Experimental Modal Analysis (EMA) Overview

What Is Experimental Modal Analysis?

Experimental Modal Analysis (EMA) is the process of determining the dynamic characteristics of a structure — its natural frequencies, damping ratios, and mode shapes — from measured vibration data. Unlike analytical methods (e.g., finite element analysis), EMA is based on physical measurements acquired under controlled excitation conditions.

EMA relies on measuring Frequency Response Functions (FRFs) that relate a known input force to the resulting structural response. By curve-fitting these FRFs with mathematical models, the modal parameters of the structure are extracted.

When to Use EMA

EMA is the preferred approach when:

You can apply controlled excitation (impact hammer, shaker, stepped-sine).
Both input (force) and output (acceleration, velocity, displacement) signals are available.
You need accurate damping estimates and scaled mode shapes.
You want to validate or update a finite element model.
You need to perform structural modification predictions.

Tip: If you cannot instrument the structure with an excitation source (e.g., a bridge under traffic, a building under wind loading), consider Operational Modal Analysis (OMA) instead.

EMA vs OMA Comparison

Criterion	EMA	OMA
Excitation	Controlled (hammer, shaker)	Ambient / operational forces
Input measurement	Required (force transducer)	Not required
Mode shape scaling	Absolute (mass-normalized)	Relative only (unscaled)
Damping accuracy	High	May include operational damping
Lab testing	Yes	Typically field testing
Algorithms	PolyProMax, LSCF, LSCE	OPolyProMax, SSI-COV, EFDD

EMA Workflow in V-Listen

The modal analysis module follows a four-step workflow, guided by the step indicator at the top of the workspace:

Step 1: Import Geometry

Load or create a structural geometry model that defines measurement point locations and connectivity. Supported formats include HMPL, STL, OBJ, and UFF. This geometry serves as the basis for mode shape visualization.

Step 2: Import Test Data

Load measured FRF data from HDF recordings or pre-computed FRF files. Map measurement channels to geometry DOFs and select reference channels. Verify data quality before proceeding to curve fitting.

Step 3: Modal Analysis

Configure the analysis algorithm (PolyProMax, LSCF, or LSCE), set the frequency range and model order, and run the curve-fitting computation. The stabilization diagram is the primary tool for identifying physical poles.

Step 4: Validation

Animate mode shapes on the 3D geometry, compute the MAC matrix to check mode orthogonality, and compare measured FRFs against synthesized FRFs. Export results for reporting or further analysis.

Key Concepts

Frequency Response Function (FRF)

An FRF describes the input-output relationship of a linear system in the frequency domain. It is typically expressed in one of three forms:

FRF Type	Response Quantity	Unit (SI)	Common Use
Receptance (Compliance)	Displacement / Force	m/N	Low-frequency structural analysis
Mobility	Velocity / Force	(m/s)/N	Vibro-acoustic applications, power flow
Accelerance (Inertance)	Acceleration / Force	(m/s²)/N	Most common in structural dynamics testing

V-Listen supports all three FRF types and can convert between them using frequency-domain integration and differentiation.

Modal Parameters

Parameter	Symbol	Description
Natural Frequency	f_n	The frequency at which the structure naturally vibrates in a given mode.
Damping Ratio	ζ	The fraction of critical damping. Typical values for mechanical structures range from 0.1% to 5%.
Mode Shape	φ	The spatial deformation pattern associated with each natural frequency.
Modal Participation Factor	L	Quantifies how strongly each mode participates in the overall response.

Quality Indicators

MPC (Modal Phase Collinearity): Ranges from 0 to 1. Values above 0.9 indicate a well-separated real normal mode.
MAC (Modal Assurance Criterion): Cross-correlation between mode shapes. Diagonal values near 1.0 and off-diagonal values below 0.1 indicate well-separated modes.
Quality Score: A composite metric combining MPC, damping consistency, and pole stability across model orders.

Important: Always verify your results using at least two independent methods — for example, MAC matrix inspection and measured-vs-synthesized FRF comparison. A single quality metric alone is not sufficient to validate modal parameters.

Excitation Methods

The choice of excitation method significantly affects the quality and applicability of EMA results:

Method	Advantages	Limitations
Impact Hammer	Fast setup, broadband excitation, no mass loading, portable.	Limited energy input, operator-dependent repeatability, risk of double hits.
Electrodynamic Shaker	Repeatable, controlled amplitude, supports various signal types (random, sine, burst).	Requires stinger attachment, adds mass to the structure, more complex setup.
Stepped Sine	Highest signal-to-noise ratio, precise frequency control.	Very slow (one frequency at a time), impractical for broadband surveys.
Random Burst	No leakage errors, good for nonlinear structures.	Requires shaker, moderate signal-to-noise ratio.

Tip: For most industrial applications, impact hammer testing is the fastest and most practical approach. Use shaker excitation when high repeatability is required or when the structure is too large for adequate hammer excitation energy.

Measurement Configurations

Configuration	Abbreviation	Description
Single-Input Single-Output	SISO	One excitation point, one response point. Simplest setup; requires roving to cover all DOFs.
Single-Input Multiple-Output	SIMO	One excitation point, multiple simultaneous response channels. Common for roving hammer tests.
Multiple-Input Multiple-Output	MIMO	Multiple shakers, multiple response channels. Required for closely spaced modes and repeated roots.

Typical Applications

Automotive body-in-white (BIW) modal characterization
Aerospace component qualification and flutter boundary prediction
Rotating machinery diagnostic modal surveys
Civil engineering structure health monitoring baselines
Consumer product vibration design optimization
FE model correlation and updating
Structural modification prediction (what-if scenarios)
Acoustic cavity modal analysis (interior noise)

Prerequisites and Preparation

Before starting an EMA campaign, ensure the following prerequisites are met:

Boundary conditions: Define how the structure will be supported during the test. Free-free conditions (structure suspended on soft springs or bungee cords) are preferred for correlation with FE models that use free boundary conditions. Fixed conditions require rigid mounting fixtures.
Sensor selection: Choose accelerometers with appropriate sensitivity, frequency range, and mass. The total added mass from all sensors should not exceed 10% of the structure's mass to avoid significant mass-loading effects.
Measurement grid: Plan measurement point locations to adequately capture all expected mode shapes. The spatial Nyquist criterion requires at least two points per half-wavelength of the highest-order mode of interest.
Excitation point selection: Choose excitation locations that do not coincide with nodal points of any mode of interest. Corner points or off-center locations generally provide good excitation coverage.