scholarly journals Benefit of Neural Network for the Optimization of Defect Detection on Composite Material Using Ultrasonic Non Destructive Testing

2021 ◽  
Author(s):  
P. Trouvé-Peloux ◽  
B. Abeloos ◽  
A. Ben Fekih ◽  
C. Trottier ◽  
J.-M. Roche
Keyword(s):  
Defect Detection ◽  
Semantic Segmentation ◽  
Detection Methods ◽  
Learning Approaches ◽  
Non Destructive Testing ◽  
Destructive Testing ◽  
Experimental Database ◽  
Out Of Plane ◽  
Non Destructive

Abstract This paper is dedicated to out-of-plane waviness defect detection within composite materials by ultrasonic testing. We present here an in-house experimental database of ultrasonic data built on composite pieces with/without elaborated defects. Using this dataset, we have developed several defect detection methods using the C-scan representation, where the defect is clearly observable. We compare here the defect detection performance of unsupervised, classical machine learning methods and deep learning approaches. In particular, we have investigated the use of semantic segmentation networks that provides a classification of the data at the “pixel level”, hence at each C-scan measure. This technique is used to classify if a defect is detected, and to produce a precise localization of the defect within the material. The results we obtained with the various detection methods are compared, and we discuss the drawbacks and advantages of each method.

A REVIEW ON EDDY CURRENT THERMOGRAPHY TECHNIQUE FOR NON-DESTRUCTIVE TESTING APPLICATION

2016 ◽  
Vol 78 (11) ◽  
Author(s):  
N. S. Rusli ◽  
I. Z. Abidin ◽  
S. A. Aziz
Keyword(s):  
Defect Detection ◽  
Eddy Current ◽  
Infrared Camera ◽  
Heat Diffusion ◽  
Non Destructive Testing ◽  
Destructive Testing ◽  
Quantitative Characterisation ◽  
Optimum Heating ◽  
Non Destructive ◽  
Eddy Current Heating

Eddy current thermography is one of the non-destructive testing techniques that provide advantages over other active thermography techniques in defect detection and analysis. The method of defect detection in eddy current thermography has become reliable due to its mode of interactions i.e. eddy current heating and heat diffusion, acquired via an infrared camera. Such ability has given the technique the advantages for non-destructive testing applications. The experimental parameters and settings which contribute towards optimum heating and defect detection capability have always been the focus of research associated with the technique. In addition, the knowledge and understanding of the characteristics heat distribution surrounding a defect is an important factor for successful inspection results. Thus, the quantitative characterisation of defect by this technique is possible compared to the conventional non-destructive which only acquired qualitative result. In this paper, a review of the eddy current thermography technique is presented which covers the physical principles of the technique, associated systems and its applications. Works on the application of the technique have been presented and discussed which demonstrates the ability of eddy current thermography for non-destructive testing of conductive materials.   

Research on Non-Destructive Testing for Defects of Steel Bar in Concrete

2013 ◽  
Vol 351-352 ◽  
pp. 143-147
Author(s):  
Jing Yang ◽  
Wei Heng Yuan ◽  
Jun Tan
Keyword(s):  
Defect Detection ◽  
Electromagnetic Induction ◽  
Experimental Result ◽  
Non Destructive Testing ◽  
Test Device ◽  
Destructive Testing ◽  
Structure Detection ◽  
Steel Bar ◽  
Non Destructive ◽  
Magnetic Scanner

Steel bar defect detection in concrete is an important content of civil engineering structure detection. Currently there are no effective methods for nondestructive testing of steel bar defects . This paper studies the application of electromagnetic induction technology for Steel bar defect detection. Firstly, the principle of electromagnetic induction technology to detect rebar are described. Secondly,an air dielectric test device was designed and Steel bar defect in the device was detected by magnetic scanner. Through analyzing we got the characteristics of scanning images from different Steel bar defects. Thirdly this experimental result was compared with detection result in concrete.Finally verify the accuracy and feasibility of this method.

A New Fuzzy Geometrical Approach to Classify Defects in Composite Materials

2012 ◽  
Vol 2 (4) ◽  
pp. 12-20
Author(s):  
Matteo Cacciola ◽  
Salvatore Calcagno ◽  
Fabio La Foresta ◽  
Mario Versaci
Keyword(s):  
Point Of View ◽  
Geometrical Approach ◽  
Non Destructive Testing ◽  
Destructive Testing ◽  
Fuzzy Distance ◽  
Reinforced Polymer ◽  
Heuristic Technique ◽  
Non Destructive

It is well known that in the Non Destructive Testing/Evaluation (NDT/E) context, Ultrasonic Echoes (UEs) and Tests (UTs) are intensively exploited to identify and characterize defects in the Carbon Fiber Reinforced Polymer (CFRP). This paper examines the localization and the classification of defects in this material from a fuzzy geometrical point of view. In particular, starting from an experimental campaign of measurements carried out in our Lab (Laboratory of Electrical Engineering & Non-Destructive Tests and Evaluations, “Mediterranea” University of Reggio Calabria), fuzzy subsethood calculus is taken into account to translate the characterization of a defect in CFRP into a sort of “fuzzy distance” among UEs. Finally, the floor is open for any questions related to the comparison with a higher computational complexity heuristic technique.

Defect Detection in XLPE Material Using Terahertz Wave-based Non-destructive Testing

2021 ◽  
Vol 71 (3) ◽  
pp. 305-309
Author(s):  
Min-Gyu BAE ◽  
In-Sung LEE ◽  
Joong Wook LEE
Keyword(s):  
Defect Detection ◽  
Terahertz Wave ◽  
Non Destructive Testing ◽  
Destructive Testing ◽  
Non Destructive

scholarly journals A NEW APPROACH TO ASSESSING THE OPERATIONAL SAFETY OF COMPOSITE MATERIALS AND PARTS OF COMPLEX STRUCTURES BASED ON ARTIFICIAL INTELLIGENCE METHODS AS A PART OF NEURAL NETWORKS AND DEEP RESULTS OF MULTI-CRITERIA COMPLEX NON-DESTRUCTIVE TESTING

2020 ◽  
pp. 18-27
Author(s):  
D. A. Akimov ◽  
A. D. Kleymenov ◽  
S. O. Kozelskaya ◽  
O. N. Budadin
Keyword(s):  
Artificial Intelligence ◽  
Neural Networks ◽  
Complex Structures ◽  
Non Destructive Testing ◽  
Destructive Testing ◽  
New Approach ◽  
Heterogeneous Information ◽  
Operational Safety ◽  
Non Destructive

The article proposes a new approach to assessing the operational safety of materials and parts of complex structures based on artificial intelligence methods based on artificial neural networks and multi-criteria complex non-destructive testing, and special mathematical and algorithmic support for systems for evaluating operational safety and predicting residual life under external influences. A method of morphological analysis of the procedures for using measurement tools for heterogeneous information with different a priori information, both about the type of characteristics and the distribution of errors in the input and output signals, has been developed. The classification of problems of measuring parameters for the integration of heterogeneous information is proposed. A macromodel of error is obtained that can be used for research purposes to minimize errors in the developed equipment or for the purpose of correcting errors during operation. A classification of methods for measuring heterogeneous information from the standpoint of probability distribution theory is proposed. Experimental testing of developed algorithms tailored aggregation of information non-destructive testing and adaptation to poorly formalized parameters, which confirmed the effectiveness of the developed methods and algorithms for assessment of structures and resource forecasting their operational reliability was carried out.

scholarly journals Effects of adjusting dynamic e-modulus to reference values of timber moisture and temperature in classification of small number of samples

2019 ◽  
Vol 71 (2) ◽  
pp. 125-135
Author(s):  
Adriana Bjelanović ◽  
Tomislav Franković ◽  
Ivana Štimac Grandić
Keyword(s):  
Reference Values ◽  
Statistical Significance ◽  
Strong Correlations ◽  
Non Destructive Testing ◽  
Destructive Testing ◽  
Static Testing ◽  
Regression Parameters ◽  
Dynamic Ultrasound ◽  
Non Destructive

Mathematical dependences are derived for non-destructive testing (NDT) and destructive testing (DT) of three timber sets, each with six beams made of soft and hard structural timber. Very strong correlations were established between elastic moduli (e-moduli) determined by non-destructive testing, from dynamic ultrasound testing with direct propagation and static testing to bending action, and the correlation of e-moduli with bending strengths. The effects of adjustment of NDT results to reference values of moisture and temperature, and statistical significance of regression parameters, were evaluated from the standpoint of use in the initial classification of a small number of samples.

Sign in / Sign up

Export Citation Format

Share Document