan,

I commented on your other post about not understanding the scope of your measurement task. Do you have any software for post processing stored/digitized data?

I have been working with some battery powered, triaxial MEMS accelerometers. They are good from DC through about 80 Hz. They have user selectable sample rates up to 160 Hz and store data onboard using a microSD card. The data is stored in ".csv" file format. Multiple units can be synchronized within a couple ten thousandths of a second, so I had adequate phase resolution to animate an ODS model for a large (5000 ton) HVAC compressor at 27 Hz.

You can also program the units to start data collection at a specific time. Each uses a "AA" battery and a standard alkaline type "AA" is good for several hours of data acquisition.

The units really "oversample" at much higher sample rate to prevent aliasing.

They can be set to 12 bit or 16 bit precision. The user can also set a high pass filter to eliminate the 1 g offset due to gravity in the axis reacting to gravity.

If you were happy with the results of a 100 m/volt g sensitivity piezo accelerometer for amplitudes these should work better.

There is a fair amount of manipulation of the data you will need to do. I import the .csv files into a spreadsheet to get segments of the files "synchronized". Then I visually compared the waveforms to get phase data, which was accurate enough for my ODS because I had a dominant frequency.

You may have to export the data files to an FFT/signal analysis program of some kind to do more processing on the digitized data, if you do not have a dominant frequency.

Let me know if you want more info about these little sensors. I will definitely be using them more for low frequency and structural stuff in the future.

Email me for info: skiph-at-machinerywatch-dot-com

http://www.machinerywatch.com