Step 2: Import Test Data

This step loads measured Frequency Response Function (FRF) data into the modal analysis workspace. V-Listen supports importing FRFs computed from raw time-domain recordings as well as pre-computed FRF files in standard formats.

Import Methods

HDF Import (HEAD Recordings)

Import time-domain recordings in HEAD .hdf format. V-Listen computes FRFs from the raw signals using its native FFT engine with configurable windowing and averaging parameters.

Mode	Description	Use Case
Single File	Load one HDF recording containing all channels.	Multi-channel simultaneous acquisition
Multi-File	Select multiple HDF files, each representing one measurement run (roving hammer or roving sensor).	Sequential roving measurements
Folder	Load all HDF files from a selected folder. Files are sorted by name and merged automatically.	Batch import of systematic test campaigns

Click Load HDF in the toolbar.
Choose import mode: Single File, Multi-File, or Folder.
Select the file(s) or folder.
Configure FRF computation parameters (see table below).
Click Compute. FRFs appear in the data preview panel.

FRF Import (Pre-Computed)

Import FRF data that has already been computed in external software. This bypasses the time-to-frequency conversion step.

Format	Extension	Description
UFF Type 58	`.uff / .unv`	Universal File Format function-at-a-node dataset. Widely supported by all major modal analysis packages.
CSV	`.csv`	Comma-separated frequency-amplitude-phase columns. First column is frequency in Hz.
HEAD HDF Transfer Function	`.hdf`	Pre-computed transfer functions stored in HEAD HDF format (result datasets).

Click Import FRF in the toolbar.
Select the FRF file(s).
V-Listen auto-detects the format and parses channel metadata.
Review the imported FRFs in the data preview panel.

Tip: UFF Type 58 files may contain multiple FRFs in a single file. V-Listen reads all datasets and lists them individually in the channel table.

FRF Computation Parameters

When computing FRFs from raw HDF time data, the following parameters are available:

Parameter	Default	Range	Description
FFT Size	4096	256 – 65536	Number of samples per FFT block. Larger values give finer frequency resolution.
Window	Hanning	Hanning, Flat Top, Rectangular, Exponential, Force	Window function applied to each block.
Overlap	50%	0% – 90%	Overlap between consecutive FFT blocks.
Averaging	Linear	Linear, Exponential, Peak Hold	Method for combining multiple FFT blocks.
FRF Estimator	H1	H1, H2, Hv	H1 minimizes output noise; H2 minimizes input noise; Hv is an optimal estimator.

Important: For impact hammer tests, use the Force window on the input channel (exponential decay on the response channel) to avoid leakage errors. The H1 estimator is recommended for hammer tests where input noise is minimal.

Channel-DOF Mapping

Each FRF channel must be mapped to a geometry point and direction (DOF) for mode shape visualization. The mapping table has the following columns:

Column	Description	Example
Channel	Channel name from the imported data file.	"Ch 1 - Accel X"
Point ID	Geometry point number this channel corresponds to.	12
Direction	Measurement direction: +X, -X, +Y, -Y, +Z, -Z.	+Z
Type	Response or Reference.	Response
Status	Mapped (green) or Unmapped (red).	Mapped

Auto-Mapping

Click Auto-Map to let V-Listen match channels to geometry points by parsing channel names for point IDs and direction indicators. The auto-mapper recognizes common naming conventions:

P12:+Z or 12Z — Point 12, Z direction
Acc_FrontLeft_X — Matched by label if labels are set in geometry

Manual Mapping

For channels that cannot be auto-mapped, select the channel row, then choose the Point ID and Direction from the dropdown menus.

Reference Channel Selection

At least one reference channel (excitation point/direction) must be designated. For MIMO (Multiple-Input Multiple-Output) tests with multiple shakers, select all reference channels.

In the channel table, set the Type column to Reference for each excitation channel.
The number of references determines the number of FRF columns available for curve fitting.
Single-reference (SISO/SIMO) tests are the most common setup for impact hammer testing.

Tip: For roving hammer tests, each measurement run has the same response point but different excitation points. V-Listen handles this automatically when using Multi-File or Folder import — the reference DOF is updated per file.

Data Preview

The data preview panel (powered by ScottPlot) displays the imported FRFs interactively:

Magnitude / Phase / Real / Imaginary: Toggle display mode using the tabs above the plot.
Linear / Log scale: Switch Y-axis scaling via the toolbar.
Overlay / Waterfall: View all FRFs overlaid or as a stacked waterfall.
Zoom: Scroll to zoom, drag to pan. Double-click to reset the view.
Cursor: Move the mouse over the plot to see frequency and amplitude readout in the status bar.

FRF Quality Check

Before proceeding to modal analysis, verify the quality of your FRF data:

Check	What to Look For	Action if Failed
Coherence	Values above 0.9 at resonance frequencies. Drops indicate noise or nonlinearity.	Increase averaging or check sensor mounting.
Reciprocity	FRF(i,j) should equal FRF(j,i) for linear structures.	Check sensor calibration and coupling.
Driving Point	Driving-point FRF (response and excitation at same DOF) must show alternating resonance/anti-resonance pattern.	Verify sensor placement and force input quality.
Noise Floor	Anti-resonance valleys should be well above the noise floor.	Increase excitation level or number of averages.
Overload	No clipping indicators in time-domain signals.	Reduce excitation level or increase sensor range.

Important: Poor-quality FRFs will produce unreliable modal parameters regardless of the algorithm used. Always inspect the coherence function and driving-point FRFs before proceeding.

Troubleshooting Common Issues

Issue	Possible Cause	Solution
FRF magnitude is unrealistically high or low	Calibration factors not applied or incorrect units	Verify sensor sensitivity values and ensure consistent units (g vs m/s2).
All FRFs show the same shape	Channels not correctly mapped or data file contains duplicates	Check channel assignments in the source recording software.
Phase is random / noisy at all frequencies	Reference channel not connected or wrong channel selected	Verify reference channel wiring and selection.
No anti-resonances in driving-point FRF	Sensor and hammer not at same DOF, or wrong FRF type	Confirm that the driving-point DOF matches in both point ID and direction.
Import fails with format error	File format mismatch or corrupted file	Verify file extension matches actual content. Re-export from source if needed.