PSI - Issue 42

Chahboub Yassine et al. / Procedia Structural Integrity 42 (2022) 1025–1032 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

1028

4

models with excellent accuracy. The architecture of the ANN model is shown in Fig. 1. It has three layers: the input layer, the concealed layer, and the output layer. Each neuron in the hidden layer gets the total output from all the neurons in the input layer: n j ij i j i V W x b = +  (9) where x i is the input to the i th neuron in the input layer, and bj is the threshold value of the j th neuron in the hidden layer. The output of a neuron in the hidden layer is calculated by applying the potential input to a transfer function. The sigmoid function shown in (10) is used as the transfer function between the input layer and the hidden layer. 1 ( ) 1 j i j V y F V e − = = + (10) For the output layer, the output σ k of the neuron k is given by the relation below: m k jk j k j W y b  = +  (11) Where m is the number of neurons in the hidden layer, W jk is the connection weight from the i th neuron in the hidden layer to the k th neuron in the output layer, y j is the output from the j th neuron in the hidden layer, and h k is the threshold value of the k th neuron in the output layer. Hamdi et al. (2010) V j presents the input to the j th neuron in the hidden layer, n refers to the total number of neurons in the input layer, W ij is the weight from the i th neuron in the input layer to the j th neuron in the hidden layer,

Fig. 1. (a) Multilayer perceptron (b) The architecture of ANN Matlab [2018]

2. Experiments and Simulations 2.1. Failure prediction of the pipeline using the direct method. The pipe's four-point bending test was reproduced with a through-wall fracture defect Fig. 2. Regarding the four-point bending test on pipes, one large-scale experiment, FP1 a, was planned on mock-ups consisting of ferritic pipes on the EDF 4-point bending test facility developed during the former STYLE project, Moinereau.D et al. (2014).