Advanced Techniques for Accurate Crack Length Monitoring in BI-Material Adhesively Bonded Joints

Abstract

Driven by sustainability goals outlined in the European Green Deal, most of the industrial sectors (i.e. automotive, aerospace and civil infrastructures) require reliable, lightweight, and durable materials. Accurate crack detection significantly extends the operational life of bonded structural components, reducing maintenance, waste, and environmental impact. This study presents acoustic emission (AE) techniques for accurately monitoring crack length in adhesively bonded joints, primarily targeting Titanium-Carbon Fiber Reinforced Polymer (Ti-CFRP) bi-material specimens, with Titanium-Titanium (Ti-Ti) joints included as a benchmark. Titanium Ti6Al4V substrates fabricated via Laser Powder Bed Fusion (LPBF) were prepared with various surface conditions: as-printed and sandblasted. The mode I fracture toughness was evaluated via Double Cantilever Beam tests, which were supported by continuous AE monitoring with high-resolution equipment capturing around 200,000 waveforms. Principal Component Analysis and machine learning techniques, including Self-Organising Maps and K-means clustering, classified AE signals into clusters associated with damage or background noise. A linear localisation algorithm tracked crack initiation and growth phases. Results validated the accuracy of AE signals to localise crack propagation under the bi-material quasi-static mode I load condition. The study highlights AE's potential for precise and sustainable structural health monitoring, informing future numerical modelling to predict joint durability.