Application of Improved b-Value and Clustering Analysis for Structural Integrity Assessment of CFRP Specimen under Tensile Loading

In the present study, the evolution of different failure mechanisms in carbon fiber reinforced polymer composites is being investigated using acoustic emission technique, unsupervised clustering technique and improved b-value analysis. The experimental part involved the realization of tensile tests of different materials, namely samples with [0/90] 2S uniaxial layer configuration and [0/90] 2S twill fabric samples. Both types of tests were monitored using one wideband acoustic emission sensor, while the tensile tests of twill fabric samples were additionally supplemented with resonant acoustic emission sensor to perform a comparative analysis between datasets from resonant/wideband acoustic emission sensor. The comparative study itself was preceded by the failure mechanisms characterization process, which has been performed on the tensile test dataset of [0/90] 2S layer configuration with the contribution of clustering technique. The subsequent analysis of the twill fabric resonant/wideband acoustic emission sensor datasets included the improved b-value technique, which relates the magnitude of fracture with the slope of the amplitude distribution. The presented results, especially in terms of the improved b-value technique applied to individual clusters, show enhanced ability to assess in more detail the actual structural integrity depending on the applied load.