Flipped Accumulative Non-Linear Single Impact Resonance Acoustic Spectroscopy (FANSIRAS): A novel feature extraction algorithm for global damage assessment

Abstract

[EN] High amplitude non-linear acoustic methods have shown great potential for the identification of micro-damage in inhomogeneous materials such as concrete. Usually, these methods evaluate non-linearity parameters related to the hysteretic behaviour from the dependence of the shifts in both frequency and damping on the amplitude of the strain. A deep understanding of the reverberation phenomena has been obtained in order to introduce a novel signal processing approach called FANSIRAS (Flipped Accumulative Nonlinear Single Impact Resonance Acoustic Spectroscopy). Traditional acoustic spectroscopy techniques, NIRAS (Non-linear Impact Resonance Acoustic Spectroscopy) and NSIRAS (Non-linear Single Impact Acoustic Spectroscopy), have been analyzed and compared with the brand new approach when providing quantitative information related to the degree of micro-cracking in thermal damaged concrete based materials. The new resonance-based algorithm demonstrates that the non-linear non-classical parameters can be determined through a single resonance frequency measurement, obtaining the expected sensitivity to internal damage. Its simplicity and robustness may be important in industrial applications. (C) 2018 Elsevier Ltd. All rights reserved.