An accelerometer calibration experiment is a systematic procedure used to determine and verify the sensitivity, frequency response, linearity, and cross-axis characteristics of accelerometer sensors against traceable reference standards. Proper calibration ensures that acceleration measurements remain accurate and reliable across applications such as vibration monitoring, structural health assessment, automotive safety testing, and inertial navigation. This article provides a comprehensive technical overview of accelerometer calibration samples, testing items, calibration methods, industry applications, and frequently asked questions.
Technical Overview
An accelerometer calibration experiment establishes the quantitative relationship between the electrical output of an accelerometer and the physical acceleration input it experiences. During calibration, the sensor is subjected to precisely controlled and known acceleration inputs, and its output response is recorded and compared against a reference standard, allowing engineers to determine key performance parameters such as sensitivity, frequency response, amplitude linearity, bias stability, and transverse sensitivity.
Calibration experiments are generally classified into two categories: primary (absolute) calibration and secondary (comparison) calibration. Primary calibration methods, such as laser interferometry and the reciprocity method, derive measurement results directly from fundamental physical quantities like wavelength, frequency, and voltage, without requiring a reference accelerometer. Secondary calibration, also known as back-to-back comparison calibration, compares the device under test (DUT) against a calibrated reference transducer mounted on a shared excitation surface.
Several international standards govern accelerometer calibration experiments, including the ISO 16063 series, which defines methods for the calibration of vibration and shock transducers. ISO 16063-11 specifies primary vibration calibration by laser interferometry, ISO 16063-21 covers vibration calibration by comparison to a reference transducer, and ISO 16063-31 addresses the testing of transverse vibration sensitivity. Additional guidance is provided by ISO 16063-13 for primary shock calibration, ISO 16063-22 for shock comparison calibration, and the ISO 5347 series for specific calibration procedures. For inertial-grade sensors, IEEE 1293 provides relevant test procedures, while national metrology standards such as the GB/T 20485 series specify equivalent calibration requirements in China.
Testing Samples
Piezoelectric Accelerometers
Piezoelectric accelerometers utilize quartz or ceramic sensing elements that generate electrical charge proportional to applied acceleration. They are widely used in industrial vibration monitoring due to their wide frequency range, high dynamic range, and robustness. Both charge-output and IEPE (Integrated Electronics Piezo-Electric) voltage-output types require periodic calibration to verify charge sensitivity or voltage sensitivity, typically expressed in pC/g or mV/g.
MEMS Capacitive Accelerometers
Micro-electro-mechanical systems (MEMS) accelerometers measure acceleration through the displacement of a suspended proof mass detected by capacitive sensing structures. These sensors respond to static (DC) acceleration, making them suitable for tilt measurement, inertial navigation, and consumer electronics. Calibration experiments for MEMS devices focus on bias, scale factor, axis alignment, and temperature-dependent drift.
Piezoresistive Accelerometers
Piezoresistive accelerometers employ strain-sensitive resistive elements, usually configured as a Wheatstone bridge, to detect acceleration-induced strain. They offer DC response and are commonly used in shock and crash testing applications. Calibration verifies bridge sensitivity, zero offset, and high-amplitude linearity.
Servo (Force-Balance) Accelerometers
Servo accelerometers use closed-loop electromagnetic feedback to maintain the proof mass at a null position, achieving high precision and excellent low-frequency resolution. These sensors are used in seismology, aerospace, and precision inclination measurement, and require high-accuracy static and low-frequency calibration.
Testing Items
A comprehensive accelerometer calibration experiment typically evaluates the following parameters:
- Sensitivity: The ratio of electrical output (charge in pC or voltage in mV) to applied acceleration, usually expressed at a reference frequency of 160 Hz or 100 Hz per ISO 16063-21.
- Frequency Response: The variation of sensitivity as a function of frequency, establishing usable bandwidth, mounted resonant frequency, and deviation limits over the specified operating range.
- Amplitude Linearity: The deviation of output from proportionality across the acceleration amplitude range, tested by applying increasing excitation levels or shock pulses.
- Transverse Sensitivity (Cross-Axis Sensitivity): The unwanted sensitivity to acceleration perpendicular to the sensing axis, typically required to be less than 5% of main-axis sensitivity and measured per ISO 16063-31.
- Bias Output (Zero Offset): The DC output at zero applied acceleration, a critical parameter for DC-responsive sensors such as MEMS and servo types.
- Zero-Bias Stability and Repeatability: The drift of the zero offset over time and across power cycles, essential for inertial navigation applications.
- Resolution and Noise Floor: The smallest detectable acceleration increment, determined by broadband noise spectral density measurements.
- Temperature Response: The change in sensitivity and bias as a function of ambient temperature, characterized inside environmental chambers.
- Hysteresis and Resonant Frequency: Evaluations of output behavior during cyclic loading and the identification of the mounted resonance, which defines the upper usable frequency limit.
Testing Methods
Static Gravity (Roll-Over) Calibration
This method exploits Earth's gravitational acceleration (9.80665 m/s²) as a reference. The accelerometer is mounted on a precision dividing head or tilt table and rotated through known angles relative to the vertical, producing sinusoidal projection of gravity on the sensing axis. This approach is suitable for DC-responsive sensors and provides absolute traceability, with typical uncertainty below 0.1% when high-precision indexing equipment is used.
Comparison (Back-to-Back) Calibration
Per ISO 16063-21, the DUT is mounted directly on top of a calibrated reference accelerometer on an electrodynamic vibration exciter. Both sensors experience identical motion, and the DUT sensitivity is derived from the ratio of outputs. By sweeping the excitation frequency, the full frequency response can be characterized. This is the most common laboratory method due to its efficiency, wide frequency coverage (typically 5 Hz to 10 kHz), and achievable uncertainties of approximately 1% to 2%.
Laser Interferometry Primary Calibration
As specified in ISO 16063-11, a laser interferometer measures the displacement amplitude of the vibration exciter surface with sub-nanometer resolution using optical wavelength as the reference. Combined with precise frequency measurement, this method yields primary sensitivity values with uncertainties as low as 0.5% or better. fringe-counting, minimum-point, and sine-approximation techniques are employed depending on frequency and amplitude ranges.
Shock Calibration
For high-amplitude dynamic verification, shock excitation is applied using a drop-ball shock machine, air gun, or Hopkinson bar. Primary shock calibration per ISO 16063-13 uses laser interferometry to measure the shock acceleration, while ISO 16063-22 describes comparison shock calibration up to very high acceleration levels. Shock calibration extends linearity verification into ranges that harmonic excitation cannot reach.
Centrifuge Calibration
A precision rotating centrifuge generates constant acceleration proportional to rotational speed squared and radial distance, enabling static calibration at levels from tens to thousands of g. This method is essential for high-range accelerometers used in aerospace and ballistic applications, and careful measurement of the effective mounting radius is critical to accuracy.
Temperature and Environmental Calibration
Sensitivity and bias tests are repeated inside climatic chambers across the rated temperature range to determine temperature coefficients. Combined temperature-vibration systems allow dynamic characterization under simultaneous thermal and mechanical excitation, per relevant sections of the ISO 16063 series and environmental test practices such as those in MIL-STD-810.
Applications
Accelerometer calibration experiments serve numerous industries and technical fields:
- Machine Condition Monitoring: Calibrated accelerometers enable accurate vibration severity assessment according to ISO 10816 and ISO 20816, supporting predictive maintenance of rotating machinery.
- Structural Health Monitoring: Bridges, buildings, and dams are monitored with calibrated sensors to detect modal changes and structural degradation over time.
- Automotive Testing: Crash-test dummies, vehicle dynamics systems, and safety validation programs rely on accelerometers with verified shock linearity and transient response.
- Aerospace and Inertial Navigation: Flight-test instrumentation and inertial measurement units require calibration of bias, scale factor, and axis alignment to meet stringent navigation accuracy budgets.
- Consumer Electronics and IoT: MEMS accelerometers in smartphones, wearables, and drone stabilization systems are calibrated in production for tilt detection, step counting, and orientation sensing.
- Seismology and Geotechnical Engineering: Seismometers and strong-motion sensors require low-frequency calibration and noise characterization for earthquake monitoring.
- Quality Assurance and Metrology Compliance: Accredited third-party laboratories and in-house metrology departments perform calibration experiments to maintain ISO/IEC 17025 conformity and measurement traceability to national standards.
FAQ
How often should an accelerometer be calibrated?
Most manufacturers and quality systems recommend recalibration every 12 months. Sensors exposed to mechanical shock, extreme temperatures, high radiation, or corrosive environments may require shorter intervals, while stable laboratory reference units may justify extended periods based on documented drift history.
What is the difference between static and dynamic accelerometer calibration?
Static calibration applies constant acceleration, typically using Earth's gravity or a centrifuge, and applies only to sensors with DC response such as MEMS and servo types. Dynamic calibration uses sinusoidal vibration or shock excitation and is required for piezoelectric sensors, which cannot measure static acceleration due to their charge-generation mechanism.
What calibration uncertainty can be achieved?
Laser interferometry primary calibration can achieve expanded uncertainties of approximately 0.5% or better, while comparison calibration typically achieves 1% to 2%. Static gravity methods can reach uncertainties below 0.1% for stable DC-responsive sensors. The achievable uncertainty depends on frequency, amplitude, mounting conditions, and instrumentation quality.
Why is transverse sensitivity important?
High transverse sensitivity causes measurement errors when vibration occurs in directions other than the intended sensing axis, which is common in real-world multi-axis vibration environments. Standards such as ISO 16063-31 specify test methods, and most precision accelerometers are designed to keep cross-axis sensitivity below 5% of the main-axis value.
How is measurement traceability ensured?
Traceability is established through an unbroken chain of calibrations linking the working accelerometer to national measurement institutes and ultimately to SI units. Accredited third-party laboratories operating under ISO/IEC 17025 issue calibration certificates documenting reference standards, environmental conditions, measured values, and expanded uncertainties.
Can damaged accelerometers be restored by recalibration?
Recalibration can quantify degradation but cannot repair physical damage. If sensitivity shift, excessive transverse sensitivity, or cracked piezoelectric elements are detected, the sensor should be returned to the manufacturer for repair or replaced, after which a full recalibration is performed before returning the unit to service.