Ultrasonic NDE of adhered T-joints using Lamb waves and intelligent signal processing

This paper examines the application of artificial neural networks to the estimation of geometrical parameters of an adhered aluminium T-joint using ultrasonic Lamb waves ( s 0 + a 1). Modulus FFTs of received signals were applied as inputs to conventional feed-forward networks, which were trained us...

Full description

Saved in:
Bibliographic Details
Published in:Ultrasonics 1996-06, Vol.34 (2), p.455-459
Main Authors: Challis, R.E., Bork, U., Todd, P.C.D.
Format: Article
Language:English
Subjects:
Acoustic signal processing
Acoustics
Analysing. Testing. Standards
Applied sciences
Exact sciences and technology
Fundamental areas of phenomenology (including applications)
Lamb waves
Metals. Metallurgy
Neural networks
Nondestructive testing
Physics
Testing for defects
Ultrasonics
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper examines the application of artificial neural networks to the estimation of geometrical parameters of an adhered aluminium T-joint using ultrasonic Lamb waves ( s 0 + a 1). Modulus FFTs of received signals were applied as inputs to conventional feed-forward networks, which were trained using the delta rule with momentum. The success rate of various network structures in recognising bond categories was studied as a function of the density of information applied to the network inputs and the number of hidden nodes in the network. An optimum network structure appears to exist that will solve a number of problems of this type.
ISSN:0041-624X
1874-9968
DOI:10.1016/0041-624X(95)00115-J