Enhancing Structural Health Monitoring in Additive Manufacturing Through Embedded Sensors, Infill Designs and Deep Learning

Abstract

Additive manufacturing (AM) enables the integration of sensors into complex structures. Embedding sensors within the structure reduces cost, allows precise placement, and protects the sensors from environmental exposure. This study investigates the performance of embedded piezoelectric transducers (PZTs) within polymer plates fabricated via AM, using varied infill patterns to influence surface wave behavior. The Surface Response to Excitation (SuRE) method was employed to excite the structure using multiple pulse width excitation (MPWE) and to monitor the resulting surface wave propagation. Signals captured from the embedded sensors were processed using the Short-Time Fourier Transform (STFT) to generate time-frequency spectrograms, which were then classified using Convolutional Neural Networks (CNNs). This approach enabled accurate estimation of both the location and magnitude of applied loads, achieving classification accuracy above 90%. The results demonstrate the effectiveness of combining embedded sensing, infill-based wave manipulation, and deep learning for structural health monitoring. This method shows strong potential for applications in biomedical, aerospace, and mechanical engineering, particularly where polymer components are used in critical functions.