322 Ad ances in Elec ical and Elec onic Enginee ing
RECENT TRENDS IN ELECTROMAGNETIC NON-DESTRUCTIVE SENSING
K. ápo á, I. áp, L. Janoušek, M. Sme ana
Uni e si y o Žilina, Facul y o Elec ical Enginee ing, Veký diel, 010 26 Žilina, Slo akia
Tel.: +421-41-5132100, Fax: +421-41-513 1519, e-mail: kla a.capo a@ el.uniza.sk
Summa y The pape deals wi h ma e ial elec omagne ic non-des uc i e es ing (eNDT) wi h emphasize on eddy cu en
es ing (ECT). Va ious modi ica ions o ECT sensing a e compa ed and discussed om he desi ed de ec ed signal
cha ac e is ics poin o iew. Excep o he op imiza ion o usual p obe coils a angemen s o he conc e e applica ions, he
new magne ic senso s as gian magne o- esis ance (GMR) and spin dependen unneling (SDT) a e p esen ed. The ad anced
ECT senso s a e cha ac e ized by hei sensi i i y, equency ange and senso dimensions.
Keywo ds: elec omagne ic nondes uc i e es ing, eddy cu en sensing, su ace and subsu ace de ec s, de ec ion p obes,
magne ic senso s, gian magne o- esis ance (GMR), spin dependen unneling (SDT).
1. INTRODUCTION
Non-des uc i e elec omagne ic es ing (eNDT) is
an e ec i e me hodology o diagnos ics in many
echnical and scien i ic applica ions. The
equi emen s o new e ec i e eNDT ools a e
connec ed wi h he wide and s ill inc easing
demands o high quali y and eliabili y s anda ds in
indus ial p oduc ion and also wi h de elopmen s o
o he echnical and scien i ic a eas, e.g. medicine,
geology, ci il and en i onmen al enginee ing, e c.
Va ious NDT echniques o m a wide g oup o
a he di e en ools, which a e based on di e en
physical phenomena and hey a e cha ac e ized by
di e en and speci ic pe o mance and applica ion
ields. F om he applica ion poin o iew he eNDT
is used in many a eas - om he inspec ion o
me allic pipes o he ae onau ical main enance and
om he localiza ion o liquids in subsoil o he
ho acic imaging o clinical diagnos ics.
One o he mos popula eNDT me hods is he
eddy cu en es ing (ECT) and e alua ion. The
p inciple o ECT can be b ie ly desc ibed by he
ollowing way. Eddy cu en coil ed by al e na ing
sinusoidal cu en (AC), o equencies in he ange
50 Hz – 10 MHz, gene a es p ima y magne ic ield
acco ding o he Ampe e´s law. This p ima y
magne ic ield induces eddy cu en s in he es ed
conduc i e ma e ial objec acco ding o he
Fa aday´s law. Then eddy cu en s gene a e
seconda y magne ic ield in he opposi e di ec ion in
ag eemen wi h he Lenz´s law. Following om
hese p ocesses he coil impedance changes in he
case o ma e ial changes, e.g. in he p esence o
impe ec ions - de ec s in he ma e ial objec . The
impedance change is measu ed, analyzed and
co ela ed wi h he de ec dimensions. The locus o
impedance change o med du ing he mo emen o
an eddy cu en p obe coil o e a es ma e ial
ha ing de ec is called an eddy cu en signal. I s
ampli ude p o ides in o ma ion abou he de ec size
and i s phase angle wi h espec o li -o gi es
in o ma ion abou de ec loca ion o dep h, Fig.1.
Eddy cu en densi y in ma e ial is no uni o m in
Fig. 1. Eddy cu en es ing a angemen .
he ma e ial dep h di ec ion. I is g ea es on he
su ace and dec eases mono onously wi h dep h
(skin e ec ) acco ding o he ela ion o s anda d
dep h o pene a ion which dec eases wi h inc easing
equency, conduc i i y and pe meabili y. I means
ha o measu ing hickness o hin su aces e y
high equencies a e o be used and on he con a y
o de ec ion o sub-su ace bu ied de ec s and o
es ing highly conduc i e (magne ic) hick ma e ials
low equencies a e o be employed.
Usually he d i ing cu en is kep cons an ( ew
hund eds o mA) and he impedance changes
occu ed due o pe u ba ion o eddy cu en s a
de ec egions a e o be measu ed. Since hese
changes a e e y small (), high p ecision AC
b idge is used, Fig. 2. The b idge imbalance is
co ela ed wi h he de ec o ma e ial cha ac e is ic
esponsible.
The ECT ins umen consis s usually o an
oscilla o ( o exci ing equency), cons an AC
supply, AC (Maxwell) b idge ci cui , ampli ie and
sc een ( o display he changes in a 2D g aph o as a
ec o ). In mode n sys ems he e a pe sonal
compu e wi h he necessa y ha dwa e (plug-in ca d)
and so wa e is used o he measu emen s,
AC
SUPPLY
R
L
PRIMARY
FIELD
INDUCED
FIELD
EDDY
CURRENTS
Li o
Dep h o
pene a ion
Recen ends in elec omagne ic non-des uc i e es ing 323
Fig. 2. AC b idge o ECT signal measu emen .
adjus men , da a s o age, analysis and managemen .
The ECT is he mos ly used echnique o de ec ing
a igue c acks and co osion in conduc i e ma e ials.
The cos o using echnology is low and i is possible
o moni o subsu ace de ec s and de ec s unde
insula ing coa ings wi hou ouching he su ace
specimen. One o he mos impo an pa s o ECT
de ice is he sensing pa called p obe which is
c ea ed ob iously by p obe coils o o he senso s.
Because sa e y-c i ical sys ems depend on ea ly
de ec ion o a igue c acks o a oid majo ailu es,
he e is an inc easing need o eddy cu en p obes
ha can eliably de ec e y small de ec s. Also he e
a e inc easing demands o p obes ha can de ec
deeply bu ied de ec s o a oid disassembling
s uc u es. The e a e also many o he applica ions
whe e ECT is success ully used, e.g. ma e ial
hickness measu emen s, coa ing hickness
measu emen s and conduc i i y measu emen s o
he ma e ial iden i ica ion, hea damage de ec ion,
case dep h de e mina ion o hea ea men
moni o ing. Acco ding o he applica ion a eas he
measu ing se -up is designed a ealized.
2. EDDY CURRENT PROBES
CONFIGURATIONS
As i was men ioned abo e he app op ia e selec ion
o p obe coil is e y impo an in ECT in o de o ge
he igh (desi ed) in o ma ion om i . The mos
common p obes used in ECT a e su ace o pancake
p obes (wi h he axis no mal o he su ace) which
a e chosen o inspec ion o pla es and bol -holes
ei he as a single elemen o an a ay, in bo h
absolu e and di e en ial modes. The enci cling
p obes a e used o inspec ion o ods, ba s and
ubes wi h ou side access and he Bobbin p obes o
p e- and in- se ice inspec ion o hea exchange ,
s eam gene a o , condense and o he s wi h inside
access, Fig. 3, [1]. These h ee ypes can also ope a e
in he send- ecei e mode wi h he sepa a e coils o
sending and ecei ing o signal and also in absolu e
o di e en ial mode.
The absolu e EC p obe consis s o a single sensing
coil o signal exci a ion and ecep ion. I is
de e mined o de ec ion o c acks as well as g adual
Fig. 3. Con igu a ions o ECT P obes.
a ia ions. Bu absolu e p obes a e also sensi i e o
li -o , p obe il , empe a u e changes, e c.
Di e en ial p obes ha e wo sensing coils wound in
opposi e di ec ion and in es iga ing wo di e en
egions o he ma e ial. These p obes a e good o
high sensi i i y de ec ion o small de ec s and hey
a e mo e immune o changes in empe a u e and
p obe wobble.
The e a e many ac o s which in luence eddy
cu en esponse om a p obe. Success ul
assessmen o laws elies on holding he o he s
cons an , o somehow elimina ing hei e ec on he
esul s. The main ac o s a e ma e ial conduc i i y,
pe meabili y, equency, geome y and he li -o .
As o he ma e ial conduc i i y is seen he g ea e
he conduc i i y he g ea e he low o eddy
cu en s on he su ace. F om he conduc i i y
measu emen s we can ge he in o ma ion abou he
ma e ial composi ion, hea ea men and wo k
ha dening, e c.
Fo he non- e ous me als he pe meabili y is he
same as o he “ ee space”, he ela i e pe meabili y
is equal o one, and o he e ous me als i has
alues se e al hund eds o mo e. Pe meabili y is
a ying s ongly wi hin he me al pa due o
localized s esses, hea ing e ec s, e c.
F equency g ea ly a ec ed he eddy cu en
esponse bu i can be con olled wi hou p oblems.
Geome ical ea u es such as cu a u e, edges,
g oo es, e c. a ec he eddy cu en esponse. The
used echniques mus ecognize his, e.g. in es ing
an edge o c acks he p obe will no mally be mo ed
along pa allel o he edge so ha small changes may
be easily seen. Whe e he ma e ial hickness is less
han he e ec i e dep h o pene a ion his will also
a ec he eddy cu en esponse.
As o he p oximi y o li -o he close a p obe
coil is o he su ace he g ea e will be he e ec on
ha coil. I means ha he li -o signal a ises as he
p obe is mo ed on and o he su ace and he
sensi i i y will be educed as he coil p oduc
spacing inc eases.
Fig. 4. Posi ions o ECT p obes abo e es ed objec .
U
R
1
R
2
R
3
L
3
Tes ed body
Inspec ion coil
Z
4
=
R
4
+
j
ω
L
4
U
ou pu
∼
324 Ad ances in Elec ical and Elec onic Enginee ing
Fig. 4 shows a ious posi ions o he p obe abo e
he es ed objec acco ding o changes o li o , il
and geome y o edge e ec s.
3. ADVANCES OF EDDY CURRENT PROBES
Eddy cu en es ing p obes usually combine an
exci a ion coil ha induces eddy cu en s in a
specimen and a de ec ion elemen ha iden i ies he
pe u ba ion o he cu en s by c acks o o he
de ec s. In o de o de ec deepe de ec s in ma e ial
objec i is necessa y o p opose he ad anced coils
con igu a ion and ollowing da a p ocessing, as i
was ealized and published e.g. in he pape s[2], [3].
The op imiza ion and c ea ion o a new p obe
modi ica ion is connec ed wi h he used ECT
echnique. The new coils p obe de elopmen used in
he emo e ield eddy cu en es ing (RFECT)
de o ed o he inspec ion o pipes ha e been
desc ibed e.g. in he pape [4]. Bu excep o coils
and hei a ious a angemen s he ECT de ec ion
elemen s can be also supe conduc ing quan um
in e e ence (SQUID) de ec o s, o solid-s a e
magne ic senso s, such as Hall e ec , luxga e o
magne o- esis ance (AMR o GMR) and spin-
dependen - unneling (SDT) senso s.
The use o low- ield solid-s a e magne ic senso s
ep esen s a signi ican ad ance o e mo e
adi ional induc i e p obes in use oday, [5]. Two
key a ibu es will open oppo uni ies o inc eased
use o eddy cu en p obes: senso cons an
sensi i i y o e a wide ange o equencies and
de elopmen o smalle senso s.
P obes ha de ec eddy cu en ields using
induc i e coils ha e less sensi i i y a low
equencies. Un o una ely, his is whe e he de ice
would ha e o ope a e o de ec deep laws. Small
sensing coils which a e equi ed o de ec small
de ec s, also ha e low sensi i i y. In con as , small,
high-sensi i i y hin ilm senso s can locally
measu e a magne ic ield o e an a ea compa able o
he size o he senso i sel / ens o mic ome e s/.
Limi a ion o con en ional eddy cu en p obes is
he di icul y o de ec ing small c acks o igina ing a
he edges o a specimen. This de ec is he mos
common ype encoun e ed in p ac ice. An example
is he c acks ha appea a ound he as ene o i e
holes in ai c a mul ilaye s uc u es. Mos
induc i e coil p obes a e sensi i e o bo h he edge
and he c acks ini ia ing om o nea he edge. The
edge c ea es a la ge signal ha obscu es he small
signal om he c ack. GMR and SDT magne ic
senso s can be o ien ed o elimina e he edge signal.
Wi h his o ien a ion he p esence o he edge
enhances he signal om he c ack.
To achie e high esolu ion o de ec ing small
su ace and nea -su ace de ec s i is necessa y o
educe he dimensions o he exci a ion coil. The
minimum leng h o a de ec able c ack is oughly
equal o he mean adius o he coil. The e ha e been
de eloped and es ed p obes inco po a ing small,
la , pancake coils o plana exci a ion coils
deposi ed on he senso subs a e.
Recen de elopmen o hin ilm magne ic
echnology has esul ed in ilms exhibi ing a la ge
change in esis ance wi h magne ic ield, [5]. This
phenomenon is called gian magne o- esis ance o
dis inguish i om con en ional aniso opic
magne o- esis ance (AMR). Whe eas AMR esis o s
exhibi a change o esis ance o up o 3%, a ious
GMR ma e ials achie e abou a 10% - 20% change
in esis ance.
GMR ilms ha e wo o mo e so magne ic
laye s o i on, nickel and cobal alloys sepa a ed by
a nonmagne ic conduc i e laye such as coppe .
Because o spin-dependen sca e ing o conduc ion
elec ons, he esis ance has maximum alue when
he magne ic momen s o he laye s a e an i-pa allel
and minimum when hey a e pa allel, [5 ].
SDT s uc u es a e a ecen addi ion o he
ma e ials exhibi ing a la ge change in esis ance. In
hese s uc u es an insula ing laye sepa a es wo
magne ic laye s. Quan um unneling h ough he
insula o allows conduc ion. The angle be ween he
magne iza ion ec o s in he wo magne ic laye s
modula es he magni ude o he unneling cu en
be ween he wo laye s. The e we e obse ed
changes o esis ance o 10% o 40% in SDT
s uc u es. The ield equi ed o maximum change
in esis ance depends on he composi ion o he
magne ic laye s and he me hod o achie ing an i-
pa allel alignmen . Values o he sa u a ion ield
ange om 100 o 10000 A/m. A he low end, his
o e s he possibili y o ex emely sensi i e magne ic
senso s.
Wi hin he ame o senso s p og ess he e ha e
been adap ed SDT ma e ials o c ea e highly
sensi i e magne ic ield senso s o use in low- ield
applica ions ha p esen ly equi e luxga e
magne ome e s. These senso s a e e y small,
equi e li le powe , and a e easily combined wi h
o he elec onics. The insula ing unnelling laye
p o ides high- esis ance senso s sui able o ba e y
ope a ion. The e can be ab ica ed ex emely small
SDT de ices (se e al ens o mic ome e on a side)
wi h high esis ance using pho oli hog aphy,
allowing dense packing o magne ic senso s in small
a eas.
The main componen s o an eddy cu en p obe
o non-des uc i e es ing a e pancake- ype coil and
an AC b idge o GMR o SDT senso s. A angemen
o coil and GMR senso o eddy cu en de ec ion
o de ec s in conduc o s is shown in he Fig. 5, [5].
When measu ing he sensing axis, i mus be kep
he GMR p obe coplana wi h he su ace o
specimen. The exci a ion ield on he coil axis, being
pe pendicula o he sensing axis o he GMR, has
no e ec on he senso . In his way, he de ec ed
ield, which is he esul o he pe u ba ion o he
eddy cu en low pa hs caused by he c ack, is
sepa a ed om he exci a ion ield.
Eddy cu en induced in he su ace o a de ec -
ee specimen a e ci cula because o he ci cula
symme y o he ield p oduced by he coil.
Recen ends in elec omagne ic non-des uc i e es ing 325
Fig. 5. Se up o ECT wi h GMR senso .
The angen ial componen o he ield c ea ed by he
eddy cu en s is ze o a he loca ion o he senso .
In p esence o de ec s, he eddy cu en s a e no
longe symme ical and he p obe p o ides a
measu e o he pe u bed eddy cu en s caused by
unde lying laws. The size o he coil is ela ed o
he esolu ion necessa y o de ec he de ec s. Fo
la ge de ec s and o deep de ec s, la ge coils
su ounding he senso s a e equi ed. Small coils
loca ed close o he specimen a e necessa y o
esol e small de ec s.
Eddy cu en s shield he in e io o he
conduc ing ma e ial wi h he skin dep h ela ed o
he conduc i i y and he equency. By changing he
equency i is possible o p obe di e ing dep hs o
he ma e ial. GMR and SDT senso s wi h hei wide
equency esponse, om DC o he MHz ange, a e
well sui ed o his applica ion. The small size o he
sensing elemen inc eases he esolu ion o de ec
loca ion while he de ec o is as e -scanned o e he
su ace. Mo e apid scans can be p e o med using an
a ay o de ec o s, [6].
Wi hin he ecen de elopmen he e we e buil
he op imized EC p obe p o o ypes o de ec and
map di e en ypes o de ec s encoun e ed in
p ac ice. They a e e alua ed p obe pe o mance on
calib a ed slo s o di e en leng hs, wid hs, and
heigh s machined in o he op su ace, bo om
su ace, o edges o specimen. The esul s a e
combined wi h he esul s ob ained on a specimen
ha con ained eal c acks a i icially g own a ound
a hole. Finally, hey demons a ed magne ic p o ile
imaging by scanning a gi en objec using a high-
esolu ion p obe, [6].
EC p obes we e es ed on su ace c acks longe
han he exci a ion coil diame e . Quali y o he
maps p oduced when scanning his ype o de ec
depend on he ela i e o ien a ion o he sensi i e
axis o he SDT o GMR sensing elemen s wi h
espec o he c ack o ien a ion. Sho su ace c acks
can also be eliably de ec ed using small exci a ion
coils. The unidi ec ional sensi i i y o GMR and
SDT senso s enables he de ec ion o c acks a and
pe pendicula o he edge o a specimen. This
disc imina ion is possible because he sensi i e axis
o he senso can be o ien ed pa allel o he edge.
Consequen ly, he ou pu signal o he senso is
caused only by he c ack.
SDT senso s a e pa icula ly a ac i e o
nondes uc i e e alua ion, low- equency
applica ions, such as he de ec ion o deeply bu ied
laws. In con as , induc i e p obes ha e poo
sensi i i y a low equencies because hey a e
sensi i e o he ime de i a i e o he magne ic ield
a he han o he magni ude o he magne ic ield
c ea ed by he law. To de ec deep c acks, i is
necessa y o use la ge diame e exci a ion coils o
inc ease he pene a ion o he eddy cu en s in he
ma e ial unde es .
4. CONCLUSION
The main cha ac e is ics o ECT p obes we e
desc ibed and compa ed in he pape . Acco ding o
he a ious equi emen s on he de ec ed signals and
o he used ECT echniques he applica ions o usual
and selec ed ad anced p obe de ices we e discussed
and mu ually compa ed. As o he las ECT sensing
ends he emphasis was pu mainly on he solid s a e
GMR and SDT senso s and hei p ope ies such as
he cons an sensi i i y in a wide equency ange,
hei small size and he possibili y o low magne ic
ields measu emen . The examples o he wide
u iliza ion o hese ypes o ECT senso s we e
p esen ed mainly o he p ac ical cases whe e he
ECT coil p obes applica ions ha e some limi a ions.
Acknowledgemen
This wo k was suppo ed by g an o he Slo ak
G an Agency VEGA, P ojec No. 1/0308/08.
REFERENCES
[1] Bli z, J.: Elec ical and Magne ic Me hods o
Nondes uc i e Tes ing, Edi ed By Chapman and
Hall, 1997, ISBN 0-412-79150-1.
[2] Janoušek, L., Gombá ska, D., ápo á, K.: A new
app oach o enhancing eddy- cu en non-
des uc i e e alua ion. Elec oScope, on-line jou nal,
h p://elec oscope.zcu.cz, Vol. 2007, No. 4, ISSN
1802-4564, 6 p.
[3] Janoušek, L., ápo á, K., Yusa, N., Miya, K.:
Ad anced p obe wi h a ay o pick-up coils o
imp o ed c ack e alua ion in eddy-cu en non-
des uc i e es ing. P oceeding o abs ac s om he
The 12 h In e na ional Wo kshop on Elec omagne ic
Non-Des uc i e E alua ion, ENDE 07, Ca di , UK,
June 2007, pp. 33
[4] Ma ek, T., Gombá ska, D., Janoušek, L., ápo á, K.,
Yusa, N., Miya, K.: Design o a emo e ield eddy
cu en p obe dedica ed o inspec ion o a magne ic
ube om i s ou e su ace. S udies in Applied
Elec omagne ics and Mechanics, Vol.28, 2007, IOS
P ess, ISSN 1383-7281, pp. 147-152.
[5] Ca uso, M.J., B a land, T., Smi h, C.H., Schneide ,
R.: A NewPe spec i e on Magne ic Field Sensing.
Senso s Magazine, Vol. 15, No. 12, 1998, pp. 34-46.
[6] Smi h, C.H., Schneide , R.W., Doga u, T., Smi h,
S.T.: Eddy-cu en es ing wi h GMR magne ic
senso a ays. Re iew o P og ess in Quan i a i e
Nondes uc i e E alua ion, Vol.23, NY, USA, 2003,
pp. 406-413.