Acoustic emission source localization in plates using fiber Bragg grating sensors

Abstract

Acoustic emission (AE) source localization serves as a core technique for structural health monitoring, yet conventional approaches suffer from limitations including reliance on multiple sensors, demand for pre-measured wave velocity, and vulnerability to environmental noise. This paper presents a novel AE localization methodology that integrates the intrinsic directional selectivity of fiber Bragg grating (FBG) sensors with time reversal (TR) processing to enable reliable two-dimensional source localization using only two sensors without any pre-measured wave velocity. The proposed method leverages the cos²θ directional response of FBG sensors, constructs a directional weighting map by matching theoretical directional selectivity with TR-processed signal amplitude ratios. This integration effectively resolves the inherent ambiguity of conventional two-sensor TR localization, converting directional selectivity from a sensing limitation into angular localization information. To enhance robustness in noisy environments, an adaptive line enhancer based on the Hurst exponent is introduced to recover signal focusing capability for large incidence angles where signal-to-noise ratio degradation typically compromises TR performance. Experimental validation on a 1000 × 1000 mm aluminum plate using pencil-lead-break sources demonstrates that the method achieves an average localization error of approximately 25 mm across various source positions, with consistent repeatability. Parametric optimization reveals that a rectangular TR window with length of three times the signal rise time maximizes energy convergence. This study demonstrates a sensor-efficient AE source localization method for plate-like structures under controlled laboratory conditions, with potential for extension to applications in aerospace, civil infrastructure, and mechanical systems where sensor placement is constrained.