Accelerometer Calibration Frequency Response

Common System Options

Frequency response function (FRF) and single point reference frequency amplitude sensitivity are the most common forms of accelerometer calibration.  However, laboratories are becoming more interested in supplemental test capabilities to ensure both the health and performance of their users' accelerometers.  The use of the new generation of robust and precise calibration actuators allows for greater control and shaping of input energy to shock and vibration calibration.  The techniques are in use around the world by various sensor companies, as well as the leading aerospace and transportation manufacturers.

Base Calibration

The primary function of an accelerometer system is accurate (low uncertainty), reliable and fast calibration of a sensor under test.  Most offer a single point (100 Hz or 159 Hz) calibration for expedience and functionality confirmation, along with the more common feature to generate a full frequency response function (FRF).  Most modern systems provide the magnitude and phase response of the test sensor.  Though once an option, precision calibration-grade airbearing exciters are now a standard to reduce lateral inputs and ensure quality system measurement uncertainties.  Automated systems can provide further time savings and convenience like a test sensor database lookup which automatically configures the test frequency ranges. 

Mounted Resonance Frequency

Many users opt to test the test sensors mounted resonance as an integral sweep to the FRF calibration process.  Ensuring that the mounted resonance remains above the manufacturer's specification can serve as a rudimentary health check to the integrity of the internal sensing element.  Physical damage to the element such as a crack will cause a decrease in stiffness and the resonance frequency to be reduced accordingly.

TEDS Read/Write

Most modern Fast Fourier Transform (FFT) analyzers now come with the capability to manage Transducer Electronic Data Sheet (TEDS) functionality as specified by the IEEE1451.4 standard.  TEDS incorporates a small memory chip inside the sensor that can be accessed over the standard two wire operation with a reverse bias voltage. The ability to write to and update the electronic storage of a sensor's sensitivity is imperative and available in modern calibration systems.

Limited Frequency Linearity / High G

For limited frequencies, an economical means of exciting and verifying performance at higher amplitudes can be accomplished via a beam structured mechanical amplifier.  Piezoelectric accelerometers of typical mass can be driven at or near the beam structures resonance at various amplitudes up to 500 g’s.  The test sensor output at the various g levels can then be checked to insure linearity.

High G Shock

Exceeding a few hundred g’s requires a specialized actuator.  Using a falling or projectile mass, accelerations can be achieved in the higher ranges of 100 – 10k g.  Modern shock calibration systems utilize pneumatically fired projectiles, various impact anvils and customizable impact surfaces for controllable impact amplitude and duration.  The adaptation of pneumatic input has provided increased repeatability/control along with reduced uncertainties.  This range of shock calibration is useful for shock sensors used in automotive crash or aerospace pyrotechnic separation.

Ultra-High G Shock

At the highest shock requirements, such as 1000’s to 100k g, a long slender Hopkinson Bar test structure is employed.  This is effective by making one dimension of the calibration structure large compared to the wavelength of the imparted shock pulse. The pulse becomes a compression strain wave in the long slender bar which serves as a mechanical wave guide to the test sensor mounted at the end.  The reference can then be derived from either integral strain measurement on the bar or laser vibrometer measurement at the end of the bar.  This range of shock calibration is useful for shock sensors measuring extreme impacts such as high g crash, pyrotechnics and penetration.

Low Frequency

One of the more significant challenges in accelerometer calibration occurs at lower frequencies (sub 5 Hertz), since mechanical exciters are stroke limited.  At the exciter's maximum stroke, the acceleration levels fall off proportionally to the square of the frequency as calibration frequency decreases.  This is further complicated by the rising 1/f resistor noise floor of the typical piezoelectric accelerometer at these very low frequencies.  Typical general purpose calibration grade exciters with stokes of a few centimeters are adequate to 5 Hertz, but are often replaced by longer stroke (tens of centimeters or longer) mechanical exciters for frequencies of a few Hertz and below.  Additional new technology has been developed utilizing a patented optical displacement based reference to provide near laser quality resolutions at frequencies as low as 0.25 Hz.

Transverse Sensitivity

Characterizing the transverse sensitivity of a test accelerometer is useful for ensuring minimal axial sensor output to off-axis input acceleration which contributes to measurement noise and uncertainty.  This measurement is accomplished by mounting a sensor under test at the end of flexible rod with lateral exciters built into the rod.  By controlling the phase of the lateral exciters and referencing to an out-of-plane biaxial reference accelerometer pair, a complete polar plot can be generated of the cross-axis sensitivity of the sensor under test.  The cross-axis sensitivity is usually displayed as a percent of the primary axis sensitivity and is usually less than a few percent.

Laser Primary Accelerometer Calibration

While not currently a common option for commercial end users, national metrology labs are most often outfitted with a primary means of accelerometer calibration.  Traditionally this technique utilized an expensive heterodyne laser vibrometer.  New technology has recently come to the market that provides both a more economical solution based on multi-pass and multipoint homodyne lasers.  These allow for integrated spatial averaging of the test sensor mounting platform, thereby reducing the time, cost and complexity of the traditional 3 point primary measurement procedure.

As you can see, the accelerometer calibration world has grown with the times and there are a number of solutions to meet most every need.  Please do not hesitate to contact me if you have any questions about this area.  We are happy to serve you.