A Neural Network Approach for Locating Multiple Defects

Abstract

A method is presented to demonstrate the use of artificial neural networks (ANNs) in providing additional information regarding defects or flaws when used in conjunction with the ultrasonic A-scan method. ANNs were employed both as pattern classifiers and as function approximators to maximise the amount of data available from the temporal A-scan signal. A steel bar was modelled in 2D using ABAQUS finite element analysis (FEA) software. A single defect was introduced to the bar, modelled as a void, and parametric studies conducted to record data with the defect at various locations. An ultrasonic Lamb wave was introduced at the top of the bar. The longitudinal wave propagated along the length of the bar and was partially reflected by the defect. Multiple cases were simulated, modelling voids between 1mm and 6mm in width in various locations. Mean displacement of all the nodes at the top of the bar was recorded throughout the simulation, and features extracted from this waveform to create the data set for the ANNs. The ANNs were trained with a percentage of the data collected, selected at random, and assessed with the remaining data. The target data for the ANNs were the depth and size of the defect. The case of two separate defects was also investigated. The procedure was carried out in the same manner as for one defect, but in this case the target data for the ANNs were the depth of the first defect and the distance between the defects. A separate ANN was employed as a pattern classifier, to determine if the reflected A-scan signal represented one or two defects. The final system was tested using previously unseen data, and provided very good results both in determining the number of defects and the size and location of the defects, even with data to which noise had been added.

Artificial neural network for random fatigue loading analysis including the effect of mean stress

“…Petrucci and Zuccarello [2] presented a fatigue damage formula in the form shown in equation (16). (16) The function are also functions of the spectral moments mi, i=0,1,2 and 4 , the parameter = / derived from the use of the Goodman's formula for accounting for the effect of mean stress. Smax is the maximum stress in the stress history, and k is a fatigue material property.…”

“…This paper presents an alternative approach to those reviewed in the foregoing. Artificial neural networks have been known to provide greater scope for non-linear generalisation and have the ability to deal with a large number of input variables than direct application of optimisation methods [15], [16], [17]. The paper presents an artificial neural network frequency based approach for the analysis of random loading fatigue problems including the effect of mean stress.…”

Locating and classifying defects using an hybrid data base

“…In previous works [2][3][4][5][6], the same methodology was followed and synthetic data was only used. In this works few experimental results were introduced.…”

“…In a previous papers [2,3,4,5,6], the dynamic response has been used for this purpose. The propagation of elastic waves was considered in [2].…”

A pattern recognition artificial neural network method for random fatigue loading life prediction

“…Artificial neural network has been known to provide greater scope for non-linear generalisation and ability to deal with a large number of input variables than direct application of optimisation methods [3], [4], [5]. Very little has however been reported in the literature on the use of artificial neural network method on problems related to random fatigue loading problems.…”