Primary accelerometer calibration is at the root of traceability for virtually every accelerometer calibration chain in the world: commercial, government, military, academic or vendor. Acceleration calibration via laser interferometry is a primary method because it is an absolute method comparing the measured vibration from a sensor under test to a constant of nature – the wavelength of laser light.
The ISO16063-11 standard discusses three alternative methods for primary calibration. Each method requires increasingly more complexity in hardware as well as algorithm/analysis, and leads towards increased functionality
such as inclusion of phase information and sub wavelength measurement resolution.
The most commonly employed technique is Method I - Laser Interferometry via
fringe counting. The benefits of this method center around the use of relatively simple electronics in the form of counters to detect interference fringes in the reflected laser beam. ISO16063-11 specifies that this
technique is only useful for a low frequency range (up to 800 Hz) but footnotes that it can be applied to higher frequencies under special circumstances – namely, the acceleration levels must be significantly increased
at frequencies above 800 Hz. This method has an extremely low uncertainty contributed mainly by mechanical components rather than the electrical components.
Drawbacks to the interferometry are that it can only be used at low frequencies at which displacement is large enough for a significant number of fringes to be counted. This technique also provides no phase information.
As with all of these techniques that will be discussed, the position of the point of laser reflection is important, preferably at the plane of mounting. However, since the shaker might not have purely planar motion, multiple
measurements may need to be made to determine or verify the planar motion of the device under test. Hence, the systems require either time consuming multiple measurements to be made or multiple laser heads. Additionally,
significant attention and investment is made to provide a electromechanical exciter which minimizes any transverse motion to the greatest extent. This is typically accomplished in the form of a new generation
air-bearing calibration grade exciter.
The ISO16063-11 - Method II - Minimum point method is extremely cumbersome and manually intensive, requiring the user to tune the acceleration amplitude to match the zero crossings of the laser. Because low transverse
calibration exciters are incapable of providing the displacements fringe counting needs at high frequencies, this method is employed in the frequency range up to 10 kHz.
The primary benefit of this technique is that by utilizing a low frequency modulating mirror and compensating mathematics, the counting methodology can be extended into the higher frequency ranges. However, certain implementations
require a exotic high g level piezoelectric shaker to attain the desired displacements at those higher frequencies. Clear drawbacks from commercialization include the complexity and durability concerns with the low frequency
moving mirror to modulate the beam. Effectively, Method II is not able to be implemented practically.
Method III - Sine approximation methods originally required much more digitization and processing; however, this has become realizable with today’s electronics. The fundamental benefits are that it can handle the complete
frequency range at reasonable acceleration levels and also measure phase information. Primary drawbacks to this technique are the increased costs associated with a general purpose laser vibrometer and sometimes acquisition
time due to speckle drop out on a non-cooperative target. However, there has been very recent advances in multi-head laser primaries which have reduced acquisition time by eliminating scanning or repositioning and also
utilizing integrated cooperative targets in the calibration reference.
The standard is very detailed in specifying and quantifying uncertainties for the various components of the systems. The single largest challenge with any system is to minimize the larger contributions to uncertainty. Transverse
motion of the mechanical exciter is a significant source of measurement uncertainty that ISO 16063 specifically identifies. Accordingly, the standard states the following limitations for transverse motion:
1 hz - 10 hz < 1%