Transient Electromagnetic-Thermal Nondestructive Testing: Pulsed Eddy Current and Transient Eddy Current Thermography

Transient Electromagnetic-Thermal Nondestructive Testing

Pulsed Eddy Current and Transient Eddy Current Thermography

Yunze He

Associate Professor, College of Electrical and Information Engineering,

Hunan University, Changsha, China

College of Mechatronics Engineering and Automation,

National University of Defense Technology, Changsha, China

Bin Gao

Professor, School of Automation Engineering, University of Electronic

Science and Technology of China, Chengdu, China

Ali Sophian

Assistant Professor, Mechatronics Engineering Department,

Faculty of Engineering, International Islamic University Malaysia,

Kuala Lumpur, Kuala Lumpur, Malaysia

Ruizhen Yang

Associate Professor, Department of Civil and Architecture Engineering,

Changsha University, Changsha, China

Table of Contents

Cover

Title page

Copyright

Preface

Acknowledgments

Abbreviations

Chapter 1: Nondestructive Testing and Transient Electromagnetic-Thermal NDT

Abstract

Part I: Pulsed Eddy Current

Chapter 2: Magnetic Sensor Based Pulsed Eddy Current for Defect Detection and Characterization

Abstract

2.1. Introduction to PEC

2.2. Magnetic sensor based PEC systems

2.3. Signal processing software

2.4. Inspection of nonferromagnetic samples

2.5. Inspection of ferromagnetic samples

Chapter 3: Hall-Device-Based PEC Features for Material Properties Evaluation and Defect Detection

Abstract

3.1. Feature extraction in the time domain

3.2. Stress measurement using time-domain features

3.3. Corrosion evaluation and development prediction

3.4. Scanning PEC for honeycomb evaluation

3.5. Scanning PEC for CFRP impact evaluation

Chapter 4: Coil-Based Rectangular PEC Sensors for Defect Classification

Abstract

4.1. Rectangular PEC sensors and feature extraction

4.2. Defect classification under different directions

4.3. Defect classification under variations of lift-off

4.4. PCA with frequency-domain responses for defect classification in multilayer structures

4.5. PCA-SVM based defect classification

Part II: Transient Eddy Current Thermography

Chapter 5: Active Thermography and Eddy Current Excited Thermography

Abstract

5.1. Active thermography

5.2. Eddy current thermography

5.3. ECPT for quantitative analysis of surface defects

Chapter 6: Heat Conduction Based Eddy Current Pulsed Thermography (ECPT) for Defect Evaluation

Abstract

6.1. Time domain quantification analysis for deep defects

6.2. Log domain quantification analysis for deep defects

6.3. Lateral heat conduction based defect evaluation

Chapter 7: Eddy Current Step or Time-Resolved Thermography (ECST)

Abstract

7.1. Principle of ECST

7.2. Numerical studies

7.3. Experimental studies

Chapter 8: Eddy Current Pulsed Phase Thermography (ECPPT) for Metal Materials Evaluation

Abstract

8.1. Basic theory of ECPPT

8.2. Finite element analysis for defect quantification

8.3. Experimental studies for defect quantification

8.4. Elimination of surface emissivity variation

Chapter 9: Volume or Inside Heating Eddy Current Thermography

Abstract

9.1. Physical principles of SHT and VHT

9.2. Numerical studies

9.3. Experimental studies

Chapter 10: Volume Heating ECT and Phase Analysis for CFRP Evaluation

Abstract

10.1. Methodology of ECVHT and phase analysis

10.2. Delamination evaluation using volume heating ECPPT

10.3. Impact evaluation using eddy current square pulse phase thermography

Chapter 11: Pulsed Inductive Thermal Wave Radar (PITWR)

Abstract

11.1. Theory of TWR

11.2. Subsurface defect evaluation and suppression of emissivity variation in steel

11.3. Experimental studies of CFRP

Chapter 12: Through Coating Imaging of Early Marine Corrosion Using ECPPT

Abstract

12.1. Steel corrosion detection and evaluation

12.2. Methodologies

12.3. Experimental studies

Part III: Physical–Mathematical Model–Based ECPT for Defect Evaluation

Chapter 13: Separation of ECPT Transient Electromagnetic–Thermal Fields

Abstract

13.1. Demand for separation of transient electromagnetic–thermal fields

13.2. Physical–mathematical time-dependent partition model

13.3. Validation of model: simulation and experiments

Chapter 14: Unsupervised Sparse Pattern Diagnostic of Defects with ECPT

Abstract

14.1. Methodology

14.2. Physics-based data analytics

14.3. Experimental studies

14.4. Results and discussion

Chapter 15: Multidimensional Tensor-Based ECPT for Wind Turbine Gear Inspection

Abstract

15.1. NDT, CM, and SHM of wind turbines

15.2. Methodology

15.3. Experimental studies

Chapter 16: Physics-Based Modeling and Pattern Mining of ECPT

Abstract

16.1. Methodology

16.2. Modeling and mining of thermal patterns in the spatial, time, frequency, and sparse domains

16.3. Results and discussion

References

Index

Copyright

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom

50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States

Copyright © 2017 National Defense Industry Press. Published by Elsevier INC. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-12-812787-2

For information on all Elsevier publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Jonathan Simpson

Acquisition Editor: Glyn Jones

Editorial Project Manager: Naomi Robertson

Production Project Manager: Susan Li

Designer: Greg Harris

Typeset by Thomson Digital

Preface

With a growing interest in safety and quality problems in aerospace, renewable energy, marine science, railway, traffic, civil, and other industrial fields, much attention has been devoted to the development of reliability-based or condition-based maintenance. Condition monitoring (CM), structural health monitoring (SHM), and nondestructive testing (NDT) techniques are the main means to lower the operation and maintenance (O&M) costs and improve reliability and availability. Transient electromagnetic–thermal (EMT)-related CM, SHM, and NDT techniques with advantages of high speed, great penetration, high sensitivity, wide spectrum, low cost, and easy quantification, are widely investigated and have become one of the most promising directions. In recent years, two transient EMT NDT methods—pulsed eddy current (PEC) and transient eddy current thermography (TECT)—have been the research hotspots. For example, in the Elsevier ScienceDirect database, PEC has been on a steady increasing phase in the past 10 years and TECT has been growing rapidly since 2011. Unfortunately, there are only a few books on PEC and TECT published in English, which hinders the spread and development of the technology and related research. In addition, there are more and more bilingual courses in many universities in China and non-English-speaking countries that would benefit from the availability of such a book. Many undergraduate and postgraduate students, as well as scientists and engineers, need a suitable book for their study and research on EMT NDT. Therefore, it is thought that it is the right time to publish a relevant book to meet the demand of readers.

This book is intended to solve several challenges in the physical principles, methods, techniques, and applications of PEC and TECT. Theoretically, multiphysics fields including EM induction, eddy current distribution, induction heating, heat conduction, thermal wave propagation, temperature field distribution, and infrared radiation are studied for defect evaluation and quantification; lateral heat conduction is used to detect parallel cracks; longitudinal heat conduction is used to detect deep defects, which are beyond the skin depth; thermal wave propagation is analyzed for phase-based defect quantification, and the physical–mathematical model is investigated to improve detection sensitivity and quantitative analysis. Methodologically, various EMT NDT techniques like eddy current pulse thermography (ECPT), eddy current step thermography (ECST), eddy current pulse phase thermography (ECPPT), and pulsed inductive thermal wave radar (PI-TWR) are investigated. Time-domain, frequency-domain, and logarithm-domain defect evaluation methods are described and analyzed; advanced algorithms, such as principal components analysis (PCA), independent components analysis (ICA), support vector machine (SVM), and multidimensional tensor decomposition are used, and image reconstruction and enhancement is used to improve detectability. In view of applications, material property variations including conductivity, permeability, and lift-off are evaluated; experimental studies for real damages including corrosion in steel, stress in aluminum, impact and delamination in carbon fiber reinforced polymer (CFRP) laminates, and RCF cracks in rail are abundant. The book is expected to fill the gap in the literature and provide a relatively comprehensive coverage of PEC and TECT, giving the readers detailed information in related theories, methods, and applications, and benefit readers in both academia and industry.

Chapter 1 introduces the classification of NDT and the classification of transient EMT NDT techniques. The book mainly consists of three parts. Part I is related to PEC. In Chapter 2, a PEC technique where a magnetic sensor is used for sensing the magnetic field is introduced. The usages of signal processing and feature extraction using PCA and wavelets on PEC signals are also presented. In Chapter 3, the influences of material properties on PEC responses are investigated and normalization technique is used to reduce the lift-off effect. Subsequently, two time-domain features representing electrical conductivity and magnetic permeability are extracted. These features are utilized to measure stress in aluminum alloys, to detect defects in honeycomb sandwich structures, to evaluate low-energy impact defects in CFRP, and to characterize atmospheric corrosion on steel. In Chapter 4, the directional PEC probe design providing uniform eddy current is presented. The PEC feature extraction techniques are then studied in both time and frequency domains. PEC frequency response optimization is investigated and used in combination with PCA to eliminate the lift-off effect and interlayer air gap and to classify defects in multilayer structures. The optimized SVM is used to build the classifier model and predict the types of defects.

Part II, which includes Chapters 5–12, is related to TECT and its methods for defect evaluation and characterization. In Chapter 5, basic concepts and classifications for active thermography and eddy current thermography are introduced. Chapter 6 is focused on defect evaluation based on longitudinal and lateral heat conduction of ECPT. Subsurface defects, parallel cracks, and natural oblique cracks in rail are evaluated. Chapter 7 is focused on ECST. Two characteristic features representing separation time were extracted from temperature responses and their monotonic relationships with defect depth were obtained for subsurface quantification. Chapter 8 introduces ECPPT for metal materials through numerical and experimental studies. Two characteristic features, namely blind frequency and minimum phase, were extracted from differential phase spectra. Based on their linear relationships with defect depth, both features can be used to measure defect depth. This chapter also experimentally demonstrates the advantages of phase information to reduce or enlarge the effect of surface emissivity variation. Chapter 9 presents volume heating thermography (VHT) and inside heating thermography (IHT) for advanced composite materials through these electromagnetic excitations. Several specific VHT/IHT methods have been developed in the forms of both (square) pulse and step analysis in the time domain and phase analysis in the frequency domain. Chapter 10 presents eddy current volume heating thermography (ECVHT) and phase analysis for delamination and impact inspection in CFRP. Chapter 11 presents PI-TWR by introducing cross correlation (CC) matched filtering in ECPT. The results illustrate a significant improvement in the dynamic range, depth resolution, emissivity variation reduction, and detectability of subsurface defects and inside delamination. Chapter 12 proposes through coating imaging (TCI) based on ECPPT. The experimental results illustrate that ECPPT has a greater potential for corrosion detection, sizing, and monitoring than laser profilometry, PEC, and microwave waveguide. Finally, power function is suitable for demonstrating development of early stage (in 6 months) marine atmospheric corrosion like it does for long-term corrosion.

Part III, including Chapters 13–16, is related to the physical–mathematical model-based ECPT for defect evaluation. Chapter 13 constructs a physical–mathematical time-dependent partition model to analyze the whole thermal transient process and considers characteristic times for separating Joule heating and thermal diffusion into four different stages. Chapter 14 proposes an unsupervised method for defect diagnosis with ECPT. The proposed method is fully automated and does not require manual selection of the specific thermal frame images for defect diagnosis by the user. The core of the method is a hybrid of a physics-based inductive thermal mechanism with a signal processing-based pattern extraction algorithm using sparse greedy based principal component analysis (SGPCA). Chapter 15 develops a physics-based multidimensional spatial–transient–stage tensor model to describe the thermooptical flow (TOF) pattern for evaluating contact fatigue damage in a helical gear. It indicates that the proposed methods are effective tools for gear inspection and fatigue evaluation. In Chapter 16, we generate a physics–mathematical modeling and mining route in the spatial-, time-, frequency-, and sparse-pattern domains. This is a significant step toward realizing the deeper insight in ECPT and automatic defect identification. This renders ECPT a promising candidate for the highly efficient and yet flexible nondestructive testing and evaluation (NDT&E) technique.

Dr. Yunze He

Associate Professor, College of Electrical and Information Engineering,

Hunan University, Changsha, China;

College of Mechatronics Engineering and Automation,

National University of Defense Technology, Changsha, China

Dr. Bin Gao

Professor, School of Automation Engineering,

University of Electronic Science and Technology of China, Chengdu, China

Dr. Ali Sophian

Assistant Professor, Mechatronics Engineering Department,

Faculty of Engineering, International Islamic University Malaysia,

Kuala Lumpur, Selangor, Malaysia

Dr. Ruizhen Yang

Associate Professor, Department of Civil and Architecture Engineering,

Changsha University, Changsha, China

October 2016

Acknowledgments

The authors work was supported in part by the National Natural Science Foundation of China under grants 61501483, 61171134, F011404, 51377015, 61401071, U1430115, 51408071, and 61527803; in part by the EU FP7 through Health Monitoring of Offshore Wind Farms (www.hemow.eu) under grant 269202; and in part by the Engineering and Physical Sciences Research Council (EPSRC) of the United Kingdom through Future Reliable Renewable Energy Conversion Systems & Networks: A Collaborative UK–China Project under grant EP/F06151X/1. The authors would also like to thank TWI Ltd., the University of Huddersfield and Newcastle University, United Kingdom for sponsoring Dr. Ali Sophian’s PhD research and study. The authors also thank ALSTOM and Prof. Grimberg in the National Institute of Research and Development for Technical Physics, Romania, for providing the experimental CFRP samples.

The authors would like to thank Prof. Gui Yun Tian at the School of Automation Engineering, University of Electronic Science and Technology of China, Chengdu, China, and the School of Electrical and Electronic Engineering, Newcastle University, Newcastle Upon Tyne, United Kingdom, and Dr. W.L. Woo at the School of Electrical and Electronic Engineering, Newcastle University, United Kingdom for their guidance and leadership in their studies and academic careers. Dr. Yunze He would like to thank Prof. Feilu Luo and Prof. Mengchun Pan at the College of Mechatronics Engineering and Automation, National University of Defense Technology, Changsha, China for supervision in Mr. Yunze He’s MA and PhD programs. Dr. Ali Sophian would like to again thank Dr. John Rudlin of TWI Ltd. and Prof. David Taylor at the University of Huddersfield for their supervision, guidance, and support during his PhD study.

The authors would like to thank their colleagues: Dr. John Wilson, Dr. Liang Cheng, Dr. Hong Zhang, Prof. Xianzhang Zuo, Dr. Libing Bai, Prof. Aijun Yin, Dr. Jianping Peng, Dr. Omar Bouzid, Dr. Abdul Qubaa, Dr. Kongjing Li, Dr. Yunlai Gao, Dr. Yizhe Wang and all at Newcastle University; Dr. Ying Tang, Dr. Yuhua Zhang, Dr. Xiangchao Hu, Dr. Junzhe Gao, Dr. Bo Liu and all at the National University of Defense Technology, who have provided fruitful discussions and advice on pulsed eddy current testing.

The authors are also grateful to the China Scholarship Council for sponsoring Dr. Yunze He to Newcastle University, United Kingdom and Dr. Ruizhen Yang to the University of British Columbia, Canada for joint study. The translation, polishing, and publication of the book are partly supported by Talent Development Special Funding of Changsha City, Hunan Province, China. The authors also thank editors in Elsevier (Dr. Simon Tian, Glyn Jones, Naomi Robertson, and Susan Li) and the National Defense Industry Press, China (Ms. Junyin Xin and Xinjuan Zhang) for their help and advice.

The authors’ deepest thanks go to their parents and families, for always supporting and encouraging their research and study.

Yunze He

Bin Gao

Ali Sophian

Ruizhen Yang

October 2016

Abbreviations

ACFM Alternating current field measurement

ADC Analog-to-digital converter

AHT Abnormal heating thermography

APG Accelerated proximal gradient

BOB Bottom of bottom

BOT Bottom of top

BPT Burst phase thermography

BSS Blind source separation

CC Cross correlation

CCMF Cross correlation matched filtering

CFRP Carbon fiber reinforced plastic

CM Condition monitoring

DAC Different absolute contrast

DAQ Data acquisition

DF Diff frequency to zero

DFT Discrete Fourier transform

DL Digital level

DWT Discrete wavelength transform

EC Eddy current

ECLT Eddy current lock-in thermography

ECPPT Eddy current pulsed phase thermography

ECPT Eddy current pulsed thermography

ECSPPT Eddy current square pulsed phase thermography

ECSPT Eddy current square pulsed thermography

ECST Eddy current step thermography

ECT Eddy current testing or eddy current thermography

ECTRT Eddy current time-resolved thermography

ECVHT Eddy current volume heating thermography

EDM Electrical discharge machine

EIS Electrochemical impedance spectroscopy

EMAT Electromagnetic acoustic transducers

EMNDT Electromagnetic nondestructive testing

EMT Electromagnetic–thermal

EVD Eigenvalue decomposition

FD Fault diagnosis

FEM Finite element modeling

FFT Fast Fourier transform

FMT Frequency modulated thermography

FN False negative

FP False positive

GFRP Glass fiber reinforced plastic

GMR Giant magnetoresistive

IC Independent components

ICA Independent component analysis

IHT Inside heating thermography

IR Infrared

LHC Lateral heat conduction

LLT Laser lock-in thermography

LOI Lift-off point of intersection or lift-off invariance

LST Line scanning thermography

LT Lock-in thermography

MAE Magnetoacoustic emission

MBN Magnetic Barkhausen noise

MCMC Markov chain Monte Carlo

MFL Magnetic flux leakage

MP Matching pursuit

MT Modulated thermography

MUT Materials under test

MWT Microwave thermography

NC Normalized contrast

NDE Nondestructive evaluation

NDT Nondestructive testing

NDT&E Nondestructive testing and evaluation

NMF Nonnegative matrix factorization

OF Optical flow

O&M Operations and maintenance

OMP Orthogonal matching pursuit

PC Principal components

PCA Principal component analysis

PEC Pulsed eddy current

PI-TWR Pulsed inductive thermal-wave radar

POD Probability of detection

PPI Preprocessed image

PPMM Physics–mathematical pattern modeling and mining

PPT Pulsed phase thermography

PSO Particle swarm optimization

PT Penetration testing or pulsed thermography

PTFE Polytetrafluoroethylene

QEC Quantitative error curve

RBF Radial basis function

RCF Rolling contact fatigue

RFID Radio frequency identification

ROC Receiver operating characteristic

SeHT Selectively heating thermography

SFPM Spatial–frequency pattern mining

SGPCA Sparse greedy based principal components analysis

SHM Structural health monitoring

SHT Surface heating thermography

SNR Signal-to-noise ratio

SPPT Square pulse phase thermography

SPT Square pulse thermography

SSFPM Sparse spatial–frequency pattern mining

SSIM Structural similarity

SSTPM Sparse spatial–time pattern mining

ST Step thermography

STPM Spatial–time pattern mining

SVD Singular value decomposition

SVM Support vector machine

TBET Tone burst eddy current thermography

TCH Through coating heating

TCI Through coating imaging

TC-PCT Truncated correlation photothermal coherence tomography

TECT Transient eddy current thermography

TN True negative

TOB Top of bottom

TOF Thermooptical flow

TOT Top of top

TP True positive

TRT Time-resolved thermography

TSR Thermographic signal reconstruction

TWR Thermal-wave radar

TWRI Thermal-wave radar imaging

UT Ultrasonic Testing

VBSPCA Variational Bayesian sparse principal components analysis

VHLT Volume heating lock-in thermography

VHPPT Volume heating pulse phase thermography

VHPT Volume heating pulse thermography

VHST Volume heating step thermography

VHT Volume heating thermography

WT Wind turbine

Chapter 1

Nondestructive Testing and Transient Electromagnetic-Thermal NDT

Abstract

Nondestructive testing (NDT) is a wide group of analysis techniques used in science and industry to evaluate the properties of a material, component, or system without causing damage. NDT technologies have a crucial role in various industrial applications in the aerospace, automotive, manufacturing, petrochemical, transportation, civil, marine, and defense industries. There are currently hundreds of NDT methods being investigated or applied in the science and technique fields, which can be categorized as optical, acoustic, electromagnetic, infrared/thermal, or radiography NDT. Transient electromagnetic-thermal NDT techniques with a lot of advantages, such as high speed, great depth, high sensitivity, width spectrum, low cost, and easy quantification, are widely investigated in various industrial fields.

Keywords

nondestructive testing

eddy current

pulsed eddy current

infrared

transient electromagnetic-thermal NDT

Nondestructive testing (NDT) is a wide group of analysis techniques used in science and industry to evaluate the properties of a material, component, or system without causing damage [1]. The terms nondestructive examination, nondestructive inspection (NDI), and nondestructive evaluation (NDE) are also commonly used to describe this technology [2]. NDT technologies have a crucial role in various industrial applications in the aerospace, automotive, manufacturing, petrochemical, transportation, civil, marine, and defense industries [3]. There are five common NDT methods: ultrasonic testing (UT), magnetic particle testing (MPT), liquid penetrant testing, radiographic testing, and eddy current testing (ECT). There are currently hundreds of NDT methods being investigated or applied in the science and technique fields, which can be categorized as optical, acoustic, electromagnetic, infrared/thermal, or radiography NDT as shown in Fig. 1.1.

Figure 1.1   Categories of nondestructive testing (NDT) and evaluation (NDE) techniques.

Electromagnetic nondestructive testing (EMNDT) is based on the physics of electromagnetism and the interaction of electromagnetic fields/waves with the materials under testing (MUT). EMNDT covers many different methods including the eddy current technique, magnetic flux leakage, magnetic Barkhausen noise (MBN), magneto-acoustic emission (MAE), magnetic particle inspection (MPI), ground-penetrating radar (GPR), Terahertz testing, etc. EMNDT methods are widely used for the evaluation of metallic materials, composites, semiconductors, ceramics, concrete, and coatings. The eddy current technique is one of the most effective EMNDT methods, and is based on electromagnetic induction. As shown in Fig. 1.2, a primary magnetic field is established around an AC-driven coil. The field, in turn, induces eddy currents in the sample. The eddy currents then induce their own secondary magnetic field, which is in the opposite direction to the primary magnetic field, at all times in accordance with Lenz’s Law. The interaction between the magnetic field generated by the coil and the magnetic field generated by the eddy currents is then examined with either magnetic sensors or coils [eddy current testing (ECT) sensor]. The development of eddy current technology has led to the introduction of the pulsed eddy current (PEC). In contrast to conventional eddy current techniques where the excitation is generally limited to a single frequency component, PEC techniques excite the excitation or driver coil with a pulsed waveform, typically in the form of a rectangular or square waveform. The response from the sensor is then a transient voltage/current signal (U/I-t) that changes quickly with time and has an obvious start and end. Looking at the Fourier transform of a step function, it is clear that it contains a continuum of frequency components. As penetration depth is dependent on operating frequency, the PEC response signal will contain information from multiple depths, which is thus equivalent to multiple frequency eddy current. Besides PEC, other EMNDT techniques could also be configured as transient modes, such as pulsed magnetic flux leakage and pulsed remote field eddy current.

Figure 1.2   Interaction of induced eddy currents with a conductive sample.

Infrared/thermal NDT involves the use of temperature and heat flow measurements as a means to predict or diagnose failure. This may involve the use of contacting or noncontacting devices, or a combination of both. Contacting devices include thermometers of various types, thermocouples, thermopiles, and thermochromic coatings. Noncontacting devices include convection devices, optical pyrometers, infrared radiation thermometers, and infrared thermal imaging equipment [4]. Infrared (IR) thermography is the main kind of thermal NDT. IR thermography has great potential and many advantages including greater inspection speed, higher resolution and sensitivity, and detectability of inner defects due to heat conduction. The common types of active IR thermography are pulsed thermography (PT), step thermography (ST) [5], lock-in thermography (LT), which is also called modulated thermography (MT) [6], pulsed phase thermography (PPT) [7,8], and frequency-modulated thermography (FMT) [9,10]. PT, ST, and PPT can be classified into transient thermal/thermography NDT, where the responses from IR cameras are transient temperature signals (T-t) that change quickly with time and have an obvious start and end.

Besides conventional optical excitation, eddy current, laser, microwave, and ultrasound have been developed as excitation sources in thermography [11]. Therefore laser thermography [12], ultrasonic thermography, and eddy current thermography (ECT) [13,14] have been developed. ECT combines advantages of eddy current testing (such as noncontact, fast, and high resolution) with IR thermography. Just like general thermography, it can be applied in terms of eddy current pulsed thermography (ECPT) [15,16], eddy current step thermography, eddy current lock-in thermography [14], and eddy current pulsed phase thermography [17]. Among them, eddy current pulsed thermography, eddy current step thermography, eddy current pulsed phase thermography [17] could be classified into transient eddy current thermography (TECT) or transient electromagnetic-thermal NDT, where the responses from IR cameras after excitation by an electromagnetic field/wave are transient temperature signals (T-t) that change quickly with time and have an obvious start and end. As a popular transient electromagnetic-thermal NDT, ECPT involves the application of a high frequency electromagnetic wave (typically 50–500 kHz) to the material under inspection, at a high current for a short period [18]. Induced eddy currents are forced to divert when they encounter a discontinuity, which leads to changes in the density of the eddy current in that area. Areas with increased density of the eddy current exhibit higher levels of Joule heating, and thus defects can be obtained from sequenced thermograms during the heating and cooling periods. The typical configuration of an ECPT system is shown in Fig. 1.3. It consists of an induction heating system that induces eddy currents in the sample under inspection and generates heat, which is recorded by an IR camera to form digital data. These digital data can then be displayed on a monitor and stored on computers.

Figure 1.3   Basic configuration of an ECPT system [19].

Transient electromagnetic-thermal (EMT) NDT techniques have many advantages including high-speed, great depth, high sensitivity, width spectrum, low-cost, and easy quantification and are widely investigated in various industrial fields. In recent years, two typical transient electromagnetic (EMT) NDT methods, PEC and TECT have been the research hotspots. This book is intended to solve some problems in the physical principles, methods, techniques, and applications related to PEC and TECT.

Part I

Pulsed Eddy Current

Chapter 2: Magnetic Sensor Based Pulsed Eddy Current for Defect Detection and Characterization

Chapter 3: Hall-Device-Based PEC Features for Material Properties Evaluation and Defect Detection

Chapter 4: Coil-Based Rectangular PEC Sensors for Defect Classification

Chapter 2

Magnetic Sensor Based Pulsed Eddy Current for Defect Detection and Characterization

Abstract

This chapter discusses one class of nondestructive testing (NDT) pulsed eddy current (PEC), where a magnetic sensor is used for sensing the magnetic field in order to detect and characterize the defects in the specimen. The chapter starts with an introduction to PEC, followed with an example of a PEC system. The usages of signal processing and feature extraction using principal component analysis (PCA) and wavelet analysis on PEC signals are then presented. Finally, its applications in the inspection of ferromagnetic and nonferromagnetic materials by using both conventional features and the PCA-based features are presented with relevant experimental results.

Keywords

pulsed eddy current

feature

sensor

principal component analysis

wavelet

defect characterization

2.1. Introduction to PEC

Being a member of the NDT family, eddy current testing (ECT) has been widely used in numerous industrial sectors, particularly where metal inspection is required. However, in addition to merely detecting defects, there has been a growing demand for classification and quantification of defects to ensure the safety and integrity of structures. Therefore more accurate techniques for the characterization of defects are needed, and this still poses a major challenge [20].

Pulsed eddy current (PEC) techniques have been considered as a relatively new ECT technique [21]. These techniques, in contrast to conventional ECT methods that use a single frequency excitation, typically use a rectangular or square waveform for the field excitation. The pulsed excitation is comprised of a spectrum of frequencies, and is sensitive to different depths in the specimen due to the skin effect. This in turn enables simultaneous detection and characterization of flaws at both the surface and subsurface. Interpretation and classification techniques are therefore required to translate the transient response from PEC sensors into useful information, and characterize detected defects. Potential advantages of PEC techniques include greater penetration, their ability to locate discontinuities from time-of-flight determinations, and their simultaneous inspection of different depths. Lack of interpretation techniques is one of the main reasons why PEC sensing is not widely used by the NDT community [22]. This difficulty can be overcome by using advancements in computing power and signal processing techniques, and a number of different approaches for interpreting the eddy current sensor response using advanced signal processing techniques have been proposed [23,24].

As embodied in the skin depth effect, the depth of penetration of an eddy current technique depends, among others, on the excitation frequency. The lower the frequency, the deeper the penetration of eddy currents into the sample. The skin depth relationship that governs the penetration is defined as:

(2.1)

where δ is skin depth (m), μ is magnetic permeability (H/m), σ is electrical conductivity (S/m), and ω is the field frequency (Hz).

In contrast to conventional sinusoidal eddy current techniques, where the excitation is generally limited to a single frequency component, PEC techniques excite the excitation or driver coil with a pulsed waveform, typically in the form of rectangular or square waveform. The frequency components of such a pulse, or a step voltage, can be determined by using a Fourier series. If the step excitation is defined as:

(2.2)

Then, using the Fourier series theorem, the excitation can be written as:

(2.3)

The series above shows that the excitation has a wide range of spectral components. The wideband of frequencies has given the technique the potential to inspect a range of depths simultaneously and it will therefore be able to offer more information compared to the conventional approach. So theoretically it can achieve what multifrequency eddy current techniques can achieve but with a greatly reduced testing time.

To obtain PEC signals, one can use an induction (pickup) coil, a magnetic sensor, or an array of these. Induction coils detect time-varying magnetic fluxes based on Faraday’s induction theory, but are unable to detect static ones. The voltage detected by an eddy current coil is directly proportional to the frequency of the excitation (ω), number of wire turns (N), the incident magnetic field amplitude (B), and the effective area intersected by the flux lines (A) as described in Eq. (2.4):

(2.4)

This equation indicates that the sensitivity of an induction coil increases with frequency. However, according to the skin depth relationship in Eq. (2.1), the higher the frequency, the lower the standard depth of field penetration (δ). On the contrary, the sensitivity of solid-state magnetic sensor devices is constant for a relatively wide bandwidth. There are different types of solid-state magnetic sensors used in PEC, the common ones being Hall devices and giant magnetoresistive (GMR) sensors. A Hall sensor is a multiterminal, solid-state device that produces a voltage proportional to the product of its input current and the magnetic field it is placed in. The output voltage is due to the opposite sign charge separation under Lorentz force. The magnetoresistive effect is evident in thin-layer sandwiched structures of ferromagnetic and nonmagnetic materials. The resistance of such a structure varies with the magnetic field it senses and it is lowest when the external magnetic field is aligned with the magnetization vectors of the sandwiched layers. By applying a DC voltage along the length of the magnetoresistive layers, a current flows, and the resistance change can be detected by reading a voltage value across the device. Fig. 2.1A shows the typical PEC response from an induction coil, a Hall sensor, and a GMR sensor.

Figure 2.1   (A) Typical pulsed eddy current (PEC) signals from an induction coil, a Hall sensor and a giant magnetoresistive (GMR) sensor [25]; (B) typical PEC signals from a Hall sensor and its difference signal.

Similarly to single frequency ECT, PEC techniques use a null or reference signal that is typically obtained from a defect-free sample. Subsequent measurement signals, which are called base signals, will be subtracted from this null signal to obtain the differential signal during inspection. When no defects exist, the differential signal should ideally be a zero signal. When a discontinuity or defect occurs in the sample, then a difference will become evident. The shape and magnitude of the differential signal infers the change in the test sample. So, in opposition to conventional ECT, the analyses of PEC signals are carried out in the time domain, as well as in the frequency domain.

This chapter discusses PEC techniques that use a magnetic sensor. The advantages of magnetic sensors over a pickup coil include the sensitivity down to zero frequency, which is very useful for the inspection of defects that are buried deeply in the sample. Consequently, for subsurface testing, magnetic sensors have the upper hand compared to pickup