It is recommended to calibrate the measuring instruments before performing any experiment. Ideally, how often do we need to calibrate an accelerometer? Are there any simple and effective methods for calibrating accelerometers in the laboratory without seeking professional assistance?
Consult the following link, It may be helpful.
https://nvlpubs.nist.gov/nistpubs/jres/041/jresv41n5p359_A1b.pdf
A simple and practical way to get the calibration factor of an accelerometer is by aligning the measuring direction with gravity. The data recorded would be the equivalent to measuring gravity acceleration, i.e., 9.81 [m/s^2]. Then, rotate the accelerometer 90°, which indicates 0 [m/s^2]. Lastly, rotate the accelerometer another 90° and the data would record a negative value of gravity acceleration. With these three points, you can correlate the voltage with real acceleration, i.e., the calibration factor of the accelerometer. There are more sophisticated ways but I find this way very easy to apply.
The method suggested by Diego Hernandez-Hernandez works for inductive and mems-type accelerometers. However, for the typical piezo-electric IEPE (ICP) type, one's low freq calibration is not so easy. One can use the single-ended or double-ended calibrator. If you have a laser Doppler vibrometer, and a modal shaker, one can calibrate the accelerometer against the laser.
Stick the accelerometer to a weight of known mass. Excite the weight with a known force using instrumented hammer or shaker. The acceleration and force will be related by F=ma. If you know the mass m, you can work out an appropriate calibration factor quite easily.
We usually perform calibrations every time we need to use it. For this, we use a modal shaker PCB 394C06.
If your calibration aims at being recognized officially, you need to comply to the ISO standard ISO 16063-1 and have your calibration made by a certified metrology reference lab! (or by the manufacturer, which generally offers this periodic calibration service)
From my experience, you should definitely calibrate your accelerometers at least whenever you do a new measurement set-up.
Why?
Not because you do have a typical drift of sensitivity but
- this is an easy way to check whether you do have any defect on the accelerometer
- you can check all of your set-up, including wiring, signal conditioning, potential electrical noise, channel labels / alignment, ...
As soon as you are doing instrumentation "in the field" and more than half a dozend of channels, this is really going to be essential and will save you a lot of time and headaches, even though it might seem to be an unnecessary waste of time at a first glance.
Perhaps it is not that important any more today with IEPE instrumentation as some potential issues in the signal chain disappeared (charge amplifiers, etc.), but i would still strongly recommend it (just for channel alignment and validating that no cables are defective).
How to do that? There are commercially available calibrators for acceleration in the market producing well-defined acceleration on a mounting surface. And yes, they work dynamically with a sine wave - typically at 159.2 Hz and 10 m/s². They can be hand-held so that you can really calibrate the complete signal chain when mounting a readily wired accelerometer even in a field installation. About this frequency range is useful as it is above DC, i.e. already produces significant acceleration while still not being too high to cause issues with added mass by most accelerometers - and going with a sinusoidal signal allows for detection of distortion and electrical noise as well.
For lab calibration, going with hammer measurements / defined mass works also well, of course. And if you are looking into specific issues / instrumentation, then you also might have a good reason to rethink your calibration needs (e.g. very low frequencies, etc.).
Joshua W.R. Meggitt Do you have experience of real test with impact hammer and accelerometers? I mean Roving hammer (EMA). I have one question in this domain.
I am using Artemis pro modal analysis software for EMA and I imported the data directly in it.
I am not sure if I am correct since I have not removed the noise.
Now the mode shapes are not smooth.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling