Audio signal data

This dataset comprises acoustic signals from a standardized tapping experiment designed for the quantitative identification of voids in concrete walls. It was collected to support research on transforming subjective acoustic inspection into an objective, data-driven process. The data acquisition involved using a rubber-wrapped hammer for consistent excitation and a high-fidelity recorder (sampling at 16 kHz) positioned at a fixed distance to capture sound pressure. The dataset includes two primary classes of samples: signals from structurally intact, normal walls and signals from walls with precisely controlled, preset voids. Provided in their original WAV format, the signals capture the distinct temporal characteristics of each condition. This dataset enables the analysis of time-domain waveforms and the calculation of autocorrelation functions (ACF), revealing key differences in signal decay rates and the presence of side-lobe peaks. It serves as a valuable resource for developing, training, and validating signal processing algorithms for non-destructive evaluation (NDE) and structural health monitoring (SHM) in civil engineering applications.