Information fusion-based maintenance strategies selection for coastal concrete bridges using recycled fishing nets

Abstract

Repurposing recycled materials for coastal concrete bridge maintenance addresses both engineering requirements and marine ecosystem preservation. This study explored the application of recycled nylon (RN) and polyethylene (PE) from discarded fishing nets in coastal concrete beam maintenance. Specifically, these fibers were incorporated into the polymer cement mortar (PCM) to produce two repair materials. Then, thirteen unrepaired or preventive/post-repaired concrete beams were exposed to the tidal zone for one or two years, followed by beam bending tests and the establishment of the corresponding parametric Finite element modeling (FEM). As an alternative to traditional heavy-parameter models, a novel lightweight crack segmentation model (LiteSqueezeSeg) was proposed for quantitatively analyzing beam surface crack properties. Lastly, experimental, simulated, and quantitative crack information were fused to comprehensively assess the beam performance. Results showed that both PCM-RN and PCM-PE effectively inhibit chloride ion penetration. The proposed LiteSqueezeSeg model (95.66% accuracy) achieved nearly identical performance to Deeplab v3 + but only with about 1/6 parameters. Fused results revealed that PCM-PE repaired beams exhibited superior mechanical performance but inferior ductility compared to PCM-RN. Moreover, both PCM-PE and PCM-RN repaired beams demonstrated that preventive repairs achieved more balanced performance than post repairs. Consequently, utilizing PCM-RN for preventive repair emerged as an optimal candidate. This study significantly contributes to the utilization of waste fishing nets in coastal concrete structure maintenance.