A Novel Differential Time-of-Arrival Estimation Technique for Impact Localization on Carbon Fiber Laminate Sheets

Abstract

Composite material structures are commonly used in many industrial sectors (aerospace, automotive, transportation), and can operate in harsh environments where impacts with other parts or debris may cause critical safety and functionality issues. This work presents a method for improving the accuracy of impact position determination using acoustic source triangulation schemes based on the data collected by piezoelectric sensors attached to the structure. A novel approach is used to estimate the Differential Time-of-Arrival (DToA) between the impact response signals collected by a triplet of sensors, overcoming the limitations of classical methods that rely on amplitude thresholds calibrated for a specific sensor type. An experimental evaluation of the proposed technique was performed with specially made circular piezopolymer (PVDF) sensors designed for Structural Health Monitoring (SHM) applications, and compared with commercial piezoelectric SHM sensors of similar dimensions. Test impacts at low energies from 35 mJ to 600 mJ were generated in a laboratory by free-falling metal spheres on a 500 mm × 500 mm × 1.25 mm quasi-isotropic Carbon Fiber Reinforced Polymer (CFRP) laminate plate. From the analysis of many impact signals, the resulting localization error was improved for all types of sensors and, in particular, for the circular PVDF sensor an average error of 20.3 mm and a standard deviation of 8.9 mm was obtained.

Analysis of Errors in the Estimation of Impact Positions in Plate-Like Structure through the Triangulation Formula by Piezoelectric Sensors Monitoring

“…The estimation of the DToA, among different techniques [ 16 , 21 , 24 , 27 , 28 ], was performed with the following three methods: Threshold, Correlation, Likelihood algorithm. …”

“…The measurement of the DToA was finally evaluated by applying the algorithm described in our previous work [ 21 ]. By estimating the rise time around the zero crossings of the selected portion of the impact signal, a fixed threshold crossing event is considered to be valid only when the corresponding rise time falls within a predetermined range.…”

“…In the literature, different strategies for locating impact positions using Lamb waves are investigated, with particular reference to the extraction of the Differential Time of Arrival (DToA) [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 ].…”

“…The aim of this work is to describe a method for the evaluation of the experimental errors of a targeted set-up, and then with the support of a theoretical analysis, the influence of the different sources of error of the impact position estimation. Without lack of generality, in this work, we present an analysis of errors referring to an acquisition system and an impacts generation tool, while for the signal processing technique we compared three methods: the traditional threshold crossing technique [ 17 , 20 ], the cross-correlation technique [ 24 , 25 ], and a third one implementing the algorithm of "likelihood" previously published by the authors [ 21 ].…”

Impact Monitoring for Aircraft Smart Composite Skins Based on a Lightweight Sensor Network and Characteristic Digital Sequences

“…Therefore, it must also meet the requirements of light weight, low power consumption and high reliability [ 11 ]. Among the existing SHM methods, the piezoelectric (referred to as PZT) sensors and guided wave based impact monitoring method is a popular one [ 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 ] because PZT sensors have a large monitoring range by constructing a monitoring network and can work in a passive way for sensing impact response signals of guided wave on a monitored structure without any power source. However, it is still difficult for a PZT sensor network and the corresponding impact monitoring system to meet the requirements mentioned above at the same time because of the following main problems: The PZT sensor network of ASCSs often contains a large number of PZT sensors.…”