Acoustic Monitoring Dataset for Robotic Laser Directed Energy Deposition (LDED) of Maraging Steel C300

<p>This dataset presents a set of acoustic signals captured during a single-bead wall experiment in robotic Laser Directed Energy Deposition (LDED) using Maraging Steel C300. The acoustic data was recorded using a high-fidelity Prepolarized microphone sensor (Xiris WeldMIC), capturing the intricate sound profiles associated with the LDED process at a sampling rate of 44,100 Hz.</p><p><strong>Laser Directed Energy Deposition:</strong></p><p>This dataset was generated with a robotic LDED process that consists of a six-axis industrial robot (KUKA KR90) coupled with a two-axis positioner, a laser head, and a coaxial powder-feeding nozzle.</p><p>&nbsp;</p><p><strong>Folder Structure:</strong></p><ul><li><strong>/sample-1</strong>: The main folder for the experiment sample.<ul><li><strong>/audio_files</strong>: Contains 4624 <strong>.wav</strong> audio files, each representing a 40 ms chunk of the LDED process sound.</li><li><strong>/annotations_1.csv</strong>: A CSV file providing annotations for the audio files, labeling each as "Defect-free", "Defective", or "Laser-off".</li></ul></li><li>audio_features.h5: extracted acoustic features in time-domain, frequency-domain, and time-frequency representations (MFCC features). Feature extraction was conducted using Python Essentia Library.</li></ul><p>&nbsp;</p><p><strong>File Naming Convention:</strong></p><ul><li>Audio files within the <strong>audio_files</strong> folder are named following the pattern <strong>sample_ExperimentID_SampleID.wav</strong>. Given that there's only one experiment and one sample, the naming will be consistent, for example, <strong>sample_1_1.wav</strong> for the first file.</li></ul><p><strong>Annotation Details:</strong></p><ul><li>The <strong>annotations_1.csv</strong> file contains detailed labels for each audio file, correlating to the conditions observed during the experiment, aiding in quick identification and analysis.</li></ul><p><strong>Experimental Parameters:</strong> The dataset reflects a controlled experiment setup with the following specifications:</p><ul><li>Geometry: Single bead wall structure</li><li>Dimensions: 90 mm * 42.5 mm</li><li>Number of layers: 50</li><li>Laser beam diameter: 2 mm</li><li>Layer thickness: 0.85 mm</li><li>Stand-off distance: 12 mm</li><li>Laser profile: Gaussian</li><li>Laser wavelength: 1064 nm</li></ul><p><strong>Process Parameters:</strong></p><ul><li>Laser power: 2.3 kW</li><li>Speed: 25 mm/s</li><li>Dwell time: 0 s</li><li>Powder flow rate: 12 g/min</li></ul><p>This dataset aims to facilitate the development and testing of acoustic-based defect detection models for real-time quality monitoring in LDED processes. It can also serve as a reference point for further research on sensor fusion, machine learning, and real-time monitoring of manufacturing processes.</p>