Performance Analysis of Acoustic Emission Hit Detection Methods Using Time Features

Acoustic emission (AE) analysis is a powerful potential characterisation method forfracture mechanism analysis during metallic specimen testing. Nevertheless, identifying and extracting each event when analysing the raw signal remains a major challenge. Typically, AEdetection is carried out using a thresholding approach. However, thoughextensively applied, this approach presents some critical limitationsdue to overlapping transients,differences in strength and low signal-to-noise ratio.To address these limitations, advancedmethodologies for detecting AE hits have been developedin the literature. The most prominently used are instantaneous amplitude, the short-termaverage to long-term average ratio,the Akaike information criterionandwaveletanalysis, each of which exhibits satisfactory performance and easeof implementationfordiverseapplications. However, their pronenessto errors in the presence of non-cyclostationary AEwavefrontsand the lack of thoroughcomparison for transient AE signalsare constraints to the wider application of these methodsin non-destructive testing procedures.In this studywith the aim of make aware about the drawbacks of the traditionalthreshold approach, a comprehensiveanalysis ofits limiting factorswhentaking in regard the AE waveformbehaviouris presented.Additionallyin a second section, a performance analysis of the main advanced representative-methods in the field is carried out throughacommon comparative framework, by analysing first, AE waves generated from a standardisedHsu-Nielsen testand second, adata frame of a highly active signal derivedfrom a tensile test.With the aim to quantify the performance with which theseAE detection methodologies work, for the first time in literature, time features as the endpoint and duration accuracies, as well as statistical metricsas accuracy, precisionand false detection rates, are studied.