Inertial device input axis testing is a critical characterization process that verifies the alignment accuracy, scale factor, bias behavior, and cross-axis sensitivity of gyroscopes, accelerometers, and inertial measurement units. By applying precisely controlled angular rate or linear acceleration stimuli along defined reference axes, accredited third-party laboratories and manufacturer qualification labs can quantify input axis misalignment and identify error model coefficients that directly determine navigation and guidance accuracy.
Technical Overview
The input axis is the fundamental sensitivity direction of any inertial device. For a gyroscope, it is the axis about which an applied angular rate produces the maximum rated output; for an accelerometer, it is the axis along which an applied specific force (linear acceleration) produces the maximum rated output. In an ideal sensor, the input axis coincides exactly with the input reference axis (IRA) defined by the mechanical datum features of the sensor or its mounting fixture. In practice, manufacturing tolerances, assembly stress, and thermal deformation cause a small angular deviation, known as input axis misalignment, which is a primary systematic error source in inertial navigation.
Input axis testing evaluates how faithfully an inertial device converts stimuli applied along its nominal input axis into output, and how much undesired sensitivity exists to stimuli applied along orthogonal axes. The measured parameters feed directly into the sensor error model, in which the output is expressed as the sum of scale factor multiplied by the true input, misalignment-induced coupling terms, bias, and noise. Because even fractions of a milliradian of misalignment can accumulate into kilometer-level position drift over long missions, input axis verification is mandatory for tactical-grade, navigation-grade, and strategic-grade inertial products, and is routinely performed according to internationally recognized test procedures issued by standards organizations such as ISO, IEC, and national metrology frameworks consistent with the GB standard series.
Testing Samples
Gyroscopes
Rate and rate-integrating gyroscopes of all major sensing technologies are submitted for input axis characterization, including MEMS vibratory gyroscopes, fiber optic gyroscopes (FOG), ring laser gyroscopes (RLG), hemispherical resonator gyroscopes (HRG), and dynamically tuned gyroscopes. Samples may be bare die-level components, hermetically sealed packages, or line-replaceable units with integrated electronics.
Accelerometers
Linear accelerometers tested for input axis performance include quartz flexure (pendulous) accelerometers used in navigation-grade systems, MEMS capacitive and piezoresistive accelerometers, and force-rebalance servo accelerometers. Package-level and module-level units are both acceptable test articles provided they define a repeatable mounting reference.
Inertial Measurement Units and Assemblies
Multi-axis IMUs containing three orthogonally mounted gyroscopes and three accelerometers are tested to evaluate not only the input axis misalignment of each individual sensor but also the orthogonality of the assembled sensor triad. Inclinometers, attitude and heading reference systems (AHRS), and inertial navigation systems (INS) may also undergo input axis verification at the system level.
Testing Items
Input axis test programs are tailored to the device grade and application, and typically cover the following measurable characteristics:
- Input axis misalignment angle relative to the mounting datum, reported for each sensitive axis
- Axis-to-axis orthogonality and skew errors within a multi-axis sensor cluster
- Scale factor magnitude, nonlinearity, asymmetry (difference between positive and negative inputs), and repeatability across power cycles and temperature points
- Bias (offset), bias repeatability between runs, in-run bias stability, and bias temperature sensitivity
- Cross-axis sensitivity (transverse sensitivity) describing the output response to stimuli applied perpendicular to the input axis
- Input axis rectification coefficients arising from nonlinear response to combined inputs
- Threshold, resolution, and dynamic bandwidth along the input axis
- Angular random walk (ARW) for gyroscopes and velocity random walk (VRW) for accelerometers, derived from noise analysis on the input axis channel
Testing Methods
Precision Multi-Position Static Tests
The device under test is mounted on a precision indexing head or a two-axis position rate table with arc-second-level angular accuracy. By sequentially orienting the sensitive axis to a sequence of defined positions relative to the local gravity vector (commonly 6-, 12-, or 24-position schedules), the accelerometer bias, scale factor, and misalignment coefficients are separated through least-squares estimation of the error model. For gyroscopes, Earth rate projection along the input axis at multiple azimuths provides a natural, highly stable reference input.
Rate Table Dynamic Tests
A precision rate table applies controlled constant angular rates over a range such as ±0.01 to ±1000 degrees per second about the device input axis. Sweeping positive and negative rate points yields the scale factor polynomial, nonlinearity, and asymmetry. Dedicated face-to-face and side-mounted orientations expose the cross-axis sensitivity by applying rate orthogonal to the nominal input axis, allowing misalignment coefficients to be solved from the observed coupling response. Temperature-controlled chambers integrated with the rate table enable scale factor and misalignment characterization across the full operating temperature range.
Centrifuge Acceleration Tests
For accelerometers requiring high-g stimulation, a precision centrifuge applies steady accelerations along the input axis as well as along deliberately introduced off-axis components. This extends the input axis characterization to the full dynamic range and identifies input axis rectification errors that only appear at elevated g-levels.
Data Processing and Error Model Identification
Raw output data are temperature compensated, de-trended, and fitted to a comprehensive linear or nonlinear error model using regression techniques. Allan variance analysis quantifies noise terms, while residual analysis validates the adequacy of the identified model. The final test report presents misalignment angles, error coefficients, their uncertainties, and pass or fail conclusions against the applicable specification limits.
Applications
Input axis test results underpin the calibration and quality release of inertial hardware across many demanding sectors:
- Aerospace and launch vehicle guidance, where misalignment compensation directly reduces attitude and position error growth
- Tactical and strategic missile guidance and control subsystems
- Marine and submarine inertial navigation requiring long-endurance drift performance
- Autonomous vehicles, drones, and unmanned systems relying on MEMS IMU calibration
- Oil and gas directional drilling (measurement while drilling) tools operating under shock and vibration
- Servo stabilization platforms, industrial robotics, mobile mapping, and geodetic surveying equipment
FAQ
What exactly is input axis misalignment?
It is the small angular deviation between the actual sensitivity axis of the sensor and its nominal input reference axis defined by the mounting interface. It causes a portion of the stimulus on orthogonal axes to leak into the measured channel, producing systematic navigation error.
Why is input axis testing performed instead of relying on manufacturing tolerances?
Machining, bonding, and thermal expansion make as-built alignment impossible to guarantee purely by tolerances. Post-manufacturing input axis testing measures the true misalignment so that it can either be compensated in software during calibration or rejected if it exceeds specification.
What equipment is required?
Essential hardware includes a precision rate table or indexing head with arc-second angular accuracy, a temperature chamber, a vibration-isolated foundation, calibrated data acquisition electronics, and for accelerometers, optionally a precision centrifuge. Traceability to national measurement standards is maintained through calibrated reference instruments.
How are misalignment results expressed?
Misalignment is typically reported in milliradians, arc seconds, or microradians for each axis pair, together with the full error model coefficients, their estimated uncertainties, and correlation information needed for downstream Kalman filter based compensation.
Which accuracy classes can be evaluated?
Consumer-grade MEMS devices through navigation-grade and strategic-grade inertial instruments can all be characterized; the required test equipment precision and environmental control scale with the device grade being verified.
Are accredited laboratories able to conduct these tests?
Yes. Accredited third-party laboratories and qualified manufacturer facilities perform input axis testing under controlled environmental conditions in accordance with recognized international and national test methodologies, issuing traceable reports suitable for design verification, qualification, and lot acceptance.