Start Date

2016 12:00 AM

Description

In a Structural Health Monitoring context certain image processing techniques allow flaws detection and localization by means of a typical set-up composed by a piezoelectric transducers network arbitrarily distributed in the structure under test. In the present work a method based on the use of three piezo electric transducers (one emitter and two receivers) is proposed, for the monitoring of plate-like structures. The method employs a guided Lamb waves propagation model including the A0 and S0 modes.

A library of simulated signals based on this model is then implemented, for every possible position belonging to a spatial matrix (grid) defined a priory. An orthogonal Matching pursuit algorithm is also included in the procedure, in order to select the data that better represent the received signals. According to these results, an estimation of the flaw position is eventually proposed. This procedure is then repeated in a different spatial grid, with new simulated signals describing a smaller region around the flaw, until a certain precision threshold imposed a priory is reached.

This precision takes into account the transducer network geometry, the flaw position as well as the Signal-to-Noise Ratio of the acquired signals. An experimental validation, based on the comparison between healthy and damaged-induced zone (artificial flaws created in a reversible way by means of magnets) is eventually proposed.

Language

en

File Format

application/pdf

Share

COinS
 
Jan 1st, 12:00 AM

Flaws Detection and Localization in a Plate for Structural Health Monitoring by Means of Guided Waves and Matching Pursuit Algorithms

In a Structural Health Monitoring context certain image processing techniques allow flaws detection and localization by means of a typical set-up composed by a piezoelectric transducers network arbitrarily distributed in the structure under test. In the present work a method based on the use of three piezo electric transducers (one emitter and two receivers) is proposed, for the monitoring of plate-like structures. The method employs a guided Lamb waves propagation model including the A0 and S0 modes.

A library of simulated signals based on this model is then implemented, for every possible position belonging to a spatial matrix (grid) defined a priory. An orthogonal Matching pursuit algorithm is also included in the procedure, in order to select the data that better represent the received signals. According to these results, an estimation of the flaw position is eventually proposed. This procedure is then repeated in a different spatial grid, with new simulated signals describing a smaller region around the flaw, until a certain precision threshold imposed a priory is reached.

This precision takes into account the transducer network geometry, the flaw position as well as the Signal-to-Noise Ratio of the acquired signals. An experimental validation, based on the comparison between healthy and damaged-induced zone (artificial flaws created in a reversible way by means of magnets) is eventually proposed.