Tracing fracture damage evolution and identifying damage patterns in cast steel using advanced acoustic emission analysis

Abstract

The application of advanced acoustic emission signal analysis techniques can facilitate the identification of fracture damage modes in ductile materials and enable the provision of fracture precursor warnings. Given the vast array of acoustic emission signal characteristic parameters, there must be a systematic and comprehensive basis for selection and a deep comparison among different fracture warning indicators. This study uses G20Mn5QT cast steel material fracture acoustic emission monitoring tests under different in-plane constraints as an example. We adopted principal component analysis and k-means++ to select acoustic emission feature parameters and cluster damage modes. Additionally, we conducted a comparative analysis of several fracture precursor warning indicators and influencing factors, including fractal dimensions, b/Ib values, and natural time analysis. Our findings show that shear and tensile fracture damage modes are present in type-I fracture tests when applying the k-means++ cluster to the rise angle-average frequency-energy parameter, but tensile fracture modes dominate. Fractal dimensions, b-values, and Ib-values show a strong trend correlation with the fracture damage evolution process. Natural time analysis can quantitatively provide the moment of fracture precursor warnings, but the sliding window length inversely affects the absolute value sizes. Through acoustic emission signal analysis, our research provides a reference for identifying and examining the evolution of fracture damage in ductile materials and engineering components.