Electromagnetic Nondestructive Evaluation XI Studies in Applied Electromagnetics and Mechanics No 11

8/17/2019 Electromagnetic Nondestructive Evaluation XI Studies in Applied Electromagnetics and Mechanics No 11

1/317


2/317

ELECTROMAGNETIC NONDESTRUCTIVEEVALUATION (XI)


3/317


4/317

Electromagnetic Nondestructive

Evaluation (XI)

Edited by

Antonello TamburrinoUniversity of Cassino, Italy

Yevgen MelikhovCardiff University, UK

Zhenmao Chen Xian Jiaotong University, China

and

Lalita UdpaMichigan State University, USA

Amsterdam • Berlin • Oxford • Tokyo • Washington, DC


5/317

© 2008 The authors and IOS Press.

All rights reserved. No part of this book may be reproduced, stored in a retrieval system,

or transmitted, in any form or by any means, without prior written permission from the publisher.

ISBN 978-1-58603-896-0

Library of Congress Control Number: 2008932761

Publisher

IOS Press

Nieuwe Hemweg 6B

1013 BG Amsterdam

Netherlands

fax: +31 20 687 0019

e-mail: [email protected]

Distributor in the UK and Ireland Distributor in the USA and Canada

Gazelle Books Services Ltd. IOS Press, Inc.

White Cross Mills 4502 Rachael Manor Drive

Hightown Fairfax, VA 22032

Lancaster LA1 4XS USA

United Kingdom fax: +1 703 323 3668

fax: +44 1524 63232 e-mail: [email protected]

e-mail: [email protected]

LEGAL NOTICE

The publisher is not responsible for the use which might be made of the following information.

PRINTED IN THE NETHERLANDS


6/317

Preface

The 12th International Workshop on Electromagnetic Nondestructive Evaluation

(ENDE’07) was held from 19th–21st June 2007. The Workshop was hosted by the

Wolfson Centre for Magnetics at Cardiff University, Cardiff, Wales, UK with sponsor-

ship from Cedrat SA, Serco Assurance, Rolls-Royce plc, Welsh Assembly Government,

Computer Simulation Technology, The Japan Society of Applied Electromagnetics and

Mechanics, Engineering and Physical Sciences Research Council. The organizersgratefully acknowledge their support.

The aim of this annual workshop is to bring together engineers and scientists fromuniversities, research institutions and industry to discuss and exchange the latest ideasand findings in basic research and development as well as industrial applications of

Electromagnetic Nondestructive Evaluation.

After the introductory welcoming remarks from Dr. David Grant (Vice Chancellorof the Cardiff University), Prof. Hywel Thomas (Head of the School of Engineering,

Cardiff University) and Prof. David Jiles (Chairman of the Workshop), the technical

program of the Workshop commenced with a plenary talk “NDE Research Makes a

Difference” by Prof. Chris Scruby, Director, UK Research Centre in NDE, ImperialCollege London, U.K. Four distinguished invited speakers discussed the challenges and

achievements in various fields of ENDE. Prof. J. Bowler (Iowa State University, USA),gave a talk titled “Integral methods for calculating the interaction of eddy currents with

cracks”. Prof. K. Miya (Keio University, Japan), presented the second invited talk on“The Start of a New Field of Electromagnetic and Mechanical Maintenance Engineer-

ing”. Prof. G. Dobmann (Fraunhofer-Institute for Non-destructive Testing, Germany),

was invited to present “Industrial Applications of 3MA – Micromagnetic Multiparame-ter Microstructure and Stress Analysis”. Finally Prof. N. Takahashi (Okayama Univer-

sity, Japan) presented an invited talk on “3D Nonlinear Eddy Current Analysis of Elec-

tromagnetic Inspection of Defects in Steel”. A total of 75 technical papers were dividedinto 30 oral and 45 poster presentations. The oral presentations were organized into 7

sessions covering a variety of topics on both theoretical and experimental aspects of

NDE in eddy currents, magnetic measurements, magnetic flux leakage, Barkhausen

methods, new methods and inverse problems for crack detection. These sessions were

chaired by experts in the field including Profs. S. Udpa, L. Udpa, D. Jiles, C. Scruby,

P. Nagy, T. Moses, G. Dobmann, K. Miya, S. Takahashi, A. Tamburrino and others.During closing remarks it was announced that the next ENDE Workshop (ENDE2008)

will be held June 10–12, 2008 in Seoul, Korea. The ENDE Workshop 2009 will be

held in Dayton, Ohio, U.S.A. The conference concluded with remarks from the chair-man Prof. David Jiles.

A total of 73 participants from 16 countries were registered for the Workshop. Theshort versions of the papers were published in the Workshop digest and 39 reviewedfull papers were accepted for publication in this proceeding. The organizers would like

to thank all the participants for their contribution and all the referees for their role in

reviewing the full papers. Lastly, the editors gratefully acknowledge the help and hard

work of Ms. Linda Clifford in putting this volume together.

Electromagnetic Nondestructive Evaluation (XI)

A. Tamburrino et al. (Eds.)

IOS Press, 2008

© 2008 The authors and IOS Press. All rights reserved.

v


7/317

This page intentionally left blank


8/317

List of Referees

J. Aldrin Computational Tools, USA

Z. Badics Rhythmia Medical, Inc, USA

J. Bowler Iowa State University, USA N. Bowler Iowa State University, USA

T. Chady Technical University of Szczecin, Poland

M. Chan Michigan State University, USAZ. Chen Xi’an Jiaotong University, China

W. Cheng Tsurumi R&D Center, JAPEIC, Japan

G. Dobmann Fraunhofer-IZFP University, GermanyY. Gotoh Oita University, Japan

R. Grimberg National Institute of R&D for Technical Physics, Romania

X. Hao University of Birmingham, United KingdomH. Huang IIU Corporation, Japan

G. Hwang Shanghai Jiaotong University, China

L. Janousek University of Zilina, SlovakiaH. Kikuchi Iwate University, Japan

S. Kobayashi Iwate University, Japan

F. Kojima Kobe University, JapanJ. Lee Chosun University, Korea

D. Lesselier Supélec, France

L. Li Tsinghua University, China

Y. Melikhov Cardiff University, United KingdomV. Melapudi Michigan State University, USA

O. Mihalache JAEA, Japan

T. Moses Cardiff University, United Kingdom

G. Ni Zhejing University, ChinaJ. Pávó Budapest University of Technology and Economics, Hungary

P. Ramuhalli Michigan State University, USAG. Rubinacci Università degli Studi di Napoli “Federico II”, ItalyO. Stupakov Tohoku University, Japan

J. Taggart Serco Assurance, United Kingdom

A. Tamburrino University of Cassino, ItalyT. Takagi Tohoku University, Japan

N. Takahashi Okayama University, Japan

S. Takahashi Iwate University, Japan

T. Theodoulidis University of Western Macedonia, Greece

G.Y. Tian University of Newcastle upon Tyne, UK

I. Tomas Institute of Physics, Czech RepublicY. Tsuchida Oita University, Japan

L. Udpa Michigan State University, USAS. Udpa Michigan State University, USA

M. Vaidhianathasamy University of Newcastle Upon Tyne, United Kingdom

S. Ventre University of Cassino, Italy

vii


9/317

M. Versaci Università “Mediterranea” di Reggio Calabria, Italy

F. Villone University of Cassino, ItalyJ. Wilson University of Newcastle Upon Tyne, United Kingdom

L. Xin Michigan State University, USA

N. Yusa IIU Corporation, JapanZ. Zeng Michigan State University, USA

viii


10/317

Organizing Committees

International Committee

Chairman • S. Udpa, Michigan State University, USA

Members • J.R. Bowler, Iowa State University, U.S.A.

• N. Bowler, Iowa State University, U.S.A.

• Z. Chen, Xian Jiaotong University, China

•

G. Dobmann, Fraunhofer Institute for NDT, Germany• H.K. Jung, Seoul National University, South Korea

• F. Kojima, Kobe University, Japan

• D. Lesselier, DRE-LSS CNRS-SUPELEC-UPS, France

• K. Miya, Keio University, Japan

• G.Z. Ni, Zhejing University, China

• J. Pavo, Budapest University, Hungary

• G. Pichenot, CEA SACLAY, France

• A. Razek, LGEP CNRS-SUPELEC-UPS-UPMC, France

• G. Rubinacci, Universita di Napoli Federico II, Italy

• S.J. Song, Sung Kwan University, Korea

• T. Takagi, Tohoku University, Japan

• S. Takahashi, Iwate University, Japan

• A. Tamburrino, Universita degli Studi di Cassino, Italy

• L. Udpa, Michigan State University, U.S.A.

Organizing Committee • Tony Dunhill, Rolls Royce Plc

• Keith Jenkins, Cogent Power

• David Jiles, Cardiff University (Chairman)

• Chester Lo, Iowa State University• Tony Moses, Cardiff University

• Ian Nicholson, TWI Ltd

• Allan Rogerson, Serco Assurance Plc

• Chris Scruby, Imperial College London

• Gui Yun Tiann, University of Huddersfield

Local Committee • Phil Anderson

• Jeremy Hall

• Eugene Melikhov• John Snyder

• Paul Williams

• Stan Zurek

ix


11/317



12/317

List of Participants

Mr. Kavoos AbbasiTohoku University, [email protected]

Prof. Purnachandra Rao BhagiIndira Ghandi Centre for AtomicResearch, [email protected]

Prof. John BowlerIowa State University, USA [email protected]

Dr. Nicola BowlerIowa State University, [email protected]

Dr. John BurdJB Consulting Ltd, UK [email protected]

Hee Jun ChenSungkyunkwan University, [email protected]

Dr. Zhenmao ChenXian Jiaotong University, China

[email protected]

Prof. Gerd DobmannFraunhofer-Institut IZFP, [email protected]

Mr. David EdgarUniversity of Nottingham & Qinetiq, [email protected]

Dr. Christiaan EgginkShell Global Solutions Int., [email protected]

Dr. Dagmar FaktorovaUniversity of Zilina, Slovak [email protected]

Mr. Fabrice FoucherCEDRAT, [email protected]

Dr. Raimond Grimberg National Inst. of R & D for TechnicalPhysics, [email protected]

Dr. Xinjiang HaoBirmingham University, [email protected]

Dr. Jeremy HallWolfson Centre for Magnetics, [email protected]

Mr. Jiseong HwangChosun University, Republic of [email protected]

Dr. Richard IrelandQinetiQ, UK

[email protected]

Dr. Ladislav JanousekUniversity of Zilina, Slovak Republic [email protected]

Dr. Mohan JayawardeneCST, UK/[email protected]

Dr. Steve JenkinsCurrent Enterprises, [email protected]

xi


13/317

Prof. David JilesWolfson Centre for Magnetics, UK [email protected]

Dr. Tonphong KaewkongkaChulalongkorn University, [email protected]

Mr. Yuchiro KaiOita University, Japan [email protected]

Dr. Hiroaki Kikuchi NDE & SRC Ueda, [email protected]

Mr. Jeremy KnoppAir Force Research Laboratory, USA [email protected]

Dr. Satoru Kobayashi NDE & Science Research Centre, [email protected]

Prof. Fumio KojimaKobe University, [email protected]

Dr. Yann Le BihanLGEP- SUPELEC, France [email protected]

Prof. Jinyi Lee

Chosun University, Republic of Korea [email protected]

Dr. Yohan Le DiraisonENS-CACHAN, France [email protected]

Dr. Dominique LesselierSUPELEC, [email protected]

Prof. Li LumingTsinghua University, [email protected]

Dr. Dmitriy MakhnovskiyUniversity of Plymouth, [email protected]

Dr. Eugene MelikhovWolfson Centre for Magnetics, [email protected]

Prof. Kenzo MiyaIIU, [email protected]

Prof. Tony MosesWolfson Centre for Magnetics, UK

[email protected]+D70

Dr. Shoichiro NagataUniversity of Miyazaki, [email protected]

Prof. Peter BN NagyUniversity of Cincinnati, [email protected]

Severine PaillardCEA, [email protected]

Dr. Manuele PapaisUniversity of Udine, [email protected]

Dr. Grégoire PichenotCEA, [email protected]

Mr. Grzegorz PsujSzczecin University of Technology,[email protected]

Mr. Brian RadtkeIowa State University, USA

[email protected]

Miss Alicia Romero RamirezSwansea University, [email protected]

xii


14/317

Dr. Cyril RavatENS-CACHAN, [email protected]

Dr. Alan RogersonSerco Assurance, UK

Dr. Adriana Savin National Inst. of R & D for TechnicalPhysics, [email protected]

Prof. Chris ScrubyImperial College, [email protected]

Dr. Gongtian ShenCSEI, [email protected]

Prof. Young-Kil ShinKunsan National University, South Korea [email protected]

Dr. Anastassios SkarlatosCEA, Francegregoire.pichenot.cea.fr

Dr. Sung-Jin SongSungkyunkwan University, [email protected]

Mr. Giuseppe Sposito

Imperial College, [email protected]

Dr. Vladamir SyaskoConstanta, [email protected]

Dr. John TaggartSerco Assurance, UK

[email protected]

Prof. Toshiyuki TakagiTohoku University, [email protected]

Prof. Norio TakahashiOkayma University, [email protected]

Prof. Seiki TakahashiIwate University, [email protected]

Prof. Antonello TamburrinoDaeimi University of Cassino, [email protected]

Dr. Alan TassinENS-CACHAN, [email protected]

Dr. Theodoros TheodoulidisUniversity of West Macedonia, [email protected]

Prof. Gui Yun Tian Newcastle University, [email protected]

Mr. Yuji TsuchidaOita University, [email protected]

Dr. Lalita UdpaMichigan State University, [email protected]

Prof. Satish Udpa

Michigan State University, [email protected]

Dr. Tetsuya UchimotoTohoku University, [email protected]

Dr. Moorthy Vaidhianathasamy Newcastle University, [email protected]

Dr. Haitao Wang Nanjing University of Aeronautics &Astronautics, [email protected]

xiii


15/317

Dr. Casper WassinkApplus RTD, The [email protected]

Mr. Chris WardUniversity of Nottingham/RWE NPower,UK [email protected]

Dr. Paul WilliamsWolfson Centre for Magnetics, [email protected]

Mr. John Wilson Newcastle University, UK [email protected]

Mr. Chen XingTsinghua University, [email protected]

Dr. En Tao Yao Nanjing University of Aeronautics &Astronautics, China [email protected]

xiv


16/317

Contents

Preface v

List of Referees vii

Organizing Committees ix

List of Participants xi

Invited Speakers

A Start of New Field of Electromagnetic and Mechanical Maintenance Engineering 3 Kenzo Miya

NDE Research Makes a Difference 10

C.B. Scruby

Industrial Applications of 3MA – Micromagnetic Multiparameter Microstructureand Stress Analysis 18

Gerd Dobmann, Iris Altpeter, Bernd Wolter and Rolf Kern

3D Nonlinear Finite Element Analysis of Electromagnetic Inspection of Defects

in Steel 26 N. Takahashi and Y. Gotoh

Magnetic Materials

Evaluation of Irradiation Embrittlement in Fe-Cu-Ni-Mn Model Alloys byMeasurements of Magnetic Minor Hysteresis Loops 37

Satoru Kobayashi, Hiroaki Kikuchi, Seiki Takahashi, Katsuyuki Ara and

Yasuhiro Kamada

Analysis of Barkhausen Noise Characteristics and Mechanical Properties onCold Rolled Low Carbon Steel 42

Hiroaki Kikuchi, Tomoki Koshika, Tong Liu, Yasuhiro Kamada,

Katsuyuki Ara, Satoru Kobayashi and Seiki Takahashi

A Bridge Between NDE and Charpy Impact Testing 46

Seiki Takahashi and Satoru Kobayashi

ND-Materials Characterization of Neutron-Induced Embrittlement in German

Nuclear Reactor Pressure Vessel Material by Micromagnetic NDT Techniques 54

Gerd Dobmann, Iris Altpeter, Melanie Kopp, Magdalena Rabung and

Gerhard Hübschen

Evaluation of Chill Contents in Flake Graphite Cast Irons Using AC

Magnetization Method 62

Tetsuya Uchimoto, Jun Matsukawa, Toshihiko Abe, Toshiyuki Takagi,Takeshi Sato, Hiroyuki Ike, Takahito Takagawa and Noritaka Horikawa

xv


17/317

Characterisation of Microstructures in Heat Treated Maraging Steel Using Eddy

Current and Barkhausen Emission Techniques 70 K.V. Rajkumar, B.P.C. Rao, B. Sasi, S. Vaidyanathan, T. Jayakumar and

Baldev Raj

Novel Acoustic Barkhausen Noise Transducer and Its Comparison withElectromagnetic Acoustic Transducer 78

John Wilson, Gui Yun Tian, Rachel S. Edwards and Steve Dixon

Modelling and Measurement of Decarburisation of Steels Using aMulti-Frequency Electromagnetic Sensor 86

X.J. Hao, W. Yin, M. Strangwood, A.J. Peyton, P.F. Morris and C.L. Davis

Assessment of Grinding Damage on Gear Teeth Using Magnetic Barkhausen Noise Measurements 90

Moorthy Vaidhianathasamy, Brian Andrew Shaw, Will Bennett and Peter Hopkins

Evaluation of Contact Fatigue Damage on Gears Using the Magnetic Barkhausen

Noise Technique 98

Moorthy Vaidhianathasamy, Brian Andrew Shaw, Will Bennett and

Peter Hopkins

Inverse Problems

3D Reconstruction of Flaws in Metallic Materials by Eddy Currents Inspections 109 Alessandro Pirani, Marco Ricci, Antonello Tamburrino and

Salvatore Ventre

Multi-Frequency Eddy Current Imaging for the Detection of Buried Cracks in

Aeronautical Structures 117

Yohan Le Diraison and Pierre-Yves Joubert

A Comparative Study of Source Separation Techniques for the Detection

of Buried Defects in the EC NDE of Aeronautical Multi-Layered Lap-Joints 125

Alan Tassin, Yohan Le Diraison and Pierre-Yves Joubert

Fundamental Feature Extraction Methods for the Analysis of Eddy Current Data 133 Jeremy S. Knopp and John C. Aldrin

Inversion of Potential Drop Data for the Reconstruction of Crack Depth Profiles 141Giuseppe Sposito, Peter Cawley and Peter B. Nagy

Automatic Classification of Defects with the Review of an Appropriate Feature

Extraction 148 Alicia Romero Ramirez, Neil Pearson and J.S.D. Mason

Microwave Nondestructive Detection of Longitudinal Cracks in Pipe with

U-Bend and Prediction of Its Location by Signal Processing 154 Kavoos Abbasi, Satoshi Ito and Hidetoshi Hashizume

Some Experiences with Microwave Investigation of Material Defects 162

Dagmar Faktorová

xvi


18/317

Modeling

Effect of Crack Closure on Quantitative ECT Inspection of Closed FatigueCracks 171

Zhenmao Chen, Noritaka Yusa, Kenzo Miya and Hideaki Tokuma

Toward the Reconstruction of Stress Corrosion Cracks Using Benchmark EddyCurrents Signals 179

Maxim Morozov, Guglielmo Rubinacci, Antonello Tamburrino,

Salvatore Ventre and Fabio Villone

Evaluation of Subsurface Cracks in Riveted Aluminium Joints Using Industrial

Eddy Current Instrumentation 187

Maxim Morozov, Guglielmo Rubinacci, Antonello Tamburrino andSalvatore Ventre

Design of a System for the Long Defects Detection with Advanced Methods for

Eddy-Currents Analysis 195 E. Cardelli, A. Faba, A. Formisano, R. Martone, F.C. Morabito, M. Papais,

R. Specogna, A. Tamburrino, F. Trevisan, S. Ventre and M. Versaci

Theory of Four-Point Alternating Current Potential Drop Measurementson a Layered Conductive Half-Space 203

Nicola Bowler and John R. Bowler

Integration of Tilted Coil Models in a Volume Integral Method for Realistic

Simulations of Eddy Current Inspections 211Theodoros Theodoulidis and Gregoire Pichenot

Modeling of Flawed Riveted Structures for EC Inspection in Aeronautics 217

S. Paillard, G. Pichenot, Y. Choua, Y. Le Bihan, M. Lambert,

H. Voillaume and N. Dominguez

Numerical Modeling of Eddy Current Nondestructive Evaluation of

Ferromagnetic Tubes via an Integral Equation Approach 225

Anastassios Skarlatos, Grégoire Pichenot, Dominique Lesselier, Marc Lambert and Bernard Duchêne

Design of Reflection Type Pulsed Eddy Current Nondestructive Testing 231

Young-Kil Shin, Dong-Myung Choi and Hee-Sung Jung

Applications

Noninvasive Characterization of Bjork-Shiley Convexo-Concave Prosthetic

Heart Valves Using an Electromagnetic Method 241 Raimond Grimberg, Shiu C. Chan, Adriana Savin, Lalita Udpa and

Satish S. Udpa

Remote Field Eddy Current Control Using Rotating Magnetic Field Transducer.Application to Pressure Tubes Examination 249

Adriana Savin, Lalita Udpa, Rozina Steigmann, Alina Bruma,

Raimond Grimberg and Satish S. Udpa

xvii


19/317

Lifetime Prediction of Pressure Tubes in PHWR Nuclear Power Plants Using

Eddy Current Data 257 Raimond Grimberg, Adriana Savin, Rozina Steigmann, Aurel Andreescu,

Nicoleta Iftimie and Marius Mihai Cazacu

Electromagnetic Non-Destructive Evaluation of Reinforced Concrete Rebars 263 Maxim Morozov, Guglielmo Rubinacci, Antonello Tamburrino

and Salvatore Ventre

Advanced Probe with Array of Pick-Up Coils for Improved Crack Evaluationin Eddy-Current Non-Destructive Testing 271

Ladislav Janousek, Klara Capova, Noritaka Yusa and Kenzo Miya

Observation of Stress Loaded Ferromagnetic Samples Using Remanent FluxLeakage Method 276

Tomasz Chady, Grzegorz Psuj and Ryszard SikoraEvaluation of Complex Multifrequency Eddy Current Transducer Designedfor Precise Flaw Depth Measurements 283

Tomasz Chady, Piotr Baniukiewicz, Ryszard Sikora and Grzegorz Psuj

Comparative Study of Coil Arrangements for the EC Testing of Small SurfaceBreaking Defects 288

Cyril Ravat, Yann Le Bihan, Pierre-Yves Joubert and Claude Marchand

Author Index 295

xviii


20/317

Invited Speakers


21/317



22/317

A Start Of New Field Of ElectromagneticAnd Mechanical Maintenance Engineering

Kenzo MIYA1

IIU, Corp., 2-7-17-7F, Ikenohata, Taitoku, Tokyo, 110-0008, Japan

Abstract. Maintenance to ensure the integrity of structures is one of the mostimportant issues in modern industry, but more so in the nuclear power industry.This is primarily because complete prevention of all degradation of industrialmaterials is not possible, at least not in a realistic way. Consequently it isextremely important to detect degradation before machines start to loose theirability to function which may lead to harmful failures. Usually there is some kindof significant interval between the start of material degradation and failure; inother words, degradation usually progresses slowly with time. The problem iswhether we can detect precursors of failure in the interval or not. If it is possible,we can have economic as well as safety related benefits because we are able tostop the machine before a functional failure and prevent an accident initiated bythe failure of the machine. The benefits are not restricted to the nuclear powerindustry because it can be applied to any other heavy industry. There are manyreasons why many conferences, such as ENDE, have been established to discuss

and promote the progress of nondestructive inspection techniques. Whereasnondestructive inspection plays an important role in maintaining structuralintegrity and the performance of nondestructive inspection techniques should beenhanced in that sense, we need to regard it as one of several componentscomposing “maintenance engineering.” Moreover, although nondestructiveinspection is a proactive measure of maintenance, it is not a predictive tool. In fact,it is effective for detecting existing defects and does not say anything about thetemporal evolution of defects. On the other hand, a condition monitoring systemcan offer significantly more useful information as mentioned above. In this paper Iwould like to introduce the concept of maintenology as a new science andtechnology in contrast to conventional maintenance engineering and to presentseveral important results on electromagnetic maintenance for nuclear power plants

as condition monitoring techniques (CMT). In particular, the introduction ofelectromagnetic maintenance is expected to play a very promising role inabnormality predictions of many dynamic machines that are required to beinspected regularly by law. Application of the technique would changeconventional wisdom in thinking that machines should be taken apart andinspected regularly based on regulations. In many cases, this TBM (time basedmaintenance) is not too conservative in achieving an optimal maintenanceapproach.

Introduction

The nuclear energy renaissance is occurring not only in advanced countries but also indeveloping ones like China, India and Brazil, due to demands for renewable energysources and to protect the earth against abnormal climate caused by the green houseeffect. Nuclear power plants are one of the greenest sources of energy, and operated



IOS Press, 2008


doi:10.3233/978-1-58603-896-0-3

3


23/317

safely, can provide energy reliably to meet base loads. Thus, there is considerableinterest in the safe and economic operation of nuclear power plants (NPPs).

Maintenance engineering is extremely important to achieve these two goals. Twoapproaches are usually pursued: time-based maintenance (TBM) and condition-based

maintenance (CBM). Optimal maintenance may be achieved by planning the best mixof the two approaches. The development of CBM techniques is greatly needed toenhance the safety level of NPPs.

In this paper we explain the possibility of using electromagnetic methods and theapplication of the technology to maintenance operations at NPP. The essential part ofelectromagnetic maintenance technology is the utilization of u x B electromagneticmotive force (u: vectored velocity, B: magnetic field). In principle, diagnostics basedon the technique can be applied to rotating machines at present although applicationcould be extended to static components in future. Theoretical issues underlying theapproach and experimental work associated with electromagnetic maintenance will be

introduced first. Numerical simulations and experiments will then be presented.

1. Construction of maintenance engineering as a new field

Maintenance issues comprise two different aspects, one is related to human behaviourand another relates to technical matters. Complicated human behaviour relating tomaintenance of machines is considered to consist of three principles, as shown inFigure 1: the selection principle, connection principle and projection principle. Theirrelationship can be understood well if we imagine a process of making a sentencewhere proper words are selected first (selection principle), words are properly arrangedto meet grammatical requirements (connection principle), and the intended meaning isrealized by the sentence. The grammatically correct arrangement of words is called a projection principle. Concept of such a view is translated into maintenance activitieslike words corresponding to various kinds of technology supported by natural science,sociology, codes and standards, the connecting principle corresponding to maintenancescheme, and projection principle corresponding to the selection of necessary techniquesfrom existing technological systems. These basic principles should be favourablyapplied to planning the maintenance scheme.

In Fig. 2, a flow chart of maintenance activities is shown schematically. They startfrom the selection of the component to be maintained followed by the selection of themaintenance method, i.e. TBM or CBM, and then finally we proceed to the inspection process. This, in general, is the most universal process.

In Fig. 3, concepts of maintenology are newly defined and explained to introduce anew approach to conventional maintenance engineering. The new approach consists ofthree principles depicted in Fig. 1; the basic form of maintenance action is shown inFig.2. This recognition may contribute to the construction of revised maintenanceengineering by applying theoretical aspects.

2. Electromagnetic maintenance engineering

Principles of electromagnetic maintenance may be called the EM method. It isassociated with the following processes:

1) Creation of a static magnetic field in a region of rotating components

K. Miya / A Start of New Field of Electromagnetic and Mechanical Maintenance Engineering4


24/317


25/317

-5.00E-07

-4.00E-07

-3.00E-07

-2.00E-07

-1.00E-07

0.00E+00

1.00E-07

2.00E-07

3.00E-07

4.00E-07

5.00E-07

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

Time(s)

U ( v )

-5.00E-08

-4.00E-08

-3.00E-08

-2.00E-08

-1.00E-08

0.00E+00

1.00E-08

2.00E-08

3.00E-08

4.00E-08

5.00E-08U: 5mmU: 10mm

U: 15mm

-5.00E-07

-4.00E-07

-3.00E-07

-2.00E-07

-1.00E-07

0.00E+00

1.00E-07

2.00E-07

3.00E-07

4.00E-07

5.00E-07

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

Time(s)

U ( v )

-5.00E-08

-4.00E-08

-3.00E-08

-2.00E-08

-1.00E-08

0.00E+00

1.00E-08

2.00E-08

3.00E-08

4.00E-08

5.00E-08

U: 5mm

U: 10mm

U: 15mm

Left scale for 5mm Left scale for 5mm

Without a holder With a holder

V)

V)

-5.00E-07

-4.00E-07

-3.00E-07

-2.00E-07

-1.00E-07

0.00E+00

1.00E-07

2.00E-07

3.00E-07

4.00E-07

5.00E-07

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

Time(s)

U ( v )

-5.00E-08

-4.00E-08

-3.00E-08

-2.00E-08

-1.00E-08

0.00E+00

1.00E-08

2.00E-08

3.00E-08

4.00E-08

5.00E-08U: 5mmU: 10mm

U: 15mm

-5.00E-07

-4.00E-07

-3.00E-07

-2.00E-07

-1.00E-07

0.00E+00

1.00E-07

2.00E-07

3.00E-07

4.00E-07

5.00E-07

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

Time(s)

U ( v )

-5.00E-08

-4.00E-08

-3.00E-08

-2.00E-08

-1.00E-08

0.00E+00

1.00E-08

2.00E-08

3.00E-08

4.00E-08

5.00E-08

U: 5mm

U: 10mm

U: 15mm

Left scale for 5mm Left scale for 5mm

Without a holder With a holder

V)

V)

Figure 7. Signals from sensor. (1. Decrease of signal amplitude by 1/10 when distance becomes longer, from5 to 10 mm. 2. Effect of a holder on signal is not significant, about 10% decrease)

-4.00E-08

-3.00E-08

-2.00E-08

-1.00E-08

0.00E+00

1.00E-08

2.00E-08

3.00E-08

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02

Time(s)

U ( v )

U: crack free

U: V carck

U: H crack

Figure 8. Ball bearings with a defect Figure 9. Signals from defect

In Figure 4, four blades rotate around an axis which is not shown in the static

magnetic field created by a permanent magnet (not shown here). Eddy currentdistribution induced in blades is shown in the figure together with a pick-up coil. If acrack is present or the blade is greatly deformed, we measure some changes in signalswhen abnormalities are present.

In Figure 5 the finite element mesh for the impeller of a pump is shown. The finiteelement method is employed to evaluate eddy currents in the impeller.

In Figure 6 ball bearings witness an orbital motion. The rotational eddy current isnot significant, but the orbital motion produces a considerable amount of eddy currentsin the balls. Thus, defects in a ball bearing may be easily detected by the method. Ifworn particles are present between bearings and steel plates, the speed of the orbital

motion will change and this change can be detected using this method. Simulationresults are shown in Fig. 7 for two cases: one is without a holder and the other is with aholder. Eddy currents in these cases are not different from each other indicating smalleffects of the holder. In the figure, “U” is a measure of distance between a bar magnetand surface of a casing. The measured signals largely depend on the distance, which iseasily estimated by the decrease in the field due to the distance. The number of signaloscillations corresponds to the number of blades.

In Fig. 8, the possibility of the detection of two notches on the surface of balls wastested by numerical simulation. The notches are in horizontal and vertical direction asshown in the figure. Results are shown in Fig.9. In the case of the vertical notch,

differences from the case without a notch are evident but are difficult to find. Thisresult can be explained from the magnitude of eddy currents due to positionaldifferences of the two notches. Therefore, it becomes essentially important to optimizethe relative position of a magnet and rotating blade.



26/317

3. Experimental results

The numerical results above were introduced on the basis of the explained concept toapply electromagnetic phenomena to defect inspection of rotating machines. The

numerical results showed that there is a valid basis for the concept. The next step is toshow experimental verification of the concept. Up to now, our research group hasconducted many experiments with a small fans and pumps. We have also measuredelectromagnetic signals generated by rotating parts of real machines. When we can seta sensor and magnet close to the rotating parts, we can successfully measure a signal.However, when access is limited, it is difficult to obtain a useful signal.

In Fig.10, measurement results are shown for the case of a fan. The signals wereobtained as a function of distance between the sensor and the rotating part. When thedistance is 5 mm, the signal is the largest and in the case of 20mm distance, the signalis the smallest. These results are from the set-up without fan casing. But, the 20mm

distance result indicates that the measurement with a set-up that includes fan casing is possible. Results with the casing are not shown here, but a sufficient number of signalswere obtained. Cyclic signals are obtained for 5 blades. There is the possibility that thesmaller signal on the inside is due to a weight.

The major purpose of this study is to verify the possible application ofelectromagnetic phenomena to proactive detection of various types of defects.Extensive studies must be conducted to find useful relations between signal changesand the condition of defects. Useful information must be provided to make engineeringdecisions on whether operation of the machine should be stopped for repair orcontinued with careful monitoring. With this in mind, we carried out an experiment

with a notched blade as shown in Fig. 11. Two curves are shown corresponding to twotypes of bar and U-magnets. It is possible to see the movement of the 5 blades from both results. One of the blades was machined with the notch and signals reflecting anexistence of the notch are observed in both results. Through a two-way excitationmethod, a crack of 1/3 the width was easily detected. Significant differences in signalswere seen for cracks 1/10 the width.

In Fig.12, three pictures of a casing of the real pump, location of a sensor and a pick-up coil, and an impeller are shown. The casing is cast iron and shows magnetic property. The rod type permanent magnet is attracted to the casing.

-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

-2.00E-02 -1.50E-02 -1.00E-02 -5.00E-03 0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02

5mm

10mm

15mm

20mm

-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

-2.00E-02 -1.50E-02 -1.00E-02 -5.00E-03 0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02

5mm

10mm

15mm

20mm

Figure 10. Fan experiment

K. Miya / A Start of New Field of Electromagnetic and Mechanical Maintenance Engineering 7


27/317

Figure 11. Detection of a defective blade

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-2.00E-02 -1.50E-02 -1.00E-02 -5.00E-03 0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02

-1.5

-1

-0.5

0

0.5

1

1.5

-2.00E-02 -1.50E-02 -1.00E-02 -5.00E-03 0.00E+00 5.00E-03 1.00E-02 1.50E-02 2.00E-02

U

In Fig. 13, measurement results are shown. A careful examination shows that thenumber of impellers is 5 and the rotation cycle is 0.02 seconds since a motor operatesat 50 Hz in the Kanto area in Japan. In addition to this, we can recognize a small

difference in the amplitude of the measured voltage corresponding to the difference insize of the 5 blades. Since the signal amplitudes are different this suggests that themethod can be applied to test for the dimension of axis and impellers, as well as theircondition of motion. The wave form is not sinusoidal and shows a small tooth. Thereason for the tooth is not clear until we see the inside of the pump after opening thecasing. This small observation indicates that the detailed structure inside a rotatingmachine will be made clear after we accumulate knowledge on dimensionalinformation and defects signals in the future.

Figure 12. Casing and impellers (real pump)



28/317

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.010 -0.005 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040

s

v

Rotation cycles are 0.02s

Blade intervals are 0.004s

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.010 -0.005 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040

s

v

Rotation cycles are 0.02s

Blade intervals are 0.004s

Figure 13. Signal with rod type permanent magnet sensor

3. Conclusion

uld like to note that there are many challenging problems in theThe author wo

diagnostics of rotating machines and that there is a strong need for developing methodsof detecting abnormality precursors as a prognosticator of functional failure. There arewell-known methods of condition monitoring. These include vibration monitoringmethods, oil analysis, thermography, etc. These methods will be useful if they are used properly; however, there are several important problems to be solved in identifying theroot causes of malfunctions occurring in a machine. Electromagnetic methods haveshown their superiority over conventional methods in being able to locate defects andin identifying conditions at specific locations in rotating parts. Extensive studies arerequired to establish the relationship between the measured signal and the nature ofdefects. Efforts are urgently needed to translate R&D efforts to industrial applications.

K. Miya / A Start of New Field of Electromagnetic and Mechanical Maintenance Engineering 9


29/317

NDE Research Makes a DifferenceC B SCRUBY

UK Research Centre in NDE, Imperial College London, SW7 2AZ UK

Abstract. This paper discusses the development of research in NDE and itsimpact on industry. Examples will be given of past research projects that have been translated into solutions to industrial inspection problems, and present daychallenges to industry and the NDE research community. Recurring themes

include the need for quantification, and physical models to give scientificunderstanding and hence improved confidence for product quality and safety-critical application. Timely technology transfer is a continuing challenge andlessons from past experience will be discussed. Finally, the author will discussfuture research strategy, including opportunities for interdisciplinary collaborationin order to integrate NDE more effectively into the engineering life cycle.

Keywords. Non-destructive evaluation, NDE, electromagnetic.

Introduction

There has been steady growth in non-destructive evaluation (NDE) research since the1950s, reflecting the increasing demand for greater safety and environmental protection by the public. NDE research has always been particularly challenging. Firstly, the NDE field makes use of multiple technologies & disciplines, including magnetic,electrical, ultrasonic, radiographic, optical, and thermographic methods. Secondly, NDE is used for a wide range of materials (from metals to plastics, ceramics andcomposites) & applications (manufacturing process control, in-service inspection,defect sizing, corrosion measurement, and degradation monitoring). NDE is used inmost sectors of industry: aerospace, power, oil & gas, transportation, infrastructure,

built environment, manufacturing, nuclear, process, defence, electronics, packaging,etc., which have vastly different products and assets to be inspected – everything fromelectronic devices and foodstuffs through to entire rail networks and huge refineries.To enhance the challenge there is a complex supply chain to take new technologiesfrom university research through to routine use, involving research and technologyorganizations (RTOs), service companies, equipment suppliers, certification andstandards authorities, trainers and consultants (Figure 1).

However, on the positive side, there is much commonality and overlap at theresearch stage in terms of techniques and generic applications. Many universities andRTOs have begun to recognize the benefits of an interdisciplinary approach to NDEresearch. Until the late 1980s much research was funded by government researchlaboratories and state-owned industries, especially in the nuclear and defence sectors.There were then changes due to privatisation that made a major impact on the fundingof NDE research. Although the value of an NDE solution to an individual company orsingle industry sector may be too small to justify large investment, there are significantopportunities for collaborative funding to increase benefit, share risk and reduce cost.



IOS Press, 2008


doi:10.3233/978-1-58603-896-0-10

10


30/317


31/317

-30

100

230

1 121Technology development

Industrial pull

Scientific push

Pioneers

Research

Product

development

Commercialisation

Trials &

evaluation

Standardisation

Acceptance

Further research &

development

Figure 2. Research and innovation process for NDE

Within the defence and aerospace sectors of industry the innovation process isoften described in terms of technology readiness levels (TRLs), as shown in Figure 3.Here the inventive phase is described as technology assessment. This may take place

in an industrial R&D laboratory or an university. Production implementation is theremit of the end-user company. The challenge is the intermediate steps from 4 to 7;there is a need not only for organizations to undertake work at TRLs 5 and 6, but alsoto interact effectively with the organizations who carry out TRLs 1-4 on the one handand TRLs 7-9 on the other. Experience has shown that the most risky part of the process is this translation from research to production. This may be because researchand production are carried out by organizations with different objectives and cultures.The NDE industry has a relatively complicated supply chain for new technology(Figure 1) because the technology is used in both production and operation, andoperators often sub-contract to service companies. There need to be routes to market

for NDE services as well as products. In either case, improved ways are needed tofacilitate the movement of new technology from left to right along the chain.

2. Case Studies of NDE Research and Innovation

There are many potential examples within the field of electromagnetic NDE, but only asmall number can be examined here with a view to understanding the process wherebya research idea is converted into an industrial technology. The examples are selectedfrom the knowledge and experience of the author, and selection or omission is not avalue judgment of the product in question.

Alternating Current Field Measurement (ACFM) [1] is a current perturbationtechnique related to, but different from eddy current testing methods. Its developmentfollowed closely the pattern described by Figure 2. Thus there was scientific push inthe form of university research at University College London (UCL) including

C.B. Scruby / NDE Research Makes a Difference12


32/317

1. Basic principles observed and reported

2. Technology concept or application vetted or demonstrated

3. Experimental proof of concept completed

4. Inspection validated in laboratory environment

5. Inspection validated in relevant environment (prototype equipment)

6. Inspection optimised for capability and speed

7. System prototype demonstration in an operational environment

8. Production system qualified through test and demonstration

9. Track record established through successful applications

Technology

Assessment

and Proving

Pre-Production

Production

Implementation

Proven

Method

D e v e l o p m e n t

Idea1. Basic principles observed and reported1. Basic principles observed and reported

2. Technology concept or application vetted or demonstrated2. Technology concept or application vetted or demonstrated

3. Experimental proof of concept completed3. Experimental proof of concept completed

4. Inspection validated in laboratory environment4. Inspection validated in laboratory environment

5. Inspection validated in relevant environment (prototype equipment)5. Inspection validated in relevant environment (prototype equipment)

6. Inspection optimised for capability and speed6. Inspection optimised for capability and speed

7. System prototype demonstration in an operational environment7. System prototype demonstration in an operational environment

8. Production system qualified through test and demonstration8. Production system qualified through test and demonstration

9. Track record established through successful applications9. Track record established through successful applications

Technology

Assessment

and Proving

Pre-Production

Production

Implementation

Proven

Method

D e v e l o p m e n t

Idea

Figure 3. Innovation process described in terms of Technology Readiness Levels

substantial modelling from 1987-9. This was followed in 1991 by application trials onmodel offshore nodes supported in part by the oil & gas industry, demonstratinggrowing industrial pull. In 1993 ACFM was licensed to TSC (a small university spin-out company), with (1993) a royalty agreement to cover the UCL developments neededto commercialize the technology. 1994-8 saw the development of an applicationsmarket, mainly in oil & gas sector. This was supported by applications R&D in both

university and company (1996-2000) with strong industrial pull. Standards for ACFMwere published in 2003, and from 2004 onwards the focus was growing the applicationmarket and moving into the rail and nuclear sectors as the technology gainedacceptance. The development was over a shorter period than some other examples,initial research to standard taking some 16 years. This was perhaps partly because therisk was reduced by the existence of an established market for eddy current technology.There was a good combination of scientific push and industrial pull for most of thedevelopment phase. But also, significantly, a strong supply chain based on closeorganizational and individual relationships was established early on.

A second example is Pulsed Eddy Currents (PEC). It is based upon pioneering

research at Argonne National Laboratory (USA) in 1950 with nuclear applications inmind. It was then picked up by the aerospace industry for detecting cracks inaluminium. In 1987 PEC was adapted by ARCO in their TEMP instrument (a pulsededdy current device) to detect steel corrosion under insulation (CUI) for the oil and gasindustry. In 1990 TEMP was licensed to RTD in Holland, being commercialized asINCOTEST. In 1994 Shell investigated PEC for detecting CUI, but found someshortcomings. However, at the same time, Shell carried out a laboratory investigationof PEC for other industrial applications. This led in 1996 to the launch of acommercial PEC system for a range of oil and gas applications, such as corrosion underfireproofing. There are no standards for PEC at the time of writing. The research from

the 1980s onwards mainly took place in companies rather than universities, whichensured a strong industrial pull through most of development of PEC and a good potential market, once the target applications had been identified [2].

Basic research into eddy currents was started at RAE Farnborough (MoD) incollaboration with Surrey University in the 1970s. In the 1980s this moved on to eddy

C.B. Scruby / NDE Research Makes a Difference 13


33/317

current research using Hall sensors, while from 1985-9 industrial pull joined scientific push and a prototype EddyScan system for crack detection in aircraft was developed.The scientific focus then moved to transient eddy currents in the 1990s [3], as furtherresearch was undertaken, first on the technique and then (1995-2000) on applications as

industrial pull increased. Activity switched to a product and the TRECSCAN systemwas developed from 2000-3, addressing aircraft inspection needs in the defence andaerospace sectors. Like ACFM and PEC, the development of transient eddy currentswas heavily dependent upon background research into electromagnetic induction andeddy currents in particular. Here the early work benefited from collaboration betweena government laboratory and a university. There was a requirement to deliver usefulresearch outputs to the MoD customer in the early days of this development, butindustrial pull appears to have strengthened progressively. This is consistent with thegrowing market for methods to prolong the life ageing aircraft.

The first three examples build on related research into eddy currents. Apart from

isolated pieces of pioneering research in various countries, the next example is of atechnology developed from first principles, and it is therefore not surprising that it isstill some way from acceptance, even though about 20 years have passed. In 1982-9some fundamental scientific studies of magnetic methods for materials and stresscharacterisation were undertaken at Oxford University and Harwell Laboratory [4].This led on to a programme of research into NDE stress measurement, 1989-93, with progressively stronger industrial pull and support from the oil and gas industry; this ledto trial applications of what became known as MAPS in 1993-5. The early applicationsdemonstrated the need for further research into the technique as well as its application(1995-00). It was found necessary to implement further refinements to the

underpinning physical model as well as to the MAPS instrument (2000-7), while at thesame time beginning to commercialise the technology for niche applications in the oiland gas and rail sectors. The early stages of the development showed the importance ofcollaboration between university, RTO and end-user industry. Industrial pull waned inthe middle period of the development and this, coupled with privatisation andsuccessive company reorganisations, may have delayed the innovation process.

The final example is of an important ultrasonic innovation, time-of-flightdiffraction (TOFD), rather than another electromagnetic technique. Maurice Silkstarted research into TOFD [5] at the NDT Centre, Harwell Laboratory (UKAEA) in1974 using ideas from previous research into neutron time-of-flight spectrometry.During the ensuing years (1974-82) the work was mainly laboratory research (scientific push). Other scientists became involved and the theory of ultrasonic diffraction wasdeveloped in 1981-83. The first major industrial application was to UKAEA’sWinfrith Reactor 1982-4. Soon afterwards (1983-4) TOFD used by UKAEA for theirnuclear defect detection trials. Almost simultaneously (1984-90) there was furtherindustrial pull when TOFD was developed for offshore and undersea use as an earlyactivity of Harwell’s HOIS project. The first commercialisation came 1982-4 whenZipscan was developed and licensed. Soon afterwards in 1984, TOFD started to beused for major industrial inspections, especially nuclear pressure vessels and offshore

structures. The first TOFD standard was published in 1993 and TOFD was soonaccepted as a mainstream NDE method across industry. TOFD took about 20 years tomove from initial research idea through to an industry standard. It is salutary towonder how long exploitation would have taken had it not been for strong industrial pull in the mid-1980s from two separate energy sectors. TOFD had little input from theuniversity research base in its early days, reflecting the strength of public sector



34/317

laboratories then, a different situation from the 21st century. Throughout the central period of development there was strong collaboration between researchers, developersand end-users, helped by national imperatives in the nuclear case and a well-establishedand effective joint industry collaborative project (HOIS) in the other. Another

important element was the utilisation of ideas and technology from other fields.A common theme in all of these and similar developments has been the need todevelop physical models of the technique under investigation. In many cases usingmodels to perform predictive or interpretive calculations was a major exercise, withmore limited computing facilities prior to the 1990s. A model was vital to understandthe amplitude variations in TOFD data, as it has been for example to interpret theelectrical signals from ACFM, TRECSCAN and PEC techniques. In the case of MAPS,modelling the effects of stress on magnetic parameters from first principles involvesvery difficult physics as well as computational power. However difficult or time-consuming, robust physical models are, in the author’s view, vital for any new

technology. They are needed, not only to interpret the data and design the best way touse it, but also to give credibility within the scientific and engineering communities.

3. Lessons to be learnt

From the author’s experience and knowledge, there are a number of barriers to or brakes that slow down effective NDE research, i.e. research whose results make a positive difference to industrial practice. Two important hurdles are, on the one handlack of understanding by researchers of industrial needs and culture, and on the otherhand lack of understanding by industry of scientific issues and research culture.Research itself, or the development process that follows can be hindered by the wronglevel of funding, more commonly too little but occasionally too much at a time whenthe team is unable to deliver what is anticipated. This leads on to the importance ofcorrectly managing expectations. Often researchers are over-optimistic about whatthey will achieve and their speed of progress in order to secure funds. When they failto meet their customer’s expectations the development may be dropped for the wrongreasons. Intellectual property (IP) can cause problems. Sponsors of research may insiston IP rights that are too restrictive and stifle creativity, while loose IP arrangements candissuade companies from investing in technology.

As already stated, a serious issue concerns gaps in the supply chain, i.e. betweenresearchers, developers and end-users. In terms of TRLs, the problems usually arisewhen progressing from level 4 to level 7. Especially during the past 10 years, membersof the NDE supply chain have undergone reorganisations, changes of status, missionand objectives. Staff have been lost, working relationships destroyed and the supplychain disrupted. Finally, human factors always play a part, the biggest hurdles beinglack of trust and poor communication for whatever reason. The Engineering Doctoratein NDE is a very encouraging recent development, already proving its worth in termsof technology transfer from research to application. Accountable to both university and

company, the student (research engineer) begins to bridge that TRL gap.What lessons can be learnt of the ingredients for successful NDE research? Thefollowing list is not exhaustive, nor is every ingredient always necessary:

1. Adequate resources - people, facilities, environment, timely funding2. Scientific excellence - creativity and lateral thinking balanced against

penetration and focus; correct fundamentals essential for robust application



35/317

3. Technology push balanced against industry pull throughout4. Interdisciplinary collaboration, networking and international links5. Import of ideas, results & technology from other fields6. Industrial pull founded on sound business needs or regulatory imperatives

7. Supportive organisational structures with long-term vision8. Viable exploitation route along “joined up” technology supply chain9. Committed individuals to champion and drive innovation process10. Trust, good communication and shared objectives within the whole “team”11. Proprietary secrecy balanced against publication and scientific credibility12. Good timing, grasp of opportunity, achievable timescales

The author is tempted to add: serendipity, since chance events often seem tounlock research and stimulate vital creative leaps; patience, since it is difficult to predict what research will produce and when.

4. Concluding Remarks

Before discussing strategy for successful research in NDE, it is important to capture the present trends in the field of NDE. Encouragingly, most recent market surveys have predicted a steady growth in use of NDE by industry. This is driven ultimately by theneeds of society for greater assurance of the safety of engineering structures. Linkedwith this is a heightened awareness of the need for environmental protection. Newmaterials & new designs are being used that require new inspection technology. Thereis also a move towards greater automation, and the willingness to invest in hightechnology solutions to reduce inspection costs, improve quality, speed up inspectionand sentencing. In some cases the move is towards reducing the need for inspectionsthat interrupt service by new strategies such as structural health monitoring (SHM).

Techniques

NDE Research

Applications

Mathematics

Electronicengineering

Physics

Mechanicalengineering

Materialsscience

Technology &science transfer

Manufacturing

& product

quality

In-service

structural integrity

& assessment

S u p p l y c h a i n

Materials

characterisation

Condition & health

monitoring

Other disciplines

Techniques

NDE Research

Applications

Mathematics

Electronicengineering

Physics

Mechanicalengineering

Materialsscience

Technology &science transfer

Manufacturing

& product

quality

In-service

structural integrity

& assessment

S u p p l y c h a i n

Materials

characterisation

Condition & health

monitoring

Other disciplines

Figure 4. NDE research and linked engineering disciplines



36/317

As a long-term strategy there is a desire to integrate NDE more strongly into theengineering life cycle, so that the benefits of NDE are weighed against reductions inthe whole life cost of an asset. To do so requires the development of linkages withmaterial science, engineering design, structural integrity and assessment. NDE sits at

the centre of a complex network of engineering disciplines as Figure 4 shows; theseneeds to be understood and exploited. Recent years have seen large changes not onlyin the large users of NDE technology, but also in the supply chain. Equipment supplyand inspection service companies have changed hands, merged and in some cases been bought by large companies. Such changes are likely to continue in the future.

Without any doubt the NDE industry will continue to encounter technicalchallenges that can only be solved by consistent, high quality research. What are likelyto be the main elements of a future NDE research strategy? The following list indicatessome of the likely priorities:

1. Modelling, always essential for scientific understanding and to give

confidence in product quality and for safety-critical applications2. Improved quantification of results, characterisation and discrimination3. More advanced data analysis, imaging and visualisation4. Greater understanding of and improvement to reliability5. Raising inspection speeds and reducing human factors through automation and

autonomous systems6. Incorporation of technological advances from other fields7. Addressing new materials, designs, difficult applications & environments8. Earlier detection of materials & structural degradation & failure9. Technologies and strategies to facilitate reliable SHM

This is a very full and challenging list. It can only be achieved through greatercollaboration & integration with other disciplines & fields. It also requires the longterm education and training of high calibre engineers for all stages of the research andinnovation process.

The UK Research Centre in NDE is a university-industry partnership that wasestablished nearly 5 years ago to harness the UK’s research base to meet these long-term challenges, building on and learning from past experience, and hopefully avoidingsome of the pitfalls. Its vision is to make a lasting difference to NDE technologythrough well-planned world-class research, and to invigorate the NDE professionthrough the provision of highly trained engineers.

To conclude, NDE research does make a difference. It is impossible to respond tothe challenges of tighter regulation, new materials and applications, new products and plant, more stringent operational conditions, longer life, higher accuracy, reliability,long terms global trends, changes in industry, rising public expectation of safe andenvironmentally secure operation, and the commercial drivers of faster, better, cheaper products without research.

References

[1] W.D.Dover, R.Collins and D.H.Michael, Phil. Trans. R. Soc. Lond . A320, (1986) 271-283.[2] P.Crouzen, Proceedings of 9th European Conference on NDT , Berlin, (2006) in press[3] S.K.Burke, G.R.Hugo and D.J.Harrison, Review of Progress in Quantitative Non-Destructive

Evaluation, 17A, (1998) 307-314[4] D.J.Buttle, C.B.Scruby, J.P.Jakubovics and G.A.D.Briggs, Proc Roy Soc Lond . A414, (1987) 469 497.[5] M.G.Silk and B.H.Lidington, Harwell Report, AERE-R7774 (1974).



37/317

Industrial Applications of 3MA –Micromagnetic MultiparameterMicrostructure and Stress Analysis

Gerd DOBMANN1, Iris ALTPETER 1, Bernd WOLTER 1, Rolf KERN1

Fraunhofer IZFP, Germany

Abstract. Micromagnetic NDT techniques like the measurement of the magneticBarkhausen noise, the incremental permeability and the harmonic analysis of thetangential magnetic field allow deriving inspection procedures to onlinemonitoring and control machinery parts and components in production processesin order to characterize mechanical properties like hardness, hardening depth, yieldand tensile strength. These types of inspection procedures continuously werefurther developed in the last two decades so that today the second generation ofsystem hard and software is in industrial use. The application is in steel industrywhere steel sheets in hot-dip-galvanizing lines were annealed after cold rolling butalso in heavy plate rolling mills where after thermo-mechanical rolling specialtextures and texture gradients can occur. An increasing number of applications are

also to find in the machinery building industry and here especially in case ofmachinery parts of the car supplying industry. Besides mechanical hardnessdetermination the measurement of residual stresses and the detection of in-homogeneities in the surface of machined parts is an inspection task. In differentcase studies the advantage to implement a micromagnetic NDE technique into theindustrial processes is discussed.

Keywords. Micromagnetic NDE, hardness, hardening depth, residual stresses,yield strength, steel industry, steel sheets, heavy plates, machinery building,automotive

Introduction

The reason to develop 3MA (Micromagnetic-, Multiparameter-, Microstructure-, andstress-Analysis), starting in the late seventies in the German nuclear safety program,was to find microstructure sensitive NDT techniques to characterize the quality of heattreatments, for instance the stress relieve of a weld. George Matzkanin [1] just had published a NTIC report in the USA to the magnetic Barkhausen noise. The techniquewas sensitive to microstructure changes as well as to load-induced and residual stresses.Therefore a second direction of research started in programs of the European steel

industry and the objective was to determine residual stresses in big forgings. Beside themagnetic Barkhausen effect also a magneto-acoustic-one became popular [2]. Thetechnique has based on acoustic emission measurements during a hysteresis cycle andwas – because of the high amplification – also sensitive to electric interference noise.Therefore the acoustic Barkhausen noise technique has never found a real industrial

1



IOS Press, 2008


doi:10.3233/978-1-58603-896-0-18

18


38/317


39/317

the upper acceptance level (blue line). The strength values are calculated by the 3MAapproach from measured micromagnetic data. The red dots indicate the selection ofspecimens taken to destructive verification tests after performing NDT. The residualstandard errors found by validation are in the range 4-7 % concerning the yield strength.

Figure 3 shows a 3MA installation in the line of a strip producer; a robot is used tohandle the transducer.

Figure 2. 3MA predicted Yield strength [5] Figure 3. 3MA probe with robot at a strip line

1.2 Heavy plate Inspection

Ongoing research is to heavy plate inspection. The steel producer asks for themeasurement of geometrical and mechanical properties, which have to be uniform

along the product length and width, especially in the case of high-value grades used inoff-shore application. Destructive tensile and toughness tests are performed by highlyqualified and certified personnel according to codes and delivery conditions. The testscannot be integrated into online closed loop control with direct feedback. To reliablytest the mechanical hardness the surface must be carefully prepared by removing scaleand decarburized surface layers and residual stresses are to relieve. The extraction ofthe test pieces and testing is very time and cost extensive. Costs in the range of severalthousands Euro per year arise in a middle-sized heavy plate plant only by destruction ofthe test pieces.

Figure 4. Tensile strength predicted by 3MA [6] Figure 5. Yield strength predicted by 3MA [6]

0

200

400

600

800

1000

1200

1400

1600

0 200 400 600 800 1000 1200 1400 1600

Tensile strength Rm [MPa] (destructive)

T e n s i l e s t r e n g t h R m [

M P a ] ( 3 M A )

0

200

400

600

800

1000

1200

1400

1600

0 200 400 600 800 1000 1200 1400 1600

Elastic limit or yield strength

Rp0.2 or REH [MPa] (destructive)

E l a s t i c l i m i t R p 0 . 2

[ M P a ] ( 3 M A )

G. Dobmann et al. / Industrial Applications of 3MA20


40/317

In case of a mother plate of several meters length the edges are usually subjected toother cooling conditions than the rest. Indeed, especially the plate ends are known tocool faster, generating an undesired increase in tensile strength Rm and yield strengthRp0.2. State-of-the-art is to cut-off the plate edges with non-conform properties based

on empirical values concerning the cut-off length. As the destructive tensile testfollows directly after the cut-off process of the edges, only the result of these tests canreveal the selection of a not appropriate cut-off length. This results in high costs due toreworking, pseudo-scrap and delayed shipment release; the European steel producersestimate their annual costs in the range of 11 million Euros. Knowing exactly thecontour of the zone with unacceptable material properties would allow an open loopcontrol of the cut-off process. Therefore heavy plate producers will replace thedestructive quality inspection of test pieces by a NDT technology [6] applying 3MA(see Figure 4 and 5). By a manufacturer-specific calibration residual standard errors of10 MPa (Rm), 20 MPa (Rp0.2), and 4HB in the Brinell hardness can be obtained. It

should mention here that in the 3MA calibration also other measuring quantities can beintegrated so far they provide other independent information, for instance elastic properties. By using ultrasonic waves propagating in thickness direction, i.e. acompressive wave excited by a piezoelectric transducer (index L) and two linearly polarized shear waves (polarized in, index SHR, and transverse, index SHT, to therolling direction) excited by a EMAT, normalized time-of-flight quantities can bederived describing crystallographic texture effects. Taking into account these quantities(tSHR /tL, tSHT/tL, (tSHR -tSHT)/tL) together with the micromagnetic parameters then aregression result is obtained again reducing the residual standard error.

2. Application in Automotive and Machinery Building Industry

2.1 Car Engine casting

To reduce the weight of the power supply unit the car combustion engines cylindercrankcases can be made of cast iron with vermicular graphite (GJV), because thismaterial in a Diesel engine allows a higher loading pressure even by reduced wallthickness. However, the service live of machining tools is during processing an engine

block made from GJV substantially smaller compared with a block from cast iron withlamellar (flake) graphite (GJL).

Figure 6. Microstructure gradient obtained in a cylinder region of a cast engine [7]

This disadvantage can be eliminated by an innovative casting technology that produces a continuous microstructure gradient in the cast iron from lamellar graphite at

GJL GJV

work on - relevant range

GJL GJV

work on - relevant range

lamellar graphite transition zone vermicular graphite

G. Dobmann et al. / Industrial Applications of 3MA 21


41/317

the inner surface of the cylinders to vermicular graphite in radial direction. Byimplementing some chemical additives into the core of the mould which can diffuse inthe cast iron during the solidification process in the mould the gradient with acontinuous transition from lamellar graphite and finally vermicular graphite is obtained.

However, the technology can only be used by the casters so far the gradient quality can be characterized and monitored by NDT. Figure 6 documents in a micrograph such agradient beginning at the left side with cast iron (inner cylinder surface) and lamellargraphite followed by a transition region and vermicular graphite on the right side.

3MA techniques always cover a certain analysing depth depending on themagnetising frequency and geometrical parameters of the magnetisation yoke, etc. Sofar the gradient has different graphite compositions within the analysing depth, 3MAquantities should be influenced. Based on measurements at an especially designedcalibration test specimen set 3MA quantities were selected to image the gradient withoptimal contrast. As reference quantity to calibrate 3MA the local thickness of the

GJV-layer was evaluated by using micrographs and optimized pattern recognitionalgorithms in the microscope. A special designed transducer head was developed toscan the cylinder surface by line scans in hoop direction and rotating the head, thenshifting the head in axial direction to perform the next line scan. Figure 7 and Figure 8show as example the coercivity images derived from the tangential field strengthevaluation[3] (HC0 in A/cm) and line scans covering an angle range of 190°.

Figure 7. Coercivity image of a reference blockmade from GJV

Figure 8. Coercivity image of block with GJV/GJLgradient

Combining different 3MA quantities in a multiple regression the thickness of theGJL layer was predicted. A regression coefficient of R 2= 0.93 and a residual standarderror of = 0.06 mm was obtained [7].

2.2 Wheel Bearing Inspection

The fixation of the inner ring of wheel bearings is performed by a wobble riveting process. As a consequence a residual stress is built up in the ring which may not exceeda limit value of about 300 MPa to get a perfect quality.

The usual technique to inspect the residual stress state is x-ray diffraction which is

destructive in nature because it requires a preparation of the test location. Furthermoreit can only be performed statistically. The 3MA technique allows a fast non-destructiveestimation of the residual stress level (Figures 9, 10). After a calibration step by usingx-ray reference values a 100% quality inspection of these parts is possible. Thecalibration procedure requires a coincidence of the 3MA and x-ray calibration positions



42/317

because residual stress varies along the circumference. That means the 3MA data haveto be recorded in a first step before the x-ray test location is prepared by etching.According to Figure 10 the residual standard error in the calibration is in the 20 MParange. Besides the residual stress additionally the surface hardness can be measured.

-50

0

50

100

150

200

250

-50 0 50 100 150 200 250

Residual Stress [MPa] (3MA)

R e s i d u a l S t r e s s [ M P a ] ( X - r a y )

Analysing Depth:

100 µm

Adj.R² = 0.941

1 = 19.0 MPa

Figure 9. 3MA-Probe at test location Figure 10. Residual stress calibration

2.3 Evaluation of Microstructure and Stress Gradients

Machined parts in most of the cases have more or less steep gradients in their properties near the surfaces. To improve the lifetime of mechanical highly stressedmachinery components the bearing areas are surface-hardened from the µm- up to themillimetre range depending on the requirements and on the hardening technology.

Fig.11. Comparison of nitrating hardening depth measured by 3MA and Nht 700 (Vickers) versus opticalresult

Additional surface finishing by grinding can superimpose surface near defects ofmicrostructure and residual stresses which can result in a part breakdown. To inspectthe production quality in many cases not only the properties immediately at the surface but also information of the properties below the surface are desired. 3MA is aneffective tool to investigate the properties near the surface as well as the range belowthe surface up to several millimetres in depth.

One example of a 3MA application in industry is the determination of nitratinghardening depth NHD of piston rings on the flank side and on the tread surface.Typical values of nitrating hardening depth are between 60 and about 100 µm. It isfound by the user that the reproducibility of the non-destructive values of hardness andhardening depth in piston rings is better than the conventional testing by a

Adj R² = 0,9151

Adj R² = 0,8211

0

10

20

30

40

50

60

70

80

0 10 20 30 40 50 60 70Optical Result [µm]

3 M A / N h t 7 0 0 [ µ m ]

3MA Value

Nht700 Value

G. Dobmann et al. / Industrial Applications of 3MA 23


43/317

metallographic Vickers hardness test (Nht700), as can be seen in Figure 11 [8]. Thereason of that behaviour seems to be the difference in the lateral resolution of theconventional and the non-destructive testing method. Due to a diameter of the 3MAreceiver coil of about 2 mm the 3MA values are covering a much larger inspection area.

Fast data evaluation by 3MA allows a complete production feedback control.The occurrence of grinding defects, e. g. in gear wheels, is a main problem sincemany years which is caused by too much heat input during the grinding process.Modern grinding tools allow much higher grinding speed compared to former machines but on the other side this can result in more defects. To get information on the qualityof grinded microstructure states the common method in industry is the nital etchingtechnique. Grinding defects are indicated by the discoloration of the surface. Thistechnique is effective as long as the surface information is sufficient to estimate thequality. But it fails if in a preceding production step defects are produced below thesurface which are covered in the next production step by a perfect finishing. Several

examples of defective gear wheels investigated by hole drilling method and x-raydiffraction have shown that in a depth of 100 µm high tensile stresses up to several 100MPa can be present whereas at the surface a perfect compressive state of severalhundred MPA has been found. These hidden defects cannot be detected by nital etching.As a consequence after some time small cracks are covering the surface due to stress-relieve even without a mechanical load.

Figure 12. Hardness calibration at various depths;

hardness values determined by 3MA versus targetvalues

Figure 13. Residual Stress calibration at various

depths; RS values determined by 3MA versus X-rayreference values

Since several years IZFP has gained experience in the non-destructive detectionand quantitative evaluation of such grinding defect gradients by 3MA in cooperationwith industrial partners and in different research and development projects [9, 10].After a calibration step 3MA can be used to evaluate different target valuessimultaneously, especially the hardness and the residual stress at the surface and inseveral depths below the surface (Figures 12 and 13). To get unambiguous resultscalibration must be done carefully. Calibration is mainly determined by well definedcalibration specimens and only valid to the target ranges available by calibration. In

most cases calibration is restricted e. g. to the material, to the actual machining parameters and even to the 3MA probe in use. If any variation occurs, its influence onthe validation of the existing calibration has to be checked and if necessary arecalibration or extension of the existing calibration has to be performed to include anydisturbances. These limitations and the calibration effort may be seen as a disadvantageof 3MA. But if an optimal calibration is developed the fast non-destructive



44/317

determination of various quality parameters which is desired concerning expensivesecurity related parts justifies this effort.

3. Conclusion

3MA is a matured technology and a wide field of applications is given. However, besides the success story we also can find critical remarks from industrial users. Theseare mainly to the calibration efforts and problems of recalibration if a sensor has to bechanged because of damage by wear. Therefore actual emphasis of R&D is togeneralize calibration procedures.

4. Acknowledgements

The authors very much appreciate to acknowledge the contribution to the result bycompanies as ThyssenKrupp Stahl AG, Duisburg, ArcelorMittal Research, Metz,Dillinger Hütte GTS AG, Dillingen, Halberg Guss GmbH, Saarbrücken, and SchaefflerKG, Schweinfurt.

References

[1] G.A. Matzkanin, et al., The Barkhausen Effekt and its Application to Nondestructive Evaluation, NTIAC report 79-2 (1979) (Nondestructive Testing Information Analysis Center, San Antonio, Texas)1-49.

[2] W.A. Theiner, E. Waschkies, Method for the non-destructive determination of material states by use ofthe Barkhausen-effect (in German), Patent DE 2837733C2 (1984).

[3] G. Dobmann et al, Barkhausen Noise Measurements and related Measurements in FerromagneticMaterials; in Volume 1: Topics on Non-destructive Evaluation series (B.B. Djordjevic, H. Dos Reis,ed

Date post:	06-Jul-2018
Category:	Documents
Upload:	jorge-perdigon
View:	240 times
Download:	0 times

Electromagnetic Nondestructive Evaluation XI Studies in Applied Electromagnetics and Mechanics No 11

Documents