Glossary term
Ground Vibration Test
Instrumented modal test used to identify aircraft structural frequencies, damping and mode shapes before aeroelastic analysis or flight flutter clearance.
Definition
methodA ground vibration test is an instrumented modal test performed on an aircraft or major structure on the ground to identify natural frequencies, damping ratios and mode shapes.
In aerospace work, a ground vibration test, often abbreviated GVT, provides measured modal data for finite-element model correlation and aeroelastic stability analysis before flight flutter clearance. It uses controlled excitation and distributed sensors to estimate frequency response functions, modal frequencies, damping and mode shapes for the tested configuration. The result is evidence for model updating and envelope expansion, not a standalone proof that flight flutter margin exists.
A ground vibration test is an experimental modal test performed on an aircraft, wing, empennage, control surface, launch vehicle or major structural assembly while it is on the ground. The test identifies natural frequencies, damping ratios and mode shapes so that structural dynamic models can be correlated before aeroelastic analysis or flight flutter clearance.
The core measurement is a frequency response function between a known excitation and a measured response:
where F(\omega) is the excitation force in the frequency domain and X(\omega) is a response such as acceleration, velocity, displacement or strain. Modal curve fitting then estimates modal frequency, damping and mode shape.
Engineering Role
GVT is important because flutter analysis depends strongly on structural frequencies, damping, mode shapes, mass distribution and boundary conditions. A visually detailed finite element model is not enough. The model must reproduce the modes that can couple with aerodynamics, control surfaces, actuators, stores or flight-control laws.
The test also provides configuration evidence. Surface freeplay, actuator stiffness, control-surface mass balance, installed equipment, fuel or ballast state, landing gear support, stores, repairs, sensor mass and boundary conditions can all shift measured modes. If the tested configuration does not match the analysis and flight-test configuration, the GVT evidence may not support the clearance decision.
Test Setup and Data Quality
A useful GVT is a controlled measurement campaign, not only a set of frequency peaks. The setup should define the support condition, aircraft configuration, excitation locations, sensor layout, coordinate system, data-acquisition settings, calibration chain and operating state of actuators and control surfaces.
Frequency response functions should be reviewed for signal quality. Poor force input, overloaded accelerometers, loose sensors, cable restraint, aliasing, poor frequency resolution, low signal-to-noise ratio or nonlinear response can create modes that look precise but are not release-quality evidence.
For measured input and output spectra, magnitude-squared coherence is often used as a data-quality indicator:
A coherence value near one does not prove the model is correct, but low coherence at a target mode is a warning that excitation, noise, nonlinearity or instrumentation should be reviewed before modal parameters are trusted.
Worked Example: Modal Correlation Check
A team compares two important modes from a finite element model with measured GVT results. The release criterion for modes used in flutter analysis is:
- frequency error no more than 5\% in magnitude;
- modal assurance criterion, or MAC, at least 0.90;
- unresolved mismatches must be closed before using the model for envelope expansion.
| Mode | Model frequency | GVT frequency | GVT damping | MAC |
|---|---|---|---|---|
| Wing first bending | 8.6\ \text{Hz} | 8.2\ \text{Hz} | 2.1\% | 0.93 |
| Wing torsion | 14.8\ \text{Hz} | 15.6\ \text{Hz} | 1.4\% | 0.88 |
Frequency error is computed as:
For the first bending mode:
The frequency error is within the 5\% limit and the MAC value is:
The first bending mode passes this simplified correlation screen.
For the torsion mode:
The frequency error exceeds the 5\% limit. The MAC value also fails:
For two real mode-shape vectors, MAC is:
If mass is judged accurate and the torsion frequency mismatch is mainly stiffness-related, a first estimate of the stiffness scale factor is:
This suggests the model is about 11.1\% low in the effective stiffness for that torsion mode, under the simplified assumption that modal mass is correct.
Engineering comment: this does not mean the engineer should blindly multiply all stiffness by 1.111. The mismatch may come from boundary conditions, joint stiffness, control-surface restraint, sensor mass, fuel or ballast representation, actuator stiffness, mode-shape pairing or mesh modeling. The result is a diagnostic lead. The torsion mode should be reviewed and the model updated before the flutter analysis relies on it.
Distinction from Related Terms
Ground vibration test is not the same as modal analysis. Modal analysis is the general study of modes; GVT is a specific aircraft or structure test campaign with instrumentation, excitation, configuration control and release documentation.
Ground vibration test is not flutter speed. GVT provides modal evidence used in flutter prediction and clearance. Flutter speed is the aeroelastic boundary inferred from analysis, test data and margins.
Ground vibration test is not a wind-tunnel test. A GVT measures structural dynamics on the ground. A wind-tunnel test measures aerodynamic behavior or aeroelastic response under controlled flow.
Ground vibration test is not finite element analysis. It is measured evidence used to check and update finite element models. A model that does not correlate with GVT should not be used for high-consequence aeroelastic release without explanation.
Ground vibration test is not flight flutter testing. Flight testing observes the aircraft in the real flight environment after ground evidence and analysis justify controlled envelope expansion.
Validation and Common Mistakes
A defensible GVT record states the configuration, mass properties, fuel or ballast state, control-surface restraint, actuator state, support condition, excitation method, sensor layout, calibration, sampling rate, anti-alias filtering, frequency range, input levels, signal quality, curve-fit method, mode pairing, MAC values, damping estimates, uncertainty and model-correlation decisions.
The correlation decision should say what changed after the test. Possible outcomes include accepting the model for a limited envelope, updating stiffness or mass properties, re-testing a poorly identified mode, narrowing flight-test points, changing abort criteria or blocking envelope expansion until configuration differences are resolved. A GVT that identifies modes but does not connect them to model updates, uncertainty and release limits is incomplete.
Common mistakes include:
- testing a configuration that does not match the aeroelastic model or flight-test aircraft;
- using too few sensors to distinguish similar mode shapes;
- quoting frequency without damping, mode shape, boundary condition and uncertainty;
- exciting too narrow a frequency range and missing a relevant mode;
- accepting a frequency match while ignoring a poor MAC value;
- treating measured damping as a universal value for all amplitudes and configurations;
- ignoring sensor mass, cable restraint, landing gear support, control-surface freeplay or actuator boundary conditions;
- updating a finite element model until it matches one mode while damaging the physical meaning of other modes.