A NEW EDDY CURRENT SURFACE PROBE FOR SHORT FLAWS
WITH MINIMAL LIFT-OFF NOISE
H. Hoshikawa, K. Koyama, and M. Maeda
Nihon University, Izumicho Narashino Chiba 275-8575, Japan
ABSTRACT. The authors have devised a new eddy current surface probe that generates minimal lift-off
noise and provides phase information on depth of flaws. The probe comprises a tangential exciting coil and
two tangential detecting coils. The authors expect that the new probe will make eddy current testing more
reliable in detecting flaws and more quantitative in evaluating depth of flaws than the conventional probes.
INTRODUCTION
Conventional eddy current probes detect flaws in the material by detecting the variation of
the exciting coil impedance or of the eddy current circulating along the exciting coil. However,
the eddy current circulating along the exciting coil changes not only by flaws but also by the
variation of the probe lift-off from the test material. The large lift-off noise makes a shambles of
the signal phase and prevents eddy current testing from utilizing the phase. Thus the signal phase
can hardly have been utilized to evaluate flaws. As a result, only the amplitude of flaw signals
has been used to evaluate flaws for the conventional surface probes. Since the signal amplitude
changes not only by the depth of flaws but also by the length and width, eddy current testing has
not been considered as a quantitative method of evaluating depth of surface flaws.
The authors already devised a couple of eddy current surface probes with minimal lift-off
noise [1-3]. However, @ probe reported last year [3] has a problem in evaluating short flaws.
Demand for more reliable eddy current testing requires more diverse probes. Thus the authors
have devised yet another eddy current surface probe combining a tangential exciting coil and
two tangential detecting coils. The experimental results have indicated that the new probe can
detect shorter surface flaws with minimal lift-off noise. The new probe generates a
figure-eight-like signal pattern as it scans over a flaw. The phase of the signal changes according
to depth of flaws with little influence from the length, width, and direction of flaws. The probe
makes it possible to evaluate depth of flaws based on the signal phase, which makes eddy
current testing more reliable and quantitative than the conventional method. Thus the eddy
current testing using the new probe provides the phase information on depth of flaws just like the
inner bobbin coil probe used for the eddy current inspection of tubing.
CP657, Review of Quantitative Nondestructive Evaluation Vol. 22, ed. by D. O. Thompson and D. E. Chimenti
© 2003 American Institute of Physics 0-7354-0117-9/03/S20.00
413
A NEW PROBE COMBINING TANGENTIAL COILS
Conventional eddy current probes suffer from large lift-off noise because they detect the
eddy current induced in the test material by the exciting coil or the perpendicular magnetic flux
to the test material surface. Since the eddy current changes drastically by the probe lift-off from
the test material, it is inevitable for conventional probes to suffer from large noise. The authors
have conceived the following idea of a new lift-off noise free eddy current probe. A probe can
be lift-off noise free if it detects the eddy current generated by a flaw and not directly by the
exciting coil. Such a probe can be realized by arranging a tangential exciting coil perpendicular
to tangential detecting coils that detect only the magnetic flux parallel to the material surface and
perpendicular to them or only the eddy current parallel to them.
FIGURE 1 shows a newly developed eddy current probe combining a tangential exciting
coil and two tangential detecting coils arranged perpendicularly to on both sides of the exciting
coil. The tangential exciting coil induces eddy current parallel to itself. Two tangential detecting
coils connected additively in series detect only the magnetic flux parallel to the material surface
and perpendicular to them or only the eddy current component circulating parallel to them. The
probe is named Plus Probe since it looks like a plus sign when seen from its top.
FIGURE 2 shows the schematic eddy current circulation at the surface of the test
material. When there is no flaw in the material, the eddy current induced by the tangential
exciting coil circulates only perpendicular to the tangential detecting coils as shown in Figure
2(a). Thus the detecting coils generate no signal because they only pick up the eddy current
component circulating parallel to them. The detecting coils generate no signal by the variation of
the probe lift-off from the test material that causes the eddy current to change in amplitude and
not in the circulating direction. On the other hand, if the test material has a flaw parallel to the
detecting coils as shown Figure 2(b), some of the eddy current circulates along the flaw and
causes the detecting coils to generate a signal. Thus Plus Probe can pick up flaws with no lift-off
noise in principle. Plus Probe can also eliminate the troublesome bridge balance procedure from
eddy current testing because it generates no signal as long as there is no flaw in the test material.
/Test material
Detecting coils
FIGURE 1. A new eddy current surface probe with minimal lift off noise that comprises tangential coils.
414
Eddy
Eddy current
current
Eddy
Eddycurrent
current
Flaw
Flaw
\
\
/
Detecting
Detecting
Detectingcoils
coils
Detecting coils
coils
(a)
(b)
(a)with
withnonoflaw
flaw
(b)with
withaaflaw
flaw
by the
the tangential
tangentialexciting
excitingcoil.
coil.
FIGURE
FIGURE2.2. Schematic
Schematiccirculation
circulationofofeddy
eddycurrent
currentatatthe
the surface
surface of
of the
the test
test material
material by
Test
Testfrequency
frequency: :20kHz
20kHz
0.2
Flaw
Flaw ::width
width 0.5mm
0.5mm
0.2
In-phase
in phasecomponent
component
0.1
0.1
In-phase
in phasecomponent
component
0.1
o
0
-0.1
-0.1
-0.1
0
-028
-0.2
20
10
0
-10
-20
20
10
0
-10
-20 -0.2
-20
-10
0
10
20
-20
-10
0
10
20
(a)
(b)
(a)Θ0 probe
probe
(b) Plus Probe
FIGURE
depth.
FIGURE3.3. Signals
Signalsobtained
obtainedby
by Θ
0 probe
probeand
andPlus
PlusProbe
Probe for
foraa surface
surface slit
slit flaw of 5mm
5mm length and 80% depth.
Figure2(b)
2(b)also
alsoindicates
indicatesthe
the flaw
flaw signal
signal pattern
pattern by
by Plus
Plus Probe
Probe as
as itit scans
scans over
over aa flaw.
flaw.
Figure
Sincethe
theeddy
eddycurrent
currentcirculates
circulatesininopposite
oppositedirections
directionsabove
above and
and below
below the
the flaw,
flaw, the tangential
Since
detectingcoils
coilsgenerate
generateaaplus
plusororminus
minus signal
signal depending
depending on
on their
their position
position with
with respect
respect to
to the
the
detecting
flaw.When
Whenthe
thedetecting
detectingcoils
coilsare
areright
right over
over the
the flaw,
flaw, they
they do
do not
not generate
generate any
any signal
signal because
flaw.
theeddy
eddycurrents
currentscirculate
circulateininthe
theopposite
oppositedirections
directions atat both
both sides
sides of
of the
the flaw
flaw and
and cancel out the
the
detectingcoil
coilsignal.
signal.Thus
ThusPlus
PlusProbe
Probegenerates
generatesaafigure
figure eight
eightlike
likesignal
signal pattern
pattern as
as itit scans
scans over
over
detecting
a
flaw.
Since
the
probe
generates
practically
minimal
lift-off
noise
and
the
probe
lift-off
a flaw. Since the probe generates practically minimal lift-off noise and the probe lift-off
influenceslittle
littleon
onthe
thephase
phaseofofflaw
flawsignals,
signals,the
thesignal
signalphase
phasecan
canbe
be used
used for
for evaluating
evaluating depth of
influences
flaws.
flaws.
EXPERIMENTALRESULTS
RESULTS
EXPERIMENTAL
Experimentswere
wereconducted
conductedfor
forPlus
Plus Probe
Probe that
that consists
consists of
of aa tangential
tangential exciting
exciting coil
coil of
Experiments
19x19mm22and
andtangential
tangential detecting
detecting coils
coils of
of 7x9mm
7x9mm22.. Each
Each coil
coil was
was wound
wound with
with 1mm
1mm22 cross
cross
19x19mm
section.Test
Testmaterials
materials were
were brass
brass plates
plates ofof 1.5mm
1.5mm thickness.
thickness. Each
Each test
test plate
plate has
has an
an electric
electric
section.
discharge machined
machined slit
slit flaw
flaw and
and each
each flaw
flaw isis different
different inin depth,
depth, length,
length, and
and width.
width. Test
Test
discharge
frequencyofof20
20kHz
kHzwas
waschosen
chosentoto make
make 1.5
1.5ofofthe
the ratio
ratio of
of the
the plate
plate thickness
thickness to
to the
the standard
standard
frequency
penetrationdepth
depthfor
forthe
thebrass
brassplate.
plate.
penetration
Figure33shows
showsin-phase
in-phasecomponent
componentofofsignals
signalsobtained
obtainedby
by Θ@ probe
probe[3]
[3]and
andPlus
Plus Probe
Probe
Figure
theyscan
scanover
overaasurface
surface flaw
flaw ofof5mm
5mm length
length and
and 80%
80% depth.
depth. The
The figure
figure indicates
indicates that
that Plus
Plus
asasthey
Probegenerates
generatessimpler
simplerflaw
flaw signal
signalthan
than Θ
@ probe.
probe.
Probe
415
Test frequency : 20kHz
Flaw : length 15mm, width 0.5mm, depth 0~80%
flaw signal
lift-off noise
3 1
s
a
§0.5
§0.5
lift-offnoise
= 0.08~0.59mm
0
0.5
1
In-phase component
0
0.5
1
In-phase component
(a) 0 probe
FIGURE 4. Flaw signals and lift-off noises in the voltage plane.
Test frequency ' 20kHz
(b) Plus Probe
Flaw ' length 15mm, width 0.5mm
0.07
£0.1 -
t»
\ \
\ \
aa
<-^^^->
1
~od
t
- flaw depth
— — 80%
-•-60%
-0.1 -----40%
—— 20%
i
.
X
'f
. flaw depth
— — 80%
-•-60%
—-40%
l
.
^V
N
\
\
\
\
\
v^ )
i
-0.1
0
0.1
In-phase component [V]
07
0
In-phase component [V]
(a) front surface flaws
(b) back surface flaws
FIGURE 5. Signal patterns by Plus Probe for flaws with different depths.
Experiments on lift-off noise were conducted by inserting thin papers of different
thickness between the probe and the test material. Figure 4 indicates that both @ probe and Plus
Probe generate far larger flaw signals than lift-off noise. Thus the new probe can detect flaws
with far higher signal-to-noise ratio than conventional probes.
Figure 5 shows signal patterns obtained by Plus Probe for front surface flaws and back
surface flaws with different depths. Figure 5(a) indicates that the signal phase advances forward
as the depth of front surface flaws increases. On the other hand, the phase lags behind as the
depth of back surface flaws increases as shown in Figure 5(b). Figure 5 also indicates that the
signal phase is quite stable against the fluctuation of the probe lift-off.
Figure 6 shows flaw signal patterns obtained by @ probe and by Plus Probe for flaws
with different lengths. Figure 6(a) indicates that the amplitude of flaw signals by @ probe
changes a lot as the flaw length decreases. On the other hand, the amplitude of flaw signals by
Plus Probe does not change much as shown in Figure 6(b). Figure 6 also shows that the signal
phase does not change much by the flaw length. Experimental results have also indicated that the
width of flaws and the angle of flaws with respect to the detecting coils change the amplitude of
flaw signals a lot but keep the phase almost constant.
416
Test frequency: 20kHz
Flaw : depth 80%, width 0.5mm
r o.i
0.1
I
a
a
o
3
-0.1
-0.1
-0.1
_L
0
_L
0.1
flaw length
— •— 15mm
—--10mm
——— 5mm
-0.1
0
0.1
In-phase component [V]
In-phase component [V]
(a) 0 probe
(b) Plus Probe
FIGURE 6. Signal patterns by & probe and by Plus Probe for different flaw lengths.
Test frequency: 20kHz
Flaw ' length 15mm, width 0.5mm
0.5,————,———
3
0
Uft-ofllmm)
—---0
—— 0.08
—— 0.16
—— 0.59
———1.05
{
-O.i
-0.1
0
0.1
In-phase component
)
0
0.5
In-phase component [V]
(a) 0 probe
(b) Plus Probe
FIGURE?. Signal patterns by 0 probe and by Plus Probe for different lift-offs.
Figure 7 shows signal patterns for a flaw of 100% depth by @ probe and by Plus Probe
when the lift-off of the probes changes from 0 to 1.05mm. The lift-off causes the amplitude to
change much while it causes the phase less.
Figure 8 shows variation of the signal phase for a 100% depth flaw vs. probe lift-off. The
figures indicate that the phase variation is only couple of degrees if the variation of the lift-off is
kept within 0.2mm. Thus Plus Probe provides eddy current testing a method of evaluating flaw
depth by utilizing the signal phase without much influence from the probe lift-off variation.
The authors have derived the flaw depth evaluation curve for the brass plate based on the
phase of flaw signals as shown in Figure 9. Thus the depth of flaws in the brass plate can be
evaluated by applying the phase of flaw signals to the curve in the figure without much influence
from the variations of flaw length and width. The relation between signal phase and flaw depth
is just the same as the one known for the tubing inspection by the inner bobbin coils probe. The
authors expect that the flaw depth evaluation method based on the signal phase will improve the
accuracy of flaw depth evaluation in eddy current testing.
417
Test frequency ' 20kHz
Flaw ' length 15mm, width 0.5mm
bfl
-8
- Plus probe
- ©probe
-20,
0
0.5
1
lift-off[mm]
1.5
FIGURE 8. Phase variations of the signal for a 100% depth flaw by lift-off.
Test frequency ' 20kHz
Flaw ' length 15mm, width 0.5mm
lOOr
-100
-50
0
50
Flaw signal phase[deg]
FIGURE 9. Flaw depth evaluation curve based on signal phase.
CONCLUSION
The experimental results have indicated that Plus Probe generates only minimal lift-off
noise and provides a way to evaluate flaw depth based on signal phase without much influence
from the length and width. The authors expect that the new probe utilizing both amplitude and
phase of the flaw signal will make the eddy current testing more reliable in detecting flaws and
more quantitative in evaluating flaws than the conventional probes utilizing only the flaw signal
amplitude.
REFERENCES
1.
2.
3.
Hoshikawa, H. and Koyama, K., "A New Eddy Current Probe Using Uniform Rotating
Eddy Current," Materials Evaluation, 56,1, 85-89 (1998).
Hoshikawa, H., Koyama, K. and Karasawa, H., "A New ECT without Lift-off Noise and
Phase Information on Flaw Depth," Review in Progress Quantitative NDE, 20A, 969-976
(2001).
Hoshikawa, H., Koyama, K. and Maeda, M., "Signal Phase Indication of Flaw Depth by a
Lift-off Noise Free Eddy Current Probe," Review in Progress Quantitative NDE, 21A,
430-437 (2002).
418