Electronique
Acoustic emission source localization in plate-likestructure
This article proposes a geometric approach foracoustic emission source (AE) localization in plate-like structure.In order to determine the arrival time of the acoustic emissionwaves for each sensor with more precision, we have used twotechniques, the first one is based on thresholding and the secondis based on the continuous wavelet transform. Experimental testhave been carried out on steel plate shown that the continuouswavelet transform allows to be improve the accuracy of theacoustic emission source localization. Voir les détails
Mots clés : acoustic emission signal, acoustic emission source location, arrival time, threshold, continuous wavelet transform
Complexity Reduction of UltrasoundSub-Ultra-Harmonic Modeling by an Input Modified Volterra Approach
Contrast of echographic images has been highly improved by the injection of microbubbles, due to their nonlinearbehavior. However, this contrast enhancement is limited by the nonlinear acoustic propagation in tissue. To overcome this drawback,sub and ultra-harmonic contrast imaging can be used, since only microbubbles can generate these components. Nonlinear modeling is aprimordial step in the analysis of microbubble signals for sub and ultra-harmonic imaging. Nonlinear models like Volterra model hasbeen applied in harmonic imaging to model harmonics optimally. However, it can model harmonics only. For sub and ultra-harmonicmodeling, a multiple input single output (MISO) Volterra has been proposed. The aim of this study is to propose a simpler alternativefor the modeling of sub and ultra-harmonics. We propose a modified single input single output (SMISO) Volterra model based on inputdemodulation. The model is tested using simulated and experimental signals. Results showed that sub and ultra-harmonics aremodeled. The number of kernels is reduced to its half using SMISO model compared to MISO model. The relative mean square errorbetween the simulated signal and the modeled signal with SMISO Volterra model is −15.8 dB and it is −60.7 dB for experimentalsignals. The computational time is reduced by a factor of 4 and 5 in simulated and experimental cases respectively. SMISO model canmake easier the sub and ultra-harmonics modeling. Voir les détails
Mots clés : modeling, sub-ultra-harmonics, SMISO Volterra, demodulation, microbubble.
Modeling of Electromagnetic Behavior of CompositeThin Layers using Genetic Algorithm
In this paper, we present a new model using the highfrequency electromagnetic simulator for several binary mixtureswhere the load is in the lossless thin film form with a permittivityof (? = 100, 200, 300, and 400) and for various thickness values ina range of 10 µm to 250 µm with respect to the host matrix. Themodel operates in a variety of frequencies from 8.2 GHz to 12.4GHz. The effective permittivity of composites is evaluated usingNicholson Ross Weir (NRW) algorithm in a rectangularwaveguide. The implementation of NRW algorithm is conductedon various samples simulated by HFSS, in order to estimate thedielectric composite behavior. Furthermore, we employ a geneticalgorithm methodology (GA) for the filling factor optimization ofthe proposed model by Mosallaei. The obtained results show agood agreement with the theoretical models, which ensure thevalidity of our proposed model for characterizing theelectromagnetic behaviour of dielectric thin films. Voir les détails
Mots clés : Thin films, electromagnetic behaviour, dielectric mixtures, Genetic optimization, mico wave.
Débruitatge De Signal De Bruit De Barkhausenen en Utilisant La Décomposition Des Modes Empirique Pour Optimisation L’évaluation et Caractérisation Des Matériaux
Dans le domaine contrôle non destructif CND utilise les paramètres électriques et magnétiques des matériaux pour la caractérisation. Parmi ces techniques l’analyse par bruit de Barkhausen, qui est une technique récente développé grâce au progrès de l’électronique. L’information contenue dans le signal reçu offre la possibilité de déterminer plusieurs paramètres dans le but d’analyser le matériau comme toute technique de CND. Le signal de Barkhausen mesuré par l’intermédiaire d’une chaîne de contrôle réalis est noyé par le bruit (bruit blanc) et l'information de l'état des matériaux peut s’être perdue. Nous proposons dans cet article une méthode de débruitage basé sur la décomposition des modes empiriques (EMD) pour optimiser l’utilisation des indicateurs scalaires (kurtosis,RMS) et fréquentielles (FFT, spectre d’enveloppe) de l’évaluation des matériaux. L’EMD décompose d’une façon adaptative un signal en une somme de composantes oscillantes s’appelle les fonctions modales intrinsèques(IMFs) par l’utilisation un processus de tamisage. Après la décomposition on utilise le seuillage doux pour éliminer les IMFs inférieure au seuil . enfin réaliser la reconstruction du signal débruité à l’aide des IMFs débruité. Avec un bon choix du seuil et de l’ondelette utilisée, on arrive à une réduction appréciable du bruit sur le signal de barkhausen et par conséquent une amélioration des indicateurs scalaires. Les résultats obtenus montrent l'intérêt du dé-bruitage pour le contrôle des matériaux par signal de bruit de barkhausen. Voir les détails
Mots clés : bruit de barkhausen, EMD, cnd
Micrographic Image Segmentation using Level Set Model based on Possibilistic C-Means Clustering
Image segmentation is often required as a fundamental stage in microstructure material characterization. The objective of this work is to choose hybridization between the Level Set method and the clustering approach in order to extract the characteristics of the materials from the segmentation result of the micrographic images. More specifically, the proposed approach contains two successive necessary stages. The first one consists in the application of possibilistic c-means clustering approach (PCM) to get the various classes of the original image. The second stage is based on using the result of the clustering approach i.e. one class among the three existing classes (which interests us) as an initial contour of the level set method to extract the boundaries of interest region. The main purpose of using the result of the PCM algorithm as initial step of the level set method is to enhance and facilitate the work of the latter. Our experimental results on real micrographic images show that the proposed segmentation method can extract successfully the interest region according to the chosen class and confirm its efficiency for segmenting micrographic images of materials. Voir les détails
Mots clés : Level set, Possibilistic C-Means Clustering, micrographic images, image segmentation.
Phased array B-scan image enhancement based oncontinuous wavelet transform and Shannon energyalgorithm
In this work, we describe a novel algorithm for ultrasonic phased array signalsenhancement, based on continuous wavelet transform using the Mexican Hat wavelet mother(CMHWT) and normalized Shannon Energy (SE). The use of signal processing algorithms indefect detection gives generally very satisfactory results. Time–frequency analysis methods aremainly used to improve the defects detection resolution. Performance improvement is confirmedwhen the proposed approach is tested with B-scan signals containing delamination closer to thefront face. This work has successfully demonstrated that the proposed method can improve thequality of ultrasound B-scan signal. Voir les détails
Mots clés : phased array, Defects enhancement, CWT, Shannon energy, Mexican hat wavelet
Detection and classification of steel defects using machine vision and SVM classifier
the importance of quality control of steel products is increasing day by day in the manufacturing industrial systems because it offers the possibility of knowing the state of the products without stopping the production line which allows the control of a defect before it becomes a complex problem and avoiding production losses. Human quality control of steel products remains tedious, fatiguing, bit fast, bit robust, dangerous or impossible, therefore the use of automated vision system can significantly improve the quality inspection process, because the machine vision technology can overcomes the majority of manual inspection problems cited above and provide an interesting solution especially, with the impressive increasing of computing power of today's computers and the good quality of images that offer the current cameras.The main objective of this research is to propose an efficient control system based on machine vision technology and SVM classifier to classify different types of steel defects. Voir les détails
Mots clés : Defects steel, machine vision, pattern recognition, HOG, GLCM, SVM classifier
Investigation of Polycrystalline silicon TFT’s electrical characteristics
Low-temperature polycrystalline silicon thin film transistors (poly-Si TFT’s) on plastic substrate have been studied because of their high performance in Active Matrix Liquid Crystal Displays (AMLCD’s) and Active Matrix Organic Light-Emitting Diode (AMOLED) applications. The purpose of this work is to simulate the impact of varying the electrical and physical parameters (the interface states, active layer’s thickness and BBT model) in the transfer characteristics of poly-Si TFT to extract the electrical parameters like threshold voltage, mobility and to evaluate the device performance. The device was simulated using the ATLAS software from Silvaco. The results show that the electrical and physical parameters of poly-Si TFT affect significantly its transfer characteristics. Voir les détails
Mots clés : Poly-Si TFT, TCAD-ATLAS, electrical characteristics
Signal Quality Improvement Using a New TMSSE Algorithm:Application in Delamination Detection in Composite Materials
This paper introduces a novel method to improvethe quality of ultrasonic phased array signals for localizingwith accuracy delamination defects. The improvementis achieved by the introduction of a new threshold for theShannon energy. In first, we have applied the threshold modifiedS-transform algorithm (TMST) in the case of ultrasoundB-scan. Thereafter, we have adapted and applied the StransformShannon energy (SSE) algorithm in the field ofultrasonic testing. At last, we have proposed a novel algorithmbased on threshold modified S-transform and Shannonenergy (TMSSE) to increase the improvement of the ultrasoundB-scan. A simulation study has been carried outsimulating a composite material containing three defects indifferent positions in order to highlight the phenomenon ofdelamination. Experimental tests were performed on a sampleof carbon fiber reinforced polymer composite material(CFRP) with a delamination defect close to the front face.Both experimental and simulated results show that the proposedmethod can improve the quality of ultrasound B-scanwhich enhances the localization of delamination defects. Voir les détails
Mots clés : Ultrasonic signal
Characterization of Structural Noise Patterns and Echo Separation in the Time-Frequency Plane for Austenitic Stainless Steels.
The aim of this study is to characterize the structural noise for a better flaw detection in heterogeneous materials (steels, weld, composites...) using ultrasonic waves. For this purpose, the continuous wavelet transform is applied to ultrasonic A-scan signals acquired using an ultrasonic non destructive testing (NDT) device. The time-scale representation provided, which highlights the temporal evolution of the spectral content of the A-scan signals, is relevant but can lead to misinterpretation. The problem is to identify if each pattern from the wavelet representation is due to the structural noise or the flaw. To solve this problem, a detection technique based on statistical significance testing in the time-scale plane is used. Information about the structural noise signals is injected into the decision process using an autoregressive model, which seems relevant according to the spectral content of the signal. The approach is tested on experimental signals, obtained by ultrasonic NDT of metallic materials (austenitic stainless steel) then on a weld in this steel and indeed enables to distinguish the components of the signal as flaw echoes, which differ from the structural noise. Voir les détails
Mots clés : Austenitic stainless steels, Structural noise, Flaw detection, Wavelet transform, Autoregressive model, Significance testing