AI-Driven Evaluation of 3D-Printed Concrete (2025-11)¶

3D-Printed Concrete (3DPC) presents structural challenges due to inter-filament voids and weak interlayer bonds. The Impact-Echo (IE) method analyzes elastic wave reflections to detect internal defects, but conventional FFT analysis has difficulty detecting near-surface voids when sampling rates are low. This study incorporates Acoustic Emission (AE) to increase the sampling rate and enhance waveform resolution, enabling the capture of complex P-wave reflections. A Deep Learning (DL) approach classifies defects from waveform and frequency-domain data, enabling the recognition of patterns in 3DPC. The proposed method allows for automated and accurate structural assessment.