Accelerometer Calibration: Methods, Standards & Intervals

Accelerometer calibration is the process of determining an accelerometer’s sensitivity — the ratio of its electrical output to the mechanical input (acceleration) it measures — across a defined frequency range, against a traceable reference standard. A calibration tells you exactly how much voltage or charge your accelerometer produces per unit of acceleration (expressed as mV/g or pC/g), and whether that sensitivity has shifted from its rated value. An accelerometer with a 5% sensitivity error doesn’t fail visibly — it just reports every vibration measurement 5% wrong.

Accelerometers are used in applications where measurement errors have real consequences: flutter testing on aircraft, structural health monitoring on bridges, crash pulse analysis in automotive testing, and machine condition monitoring in industrial plants. If you’re not familiar with how these sensors work before diving into calibration, our introduction to how accelerometers work provides the foundation.

What Accelerometer Calibration Measures

A complete accelerometer calibration covers several parameters, not just the single sensitivity number printed on the instrument:

Sensitivity

The primary output of calibration is the complex sensitivity — the magnitude and phase of the accelerometer’s output relative to input acceleration, measured at one or more reference frequencies. For most calibrations, the reference frequency is 159.2 Hz (1000 rad/s), chosen because it falls in the flat portion of most accelerometers’ frequency response. Sensitivity is expressed in mV/g (voltage-output sensors) or pC/g (charge-output sensors).

Frequency Response

A frequency response calibration measures sensitivity across a range of frequencies — typically from a few hertz to the instrument’s upper frequency limit. Deviations from flat response are documented as amplitude frequency response errors. For vibration measurements that span a wide frequency range, knowing the full frequency response is essential; a sensor flat at 159 Hz but rising at 2,000 Hz will overreport high-frequency components.

Phase Response

The phase relationship between the input vibration and the sensor’s electrical output matters for applications involving signal timing, mode shape analysis, or multi-channel vibration measurements where channels must stay in phase alignment.

Transverse Sensitivity

An accelerometer is designed to be sensitive along one axis. Transverse sensitivity is the unwanted response to acceleration applied perpendicular to that axis. High-quality accelerometers have transverse sensitivity below 3–5%, but calibration quantifies it so measurements can be corrected when cross-axis vibration is present.

Mounted Resonance Frequency

The resonance frequency determines the upper usable frequency limit. It shifts depending on how the accelerometer is mounted — stud-mounted resonance is higher than adhesive-mounted, which is higher than magnetic-mount. Calibration can verify the resonance frequency under a defined mounting condition.

Standards and Traceability Requirements

Accelerometer calibration is governed by the ISO 16063 series, which defines methods for vibration and shock measurement transducer calibration:

• ISO 16063-1: Basic concepts (vocabulary and methodology framework)
• ISO 16063-11: Primary vibration calibration by laser interferometry
• ISO 16063-12: Primary vibration calibration by the reciprocity method
• ISO 16063-21: Vibration calibration by comparison to a reference transducer
• ISO 16063-31: Testing of transverse vibration sensitivity

ANSI/ASA S2.11-2019 (Methods for the Experimental Determination of Mechanical Mobility) and ANSI/ASA S2.1-1975 (R 2017) also apply in U.S. contexts.

For calibration to satisfy ISO/IEC 17025 requirements — which is the baseline for regulated industries, defense, and aerospace — the calibration lab must maintain reference accelerometers with current NIST-traceable calibration certificates, documented uncertainty budgets for each calibration procedure, and environmental controls appropriate to vibration calibration (typically vibration isolation tables and temperature-controlled labs).

Traceability works the same way it does for any other measurement discipline: the reference accelerometer used in your calibration must have been calibrated by a higher-level lab (often using laser interferometry or against an NMI-calibrated standard), with an unbroken chain of calibration certificates back to NIST. The calibration certificate you receive should identify the reference standard used and its calibration date. For more on how this traceability chain works, see our overview of working standards vs. reference standards.

Calibration Methods: Back-to-Back, Absolute, and Reciprocity

Three fundamentally different methods exist for accelerometer calibration, each defined by ISO 16063. The method used determines the achievable accuracy and what reference equipment is required.

Back-to-Back Comparison Calibration (ISO 16063-21)

This is the most widely used method in commercial calibration labs. The accelerometer being calibrated (UUT) and a calibrated reference accelerometer are mounted back-to-back on a vibration exciter — a precision electrodynamic shaker. Both sensors experience identical motion. The ratio of their output signals, combined with the reference accelerometer’s known sensitivity, gives the UUT’s sensitivity.

The accuracy of this method depends on the reference accelerometer’s calibration uncertainty — typically at the 1–3% expanded uncertainty level for a well-maintained lab. The method is practical, efficient, and traceable through the reference accelerometer’s calibration chain. It’s the right choice for most commercial, industrial, and test and measurement applications.

Laser Interferometry / Absolute Calibration (ISO 16063-11)

Absolute calibration uses laser Doppler interferometry to directly measure the displacement or velocity of the vibrating reference surface, with no reference transducer needed. The accelerometer output is compared directly to the laser-measured motion, providing calibration that is traceable to the SI unit of length rather than to another transducer.

Expanded uncertainty from this method can reach 0.1–0.5% — significantly better than back-to-back comparison. It’s the method used by national metrology institutes (NMIs like NIST) to calibrate the primary reference accelerometers that labs use for back-to-back calibrations. For most calibration needs, absolute calibration is a higher level of service than required; it’s specified when very low measurement uncertainty is needed or when the customer is itself a calibration lab building a reference standard.

Reciprocity Calibration (ISO 16063-12)

Reciprocity is a primary calibration method that uses no reference transducer at all. It derives sensitivity from basic electromechanical reciprocity relationships using three transducers in a defined measurement sequence. This is highly accurate but operationally complex and is mainly used in NMIs and specialized research labs. It’s not offered by most commercial calibration providers.

The chart below summarizes the typical expanded uncertainty range for each method — a useful reference when deciding which calibration tier your application requires. Lower uncertainty means higher accuracy, but also greater cost and complexity.

Calibration Intervals and When to Recalibrate Outside the Schedule

The standard calibration interval for accelerometers is 12 months. This applies to instruments in routine service under normal conditions — vibration measurements on standard industrial equipment, product testing, NVH analysis, and similar work. The 12-month interval is consistent with the general test equipment intervals required by ISO 9001, AS9100D, and most calibration management programs.

Several factors warrant shorter intervals:

• Harsh environments: Accelerometers used in field environments — outdoor structural monitoring, downhole vibration measurement, high-temperature industrial processes — experience greater thermal cycling and mechanical stress than bench instruments. Quarterly or semi-annual calibration may be appropriate.
• High-shock exposure: A piezoelectric accelerometer that has experienced a severe shock event (over 1,000 g, or even a hard drop on a concrete floor) should be recalibrated before returning to service. High-shock events can permanently alter the piezoelectric element’s sensitivity or fracture the crystal.
• Long-term measurements: For continuous monitoring installations where the accelerometer will be deployed for months without access, calibration before installation and verification after removal are both warranted — you need to know the instrument was in calibration at both ends of the deployment window.

Beyond the scheduled interval, recalibrate immediately after:

• Any drop or impact to the sensor body
• Exposure to temperatures outside the rated operating range
• Overrange events that may have saturated or damaged the sensing element
• Cable or connector repair (introduces uncertainty about signal integrity)
• Measurement results that are inconsistent with historical data or physically unexpected

Your calibration certificate should include both the as-found sensitivity (before any adjustment) and the as-left sensitivity (after adjustment, if applicable). The as-found data is what determines whether a nonconformance event is needed — if the as-found sensitivity was outside your program’s acceptance criteria, measurements made since the last good calibration may be affected. This connects directly to the information your calibration certificate must contain to support this kind of assessment.

What to Look for in an Accelerometer Calibration Service

Not every calibration lab is equipped for vibration transducer calibration. The equipment requirements — precision vibration exciters, reference accelerometers with traceable calibration, signal conditioning electronics, and vibration-isolated workspaces — are more specialized than those for general electronic or dimensional calibration. When selecting a provider, verify:

• Accredited scope covers your frequency range and sensor type: The lab’s ISO/IEC 17025 accreditation scope should explicitly list accelerometer calibration by comparison (ISO 16063-21 or equivalent), the frequency range covered, and the sensitivity range. Confirm that your specific accelerometer type (IEPE, charge-mode, MEMS) is within the scope.
• Reference accelerometer calibration is current: Ask for the calibration certificate of the reference accelerometer used. It should show NIST-traceable calibration with an uncertainty that delivers an adequate TUR for your instrument’s tolerance.
• Certificate includes frequency response data: For applications involving measurements across a frequency range, the calibration certificate should report sensitivity at multiple frequencies — not just the single 159.2 Hz reference point. A single-point calibration is adequate for narrow-band applications but insufficient for broadband vibration analysis.
• As-found data is reported: Essential for out-of-tolerance event management and calibration interval optimization.

Micro Precision’s accelerometer calibration services cover a range of sensor types and frequency bands. For questions about your specific instruments, required frequency ranges, or turnaround options, contact our team before sending equipment in.