Data for the publication "Weak adhesion detection – Enhancing the analysis of vibroacoustic modulation by machine learning."

The data is organized as a ZIP Archive split in 5GB parts. To use the Data, all files have to be downloaded and extracted together. The Archive contains 4 folders.
1. Dynamic-Test-Data
2. Modulation_measurements
3. Pictures
4. Tensile test

1. Dynamic-Test-Data
This folder contains the vibrational data from the hydraulic testing machine. It is not shown in the publication but was used to validate the correctness of the measurements.

### 2. Modulation_measurements
This folder contains the vibroacoustic measurements for all samples. There are subfolders in which the data files of the different combinations between the piezoceramics have been saved as '.mat'.
Each File contains four arrays.
+ freq ... Frequencies used for the high-frequency excitation 201000 Hz to 220000 Hz in steps of 500 Hz (Low frequency is 5Hz due to the limitations of the pulsing machine.)
+ VAM ... Measurement signal, from which the modulation is calculated. It contains 2s with a sampling rate of 2e6 MSa for each high frequency.
+ Chirp_mod ... Contains the measurement of a linear chirp ranging from 1Hz to 300kHz in 5 seconds. Here the Frequency-Response-Spectrum can be evaluated. The sampling rate was set to 1e6 to save data.
+ pulser .... 2 seconds of measurement with a sampling rate of 1e6 MSa where just the hydraulic testing machine is running. To evaluate if there are any differences in the machine. This data was not.
used for the publication.


### 3. Pictures
This folder contains pictures of all samples in different stages.
+ C-Scans of the bonds
+ Fracture Surfaces after the tensile test
+ Trough - light photos of the bonded plates
+ Through-light images of each cut-out specimen

### 4. Tensile Test
This folder contains the tensile-test data for every sample measured at a Zwick Z100. Exemplary pictures of the testing are in the pictures folder.

Questions can be directed to benjamin.boll@tuhh.de or robert.meissner@tuhh.de

Measurements are saved in the Matlab ".mat" format but can also be extracted with python.