Maximium Entropy Deconvolution of four Sensor Acoustic Emission Maps in Aerospace Structures

Date
1999
Authors
Wells, Paul
Rose, Rob
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Authors
Research Projects
Organizational Units
Journal Issue
Series
Abstract

Acoustic emission measurements allow the direct detection and localisation of fatigue damage in complex primary aircraft structures. We have previously shown [1] how a simple acoustic dispersion model can largely explain the systematic and random errors observed in the spatial localisation of pencil break measurements from four sensors on an aircraft panel. In this paper we use a maximum entropy technique and the dispersion model to calculate a map of possible source positions from a map of individual acoustic emission events. The maximum entropy algorithm [2–5] adapts an algorithm (used for analysing Hubble Space Telescope data [2]) by using the simple dispersion inspired model to construct a point spread function that varies across the test area. The results of testing the algorithm on an acoustic emission location map from a fatigue crack on a test specimen with a realistic geometry are shown. The algorithm runs on a PC and with a 49×49 image runs overnight showing good convergence after 400 iterations.

Description
Keywords
Citation
DOI