Wavelet signal analysis for the detection of sand excess defects in hardened concrete

Excessive sand in concrete undermines structural integrity by increasing porosity and reducing durability, necessitating non-destructive quality control methods. This study introduces a wavelet-based approach using discrete wavelet transform multi-resolution analysis (DWT-MRA) to detect sand excess defects. Cylindrical specimens (16×32 cm) of reference concrete (350 kg/m³ cement) and sand-excess concrete (30% surplus sand) were analyzed via ultrasonic testing (Pundit PL-200), with signals processed in MATLAB. Results identified distinct patterns: reference concrete showed stable signals (70% in categories C1–C4), while sand-excess specimens exhibited anomalies (60% in C7–C10) linked to interfacial defects and porosity. Signal decomposition revealed longitudinal amplitude spikes and transverse attenuation in defective samples, correlated with wavelet coefficients. Recurring values (-32.640/32.624 vs. 31/47) served as diagnostic markers, validated through redundancy analysis. The method’s computational efficiency, bypassing machine learning, enables real-time defect detection, offering a scalable, non-destructive solution for structural health monitoring and quality assurance in concrete production.