ADV ANCED MANUAL FOR EDDY CURRENT TEST METHOD

T ABLE OF CONTENTS

CHi\PTER 1 - SCOPE AND INTRODUCTION PAGE

1.1 SCOPE 1
1.2 EDDY CURRENT TESTING 1
1.3 HISTORICAL PERSPECTIVE OF EDDY CURRENT TESTING 2
1.4 ORGANIZATION OF MANUAL 2

CHAPTER 2 - EDDY CURRENT FUNDAMENTALS

2.1 BASIC EQUIPMENT 5

2.2 GENERA TION OF EDDY CURRENTS 6

2.2.1 Introduction 6
2.2.2 Magnetic Field Around a Coil 6
2.2.3 Equations Governing Generation of Eddy Currents 8

2.3 FUNDAMENTAL PROPERTIES OF EDDY CURRENT FLOW 10

2.4 SKIN EFFECT 11

2.4.1 Standard Depth of Penetration 12

2.4.2 Depth of Penetration in Finite Thickness Samples 13
2.4.3 Standard Phase Lag 14
2.4.4 Phase Lag in Finite Thickness Samples 15

2.5 SUMMARY 16
2.6 WORKED EXAMPLES 17
,
2.6.1 Standard Depth of Penetration and Phase Lag 17
1

CHAPTER 3 - ELECTRICAL CIRCUITS AND PROBE IMPEDANCE

3.1 INTRODUCTION 18
3.2 IMPEDANCE EQUATIONS AND DEFINITIONS 18
3.3 SINUSOIDS, PHASORS AND ELECTRICAL CIRCUITS 20
3.4 MODEL OF PROBE IN PRESENCE OF TEST MATERIAL 22
3.5 SIMPLIFIED IMPEDANCE DIAGRAMS 23

3.5.1 Derivation of Probe Impedance for Probe/Sample Combination 23

3.5.2 Correlation Between Coil Impedance and Sample Properties 26

3.6 SUMMARY 28
3.7 WORKED EXAMPLES 29

3.7 .1 Probe Impedance in Air 29

3.7.2 Probe Impedance Adjacent to Sample 30
3.7 .3 Voltage-Current Relationship 30
CHAPTER 4 - INSTRUMENTATION

PAGE

4.1 INTRODUCTION 31
4.2 -BRIDGE CIRCUITS 32

4.2.1 Simple Bridge Circuit 32

4.2.2 Typical Bridge Circui t in Eddy Current Instruments 33
4.2.3 Bridge Circui t in Crack Detectors 34

4.3 RESONANCE CIRCUIT AND EQUATIONS 35

4.4 EDDY CURRENT iNSTRUMENTS 36
4.4.1 General Purpose Instrument (Impedance Method) 36
4.4.2 Crack Detectors 38
4.4.3 . Material Sorting and Conductivity Instruments 40

4.5 :SENn~RECEIV£ EDDY CURRENT SYSTEMS

4.5.1 Hall-Effect Detector 42

4.5.2 Send-Receive Coils and Lift-Off Compensation 43
4.6- MUL TIFREQUENCY EQUIPMENT -44
4.7 PULSED EDDY CURRENT EQUIPMENT 45
4.8 'SPECIAL TECHNIQUES 46
4.9 RECORDING EQUIPMEN-T 46
4.9.1 Frequency Response 48
4.10 SUMMARY . 4&
4.11 WORKED EXAMPLES 49

4.11.-1 lmpedan-ceat Resonance

CHAPTER 5 - TESTING WITH SURFACE PROBES

5.1 INTRODUCTIQN- 51
5.2 -SURFACE :PRQBES 51

5.2.-1 Probe Types 51

5.2:2 Directional Properties 55

5.2.2.1 Sensitivity at Centre ofa Coil 55

523 Probe Inductance 57

5.3 PARAMETERS AFFECTING SENSITIVITY TO DEFECTS 60

5.3.1 Sensi ti vity with Lift-off and Defect Depth 60

5.3.2 Effect of Defect Length 61
I .1 1I'j 1 KUUUL ttUN 98
7.2 PROBES FOR TUBES AND CYLINDRICAL COMPONENTS 98

T.2.1 Probe Types 98

7.2.2 Comparing Differential and Absolute Probes 100
7.2.3 Directional Properties 102
7.2.4- Probe Inductance 103
7.2.5 Probe-Cable Resonance 105
7.3 IMPEDANCE PLANE DIAGRAMS 106

7.3.1 Solid Cylinders, 107

Z.3.1.1 Sensitivity in Centre of-a Cylinder 109

7.3.2 Tubes lil

7.3.3 Characteristic Frequency' for-Tubes 112
7.3.4 Computer Generated Impedance Diagrams 115

7..4 CHOICE OF TEST FREQUE-NCY 116-

7.4.1 Test Frequency for -Solid Cylinders 116

7.4.2 Test FrequencyJor Tubes 116

7.5 PROBES FOR DETECTING CIRCUMFERENTIAL CRACKS 117

7.6 SUMMARY 119
7.7 WORKED EXAMPLES 120
7.7.1 Calculate fffg to operate at knee location, for a cylinder 120
T.7.2-(a) Calculate optlmum test frequency for tube inspection 121
(b) Determine operating point for above frequency- 121
(c) Calculate-frequency to discriminate ferromagnetic indications 1"21

CHAPTER 8. --TUBE TESTING - SIGNAL ANALYSIS

8-.:1 INTRODUCTION 123

8.2 EDDY CURRENT SlGNALS 123

8.2.1 Defect Signal Characterisitics 123

8.2.2 Effect of Test Frequency 125
8.2.3 Calibration Tubes and Simple_ Defects 128
8.2.4 Vectorial Addition and Defects at-Baffle Plates 1.33
8.2.5 Tube Inspection at Tubesheets 137
8.2.6 Testing Tubes with Internal Suriace Probes 138

8.3 ANOMALOUS EDDY CURRENT SIGNALS 140

8.3.1 Ferromagnetic Inclusions and Deposits 141

8.3.2 Conducting Deposi ts 144
 PAGE

5.4 COMPARAISON BETWEEN SURFACE AND THROUGH-WALL INSPECTION 62

5.5 IMPEDANCE GRAPH DISPLAY 64

5.5.1 Effect of Resistivity 67

5.5.2 Effect of Permeability 67
5.5.3 Effect of Thickness 67
5.5.4 Effect of Frequency 67
5.5.5 Effect of Probe Diameter 57
5.5.6 Comparison of Experimental and Computer Ii11pedance Diagrams 68

5.6 CHARACTERISTIC PARAMETER 69

5.7 DEFINITION OF "PHASE" TERMINOLOGY 72
5.8 SELECTION OF TEST FREQUENCY 73

5.8.1 Inspecting for Defects 73

5.8.2 Measuring Resistivity 74
5.8.3 Measuring Thickness 77
5.8.4 Measuring Thickness of a Non-Conducting Layer on a Conductor 78
5.8.5 Measuring Thickness of a Conducting Layer on a Conductor 79

5.9 PROBE-CABLE RESON ANCE 80

5.10 SUMMARY 81
5.11 WORKED EXAMPLES 82

5.11.1 Effective Probe Diameter 32

5.11.2 Characteristic Parameter 82

CHAPTER 6 - SURFACE PROBE SIGNAL ANALYSIS

6.1 INTRODUCTION 83
6.2 EDDY CURRENT SIGNAL CHARACTERISTICS 83

6.2.1 Defect Signal Amplitude 83

6.2.2 Defect Signal Phase 85

6.3 EFFECT OF MATERIAL VARIATIONS AND DEFECTS IN A

FINITE THICKNESS 87
6.4 COIL IMPEDANCE CHANGES WITH DEFECTS 90

6.4.1 Surface Defect Measurement 90

6.4.2 Subsurface Defect Measurement 90

6.5 COIL IMPEDANCE CHANGES WITH OTHER VARIABLES 91

6.5.1 Ferromagnetic Indications 91

6.5.2 Electrical Resistivity 93
6.5.3 Signals from Changes in Sample Surface Geometry 'n

6.6 CALIBRATION DEFECTS 94

6.7 SUMMARY 97
 PAGE

8.4 MUL TIFREQUENCY EDDY CURRENT TESTING 146

8.4.1 Background 146

8.4.2 Multifrequency Testing of Dented Tubes 148

8.5 SUMMARY 152

CHAPTER 9 - METALLURGICAL PROPER TIES AND TESTING FERROMAGNETIC

MATERIALS

9.1 INTRODUCTION 154

9.2 ELECTRICAL CONDUCTIVITY 154

9.2.1 Factors Affecting Resistivity 154

9.2.2 Material Sorting by Resistivity 157

9.3 MAGNETIC PROPERTIES 159

9.3.1 Magnetic Hysteresis 159

9.3.2 Magnetic Permeability 160
9.3.3 Factors Affecting MagneticPermeabili ty 162
9.4 TESTING MAGNETIC MATERIALS 165
9.4.1 Simplified Impedance Diagrams 165
9.4.2 Impedance Diagrams 167
9.4.3 Material Sorting by Magnetic Permeability 169
9.4.4 Testing for Defects in Magnetic Materials 169
9.5 SUMMARY 175
9.6 WORKED EXAMPLES 175

9.6.1 Calculate Conductivity 175

9.6.2 Calculate Magnetic Permeability 176
9.6.3 Calculate Standard Depth of Penetration 176

CHAPTER 10 - SUPPORTING INFORMATION

10.1 NOMENCLATURE 177

10.2 DEFINITIONS 178
10.3 ABBREVIA TIONS FOR NONDESTRUCTIVE TESTING TERMS 186
10.4 REFERENCES 187
10.5 INDEX 188
 ADVANCED MANUAL FOR:
EDDY CURRENT TEST METHOD

CHAPTER 1 - SCOPE AND INTRODUCTION

1.1 SCOPE

This manual covers the principles of the eddy current Inethod of nondestructive
testing including relevant electromagnetic theory, instrumentation, testing
techniques and signal analysis.

It is intended to act as;

a source of educational material to persons who are seeking certific-1tion

according to the requirements of 48-GP-13M - Certification of Nondestructive
Testing Personnel (Eddy Current ~v1ethod), and

as a guide and reference text for educational organizations and training centres
that are providing or planning courses of instruction in Eddy Current Testing.

Note that the degree of scientific detail in this manual is pri:narily directed towards
Level II and Level III certification applicants. It is assumed that the reader is familiar
with basic eLectrical theory and the elements of algebra and calculus. Many sui table
textbooks and reference manuals are available to Level I applicants to be used prior
to or in conjunction with this publication. The handbook referenced in item 5 of
section IO.1f is particularly suited to the requirements for Level I expertise.

1.2 EDDY CURRENT TESTING

Eddy current testing (ET) is a nondestructive test technique based on inducing

electrical clJrrents in the material being inspected and observing the interaction
between those currents and the material. Eddy currents are generated by
electromagnetic coils in the test probe, and monitored simultaneously by measuring
probe electrical impedance. Since it's an electromagnetic induction process, direct
electrical contact with the sample is not required; however, the sample Inaterial has
to be conductive.

Eddy current testing is a versatile technique. It's mainly used for thin materials; in
thick materials, penetration constraints limit the inspected volulne to thin surface
layers. tn addition to flaw inspection, ET can be used to indirectly measure
mechanical and metallurgical characteristics which correlate with electrical and
magnetic properties. Also, geometric effects such as thickness, curvature and probe-
to-material spacing influence eddy current flow and can be measured.

The large number of potentially signif icant variables in ET is both a strength and a
weakness of the technique since effects of otherwise trivial parameters can mask
important information or be misinterpreted. Virtually everything that affects eddy
 - 2-

current flow or otherwise influence probe impedance has to be taken into account to
ootai:) reliable r esults. Thus, credible eddy current testing requires a high level of
o~erator training and awareness.

l.3 I-USTORICAL PERSPECTIVE OF EDDY CURRENT TESTING

Electromagnetic testing - the interaction of magnetic fields with circulating

elec trical curents - had its origin in 1831 when M. Faraday discovered
electromagnetic induction. He induced current flow in a secondary coil by switching
d battery on and off. D.E. Hughes performed the first recorded eddy current te:;t in
1879. He was able to distinguish between different metals by noting a change in
excitation frequency resulting from effects of test material resistivity and magnetic
permeaoili ty.

Initially, the extrerne sensitivity to many material properties and conditions made ET
difficul t and unreliable. Figure 1.1 illustrates this point. It took until 1926 before the
firs t eddy current instrument was developed to measure sample thickness. By the end
of World War II further research and improved electronics made industrial inspection
possible, and many practical instruments were developed. A major breakthrough came
in the 1950's when Forster developed instruments with impedance plane signal
displays. These made it possible to discriminate between different parameters,
though the procedure was still empirical. During the 1960's progress in theoretical
and practical uses of eddy current testing advanced the technology from an empirical
art to an accepted engineering discipline.

During that time, other nondestructive test techniques such as ultrasonics and
radiography became well established and eddy current testing played a secondary
role, mainly in the aircraft industry. Recent requirements - particularly for heat
exchanger tube inspection in the nuclear industry - have constributed significantly to
further develop,nent of ET as a fast, accurate and reproducible nondestructive test
technique.

Until recently, eddy current testing was a technology where the basic principles were
known only to researchers, and a "black box" approach to inspection was often fol-
lowed. The authors' objective in compiling this manual is to draw upon research,
laboratory and industrial inspection experience to bridge that gap and thereby permit
the full potential of eddy current testing to be realized.

1.4 ORGANIZATION OF MANUAL

The presentation moves from theory (including a review of basic electrical concepts)
to test methods and signal analysis. Simplified derivations of probe response to test
parameters are presented to develop a basic understanding of eddy current test
principles. Thus, eddy current signals are consistently illustrated on impedance plane
diagrams (the display used in modern eddy current test instruments) and to aid
explanation, the parameter "eddy current phase lag" is introduced.

Since probes playa key role in eddy current tes ting, technical aspects of probe desien
are introduced with pertinent electrical impedance calculations. While knowledge of
basic electrical circuits is required for a complete understanding of eddy current test
j - 3-

principles, a good technical base for inspection can still be obtained if sections of this
manual requiring such a background are skipped.

From an applications point of view, the material in this manual provides an inspector
with the necessary background to decide:
1) what probe(s) to use,
2) what test frequencies are suitable,
3) what calibration defects or standards are required for signal calibration and/or
simulation,
4) what tests are required to dif ferentia te between significant signals and false
indications,
5) how to estimate depth of real defects.

To supplement theory, practical examples are presented to develop proficiency in

performing inspections, and to illustrate how basic principles are applied to diagnose
real signals.

It may be useful to review sections 10.1, 10.2 and 10.3 before continuing and to refer
to these sections as necessary while reading this manual.
 - 4-

'..,,-- ,.?

THERE MUST BE DEFECTS IN

THESE TUBES SOMEWHERE -
I 5,4W SQUIGGLES ON THE EDDY
CURR£Nt SCREEN!

FIGURE 1.1
Misinterpreted Signals
 ·
!
j

....!

_i
!

- I
i

-I
-I
I

-1 j

~l
:

- ;
 - 5-

CHAPTER 2 - EDDY CURRENT FUNDAMENTALS

2.1 BASIC EQUIPMENT

Basic eddy current test equipment consists of an alternating current source

(oscillator), a probe containing a coil connected to the current source, and a
voltmeter which measures the voltage change across the coil, as shown in Figure 2.1.

OSCILLATOR VOL TMETER

CURRENT t PROBE

PROBE
MOVEMENT
..
CRACK
/
TEST PLATE

FIGURE 2.1
Eddy Current Test Equipment

The oscillator must be capable of generating a time varying (usually sinusoidal)

current at frequencies ranging from about 1 kHz (1000 cycles per second) to about
2 MHz (2,000,000 cycles per second). Oscillators which operate at higher or lower
frequencies, or with pulsed currents, are used for specialized applications.

The colI within the probe is an insulated copper wire wound onto a suitable form. The
wire diameter, the number of turns and coil dimensions are all variables which must
be determined in order to obtain the desired inspection results. Coll variables are
discussed in later chapters.
 - 6-

Depending upon the type of inspection, an eddy current probe can consist of a single
test coil, an excitation coil with a separate receive (sensing) coil, or an excitation
coil with a Hall-effect sensing detector, as shown in Figure 2.2.

VOLT ME TER VOL T.'"IE TER VOLTME TER

OSCILLATOR OSCILLATOR OSCILLATOR

SENSING HALL
COIL DETECTOR

COIL
EXCITATION EXCITATION
COIL COIL

(A) SELF-[NnUCTANCE (B) SEND-RECEIVE (e) MAGNETIC REACT[ON

FIGURE 2.2
Eddy Current Inspection Systems

The voltmeter measures charges in voltage across the coil which result from changes
in the electrical conditions and properties of the conducting material tested and/or
changes in relative position between the coil and the material tested. This voltage
change consists of an amplitude variation and a phase variation relative to the
current passing through the coil. The reason for amplitude and phase changes in this
voltage is discussed in Chapter 3.

2.2 GENERA nON OF EDDY CURRENTS

2.2.1 Introduction

In this section the topic of the magnetic field surrounding a coil carrying current is
introduced together with the mechanism by which eddy currents are induced and how
they are measured.

2.2.2. Magnetic Field Around a Coil

Oersted discovered that whenever there is an electric current, a magnetic field

exists. Consider electric current directed along a wire, a magnetic field is created in
such a direction that if your right-hand thumb points in the direction of current, your
curled fingers point in the direction of the magnetic field. This is the "right-hand
rule".
 - 7 -

:;ssociated with a magnetic field is magnetic flux density. It has the same direction
as the magnetic field and its magnitude depends upon position and current. It is
therefore d. field vector quantity and is given the symbol B. Its units in the 5I system
is the tesla (T) or webers per square metre(Wb/m 2 ).

The B-field distribution around a long straight wire is shown in Figure 2.3(a). In
Figure 2.3(b) the B-field in the axial direction of a single turn is shown as a function
of radius. As more windings are added, each carrying the same current, the flux
densi ty rapidly increases and its associated distribution is altered.

B B

r r

;1
I
t t j
I I
Magnetic. F.<.eld I
I!
cp

(a) Straight Wire (b) Single Turn Coil

CU!tJtenX Flowing into page.

FIGURE 2.3
Magnetic Flux Distribution

Flux density varies linearly with electric current in the coil, i.e., if coil current
doubles, flux density doubles everywhere. The total magnetic flux,4>p .contained
within the loop is the product of B and the area of the coil. The unit in the 51 system
for magnetic flux is the weber (Wb).
 - 8-

2.2.3 Equations Governing Generation of Eddy Currents

In any eiectrical circuit, current flow is governed by Ohm's Law and is equal to the
driving (primary circuit) voltage divided by primary circui t impedance.

I ::0 V /Z (2.1)
p p p
The eddy current coil is part of the primary circuit. The current passing through the
coil normally varies sinusoidally with time and is given by:

I - I sin(wt) (2.2)
p 0

where ,lois the peak current· value in the circuit and w (omega) is the frequency in
radians!s (w equals 2 'IT fv.ihen f is frequency in hertz).

From Oersted's discovery, a magnetic flux (4l p ) exists around a coil carrying
current (see Figure 2.4) proportional to the number of turns in the coil (N p) and the
current ( I p ) •

4lp (% N I (2.3)
P P

\ PROBE
(primary
circuit)

SAMPLE

(secondary
circuit)

FIGURE 2.4
Coil Carrying Alternating Current Adjacent to a Test Sample

Faraday's Law states that a voltage (V s) is created or induced in a region of space

when there is a changing magnetic field. When we apply this to our coil,

V
s - -
(2.4)
 - 9-

d¢
where -df is the rate of change in ¢ p with time.

Since coil current varies sinusoidally with time, total magnetic flux in the coil also
varies sinusoidally,

<Pp '" <Po sin(wt)

where ¢ 0 is the magnetic flux corresponding to 10 ,

The induced voltage as described by equation 2.4 results in

v s - - Np w¢ 0 cos(wt) (2.5)
which also varies periodically with time. If we bring the coil close to a test sample,
Ohm's Law states that if there is a driving voltage (V &) and the sample's
impedance is finite, current will flow,

1
s
II: V /Z
5 S
(2.6)
where I s is current flowing through the sample, V 8 is induced voltage and Z s is
the sample's impedance or opposition to the flow of current.

These induced currents are known as eddy currents because of their circulatory paths.
According to Lenz's Law, they, in turn, generate their own magnetic field which
opposes the primary field,

¢S a: - I (2.7)
s

and (2.S)
where ¢E is the equilibrium magnetic flux surrounding the coil in the presence of a
test sample.

The flow of eddy currents results in resistive (Ohmic) losses and a decrease in
magnetic flux. This is reflected as a decrease in probe impedance. In equation form,

(2.9)

and V c Z1 (2.10)
P

Equation 2.9 indicates a coil's impedance is a function of the magnetic field

surrounding it and in turn the magnetic field is governed by induced current in the
specimen (equations 2.8 and 2.7). The relations between probe impedance and sample
properties will be derived in Chapter 3.

To summarize, flux is set up by passing alternating current through the test coil.
When this coil is brought close to a conducti ve sample, eddy currents are induced. In
addition, the magnetic flux associated with the eddy currents oppose the coil's
magnetic flux, thereby decreasing net flux. This results in a change in coil impedance
and voltage drop. It is the opposition between the primary (coil) and secondary (eddy
 - 10 -

current) fields that provides the basis for extracting information during eddy current
testing.

It should be noted that if a sample is ferromagnetic, equation 2.9 still applies but the
magnetic flux is strengthened despite opposing eddy current effects. The high
magnetic permeability of ferromagnetic materials distinguishes them from non-
ferromagnetic materials and strongly influences eddy current test parameters.

Ferromagnetic specimen inspection is discussed in Chapter 9 and unless specified the

rest of the manual is restricted to non-ferromagnetic materials.

2.3 FUNDAMENTAL PROPERTIES OF EDDY CURRENT FLOW

Eddy currents are closed loops of induced current circulating in planes perpendicular
to the magnetic flux. They normally travel parallel to the coil's winding and parallel
to the surface. Eddy current flow is limited to the area of the inducing magnetic
field.

Test frequency determines depth of penetration into the specimen; as frequency is

increased, penetration decreases and the eddy current distribution becomes denser
near the specimen's surface. Test frequency also affects the sensitivity to changes in
material properties and defects.

Figure 2.5(a) shows the algebraic relationships and Figure 2.5(b) the oscilloscope
display of eddy current and magnetic field distribution with depth into the specimen.
Both the eddy currents and magnetic flux get weaker with depth because of "skin
effect". In addition to this attenuation, the eddy currents lag in phase with depth.
Eddy currents' phase lag is the key parameter that makes eddy current testing a
useful NOT method. The parameters skin depth and phase lag are discussed in the
next section.
J •¢

COil 4>' ¢. S in(wt)

) ; -) S sin (wt )

~v~'-~' ",<W',
];-ls(x)sln(wt-E)

~~, -~, "I,,"'w'·a I ~:-----7"'----->oo.,,.-

\
(a) (b)

FIGURE 2.5
Eddy Current and Magnetic Flux Distribution With Depth Into a Conductoc
 - 11 -

2.4 SKIN EFFECT

Eddy currents induced by a changing magnetic field concentrate near the surface
adjacent to the excitation coil. The depth of penetration decreases with test
frequency and is a function of electrical conductivity and magnetic permeability of
the specimen. This phenomenon is known as the skin effect and is analogous to the
situation in terrestrial heat conduction where dally surface temperature fluctuations
are not appreciable below the earth's surface. Skin effect arises as follows: the eddy
currents flowing in the test object at any depth produce magnetic fields which oppose
the primary field, thus reducing net magnetic flux and causing a decrease in current
flow as depth increases. Alternatively, eddy currents near the surface can be viewed
as shielding the coil's magnetic field thereby weakening the magnetic field at greater
depths and reducing induced currents.

The equation for flow of induced currents is

2 elJ (2.11)
'V J c: OlJ at
where J is current density, a is conductIvity, lJ is magnetic permeability and 'i/ 2
Is a differential operator of second order.

For a semi-infinite (thIck) conductor the solution to the above equa tion is

J
x
IJ 0 - e -6 sin(wt-S) (2.l2a)

where J x I J 0 is the ratIo of eddy current density Jx at depth x to the surface

density J o , and e = 2.718 Is the base of natural logarithms. 8 is given by x/o
where c - (11" f lJ O )-1/2,
the standard depth of penetration (see next section).

Equation 2.l2(a) can be separated into two components:

J /J a: e -x/6 (2.l2b)
x 0

which describes the exponentIal decrease in eddy current density with depth, and

(2.l2c)

denoting the Increasing time or phase lag of the sinusoidal signal with depth.
 - 12 -

2.4.1 Standard Depth of Penetration

Figure 2.6 illustrates the change in eddy current density in a semi-infinite conductor.
Eddy current density decreases exponentially with depth. .

A-. sin (wt)

't'Q

\ II \ II \ \
0.2 0 .4 0 .6 0 .8·1 . 0

( .37 )

10

J x' ftIx

8=50~P ,mm
f ).tr

FIGURE 2.6
Eddy Current and Magnetic Flux Distribution With Depth in a Thick Plate

The depth at which eddy current density has decreased to lie or 36.8% of the surface
density is called the standard depth of penetration. The word 'standard' denotes plane
wave electromagnetic field excitation within the test sample (conditions which are
rarely achieved in practice). The standard depth of penetration is given by

o• 50/p/f~
r
• mm (2.13a)

or is a 2/p/f~ t inches (2.13b)

r

where P is electrical resistivity in microhm-centimetres, f is test frequency in hertz,

and \..I r is relative magnetic permeability (dimensionless)*.

* See Chapter 9 for a description of electrical and magnetic properties. \..I r - ).16 t

incremental permeability, at zero biassing magnetization flux.

 - 13 -

The skin depth equation is strictly true only for infinitely thick material and planar
magnetic fields. Using the standard depth, 6 , calculated fr o m the above equation
makes it a material/test paramete r rather than a true measure of penetrati on .

2.4.~ Oepth o f Penetration in F ini t e Thickness Samples

Sensitivity to defects depends on eddy current density at defect locati on. Although
eddy currents penetrate deeper than one standard depth o f penetration they decrease
rapidly with depth. A t two standard depths of penetration (2 C; ), eddy current
density has decreased toe 1 / e) 2 or 13.5% of the surface density. At three depths
( 36) the eddy current densi ty is down to only 5% of the surface density. However,
one should keep in mind these values only apply to thick samples (thickness, t > 56)
and planar magnetic excitation fields. Planar field conditions require la rge diameter
probes (diameter> lOt) in plate testing or long coils (length > 5 t) in tube
testing. Real test coils will rarely meet these requirements since they would posses s
low defect sensitivity. For thin plate or tube samples, current density drops off less
than calculated from equation 2.12(b) as shown in Figure 2.7. For solid cylinders the
overriding factor is a decrease to zero at the centre resulting from geometry effects
as shown in Fig. 2.7(c) and discussed in Section 7.3.1.

One should also note, that the magnetic flux is attenuated across the sample, but not
completely. Although the currents are restricted to flow wihtin specimen boundaries,
the magnetic field extends into the air space beyond. This allows the inspection of ·
multi-layer components separated by an air space.
t
1.0 ~ : 0.01
t
--..:::--._ /'" '8 : 0. 0 I
.8 t .8
'8 = 1.0 . . . . . . . . .!. = 0
JI
.6
t
.6 " 7
.
.4
'8 : 1.5
Jo J .4
t o rI
t
'8 : 2.0 - : 0.8
.2 .2 r.
&.2.0
0 o~~ __~~__~~
0 .2 .4 .6 .8 1.0 o .2 .4 .6 .8 1.0

-x I

t
(a) PLATE (LARGE cal L. 0 ) lOt) (0) TUBE (LONG ENCIRCLING Call. I )5t)

- - - EQUA T I ON 2 12 (b) J o = EDDY CURRENT DENSITY AT SURfACE

- - ACTUAL
J 1 OR J r = EDDY CURRENT DENSITY AT LOCATION
I .0 r - - - - - - - - - - , OR r BELa. THE SURfACE

r
I
O
r, ,
.8 a: 0.01 r><J t><J
vaZ270m - \
PLATE GEOWE TRY

.2

o .2 .4 .6 .8 1.0
t:::::::::: ;
TUBE AND ROO GEOWETRY
( r, : 0 FOR ROO)
(e) ROO (E ~C IRCLI ~G COIL. f > 5 ro)

FiGURE 2.7
Eddy Current Distribution With Depth in Various Samples
 , . ...... .
samples, this depth is about three standard depths of penetration. Unfortunately, for
:1Iost components and practical probe sizes, this depth will be less than 35 the eddy
currents being a ttenua ted more than predicted by the skin depth equa tion. The ef fect
of probe diameter on the decrease in eddy current density or defect sensitivity with
depth is discussed in Section 5.3.1.

2.4.3 Standard Phase Lag

The signal produced by a flaw depends on both amplitude and phase of the currents
being obstructed. A small surface defect and large internal defect can have a similar
effect on the magnitude of test coil impedance. However, because of the increasing
phase lag with depth, there will be a characteristic difference in the test coil
impedance vector. This effect allows location and extent of a defect to be
determined.

Phase lag is derived from equation 2.12(c) for infinitely thick material. It represents
a phase angle lag of x / 6radians between the sinusoidal eddy currents at the surface
and those below the surface. it is denoted by the symbol [3 (beta) and is given by:
B • x/a radians (2.14a)

or B • x/ 6 x 57 .d e g r e e s (2.14b)

where x is distance below the surface in the same units as 6.

cpo sin (wt)

r-----f-----'----- {3 (DEGREE S)

x x
{3: x 51, DEGREES
8 8

FIGURE 2.&
Eddy Current Phase Lag Variation With Depth in Thick Samples
 .;. 15 -

When x is equal to one star.dard depth of penetration, phase lag is 57° or one radian.
This means that the eddy current flowing below the surface, at one standard depth of
penetration, lag the surface currents by 57°. At two standard depths of penetration
they lag the surface currents by 114°. This is illustrated in Figure 2.8.

2.[1.4 Phase Lag in Finite Thickness Samples

For this samples, eddy current phase decreases slightly less rapidly with depth than
stated above. See Figure 2.9(a), (b) and (c) for the plots of phase lag with depth for a
plate, tube, and cylinder, respectively. The phase lag illustrated in these plots does
not change significantly with coil diameter or length. For thick samples and practical
probe sizes, equation 2.14 is sufficiently accurate.

O°c----------------.

20' .......
........
40' 40·
.......
........ } ~ = 0.1
II
f31 f3r
60· 6 O·

80· ao·
r1
- = 0.8 "- }i = 1.4
r Q

I 0 O' L---L_.....L-_.L-..--I.~-.J 100'

o .2 .4 . 6 .8 1.0 0 .2 .4 .6 .8 1.0

(a) FLArE (b) TUSE

o· 1><1 t><J
'i' JZZZZZZZZZ e
20' PLATE GEOMETRY
I~ = 1.0
II
40·
I..!. = 1.4
f3 r Il
60·
~ =I 0
8
8 O· 10
1..°=
Il 2 . 0
"8 =2
100· !.o
15 = 2 • I
0 .2 .4 .6 .8 I .0 TUBE A~O ROO (r 1 = 0) GEOMETRY

PHASE WITH DEPTH" OR I,RELATIVE

TO SURFACE CURRE~T

ACTUAL CURVES
(c) ROG
CALCULATED. EOUlTlON 2 14 (Il )

FIGURE 2.9
Eddy Current Phase Lag in Various Samples
 - 16 -

Phase lag can be visualized as a shift in time of the sinusoidally varying current
flowing below the surface. This was illustrated in Figure 2.5. Phase lag plays a key
role in the analysis of eddy current test signals. It will be used throughout the manual
to link theory and observations. It should not be misinterpreted or confused with the
phase angle between voltage and current in AC theory. Both the voltage and current
(and magnetic field) have this phase shift or lag with depth.

2.5 SUMMARY

Eddy current testing is based on inducing electrical currents in the material being
inspected and observing the interaction between these currents and the material.

This process occurs as follows: When a periodically varying magnetic field intersects
an electrical conductor, eddy currents are induced according to Faraday's and Ohrn's
Laws. The induced current (known as eddy currents because of their circulatory
paths) generate their own magnetic field which opposes the excitation field. The
equilibrium field is reduced resulting in a change of coil impedance. By monitoring
coil impedance, the electrical, magnetic and geometrical properties of the sample
can be measured. Eddy currents are closed loops of induced current circula ting in
planes perpendicular to the magnetic flux. They normally travel parallel to the coil's
winding and parallel to the surface. Eddy current flow is limi ted to the area of the
inducing magnetic field.

Depth of penetration into a material depends on its electrical resistivity, magnetic

permeability and on test frequency. The basic equation of ET is the standard depth of
penetration given by

C EO 50/p/f'oJ • mm (2.13a)
r

where p is electrical resistivity, microhm-centimetres;

f is test frequency, hertz;
and 'oJ r is relative magnetic permeability, dimensionless.

It states that in thick materials eddy current density decreases to 37% of the surface
density at a depth of one standard depth of penetration. In most eddy current tests,
especially with surface probes, the actual eddy current density (at a depth equal to
that calculated by equation 2.13a) is much less than 37%.

Eddy currents also lag in phase with depth into the material. Phase lag depends on the
same material properties that govern depth of penetration and is given by

B '"' x/a .. , radians (2.14a)

50/p/fu
r
where x is distance below surface, mm.

Phase lag is the parameter that makes it possible to determine the depth of a defect.
It also allows discrimination between defect signals and false indications. It is the key
parameter in eddy current tes ting.
 - 17 -

2.6 WORKED EXAMPLES

2.6.1 Standard Depth of Penetration and Phase Lag

PROBLEM:(a) Calculate the standard depth of penetration in a thick 304 sst

sample, at a test frequency of 100 kHz.

(b) Determine the eddy current phase lag at a depth of 1.5 mm in

304 sst at 100 kHz.

SOLUTION: 304 sst properties: P =72 microhm - centimetres

and 11 r - 1

(a) from equation 2. 13 (a),

6 - 50 72
3
100 x 10 xl

1/2
4
- 50(7.2 x 10- ) - 1. 3 mm

I
Therefore the standard depth of penetration is 1.3 mm.
)
(b) from equation 2.14(b),

B - x/o x 57

s 1.5 x 57 - 64 degrees
1.3

Therefore the phase lag is 64 degrees.

 - 18 -

CHAPTER 3 - ELECTRICAL CIRCUITS AND PROBE IMPEDANCE

3.1 INTRODUCTION

Eddy current testing consists of monitoring the flow and distribution of eddy currents
in test material. This is achieved indirectly by monitoring probe impedance during a
test. An understanding of impedance and associated electrical quantities is therefore
imperative for a fundamental appreciation of eddy current behaviour.

The first two sections review the electrical quantities important in eddy current
testing. This is followed by presentation of a model of a test coil coupled to test
material and its equivalent electrical circuit. The equivalent'circuit approach permits
derivation of simplified impedance diagrams 'to show the effect of test and material
parameters on probe impedance in graphical form. Once the simple impedance
diagram concepts of this chapter are understood, the more complex diagrams of
subsequent chapters should present little difficulty.

3._ IMPEDANCE EQUATIONS AND DEFINITIONS

All information about a sample is obtained through changes in electrical

characteristics of the coil/sample combination. Therefore a basic understanding of
electrical quantities is important for eddy current inspection.

RESISTANCE: (symbol: R, units: ohm, n)

Opposition to the flow of electrical current is called resistance. It is

constant for both direct and alternating current. The electrical
component is called a resistor.

v =IR Ohm's Law (3.1)

where, V is voltage drop across resistor (volt), and

I is current through resistor (ampere)

INDUCTANCE: (symbol: L, units: henry, H)

The property of an electric circuit by virtue of which a varying

current induces an electromotive force in that circui t (self) or in a
neighbouring circuit (mutual) is called inductance. The electrical
component is called an inductor.

L _ total flux linkages

- current through coil (3.2a)

N¢l
L ,. --.e. (3.2b)
I
2
- kl (N A/O (3.3)
 - 19 -

where, N is number of coil turns

¢lp is magnetic flux (weber)
I is current (ampere)
kl is a geometric factor
2
A is coil's planar surface area (mm )
1 is coil's axial length (mm)

The self-inductance of a coil is proportional to coil windings squared (N 2) and

planar surface area (A), and inversely proportional to coil length (1).

I:--.JDUCTIVE REACTANCE: (symbol: XL' units: ohm, f2)

Opposition to changes in alternating current flow through a coil is

called inductive reactance.

xL ,. wL (3.4a)

or XL = 27TfL (3.4b)

where, f is frequency of alternating current (hertz), and w is angular

frequency (radians/second)

CAPACITIVE REACTANCE (symbol: Xc' units: ohm, n)

Opposition to changes in alternating voltage across a capacitor is

called capaci ti ve reactance.

Eddy current coil capacitive reactance is normally negligible.

However, capacitance can be important when considering impedance
of probe cables (Section 5.9 and 7.2.5).
1
Xc - 2 TT f C (3.5)

where, C is capacitance (farad)

IMPEDANCE: (symbol: Z, units: ohm, n)

The total opposition to alternating current flow is called

IMPEDANCE. For a coil,

I z I - ~ R2 + X~ (3.6)

XL
and e • Arctan (3.7)
R

where Iz I is magnitude of Z, and 9 is phase of Z (described in next section).

 - 20 -

3.3 SINUSOIDS, PHASORS AND ELECTRICAL CIRCUITS

In a direct current (DC) circuit, such as a battery and light bulb, current and voltage
are described completely by their respective magnitudes, Figure 3.l(a). Analysis of
alternating current (AC) circuits is more complex. Since voltage and current
a 'nplitude vary with time, the phase (or time delay) relationship between the;-n must
also be taken into account. A typical AC circuit, an inductor in series with a resistor,
is presented in Figure 3.1 (b). This is a simplified ~odel of a probe asseonbly: the
inductor is the reactive part of the assembly (coil) while the resistor models both coil
wire and cable resistance. Figure 3.l(c) shows voltage across the inductor (V L)
leads the current (I) by 90°, while voltage across the resistor (V R) is in phase with
current. Since the curren t is common to both inductor and resis tor, it is possible to
use current as a point of reference. Hence, we deduce the voltage across the inductor
leads the voltage across the resistor by 90°.

If one measures the voltage drop, V T. across both the inductor and resistor, we
find V T leads current ( 0 r V R) by an angle less than 90°, as shown in Figure 3.l(d).

To evaluate the total voltage Vr • we add vectorially the separate voltages V R and vL •
V
T
I: VR + VL (3.8)

- I(R + jwL) (3.9a)

where j is a mathematical operator (rotates a vector CCW by 90°)

or VI - IR sin ( wt+O) + j IwL sin ( w t + n/2) (3.9b)

Representing voltage waveforms as in Figure 3.1(d) or equation 3.9(b) can be both

time consuming and confusing. A simpler voltage representation is available by means
of phasor diagrams. In phasor diagrams the voltage is represented by its peak value
(amplitude) and phase shift (e) relative to the current. The two terms in equation
3.9(b) both contain an amplitude and phase shift so they can be represented by
phasors. The first term's amplitude is IR and its phase shift is O. The amplitude of
the second term is IwL and its phase shift is TT 12 or 90°. Each phasor can be
represented by an arrow starting at the origin. The phasor's amplitude is indicated by
the length of the arrow OP and the phase shift by the direction of the arrow, see
Figure 3.l(e). Phasors are displayed graphically with the resistive component
(V R)' having a phase shift e - O. along the horizontal axis. As e increases the
phasor rotates counter-clockwise. The reactive component (V L)' having a phase
shift e - 90°, will be represented along the vertical axis.

Current is common to both voltage components and since V-IZ, the voltage graph of
Figure 3.l(e) can be converted to an impedance graph display, as in Figure 3.1(0. If
this approach is applied to eddy current testing, it is found that any changes in
resistance or inductive reactance will cause a change in the position of the end of the
vector (point P) which represents the total impedance vector.
 - 21 -

DIRECT CURR£1H

v • IR

CUUENT AND VOLTAGE CAN IE

C[sC"18EO iY !'.Ar,HITUOE ONLY

. Co) DIRECT CU •• ENT ClaCUI T

AL T E~~T I ~G CURRENT

CURRENT ~UST BE DESCRIBED BY

AMPLITUDE AND PHASE

Cb) ALTERNATING CURAENT CIRCUIT

eel CRT DISPLAY OF VOLTAGE AND CUa.fMT

P
A l - - - - - r,) wL.-+----

INoue TI VE
90°-,
REACTANCE.
... L

o R
~ ~. ~ I Sr~:1r.F . ~.

(t) VOL T A ~ E GPA(.H DI SPLAY OF PH. SQAS

FIGURE 3.1
Representation of Direct Current and Alternating Current Circuit Parameters
 - 22 -

To obtain the reactive and resistive components from this graph requires knowledge
of trigonometry.

Reactive component: XL - wL • I zI sin e (3.10)

Resist ive component: R

- Iz I cos e 0.11)

Amplitude of impedance: IzI • ~R2 + XL 2

Phase angle: e - Arctan XL/R (3.7)

Note the x axis component represents pure resistance (phase shift = 0°) while the y
axis component represents pure inductive reactance (phase shift = + 90°). In these
calculations it is assumed coil capacitance is negligible .

. 3.4 MODEL OF PROBE IN PRESENCE OF TEST MATERIAL

The test probe contains a coil which when placed on or close to a test sample can be
considered as the primary winding of a transformer. The field created by alternating
current in the coil induces eddy currents in the test sample which acts as a single
turn secondary winding, NS - I, Figure 3.2(b). Eddy currents align to produce a
magnetic field which tends to weaken the surrounding net magnetic flux ¢ P'
according to Lenz's Law.

( a)

(b)

/c-------t-)~ SECONDARY
I 11 ~ V 'b' RECEIVE COIL
(c) ':..'- - ------:?
FIGURE 3.2
Model of a Coil with Test Object
 - 23 -

There are two methods of sensing changes in the secondary current, Is' The
"impedance method" of eddy current testing consists of monitoring the voltage drop
across the prirnary coil (V P - I pZp)' The impedance Zp is altered by the
load of the secondary of the transformer. Therefore, changes in secondary resis tance, Rs'
or inductance L s can be measured as changes in · V p •

The "send-receive" method of eddy current testing uses two coils. Eddy current flow
in the sample is altered by defects and these variations are detected by monitoring
the voltage across a secondary recei ve coil, see Figure 3.2(c).

3.5 SIMPLIFIED IMPEDANCE DIAGRAMS

3.5.1 Derivation of Probe Impedance for Probe/Sample Combination

We now consider how changes in the test sample affect coil impedance on the
impedance graph display.

From the previous section the probe and test sample can be modelled as a
transformer with a multi-turn primary (coil) and single turn secondary (sample),
Figure 3.3(a). This circuit can be simplified to an equivalent circuit where the
secondary circuit load is reflected as a resistive load in parallel with the coil's
inductive reactance, Figure 3.3(b). This circuit is an approximate model of a real coil
adjacent to a conductor. It is assumed that all of the magnetic flux from the primary
coil links the test sample; the coupling is perfect (100%). It is also assumed that there
is no skin depth attenuation or phase lag across the sample thickness.

(a) ~DEL OF A COIL IUD nST

SAMPLE

(b) EQUIVALENT PARALLEL CIRCUIT

(e) EQ UIVALENT SERIES CIRCUIT

FIGURE 3.3
Equivalent Circuits
 - 24 -

. ,1e equivalent circuit concept can be used to obtain simplified impedance diagrams
applicable to eddy current testing. These diagrams serve as an introduction to the
more detailed diagrams which include variations caused by the skin effect. The
coil/sample circuit model can be transformed into the simpler series circuit by the
following mathematical manipulations. The load resistance Rs can be transfered
from the secondary back to the primary winding be multiplying it by the turns ratio
squared, (N P / N s ) 2, Figure 3.3(b).

The total impedance of this parallel circui t can be evaluated and transformed into an
equi valent series circui t as follows:

Z l Z2
Z •
P Zl + Zz
2
where Z1 - NpRs

and Z2 - jX o '

where Xo· wLOt coil inductive reactance 1n air.

Therefore -
which transforms to

(3.l2a)
( NZR ) 2 + ('X ) 2
p S 0

.Iis can be viewed as a series combination, in the primary circuit, of resistance RL

and inductive reactance X p or

(3.l2b)

The series circuit in Figure 3.3(c) is therefore fully equi valent to the parallel one of
Figure 3.3(b). Rp can be considered as coil wire and cable resistance while Zp·R L+ jX p
is the total impedance of the probe/sample combination.

When the probe is far from the sample (probe in air), Rs is very large and by
substituting Rs • 00 into equation 3.12a results in
 - 25 -

The above results can be obtained by removing component N ~ R 8 from

Figure 3.3(b), since R 8 - ex> implies an open circui t.

One last transformation in the equation is required before impedance graphs can be
obtained. Equation 3.12(a) can be simplified by setting
Co - XoG

where G is equi valent circui t conductance.

Substi tution in 3.12(a) yields

xo·c 0 x
zp + j 0
2 2
1 + C
o
1 + Co

Normalizing with respect to X o ' the coil's inductive reactance when far removed
from the sample (coil in air) results in

z C
.....£. -
o + j _l_~ (3.13)
X 2
o 1 + c
o

By varying Co' in equation 3.13, from 0 to infinity the impedance curve of

Figure 3.4 is obtained. The impedance locus is that of a semi-circle with center at Xp/Xo-~
and RL/Xo - 0; its radius is 1/2. With the help of equation 3.13 and Figure 3.4,
impedance changes can be related to changes in the sample characteristics.

P (OPERATING POINT)

NORMAL! ZED
INDUCTIVE
REACTANCE

co • co
NORMALI ZED RESISTANCE

FIGURE 3.4
Impedance Graph Display
 - 26 -

5.5.2 Correlation Between Coil Impedance and Sample Properties

The effect of test parameter variations on probe impedance can be derived from
equation 3.13. Each paramater is substituted in turn into Co-Xo/N~Rs; if an
increase in the parameter results in an increase in Co. the operatmg point (position
on impedance diagram) moves DOWN the impedance curve, if Co decrease, the
operating point moves UP the impedance curve. These correlations are useful in
obtaining a qualitative appreciation of the effect of the various test parameters. It is
also useful to know that probe/sample effects can be derived from the simple
equivalent parallel circuit where the sample is treated as a resistor in parallel with
an inductor (coi!). The complete effect can then be obtained by adding the effect of
'phase lag', which will be treated in later chapters.
Study of equation 3.13 reveals the following:

1. An in'crease in R 8 results in a decrease in Co' Therefore an increase in

resistance to eddy current flow moves the operating point, P, UP the impedance
curve (along the semi-circle), see Figure 3.5(a).

2. Rs - pi/A

where, p is electrical resistivity, .2. is eddy current flow distance and A is cross-
sectional area to current flow.

Therefore, p - constant x Rg

An increase in electrical resistivity will move the operating point UP the

impedance curve. The opposite is true for an increase in electrical conductivity.
See Figure 3.5(a).

3. For thin wall tubes or plates of thickness t,

Rs - p.2./A - prrD/tw

and for constant probe or tube diameter, 0, and coil width, w,

Rs - constant/t

An increase in tube wall (or plate thickness) will move the operating point
DOWN the impedance curve, see Figure 3.5(b).

4. -
for constant sample properties.

An increase in test frequency will move the operating point DOWN the
impedance curve, see Figure 3.5(c).
 - 27 -

5. L o • constant x D2.• probe inductance increases proportional to probe

or tube diameter squared.

Also Rs - P'ITD/tw - constant x D.forconstantthickness,t,andcoilwidth,w.

SubstitutingL o and Rsinto Co - WLo/N~Rs results in Co·constant x D.
An increase in probe diameter or tube diameter will move the opera ting point
DOWN the impedance curve, see Figure 3.5(d).

6. In the equivalent circuit derivation perfect coupling was assumed for sake of
simplification. However, it can be shown that when mutual coupling between
coil and sample is decreased, the impedance point traces smaller semi-circles
as Co increases from 0 to infinity, see Figure 3.5(e).

~
r
o

_+-""'-__ _ o _ -_ _ -.-.- - - -- ---.-. -

O. S O. S
(. ) (b ) ( c)

D
TUSE

"'\. - DSURFA CE
\ \ PR03E

DECREaSING FILL FACTOR

OR I N(R[ASIN(; LIFT-orr

O.S 0 , o . ~

(d) (. )

FIGURE 3.5
Simplified Impedance Diagrams
 - 28 -

,).0 SUMMARY

The impedance method of eddy current testing consists of monitoring the voltage
drop across a test coil~ The impedance has resistive and inductive components; the
impedance magnitude is calculated from the equation

(3.6)

and the impedance phase is calculated from

wL
a - Arctan -
R • degrees (3.7)
The voltage across the test coil is V = IZ where I is the current through the coil and Z
is the impedance.

A sample's resistance to the flow of eddy currents is reflected as a resistive load and
is equivalent to a resistance in parallel to the coil inductive reactance. This load
results in a resistive and inductive impedance change in the test coil. Coil impedance
can be displayed on normalized impedance diagrams. These are two-dimensional plots
with the inductive reactance displayed on the vertical axis and resistance on the
horizontal axis as in Figure 3.6.

NORMALIZED OPERATING POINT

INDUCTANCE
REACTANCE

/ <T,t,I,D

NORMALIZED RESISTANCE.

FIGURE 3.6
Impedance Graph Display
 - 29 -

With this display we can analyze the effect of sample and test parameters on coil
impedance. The equivalent circuit derivation of coil impedance is useful for a
qualitative understanding of the effect of various test parameters. It is valid only for
non-ferromagnetic material and for the condition of no skin depth attenuation or
phase lag across the sample. (Ferromagnetic materials will be covered in Section
9.4 ).

Note that all test parameters result in a semicircle display as they increase or
decrease. A resistance increase to the eddy current flow or increase of sample's
electrical resistivity moves the operating point UP the impedance diagram, i.e.,
increase in coil inductance and a change in coil resistance.

An increase in a sample's electrical conducti~ity, thickness or tube diameter, moves

the operating point DOWN the impedance curve. An increase in test frequency or
probe diameter also moves the operating point DOWN the impedance curve. Although
not shown in the above figure, a decrease in fill-factor or increase in lift-off results
in a decrease in semicircle radius and smaller change in coil impedance.

In some test requirements it is advantageous to operate at specific locations on the

impedance diagram. By choosing the appropriate test parameters this is usually
possible.

3.7 WORKED EXAMPLES

3.7.1 Probe Impedance in Air

PROBLEM: An eddy current test is carried out at a test frequency of 50 kHz. Coil
resistance is 15 ohms while its inductance is 60 microhenries.
a) What is the inductive reactance of the test coil?
b) What is the impedance of the test coil?
c) What is the angle, e , between the total impedance vector and the
resistance vector?

SOLUTION:

a) XL· 2 'IT f L - (2 'IT) (SOx 10 3) ( 6 0 x 10 - 6 )

XL • 18.8 ohms

Iz I
-- -V
b) ,fR2+ (2 iTfL)2 1S2 + (18.8)2
IZ I 24.1 ohms

2'ITfL ,.. 18.8

c) e - Arctan R
Arctan
15
.. Arctan 1. 253

e • 51.4 degrees
 - 30 -

j.,.2 Probe Impedance Adjacent to Sample

PROBLEM: An eddy current test is carried out on brass using a surface probe at
50 kHz. Coil resistance in air is 15 ohms and its inductance in air is 60
microhenries. Probe impedance with the probe on the brass sample is
measured as Z p = 24.5 ohms and e = 35 degrees.

Calculate: a) Xp' inductive reactance

and b) R L , resistive load

SOLUTION: a) X - Z sin 8
p p
a
24.S sin 35 - 14.1 ohms
2 ,
b) RL - Zp cos8 - ROC }'v./J ~
0
- 24.S cos 35 - 15.0 - 5.1 ohms f.
J.I .3 Voltage - Current Relationship

PROBLEM: For the above probe impedance problem calculate voltage drop across
the probe if test current is 100 milliamperes.

SOLUTION: Probe impedance I Z I = 24.5 ohms

Ohm's Law states that V = I I'z I

therefore, V =(0.10) (24.5) = 2.45 volts.

Voltage across the probe is 2.45 volts.

 - 31 -

CHAPTER 4 - INSTRUMENT A nON

4.1 INTRODUCTION

/\11 the information about a test part is transmitted to the test coil through the
magnetic field surrounding it. The impedance eddy current method monitors voltage
drop across the primary coil, Vp - I pZp; as coil impedance changes so '.viii the
voltage across the coil if current remains reasonably constant. The send-receive eddy
current method monitors voltage developed across a sensing coil (or Hall effect
detector) placed close to the excitation coil, see Figure 2.2.

In most inspections, probe impedance (or voltage) changes only slightly as the probe
passes a defect, typically less than I %. This small change is difficult to detect by
measuring absolute impedance or voltage. Special instruments have been developed
incorporating various methods of detecting and amplifying small impedance changes.

The main functions of an eddy current instrument are illustrated in the block diagram
of Figure 4.1. A sine wave oscillator generates sinusoidal current, at a specified
frequency, that passes through the test coils. Since the impedance of two colIs is
never exactly equal, balancing is required to eliminate the voltage difference
between them. Most eddy current instruments achieve this through an AC bridge or
by subtracting a voltage equal to the unbalance voltage. In general they can tolerate
an impedance mismatch of 5%. Once balanced, the presence of a defect in the
vicinity of one coil creates a small unbalanced signal which is then amplified.

r-----(j)
r---~~--~ D.C. METER
PHASE
SENS I TIVE Y
AC TO DC I--~-~
PHASE
AMPLl FI ER 1-----4 CONVERTOR
(PLUS X ROTATION
FILTERING) 1---'--+
~r-----"~

BALANCE x-Y
MONITO R

OSCILLATOR
TRANSFORMER

FIGURE 4.1
Block Diagram of Eddy Current Instrument
 - 32 -

Since the sinusoidal unbalance voltage signal is too difficult and inefficient to
analyse, it is converted to a direct current (DC) signal retaining the ~mplltude and
phase characteristics of the AC signal. This is normally achieved by resolving the AC
signal into quadrature components and then rectifying them while retaining the
'l ppropria te polarity. In general purpose instruments, these signals are normally
displayed on X- Y monitors. Si mpler instruments, such as crack detectors, however,
have a rneter to display only the change in voltage amplitude. To decrease electrical
instrument noise, filtering is used at the signal output; however, this decreases the
frequency response and thereby restricts the inspection speed.

The most troublesome parameter in eddy current testing is lift-off (probe-to-

specimen spacing). A small change in lift-off creates a large output signal. The
various methods used to decrease this effect are discussed in the individual sections
on specific eddy current instruments.

4.2 BRIDGE CIRCUITS

. Most eddy current instruments use an AC bridge to sense the slight impedance
changes between the coils or between a single coil and reference impedance. In this
section the important characteristics of bridge balance are discussed.

4.2.1 Simple Bridge Circui t

A common bridge circuit is shown in general form in Figure 4.2, the arms being
indicated as impedance of unspecified sorts. The detector is represented by a
voltmeter. Balance is secured by adjustments of one or more of the bridge arms.
Balance is indicated by zero response of the detector, which means that points A and
C are at the same potential (have the same instantaneous voltage). Current will flow
through the detector (voltmeter) if points A and C on the bridge arms are at different
voltage levels (there is a difference in voltage drop from B to A and B to C). Currerlt
may flow in either direction, depending on whether A or C is at higher potential.

Aof------1

FIGURE 4.2
Common Bridge Circuit
 - 33 -

If the bridge is made up of four impedance arms, ha ving induc tive reactance a nd
resistive components, the voltage from B to /\ mus t equal t he voltage from B to C in
both amplitude and phase for the bridge to be bal a nced.

,At balance ,

and I 12 3 - I 224

from which the following relationship is obtained:

( 4.1)

Equation 4.1 states that the ratio of impedances of one pair of adjacent arms must
equal the ratio of impedances of the other pair of adjacent arms for bridge balance.

If the bridge was made up of four resistance arms, bridge balance would occur if the
magnitude of the resistors satisfies equation 4.1 (with z 1 replaced with R l,etc.).
However, if the impedance components are eddy current probes consisting of both
inductive reactance and resistance, the magnitude and phase of the impedance
vectors must satisfy equation 4.1.

In practice, this implies the ratio of inductive reactance of one pair of adjacent arms
must equal the ratio of inductive reactance of the other pair of adjacent arms; the
same being true for the resistive component of impedance.

F igure 4.2 and equation 4.1 can be used to illustrate the characteristics 'figure 8'
signal of a differential probe. If

> point C is at a higher potential than point A.

This implies that when Z 1 increases (i.e., coil moving across a defect)
with 22 ,Z3& Z4constant, the bridge voltage unbalance increases, and the opposite
happens when 23 increases. It is this bridge unbalance characteristic that results
in a plus-m inus or 'figure 8' signal as the differential probe moves across a loc alized
defect. This signal occurs independent of whether the two coils are wound in
opposition or in addition.

4.2.2 Typical Bridge Circuit in Eddy Current Instruments

F ig ure 4.3 illustr a tes a typi c al AC bridge used in eddy curren t instrument s . It is
similar to the bridge in Figure 4.2 e xcept for two additional arms. In this bridge the
probe coils are placed in paraJJel with variable resistors. The balancing, or rnatching
of voltage vector phase and amplitude, is achieved by varying these resistors until a
null is achieved. Potentiom e ter R2 balances the reac tive compon e nt of the coil s t o
ma ke the phase angle of e ach coil circuit equal. Potentiom e t e r R 1 balances t he
resuJ tant vol tage wi th an equal vol tage arnplitude to null the ins tanta neous vol tage
betw e en R 1 and R 2 •
 - 34 -

TEST COIL

R\

CO I L

FIGURE 4.3
Typical Bridge CirOlit Used in Eddy Current Instruments

The inductive voltage drop across each coil is equalized by controlling the current
passing through the coils. This is done by varying potentiometer R 2' However, when
the test coil inductance differs significantly from reference coil inductance,
potentiometer R2 will have to be rotated to one extremity. This means less current
passes through one coil making it less sensitive than the other coil. When this occurs,
a distorted (unsymmetrical) signal results if a differential probe is used. In addi tion,
the common cable lead carries the unbalanced current, resulting in cable noise,
especially if the cable is not properly shielded and grounded.

In the Figure 4.3 circuit, the output voltage for large (> 10%) off-null {off-balance}
condi tions is a nonlinear function of the change in coil impedance. However, for
defect detection, close to balance, this discrepancy is small.

L .3 Bridge Circuit in Crack Detectors ·

Portable eddy current instruments are often used to inspect for surface defects. A
typical crack detector circuit is shown in Figure 4.4. An oscillator supplies an
alternating current to an AC Bridge, containing a single eddy current probe coil as
one arm of the bridge. A capacitor is connected in parallel with the coil so the L-C
(inductance-capaci tance) circuit is near resonance. When the coil is placed on a test
sample, the bridge is unbalanced and the pointer on the meter swings off-scale. The
bridge can be balanced by adjusting potentiometer R 1 •
 - 15 -

VARIABLE
RESISTOR

OSCI LLATOR

FIGURE 4.4
Simplified Circuit of Crack Detector

4.3 RESONANCE CIRCUIT AND EQUATIONS

Probe-cable resonance must be considered when operating at high test frequencies

and/or using long probe cables. In addition, crack detectors are purposely designed to
opera te close to resonance. This section contains basic information about resonant
(tuned) circuits.

If a capacitor is connected in parallel with the test coil (inductor), there is a unique
frequency at which the inductance-capacitance (L-C) circuit resonates. At this
frequency the circuit is said to be tuned. Under this condition the output voltage, for
a given rneasurement, is maximum. A capacitor in parallel with the eddy current
probe converts the circuit of Figure 3.3(c) to that of Figure 4.5.

FIGURE 4.5
Parallel LC CirOJit
 - 36 -

At resonance,

z - (4.2)
R 2 +(X_X)2
P c
hence z - ex> when R .. 0

If resistance, R, is negligible compared to xp and Xc resonance occurs when

xp - X
c 0 r wL - l/wC (4.3a)

or w - l/ILC (4.3b)

Since w - 2 TT f. resonant frequency is

1
f r - (4.4a)

where L is coil inductance in henries and C is cable capacitance in farads.

When resistance, R, is significant,

(4.4b)

x
Q _ --E.
where , quali ty factor.
R

The resonant frequency of a practical parallel resonant circui t ( RIO) is the

frequency at which the reactive power of the inductance and capacitance are equal,
or the total impedance appears as pure resistance.

4.4 EDDY CURRENT INSTRUMENTS

General instru"ent functions were described using the block diagram of Figure 4.1. In
this section specific instruments are covered. It answers the questions: What is the
test frequency? How is lift-off compensated for? How is balancing achieved? What
type of outputs do they have?

4.4.1 General Purpose Instrument (Impedance Method)

Figure 4.6 shows a typical eddy current instrument with various control functions.
FREQUENCY control sets the desired test frequency. Frequency is selected by
continuous control or in discrete steps from about I kHz to 2 MHz. The coils'
impedances are normally balanced using an AC bridge circuit. These bridges require
two coils on adjacent bridge arms such as arms No.2 and No.4 in Figure 4.3. Coil
impedance must be compatible with internal bridge impedance.
 - 37 -

FERRITE

I CARBON STEEL

I ~ONEL

~ SST T YPE 304

~ "'" LEAD

\~ BRASS
ALUM.! N I UM
COPPER STORAGE MONITOR

@ (0'''0)
OUTPUT F.REQUENCY
SELECTOR
8 ~E

©
PROBE
CONNECTOR
f;;\
V

FIGURE 4.6
Typical Eddy Current Instrument With Storage Monitor

Most bridges can tolerate a coil impedance between 10 and 200 ohm:;. The
BALANCING controls, labelled X and R in some instruments, are potentiometers
R1 and R 2 in Figure 4.3. They match coil impedance to achieve a null when the probe
is in a defect free location on the test sample. Some instruments have automatic
balancing.

The bridge output signal amplitude is controlled by the GAIN control. In some
instruments it is labelled as SENSITIVITY. It controls the amplifier of the bridge
output signal, shown in Figure 4.1, and therefore does not affect current going
through the probe. However, some instruments control amplification by varying
current through the coils. This is undesirable because it could cause coil heating, and
when testing ferromagnetic materials the magnetization level changes, resulting in
signal distortion and non-repeatable signals.

Following amplification of the bridge unbalance signal, the signal is converted to

direct current signals. Since the AC signal has both amplitude and phase it is
converted into QUADRA TURE X and Y components. The quadrature components of
the bridge output are generated by sampling the sinusoidal signal at two positions 90°
apart (one-quarter wave) on the waveform (or by using electronic multipliers). The
DC voltage values (amplitudes) constitute the X and Y quadrature components. if
phase is taken relative to the resistive voltage component, then the X quadrature
component is RL (orV R) and the Y component, XL (orV L), in equation 3.12(b) or Figure
3.4. We now have an efficient way of analyzing bridge unbalance signals.
 - 38 -

Eddy current instruments do not have a phase reference. To compensate for this, they
have a PHASE SHIFT control (phase-discrimination contro!). Normal impedance
diagram orientation with inductive reactance displayed vertically (+ Y) and resistive
horizontally (+ X) can be obtained experimentally. This is achieved by adjusting the
PHASE control until the signal from a probe approaching a ferrite sample (high \.l and
very high P) displays a vertical (+ Y) signal indicating an increase in probe inductive
reactance, see Section 5.5.6 for examples. PHASE control can also be used to
minimize the effect of extraneous signals such as lift-off. The X-Y signal pattern is
rotated until the lift-off signal is horizontal (X). Thus any vertical (y) channel signal
indicates defects, thickness variations, etc., wi th little effect from probe wobble.

The output signal is normally filtered internally to decrease instruinent or system

noise. This decreases frequency response of the instrument and reduces the maximum
inspection speed; at faster inspection speeds signal distortion results. Instruments can
have a frequency response of 30 to 1000 Hz, although 100 to 300 Hz is most common.
At 300 Hz, the maximum attainable tube inspection speed, to detect an abrupt defect
without signal distortion, is about 0.25 m/s.

Signals are commonly displayed on X-V storage monitors with the X component on
. the horizontal axis and the Y component on the vertical axis. The writing speed or
frequency response is greater than I kHz for a storage CRT.

Analysis of recorded signals is normally done visually. The storage monitor display in
Figure 4.6 shows the change in coil impedance as a surface probe was placed on
various test samples illustrating the effects of resistivity, permeability and lift-off.

In the "impedance" method of eddy current testing, the flow of eddy currents is
monitored by observing the effect of their associated electromagnetic fields on the
electrical impedance of the inspection coil(s). This impedance includes coil wire and
cable resistance,
z - -1"'(~--+-R-i-+-R--bl-)-2+-CW-L-)2 (4.5)
" - 1 . " , re ca e

Coil wire and cable resistance increase linearly with temperature according to
R D R (1~T)
o

where Cl is temperature coefficient of resistance

and 6 T is change in temperature.

If the probe andlor cable experience a change in tempera ture during inspection, the
output signal from the eddy current instrument changes; this is normally referred to
as temperature drift.

4.4.2 Crack Detectors

A typical crack detector circui t was shown in Figure 4.4. Crack detector probes
contain only one coil, with a fixed value capacitor in parallel with the coil to form a
resonant circuit. At this condition the output voltage, for a given change in coil
impedance, is maximum. The coil's inductive reactance, XL' must be close to the
capacitive reactance, Xc' In most crack detectors this is in the range of 20 to
100 ohms.
 - 39 -

Crack detectors that operate at or close to resonance do not have selectable test
frequencies. Crack detectors for non-ferromagnetic, high electrical resistivity
materials such as Type 304 stainless steel typicaily operate between 1 and 3 MHz;
those for low resistivity materials (aluminum alloys, brasses) operate at lower
frequency, normally in the 10 to 100 kHz range. Some crack detectors for high
resistivity materials can also be used to inspect ferromagnetic materials, such as
carbon steel, for surface defects. Normally a different probe is required; however,
coil impedance and test frequency change very little.

PROBE WITH LIFT-OFF =0.1 ",.

/ PROBE WITH LI FT-OFF =0 ......

METER

OU TPU T

SAMPLE WITH DEFECT

.
\
0.8 0.9 1.0 1.1 1.2

f
OSCILLATOR FREQUEIICY.
f
r

FIGURE 4-.7
Meter Output with Varying Oscillator Frequency

Crack detectors have a meter output and three basic controls: balance, lift-off, and
sensitivity. BALANCING control is performed by adjusting the potentiometer on the
adjacent bridge arm, until bridge output is zero (or close to zero). GAIN control
(sensitivity) adjustment occurs at the bridge output. The signal is then rectified and
displayed on a METER. Because the signal is filtered, in addition to the mechnical
inertia of the pointer, the frequency response of a meter is very low (less than
10 Hz). LIFT-OFF CONTROL adjusts the test frequency (by less than 25%) to operate
slightly off resonance. In crack detectors the test frequency is chosen to minimize
the effect of probe wobble (lift-off), not to change the skin depth or phase lag. The
set-up to compensate for probe wobble can be described with the help of Figure 4.7.
Frequency is adjusted by trial-and-error to obtain the same output signal on the
meter with the probe touching the sample and at some specified lift-off (normally
0.1 mm). At this frequency a deep surface defect will give a different reading on the
meter, as shown in Figure 4.7.

However, the meter output is a complex function of signal phase and ampli tude, and
cannot be used to reliably measure depth of real defects. Nor can they be used to
distinguish between real and false indications such as ferromagnetic inclusions.
 - 40 -

4.4.3 Material Sorting and Conductivity Instruments

Material sorting, or conductivity instruments, have a precalibrated meter output and

have a unique way of compensating for lift-off. Instruments for sorting of high
resistivity materials (Type 304 stainless steel) use a fixed, high test frequency
normally between 200 and 500 kHz, and those for low resis tivity materials (alum inum
alloys), a low test frequency, between 20 and 100 kHz. They incorporate AC bridges
and normally have two coils (one as reference). Coil impedance is in the range of 20
to 100 ohms. They either have bridge balancing or a zeroing control, to keep the
signal on scale. GAIN CONTROL or sensitivity adjustment occurs at the bridge
output. The signal is then rectified and displayed on a METER.

LIFT-OFF compensation is normally pre-set. Figure 4.8 explains how the probe-
wobble (Jift-off) signal is eliminated. The bridge is purposely unbalanced (by pre-set
internal adjustment)* such that the unbalance point, P, is at the centre of curvature
of the lift-off impedance locus, AB. The instrument meter reads a voltage
proportional to the distance, PBI or P A', from the chosen unbalance point to the
impedance curves. The amplitude of this voltage remains constant with probe wobble
but changes significantly for wall thickness (and resistivity) variations. In fact any
signal that traces an impedance locus different from lift-off will change meter
output.

POE5ET UNBALANCE

I
I
I
/1 ~ E5 1:)rlvIT ·(

0(1---- - - ~~SISTANCE

FIGURE 4.&
Unbalanced Bridge Method Showing Selection of Operating Point

.:.. 'This is achieved by subtracting a signal equal to OP from the signal OA.
 ;

....:.1 - 41 -

With this type of instrument only the magnitude of the impedance change is
measured. This instrument is effective for conductivity and wall thickness
measurement (and deep defects) and is simple to operate. It has only two basic
controls: balance and sensitivity.

4. 5 SEND-RECEIVE EDDY CURRENT SYSTEMS

The "send-receive" eddy current method eliminates the temperature drift sensed by
general purpose instruments. The f low of eddy currents is monitored by observing the
effect of their associated electromagnetic fields on the voltage induced in an
independant coiI(s), Figure 4.9. The excitation or primary coil is driven with a
sinusoidal current with constant peak-to-peak amplitude to obtain a constant
magnetomotive force, .

~ ~ N I «N I sin wt (2.3)
p p p p 0

EICITATIOII r IECEIVE COILS

t
v.

TEST ARTICLE

FIGURE 4.9
Send-Receive Circuit
 - 42 -

This makes the exci tation magnetic flux ¢pindependent of primary coil resistance.
The secondary or receive coi1(s) is conneCted to a high input impedance amplifier,
hence the induced voltage Vs is not affected by receive coil resistance.
d¢
V ~ N --R ~ N w¢ 0 cos wt (2.5)
s p dt p

The wire resistance of both the excitation and receive coils can change, because of
temperature, without affecting the output signals; temperature drift has thus been
eliminated. Temperature independence makes this method useful for measuring
resistivity, wall thickness and spacing between components. It has no significant
advantage over the impedance method for defect detection, except in the through-
wall transmission system discussed in Section 5.4.

4.5.1 Hall-Effect Detector

Most send-receive circuits consist of one excitation (or driver) coil and one or more
receive (or pick-up) coils.

However, the induced magnetic flux ¢ s can be measured with a Hall-effect detector
rather than by monitoring the induced voltage V s across a pick-up coil, see Figure
2.2b and 2.2c.

4>
I

FIGURE 4.10
Hall Detector Circui t

The induced voltage in a pick-up coil is proportional to the time ra te of change of the
magnetic flux and therefore is proportional to the test frequency,

V a: f
Pick-up
 - 43 -

The Hall detector instead responds to the instantaneous magnitude of the magnetic
flux, <Po.

This means the output voltage is independent of test frequency, making it useful for
low frequency inspection (especially if the detector has to be smail).

The Hall detector works as follows: When direct current is passed through a Hall
element, volta ge (electric potential) is produced, perpendicular to current flow, see
F ig ure 4.10. This voltage is proportional to the component of magnetic flux
perpendicular to the element and the element surface area. This voltage is NOT from
a change in element resistance. Hall elements as small as 1 mm square are
commercially available.

4.5.2 Send-Receive Coils and Lift-Off Compensation

Ge ne ral purpose "send-receive" instruments are similar to "impedance" instruments,

as described in Section 4.4.1. The main difference is the method of balancing because
of the different coil configuration. Most send-receive circuits consist of one
excitation coil and two receive coils positioned symmetrically inside or outside the
excitation coil. They can either be differential where both coils sense the test
specimen or absolute where only one coil senses the test specimen, as shown in Figure
4.9. Although coil impedance is not important in send-receive instruments, the
induced voltage is a function of number of windings and test frequency. Therefore
their inducti ve reactance tend to be similar to coils used in impedance instruments.

The sensing coils are wound in opposition so the excitation field induces no net
voltage in the receive coils when they both sense the same material. In the presence
of a defect, the voltage changes as each coil moves over it. Figure 4.9 illustrates a
surface reflection type probe where both excitation and pick-up coils are on the same
side of the test sample. However, the excitation coil and pick-up coils can be placed
on opposite sides of the sample; the method is referred to as through-wail
transmission. The two methods are compared in Section 5.4.

The output signals in most send-receive instruments are the quadrature components
of the secondary voltage. However, in some special purpose instruments, one output
signal is proportional to amplitude and the other to phase of the secondary voltage
(relative to primary voltage). They compensate for LIFT-OFF as follows: if coil-to-
sample spacing varies there is a large change in amplitude of the secondary voltage
but little change in phase. The phase shift between the secondary and primary
sinusoidal voltages is measured at a voltage level V 0 slightly larger than zero,
Figure 11-.11. At this voltage the sinusoidal voltages have the same phase shift for zero
lift-off as for maximum (perhaps 0.1 mm) lift-off. The voltage discriminator in these
phase-shift measuring eddy current instru;nents trigger on the V0 voltage point,
and therefore, the output signal for lift-off between 0 and 0.1 mm is minimized.
Measurement of resistivity, wall thickness or deep defects can be made without lift-
off noise.
 - 44 -

v( t )
PROBE SIGIUL. LlFT-OFF=O

PRO 8 E S I G. A. L. L 1FT - 0 F F = O. i ••

PROBE SI GIAL. DEFECT I. TEST Ani CLE

2rr
eIo'8
1

FIGURE 4.11
Secondary Voltage Waveform foe Various Test Conditions

)t.6 MUL TIFREQUENCY EQUIPMENT

The eddy current NDT method is sensitive to many test parameters, making it very
versatile. However, one is usually only interested in a single parameter such as
defects. Insignificant parameters such as changes in electrical or magnetic
properties, the presence of dents or support plates in tube inspection and lift-off in
surface probe inspection can mask defect signals. The multifrequency eddy current
method was developed to eliminate the effect of undesirable parameters.

The response to various anomalies changes with test frequency. This allows a means
of discriminating against unimportant changes. In multifrequency instruments, two or
more frequencies are used simultaneously (through the same coi1(s». Coil current
consists of two or more superimposed frequencies, i.e., the coi1(s) is excited with
more than one test frequency simultaneously. A three-frequency multifrequency
instrument acts the same way as three separate (single-frequency) eddy current
instruments. Band-pass filters separate the signals at each frequency. The
discrimination or elimination process is accomplished by combining the output signals
(DC signals) from individual frequencies in a manner similar to simultaneous solution
of multiple equations. The elimination of extraneous signals is achieved by matching
the signal at two test frequencies and subtracting. This process is continued for other
unwanted signals using other test frequencies until the final output consists of only
the defect signal. A discussion of inspection results with multi-frequency is covered
in Section 8.4.
 - 45 -

\tlultifrequency instruments have the same controls and functions as general purpose
"impedance" type instruments, described in Section 4.4.1, with the addition of mixing
modules. These modules are used to combine or subs tract the output signals from
each combination of frequencies.

4.7 PULSED EDDY CURRENT EQUIPMENT

Faraday's Law states that eddy currents are induced in a conductor by a varying
magnetic field. This magnetic field can be generated by passing sinusoidally varying
current through a coil. However, the clirrent can be of other waveforms such as a
train of pulses. This method works only on the send-receive principle where the flow
of eddy currents is monitored by observing the effect of their associated
electromagnetic fields on the induced voltage of the receive coiI(s). The voltage pulse
is analyzed by observing its amplitude with time, Figure 4.12.

To compensate for LIFT-OFF, the voltage is sampled at a preset time, t l ' When
the waveform is triggered (measured) at time t l ' the voltage for zero lift-off and
maximum lift-off is the same, whereas the voltage waveform in the presence of a
defect is different. This method is quite similar to the send-receive method described
in Section 4.5.3. Therefore, by measuring the voltage at the appropriate crossing
point, lift-off effects can be drastically decreased.

Y( t)
DEFECT I. TEST ARTICLE

LIFT-OFF=O.l ca.

LIFT-OFF=O

y 2 -A--------t-

FIGURE 4.12
Voltage Across a Pulsed Eddy Current Pick-Up Coil as a Function of Time
 - 46 -

The pulsed eddy current method offers another advantage. The pulsed driving current
produces an inherently wideband frequency spectrum, permitting extraction of more
selective information than can be determined from the test specimen by a single
frequency method. Unfortunately, there is at present no commercially available
instrument that operates on this principle.

4.8 SPECIAL TECHNIQUES

Two old methods used to measure large coil impedance variations (greater than 5%)
are the ELLIPSE and SLIT methods. These methods analyse the AC signal directly on
an oscilloscope (without converting it to DC). They were mainly used for material
sorting. They are obsolete methods and a detailed description is not warranted in this
manual; a full description is contained in Reference 5.

Another technique, MODULATION AN AL YSIS, is also described in Reference 5. It

works on the same principle as "frequency spectrum analysis" where a discrete
frequency component of a waveform can be analysed without interference from lower
or higher frequency noise. The inspection must be performed at constant speed (in
fact it only works if there is relative motion between coil and sample). It is used in
production-line testing at speeds up to 2 m/s or higher. It is a very specialized and
complicated method and a detailed description is not warranted in this manual.

4.9 RECORDING EQUIPMENT

During inspection, eddy current instruments and recording equipment are typically
connected as in Figure 4.13. The eddy current signal is monitored on a storage CRT
{cathode ray tube} and recorded on X- Y and two-channel recorders. Recording on an
FM tape recorder for subsequent playback is also common.

The important characteristic of these recording instruments is FREQUENCY

RESPONSE, or speed response, which limits inspection speed. Section 4.4.1 indicated
general eddy current instruments have a frequency response of 100 to 300 Hz,
limiting the inspection speed to 0.25 m/s. To be compatible, recording instruments
must have the same or higher frequency response.
 - It 7 -

x-v
STORAGE
MONITOR
I

I
7
...., -' 2 , I

.-s- I 1><'1 I I
EDDY CURRENT
INSTRUMENT ~ PROBE
X~ (Y

)
X Y
( 0
X
:I )

Y
( 0
XY
X-V RECORDER 2-CHANNEL FM TAPE
CHART RECORDER RECORDER

FIGURE 4.13
Block Diagram of Eddy Current Monitoring Equipment

x- Y Recorders
Signal analysis for signal discrimination and defect depth estimation is normally done
on X-Y signal patterns. The CRT storage monitors have a frequency response of at
least 1 kHz and therefore do not restrict maximum inspection speed. However, to
obtain a permanent visual record of the signal, it must be recorded on X-Y recorders.
The fastest recorders have a speed of response of 8 Hz for small signals. This
drastically limits inspection speed if used on-line. It is therefore only used in the
laboratory or to record playback from tape recorders (this is done by recording at the
highest tape speed and playing back at the lowest, a factor of 8:1 for most tape
recorder). One solution to on-line recording of X-Y signals is to paragraph the CRT
display; however, this is not practical for recording many signals.

Another solution is to use storage monitors with hard copy (paper output) capability.
These exist commercially but require custom-made control units. They have a
frequency response of 1 k Hz or higher.

Strip Chart Recorders

Recording X and Y signal components against time is useful in locating defects and
determining their length.
 - 48 -

Common two channel ink-pen strip chart recorders have a speed response of
approximately 100 Hz. At maximum inspection speed (0.25 m/s) the recorded signal
will decrease in amplitude and be slightly distorted.

Ink-ejection strip chart recorders have a speed response of 1 kHz. These recorders
are not readily available in North America and use a lot of paper.

Ultraviolet light recorders have a speed response higher than 1 kHz, but require
special paper. These recorders are rarely used in eddy current testing.

FM Tape Recorders

Tape recorders allow storage of eddy current signals (on magnetic tape) for
subsequent retrieval. They have a frequency response proportional to recording speed.
The lowest recording speed is 24 mm/s (15/16 ips) giving a frequency response of
300 Hz, and the fastest, 380 mm/s (I5 ips), will respond to 4.8 kHz.

Frequency Response

Eddy current instruments and recording instrumentation have limited frequency

response. This means they require finite time to respond to an input signal.
Frequency response, sometimes called speed of response, is defined as the frequency
at which the output signal falls to 0.707 (-3 dB) of the maximum input signal.

A test coil with an effective sensing width w passing over a localized defect at a
speed s will sense the point defect for a duration of w/s seconds. This signal is
approximately equal to one wavelength with a frequency

f - s/w hertz (4.6)

where s is speed in mm/s and w is width in mm.

For example, at a probe speed of 0.5 m/s and probe sensing width of 2 mm,
f = 250 hertz. If the instrumentation has a frequency response of 250 hertz, the
output signal is reduced to 0.707 the input signal and the X- Y signal is distorted. If
the instrumentation frequency response is 500 hertz, the output signal decreases only
slightly. For this example, the eddy current instrument should have a frequency
response equal to or greater than 500 hertz to obtain undistorted signals. Or
in versely, if the instrument frequency response is only 250 hertz, the maximum
inspection speed should be reduces to 0.25 m/s.

4.10 SUMMARY

Basic eddy current equipment consists of an alternating current source (oscillatod,

voltmeter and probe. When the probe is brought close to a conductor or moved past a
defect, the voltage across the coil changes and this is read off the voltmeter. The
oscillator sets the test frequency and the probe governs coupling and sensitivity to
defects.
 For effective purchase or use of an eddy current instrument, the following
information is needed:
(a) type of instrument: impedance, send-receive, crack detector, etc.
(b) type of outputs: single (meter) or quadrature (X- y) component outputs
(c) test frequency
(d) type of lift-off compensation.

\1ost eddy current instruments use an AC bridge for balancing but use various
rnethods for lift-off compensation. Send-receive instrument should be used for
accurate absolute measurements in the presence of temperature fluctuations.
Multifrequency instruments can be used to simplify defect signals in the presence of
extraneous signals.

Eddy current instruments and recording equipment have a finite frequency response
limiting the inspection speed to normally 0.25 m/s.

Most instruments tolerate probe impedance between 10 and 200 ohms.

Crack detectors operate close to coll-cable resonance. The resonant test frequency is
given by

f - 1/2rr/LC (4.4a)
r
where L is coil inductance in henries and C is cable capacitance in farads. The lift-
off signal is minimized by adjusting the frequency (slightly off resonance) until zero
and a small probe lift-off gives zero output signal. High test frequencies are normally
used to inspect for shallow defects in high resistivity or ferromagnetic materials.
Low test frequencies are used for detecting deep defects or inspecting good
conductors. Crack detectors have a meter output, and cannot be used to reliably
measure defect depth.

4.11 WORKED EXAMPLES

4.11.1 Impedance at Resonance

6
PROBLEM: In a parallel L-C circuit, inductance is 80 x 10- henries,
capaci tance is 5 x 10- 9 farads and resistance is negligible.
Calcula te (a) resonant frequency, (b) inductive reactance and (d
capacitive reactance.

SOLUTION:

f ,. 1
(a)
r
(4.4a)
1
- 252 kHz
2rr ~( 80 x 10- 6 ) (5 x 10- 9 )
--, ii
--4 - 50-

(b) Inductive Reactance, XL - 2nfL (3.4b)

3 6
-- i XL = 2nx252xl0 x80xlO- - 126 . 5 ohms

(c) Capacitive Reactance, X

c
:a 1/2nfC (3.5)
1
x - - 126.5 ohms
C 2 n x 252 x 10 3 x 5 x 10- 9

- j
i
j

-I
I

-i
I
j
 - 51 -

CHAPTER 5 - TESTING WITH SURF ACE PROBES

5.1 INTRODUCTION

The goal of this chapter is to present a practical approach to eddy curren t inspections
using surface probes. The emphasis is on test variables such as test frequency, probe
size and typ~; these are normally th.e only var.iables an ins~ector .h~at his co~trol.
These selectlOns are usually determIned by skin depth conslderatlOns, defect sIze, and
probe size.

;~
Impedance graphs and the Characteristics Parameter are included because they are
tools that an inspector should not be without. A thorough understanding o'f impedance
graphs is essential to manipulate test conditions to minimize and/or to cope with
undesirab~ 'test variables. Erroneous conclusions are often made by persons who do
...• ;\.
not have a working knowledge of impedance graphs .

The scope of the approach to an eddy current inspection can be very broad; a
successful outcome usually depends on the proper approach. When planning an
inspection the first questions that must be answered before proceeding are; For what
type of defects is the inspection being conducted? If the expected defects are crack,
how big are they? Do they have directional properties? What is the minimum
acceptable defect size? Does the material have ferromagnetic properties? Other
variables will, of course, influence the test but these questions must be answered in
,
order to select an appropriate probe size and test frequency.
, .!""

5.2 SURF ACE PROBES

The eddy current probe plays two important roles: it induces eddy currents, and
senses the distortion of their flow caused by defects. Sensitivity to defects and other
variables in the test article can be affected by probe design. This is achieved by
controlling direction of eddy current flow, by controlling the coil's magnetic field,
and by selecting an appropriate coil size. The effects of undesirable material
variations and/or variations in probe to test article coupling (lift-off) can often be
decreased by using multiple coils.

A surface probe, as the name implies, is used for inspecting surfaces, flat or
contoured, for defects or material properties. Defects can be either surface or
subsurface. (Surface defects are those that break through, or originate at the surface
- typically cracks, voids, or inclusions: a subsurface defect does not break the surface
and is therefore not visible).

Other names used for variations of surface probe designs are pancake probe, f~!
prQJ4e, sR£ing p~~.~ or coil, s2inniI]g probe, and pencil probe. . - - ------..- .-. -

5.2 .1 Probe Types

Simple Probes

Surface probe designs can vary from a simple, single coil attached to lead wires, to
cornplex arrangements, as shown in Figure 5.1. ~v1ost eddy current instruments require
two similar coils to satisfy their AC bridge network as discussed in Chapter 4-. If only
one coil senses the test material, it is an absolute probe; if both coils sense the test
 ...
·

- 52 -

I ~'
n'(· liEF aJlL Ittf Call

nu, If CO tE HST COIL

rEST COil /

l.ICI)itHltI t(ST
J(Sl ""lICl£ TES T urlell UT l eLE

( .) CO)

tEFEtE.e!
COIL
--
IEII COIL

TBT lITleli

(') (t)

FIGURE 5.1
Surface Probes

material, it is a differential probe. The simple probe in Figure 5.l(a) is therefore

undesirable because a second coil or electrical device with similar impedance will be
necessary for bridge nulling. An exception would be in the use of Crack Detectors;
these instruments operate with an internal balancing circuit (see Section 4-.2.3).

A better arrangement is shown in the pencil probe of Figure 5. I(b). This probe
"Ilcorporates a second coil (reference) mounted far enough from the test article that
•t will not be influenced by it. The two coils have the same impedance when the probe
is balanced in air, but will change relative to each other when the test coil is coupled
to a sample. However, the degree of coupling is usually small because of the inherent
small size of pencil probes so the coils still match well enough for most instruments
over a reasonable frequency range. The probe shown has ferrite cores; ferrite is used
for three reasons:

1. higher inductance from a given coil size,

2. small surface area in contact with the material,
3. the coil can be further from the contact surface providing greater wear
protection.

A further improvement in reference coil arrangement is shown in Figure 5.l(c); it is

attached to a disc whose properties are similar to the test material. With this
arrangement the relative impedance of the two coils will not be affected by test
frequency.
 - 53 -

The probe shown in Figure 5. L(d) is a spring loaded type designed to minimize lift-off.
The shoe provides a broad area for squarely positioning the probe on a flat surface,
while the spring maintains probe contact at constant force.

Figure 5.1 (e) shows a probe used for inspecting large diameter tubing. The probe can
be rotated and/or moved axially. The design shown incorporates a replaceable wear
cap.

Other 'Probe Designs

A multi-coil array as shown in Figure 5.2(a) is useful for inspecting tubes. This type
of probe could detect defects that would not be detected by a conventional
circumferential coil (discussed in Section 7.5).

TEST TUBE TORROIOAL REFERENCE COIL

PROBE CENTERING DISCS

, EST CO ILS FEIIR ITE

CORE

(c)
(iI)
DIFFERENTIAL SURFACE PROBE
WULTI SURFACE - COIL PROBE

FE RROIIAGNET I C
CORE

COILS
COIIPENSATING
COIL

~~~~WAGHETIC fiELD SENS INC CO IL

(b) (d)

GAP PROBE II fT -OFF COWPENSl TINC PROBE

FIGURE 5.2
Special Surface Probes

A gap probe, Figure 5.2(b), uses ferromagnetic material to shape the magnetic field.
The field is confined by the core causing eddy currents to flow in circular loops
perpendicular to the flux lines.
 - 54 -

.4... differential configura tion is shown in Figure 5.2(c); the two coils are placed side-
by-side. Both coils have high sensitivity to localized variations but tend to cancel out
the effect of lift-off, gradual material variations, or ambient temperature changes.

A lift-off compensating probe is shown in Figure 5.2(d); this probe combines the
signals from two coils to effectively rotate the defect signal relative to the lift-off
signal. Therefore, even on "rough" surfaces, shallow defects can be detected.

SEND TEST ARTICLE

CO I l
(OR I V( R co I L )
./
~"".'. ""

( .)

PICK-UP COilS
( ... OUND OPP0SINr,
EACH OTHER)

lEST TEST AR IICl E

A~T IClf
ELE(TnIC~L cnN~ECTIDNS

ORIVER COil (t>

(b)

FIGURE 5.3
Send-Receive Probes

Send-Receive Probes

Figure 5.3(a) shows a through-transmission probe arrangement. Current flowing in the

SEND coil produces a magnetic field, part of which is transmitted through the test
article. The field is detected by the RECEIVER coil, inducing a voltage. There will be
no signal variation from the receiver coil when a defect-free test article is moved
anywhere between the two coils as long as the coil-to-coil spacing remains constant.

Figure 5.3(b) shows a reflection-type probe arrangement. The probe consists of a

large send coil which generates a field, and two small receiver coils wound in
opposite directions, as mirror images to one another, as shown in Figure 5.3(c). With

- -- .- - - - -- - - - -- -
 - 55-

the probe in air, net output is zero. However, if one end is placed near a test article,
the field differs at the two ends, and a net voltage appears across the two coils.

5.2.2 Directional Properties

Eddy currents are closed loops of induces current circulating in a plane perpendicular
to the direction of magnetic flux. Their normal direction of travel is parallel to the
coil winding and parallel to the surface. See Figure 5.4.Pancake type surface probes
are therefore insensitive to poor bonding of coatings and flaws parallel to the surface
of a sample.

COIL

EDDY CURRENT FLO'S PARALLEL TO COIL .INDINGS

POOR SENSITIVITY TO LAMINATIONS

SURFACE CRACK
IN PLATE
COIL

ZERO SENS I TI VI TY LOI SENSITIVITY IIAXIMUM SENSITIVITY

'T CENTRE OF COIL PARALLEL TO 'INOINGS ACROSS "NOINGS

FIGURE 5.4
Directional Properties of a Surface Probe

When testing for flaws such as cracks, it is essential that the eddy current flow be at
a large angle (preferably perpendicular) to the crack to obtain maximum response. If
eddy current flow is parallel to the defect there will be little or no disruption of
currents and hence no coil impedance change.

When testing for flaws parallel to the surface, such as laminations, a horseshoe
shaped probe (a gap probe with a very large gap) may have reasonable sensitivity.

5.2.2.1 Sensitivity at Centre of a Coil

Probe impedance changes with coil diameter, as will be discussed furth~r in Section
5.5. A simplified derivation of this diameter effect is derived below, for the case of
no skin depth attenuation or phase lag and long coils. From Faraday's Law,

v
s -- Nd<P
dt
 ct> - BA

where r is radial distance from centre of probe;

therefore,
vs --
or
vs (l r2

4>

~I~~
N(turns)

~,r
I i i
, I I
LJ

I Ac
Resistance to flow of current is proportional to flow path length and resistivity and
inversely proportional to cross-sectional area, Ac'

R
s ---- 2lTrr
A
c
2lTrp
unit depth x unit width

or

Since I - V /Z by Ohm's Law

s s

and , at low test frequency

and no skin depth effect,

therefore,
V 2
_ - s a: r
I
s R r
s

or I a: r
s
since ct> a: - I from Lenz's Law, it follows
s s
that ct>s a: r
 - 57 -

Therefore, eddy current flow and its associated magnetic flux are proportional to
radial distance from the centre of a coil. Hence no current flows in the centre (r = 0)
and there is no sensitivity to defects at the centre of a coil.

5.2.3 Probe Inductance

The factor governing coupling and induced voltage in test material is the magnetic
flux surrounding the coil. The total magnetic flux ( <Pp ) is proportional to probe
inductance (L) and current (I), i.e., <P ex L I. In most eddy current ins truments
excitation current is kept reasonably ~onstant (in the milliampere range) but probe
inductance could vary by a factor of one thousand. The most important aspect of
inductance is that probe impedance, which is a function of inductance, must be
compatible with the instrument and signal cable,

Iz 1 ., ~R2 + X~ and e _ Arctan X.;

where XL - 2 TT f L when f is in hertz, L in henries and R is coil wire resistance in

ohms.

TABLE 5.1 SURFACE COIL IMPEDANCE

D
o
= 1.6 mm D
o
= 3.2 mm Do = 6.3 mm Do = 12.7 mm Do = 25.4 mm
L = 0.27 ~H L = 0.54 uH L= 1.1 UH L=2.1lJH L=4.3uH
R = 0.2 n R = 0.1 n R = 0.05 n R=0.0211 R=O.Oln
N = 21
40 AWG 34AWG 28 AWG 22 AWG 16 AWG
(0.080 mm) (0.16 mm) (0.32 mm) (0.64 mm) (1.3 mm)

L = 1.5 L =3.0 L = 6.1 L = 12 L = 24

R =1 R = 0.5 R = 0.3 R = 0.1 R = 0.06
N = 50
43AWG 37 AWG 31 AWG 25 AWG 19 AWG
(0.056 mm) (0.11mm) (0.23 mm) (0.45 mm) (0.91 mm)

L = 5.8 L = 12 L = 23 L = 47 L = 94
R =4 R=2 R = 1 R = 0.5 R =0.3
N = 98
46AWG 40 AWG 34 AWG 28 AWG 22 AWG
(0.040 mm) (0.080 mm) (0.16 mm) (0.32 mm) (0.63 mm)

L = 11 L = 23 L = 45 L = 90 L = 180
R =9 R=3 R =2 R = 0.9 R = 0.5
N = 136
48AWG 41 AWG 36 AWG 29 AWG 23 AWG
(0.031 mm) (0.071 mm) (0.13 mm) (0.29 mm) (0.57 mm)

L = 24 L = 49 L = 97 L = 195 L = 390
R = 17 R = 8 R=4 R=2 R = 1
N = 200
49 AWG 43AWG 37 AWG 31 AWG 25 AWG
(0.028 mm) (0.056 mm) (0.11mm) (0.23 mm) (0.45 mm)
 - 58 -

The self-inductance of a long coil (solonoid) can be calcula ted from the equa tion

(5.la)

Lo is self-inductance in henries
where ~r is relative permeability of core (normally = 1.0)
A is coil's planar surface area, m iII i met res 2
£. is coil length, millimetres.

This formula is a good approximation for coils of length/diameter ratio greater than
10.

For a short coil, end effects will reduce inductance because of lower flux at call
ends. The N2 term remains since N enters in N ¢p (total number of flux linkages) and
again since 4>p itself is proportional to N. The following approximate equation can be
used to calculate inductance of short coils:

L .- 4'TTlJ -r 2
N ( In
Br - 2) 10- 10 (S.lb)
0 r K

Do + Di
-
where r is mean coil radius - 4
, mm

and

Most eddy current instruments will operate over a fairly broad range of probe
impedance (and probe inductance) without substantial reduction in signal-to-noise
ratio and signal amplitude. An instrument input impedance of 100 ohms is typical,
although any impedance between 20 and 200 ohms is generally acceptable, unless test
frequency is too close to probe-cable resonance; see Section 5.9. Exact probe
inductance calculations are therefore not essential. To facilitate impedance
calculations, Table 5.1 has been prepared. This table lists coil inductance and
resistance (with probe away from test material) for various outside diameters and
number of coil turns, keeping both the inside diameter and coil length equal to 0.2
times the outside diameter. Wire diameter is chosen to fill available coil cross-
sectional space. Using this table and the knowledge that inductance,

~ · N:ii2] (5.2)
where N is number of turns of wire and D is aver age coll diameter, one can usually
make a reasonable estimate of wire size and number of turn s required to a chieve a
particular inductance.
 - 59 -

L.U 1 . 0c---r---r---.---r---r---r---.---r---.---r---.-~
CJ
:::J
f- O . 8~~~--~--~--r---r---~--~
r-°l-+~ ----1__~t
-I
0..
::::E
~

-I
o.6 f----+*-"<--+~-+----+----t---+-- ~-.rr----'r----r;~"7
~
tS<J t>< 1__

/
T
~ DEFECT
Z
c:,.:, o.4 r----t-'r-;--+__~+_--_P"_c--+_---+ 2 mm 0 EEP, I 2 5 mm LON G u... u...
CJ
V'l

.-
U
><
CI
II o 3 r----t--;-t-T--+_~+__--t__"""c-+__--+_--_t_--+_--_t_-+_t:
:>-
L.U
u...
L.U
"-
>< -
:>-
CJ
CI
L.U
N D = 5 mm
-J
0<1:
::E
0::
CJ
Z
o 1 ~__~__~__~__~__~__~__~__~__~__~~~~
o 0 .5 1.0 1.5 2.0 2. 5 3 .0 3 .540 45 50 5. 5 60
(a) LIFT-OFF DISTANCE (mm)

L.U
1.0
a
:::J
.....
-I
Q..
0 .8 ~
\\ ~
IDI
L><J t><I -Lx
::E
0 .6 / // //
~/
0<1:
\~\

'"
'\ ~ \\.
DEFECT~
/// ~ t
-J

~
0<1:
Z :x:
c:,.:, 0 .4 ~-
U)

.....
C>

" 0.3
>C
\\ \ "~ 2 mmDEEP, 12.5 mm LONG l.U
CI
U
l.U
u..
l.U
CI
:>-
'-.
><
::>-
0.2
\\~ \
'\
\
f,---D .25~ ~
~\ ""~
CI
L.U
N D = 5 mm \

\\ ,\
-I
<
\
~
::E
0:::
0
Z
0.1 \ \ .......
o 0 .5 10 15 2 0 2 .5 3 .0 3.5 4 .0 4 .5 5 0 5 .5 6.0
(b) SUBSURFACE DEFECT DEPTH (mm)
8 = 10 mm
8 = 2 mm

FIGURE 5.5
Decrease in Sensitivity with
(a) Lift-off (b) Defect Depth
 - 60 -

5.3 PARAMETERS AFFECTING SENSITIVITY TO DEFECTS

During eddy current inspection one must be aware of the limi tations of the technique
and should take maximum advantage of its potential. Although sensitivity to deep
surface defects is excellent, sensitivity to deep subsurface defects is very poor. A
subs urface defect only 5 mm from the surface is considered very deep for eddy
current test purposes.

There are two factors that contribute to this limitation. The skin depth effect causes
eddy currents to attenuate with depth depending on the mate~ial properties and test
frequency. This effect is normaIIy minor and can be controlled (within limits) by
reducing test f.requency. The predominant effect (rarely mentioned) is the decrease in
magnetic flux, and consequently eddy current density, with depth because of the
small diameter of most practical probes. One can increase penetration by increasing
probe diameter, but as a consequence sensi ti vi ty to short defects decreases. One
could optimize sensitivity if defect length is known; however, the maximum depth of
detectability is still very small. Unlike ultrasonic inspection where a defect is
detected many transducer diameters away, eddy current testing is limited to
detecting defects at a depth of less than one probe diameter. It is this effect of probe
diameter that limits most volumetric eddy current inspection to materials less than
5 mm thick. In following subsections, limitations are discussed and empirical
examples presented.

5.3.1 Sensitivity with Lift-off and Defect Depth

There is a decrease in sensitivity to defects as a coil is moved away from the surface.
This is caused by the decrease in magnetic flux density with distance resulting from
finite probe diameter. Figure 5.5(a) shows the extent of this decrease for three
probes of different diameters. Note, for example, the sensitivity of the smallest
probe (5 mm diameter) decreases a factor of four when moved about 1 mm from the
surface.

This loss of sensi tivity with distance also apply to defects in a solid, in addition there
w ill be a dec~ease due to skin depth attenuation.

Figure 5.5(b) illustrates the decrease in signal amplitude with subsurface defect depth
without skin depth attenuation (solid lines) and with skin depth attenuation (dashed
lines). With large skin depths (low test frequency) the decrease in subsurface defect
sensitivity with depth is similar to the decrease in sensitivity with distance for
surface defects shown in Figure 5.5(a). This implies magnetic flux density decreases
with distance from the coil in air as in a solid (without skin depth attenuation).

At a typical test frequency, where one skin depth equals defect depth (<5 '" 2 rom
for the dashed lines in Figure 5.5(b», a further decrease, by about a factor of three,
in signal amplitude at x = 2 mm is attributed to skin depth attenuation. This occurs
since at one skin depth eddy current density is 37% of surface eddy current density.

The decrease in defect sensitivity with depth in a finite thickness sample, without
skin depth attenuation, is approximately the same as in an infinitely thick sample.
However, with skin depth attenuation, defect sensitivity decreases less rapidly than
the dashed lines in Figure 5.5(b); the curve would fall somewhere in between the
dashed and solid lines.
 JU,"}-,1<:: LIIILKIIt::s:s, one cannot IJse pro Des with coil diameter much larger than sample

thickness (because of loss in sensitivity with defect length, Figure 5.6). Eddy current
testing with surface probe is therefore normally li~ited to thicknesses less than
5 mm.

5.3.2 Effect of Defect Length

Eddy current flow is limited to the area of the inducing magnetic field which is a
function of coil geometry; defect sensitivity is proportional to coil diameter in a
surface probe, and to gap width in a horseshoe probe. As a general rule, probe
diameter should be equal to or less than the expected defect length. The effect of
probe diameter and defect length is shown in Figure 5.6. For example, when defect
length equals probe diameter, the signal amplitude ranges from one-third to two-
thirds of the amplitude for an infinitely long crac:,: depending on probe diameter and
test frequency.

The sensing area of a probe is the area under the coil plus an extended area due to
the magnetic field spread. The effective diameter, De f f ~ of a probe is
approximately equal to the coil diameter, 0 c' plus four skin depths,
Deff - Dc + 46 (5.3)

At high frequencies the 4 6 term will be small and the sensing diameter can be
assumed to be about equal to coil diameter, but at low test frequencies the magnetic
field spread can be significant. In this case it is common to use ferrite cups to
contain the field. This results in a concentrated field without affecting depth of
penetration.
 - 62 -

100\ j--- - - - - - -

1 MHZ

8
; 'j;;.: m f----

I _ PROBE DllWETlR

1.3 II1II\ PROBE OllWElER

3 I MH z ~ 0 . 36 IMI

~IOO ~Hz ~ 1.16 ..."

10 12 14 16 IB

£1:* HOTCH LENGTH , nn

FIGURE 5.6
Effect of Defect Length

5.4 COMPARISON BETWEEN SURFACE AND THROUGH-WALL INSPECTION

The major limitation of conventional eddy current methods is lack of penetration.

Figure 5.7(a) illustrates typical results obtained with the conventional eddy current
method, where the probe is placed on one side only of a 4 mm thick, 100 mm
diameter tube. Test frequency is 30 kHz and skin depth, 6 -1. 7 mm. Note the
drastic decrease in signal amplitude and the signif icant phase rotation of the defect
signals with depth. A defect has to be long and very deep before it can be picked up
from the opposite side of the tube wall. This decrease in sensitivity with depth is due
to both finite probe size and skin depth attenuation.

Figure 5.7(b) illustrates typical results obtained with through-wall transmission

equipment where excitation and receive coils were located directly opposite each
other across the wall. The probes were conventional absolute pancake type surface
probes. The output signal appears as a 'figure 8' because the signal was filtered
(differentiated).
 lWFLI TUDE Of OHECr 51 ~AL. I COwPCNE~1T

1 vOL 7 I

I. O. (.RoovE
G.D GROCVE
J.g mm DE~P 0.8 em DEEP
HOLES. 0.8 mm CIA. 13 m~ LONG 13 LOI~G
13 mm LOt~G /hrn

X-y DISPLAY OF DEFECT SIGNALS

(a) Conventional Surface Probe Results

\
25% 50% 75%,
O.D. v 1.0.
GROOVE HOLES GROOVE
0.8 rrm DEEP 0,8 rrrn DIA 0.8 mn DEEP

AMPL I TUDE OF DEFECT SIGNALS, Y COMPONENT x-v DISPLAY OF DEFECT SIGNALS

(FILTERED)

(b) Through-Wall Transmission Results

UL 5. '_,_:_---"C_o_m_,..J.C.Paring Conventional and Through-Wall

Transmission Techniques
 - 64 -

The Y -ampli tude presentation in Figure 5.7(b) shows defect signal amplitude does not
change significantly with defect depth. It is important to note the phase of the
signals does not change with defect depth when using the through-transmission
method as shown in the X- Y display.

IMPEDANCE GRAPH DISPLAY

Impedance graphs are an indispensable aid in eddy current inspections. An

understanding of these graphs provides an operator a clear picture of all variables and
the ability for appropria te action to minimize effects of adverse conditions.

All information about the test article is transmitted to the test coil via the magnetic
field. The variation of the magnetic flux, ¢ • with time induces a voltage, V, across
the test coil which, by Faraday's Law, depends on the magnitude and rate of change
of ¢> and on the number of turns in the coil, N

N~ (2.4)
dt
.. - LdI/dt since ¢ OK LI/N .

The variation in amplitude and phase of this voltage vector indicates the condition of
the test article. The voltage vector can be resolved into the two quadratures, in-
phase V O. and out-of-phase V 90 • Since V =IZ and I is kept approximately
constant, the voltage graph can be replaced with the impedance graph, as discussed in
Section 3.3.

Impedance depends not only on test article variables but also on probe parameters.
The probe parameters are coil diameter, number of turns, length, and core material.
The instrument parameter that affects impedance is test frequency (since
f ex: d ¢ / d t ). To overcome the necessity of plotting im pedance graphs for each
test coil, probe impedance is normalized. The graphs can then be used to study the
effect of test article variations without dependence of probe details.

The inductive reactance component is normalized by dividing by the product of

frequency and coil inductance (wL o ) when the probe is away from test material (in
air).

- --
wL
wL
o

where w is angular frequency, radians/second

L is inductance, henries
Lo is inductance of coil in air, henries
XL is reactance, ohms
Xo is reactance of coil in air, ohms
 - 65 -

AIR
TEST ARTI CLE
INDUCTIVE
REACTANCE

wL
o

TEST ARTICLE
PI AIR
'\.
'\. TEST ARTICLE
~
wL o

I' R DC
RESISTANC

(a) BEFORE NORMALlZ.~TION (b) AFTER NORMALIZATION

FIGURE 5.8
Coil Impedance Display

The resistive component is normalized by subtracting coil wire and cable resistance,
RDC • and then dividing by wLo •

where RL is coil resistive load due to eddy currents in test material.

The normalized components XL/Xo and RL/XO are dimensionless and independent
of both coil inductance and coil wire and cable resistance. Changes in the normalized
parameters indicate variations in eddy current flow into the test article only. Figure
5.8 displays probe impedance before and after normalization. Changes in the test
article are reflected by a change in impedance point P. Figures 5.9 to 5.11 are
normalized coil impedance graphs, produced by computer simulation, showing the
change in the point P for the following sample variables: electrical resistivity
permeability, and thickness. Figures 5.12 and 5.13 show effects of test frequency and
co it diameter.
 - 66 -
1.0 1.6
- - - - - CONSTANT PER"EAB I LITY. 1',
- - - - - CONSTANT RESISTIVITY. P

0. 9 1.~

~ 1.1
u
ii ___
.....
~+'- p 700"n·,.
u

1.0

...'"
p 110
0
~
TOT 'REQUEHcY - 50 kHz
,.,- 1 ;t o. s p 5l
7 (BRASS) THICK PLATE ~
0
%
p 1.7

0. 5 0.6 .FREQUENCY • 50 kHZ

LI FT-OFF, .~
LI FT-OFF • 0
THICK PLATE

a. o+-______________________ ~ ____
0.0 0 .1 0.2 o. l o. ~ o 0.2 0.4 0.6

~ORMlIZED RESISTANCE NORMLIZED RESISTANCE

FIGURE 5.9 FIGURE 5.10

Impedance Graph-Resistivity Effect Impedance Graph-Permeability Effect

rlO_~
Z
III .,

lItt;;~·

1.0 1.0

J.

,
0. 9 0.9
\ THICKNESS. .~
FREQUENCY. kHz

0.8
\
\
\ 0.8

P tI .0) '"ua:
...«
I u
0. 7
LEAD h. ..."" 0.7
,.n· em
"..,... '"c RES I ST I VI TV • 5l

..J
I
I . 05 ...... ,.. r
LI FT-OFF • 0
• I

~ 0.6 ..J 0.6

", . 0
<r I FREQUENCY • 50 kNI ~
0 THICX PLATE
z / 1.0 "0z
0.5
0. 1 1I rT-OF' - 0.5
50 00

O. c+---____--__- r__________ ___

~
O . O+-----~----__----~----_T----
0.0 0.1 0 . 2' 0.3 0.4 0.0 0.1 0.2 0.3 0.4
NORKALIZED RESISTANCE
1I0R,.... L1ZED RESISTANCE

FIGURE 5.11 FIGURE 5.12

Impedance Graph-Thickness Effect Impedance Graph-Frequency Effect
 - 67 -

5.5.1 Effect of Resistivi ty

Figure 5.9 shows the effect of e!.ectrical resistivity for a range of conducting
materials. The impedance point moves up the curve with increasing resistivity.
Im pedance points for step chan ges in coil to test article spacing between zero and
infini ty are also included. Note that a slnall increase in spacing Wit-off) produ ces a
large impedance change. This results from decreased magnetic flux coupling to the
sample. There would be a relatively larger effect on the impedance of a small coil
than on the impedance of a large coil for the same change in spacing.

·5.5.2 Effect of Permeability

Note in Figure 5.10 there is a large impedance increase for a small increase in
permeability. Small permeability changes can obscure other test variables.

5.5.3 Effect of Thickness

Figure 5.11 traces the impedance point path as sample thickness decreases fro;n
infinity to zero. As test material becomes thinner, causing increased resistance to
eddy currents, the impedance point moves up the curve. This was also the case in the
resistivity graph, Figure 5.9. This implies that any condition causing an increase in
resistance to flow of eddy currents, cracks, thinning, alloying elements, temperature,
etc., will basically move the impedance point up the curve towards the probe
impedance in air, XL/Xo·I.

The impedance curve in Figure 5.11, from the knee down, makes a reversal swirl as
the probe moves across a conductor with increasing thickness. This is due to skin
depth and phase lag effects which overshadow all basic movements of the impedance
point.

5.5.4- Effect of Frequency

Figure 5.12 shows the effect of test frequency (an instrument parameter). As
frequency is increased, eddy currents are sampling a thinner layer close to the
surface (skin depth effect, discussed in Chapter 2). When frequency is decreased eddy
currents penetrate deeper into the material and the impedance point moves up the
curve.

Towards the upper end of the curve, impedance is mainly composed of resistance
which has a great dependency on temperature, both in the test article and in coil wire
resistance (although the latter does not appear on this normalized curve). It is
therefore desirable, when possible, to operate near the knee of the curve say,
20 to 200 kHz in this example.

5.5.5 Ef fect of Probe Diameter

Figure 5.13 shows effect of coil diameter (a probe parameter). Note increasing coil
diameter moves the impedance point down the curve, sim ilar to increasing frequency.
When test conditions dictate use of a low frequency, the operating point can often be
brought down the curve to the desired knee region b y increasing coil diameter
(pr ov ided test conditions will permit a large probe).
 - 68 -

1. 0

0 .9

0 .8

LU
u
Z
01[
..... 0 7
u
Dj 0 . 2 Do
cr
LU
D:: f 0.2 Do
Cl
w 0 .6
N -D=_o_,
D + O·
-
oJ
01[
:I( 2
D:: mm
0
z 0 .5

Frec:uency 50 k:iz
0 .4

0
NOR~ALIZED RESISTANCE

FIGURE 5.13
Impedance Graph-Surface Coil Diameter Effect

Comparison of Experi mental and Computer Impedance Diagrams

The impedance graphs shown in Figure 5.9 to 5.12, produced by computer simulation,
can be verified using a standard eddy current instrument. Figure 5.14- shows probe
response to various test variables: resistivity, permeability, lift-off, and test
frequency. The sold lines are output voltage traces generated by varying probe-to-
test article spacing (lift-off) from infinity to contact with various conducting
samples, while keeping test frequency constant a t 10 kH z , and again at 100 kHz. The
dashed lines, connecting the points when the probe was in contact with the samples,
were sketched in to show the similarity between these graphs and the normalized
impedance graphs in the preceding section. Note that the points move down the curve
with increasing conductivity and also with increased frequency. For example, the
operating point for 304 sst moved from the top of the impedance diagram at 10 kHz
to near the knee at 100 kHz.
 - 69 -

1'1
t FERRITE
t2J1~
w:~SAloIPlEI 0};' T
I
LIFHJFF
I
FERRITE

. ~ /-:
(p
I RON
liONEL 400

MONEL 4 CO
AIR
AIR

INOUCTIVE I NDUCT I VE
REACTANCE REACTANCE
304 SST
Cu

f=100kHz
f = 10 kH z

/
R RESISTANCE R RESISTANCE

(a) ( !J )

FIGURE 5.1'1
Probe Response to Various Test Parameters at Two Frequencies

5.6 CHARACTERISTIC PARAMETER

In Section 5.5 impedance graphs were normalized to make test article parameters
independent of probe properties such as inductance. Another method, proposed by
W.E. Deeds, C.V. Dodd and co-workers, combines frequency and probe diameter
with test material parameters, to form one characteristic parameter (2).

p -2
c - r w ~
r
a (5.4)
 - 70-

where r is mean coil radius

W is angular frequency
\.1 r is relative magnetic permeability (=1.0 for nonmagnetic materials)
and a is electrical conductivity.

Using this characteristic parameter, one impedance graph can be plotted to describe
four test parameters with Peas the only varia ble .

1.0 ~~-+--
1.1.1
U
:z
I I I
~ 0 . 9 14-t--c'r+-f"t--+--"tt--""oc---t- - - - ,1W fL CT =CON STAN T

~ lIFT·OFF CONSTANT
1.1.1
~ ; = COIL WEAN RADIUS
~ 0.8 ~I = LIFT {)FF/r
~ w = ANGULAR FREQUENCY
5
:z
0.7 fL = WAGNET-I C
PERI.IEABILITY
c
~ 0 • 6 boo-'f--+'r---+-~=t- CT = EL ECTRIC AL
~ CONDUCTIVITY
:IE
gz: 0.5

O.4~ __~__~____~__~__~____~~
o 005 0 . 10 0.15 0.20 0.25 0.30 0.35

NORMALIZED RESISTANCE

FIGURE 5.15
Impedance Diagram with Characteristic Parameter, P
c

Consider Figure 5.15. The solid lines are generated by starting with Pc equal to zero
and increasing the value to infinity (while holding coil to test article spacing
con stant). The dashed lines are generated by starting with the coil infinitely far
away from the test article and bringing the coil closer until it contacts (while holding
Pc constant). Note the similarity between these curves and the impedance graphs
shown in preceding sections (the horizontal scale is twice the vertical scale).
 - 71 -

The usefulness of the characteristic parameter is that it provides a modelling

parameter. Conditions of similarity are met when
-2 -2
r r
1 wl~rlol '"' 2 W2lJr202
or
-2
r
1 wl~rlP2
.. -2
r
2 W2lJr2Pl

Test 1 Test 2

1(vJ
V

PROBE l 0

lW i -t-~
SPECIMEN (D-G)

STORAGE
I OSC I LLOSCOPE
I OISPLAY
I
I
I
--~ NOMEN CLA TU RE

V • VOLTAGE
R,
\ • CURRENT

' w • ANGULAR FREQUENCY

(w= 2"f)

La PROBE INOUCTANCE
IN AIR
Roc· FROSE WIRE & CABLE
DC RESISTANCE
R, ·SFHI!.IEN At:. ~SISrAtU

SUBseRI prs:
T • TOTAL
L • INDUCTANCE
R • RES I STANCE
P PRIMARY
I,.
S . SECONDARY

VI

FIGURE 5.16
Coil Impedance/Voltage Display
 - 72 -

Test conditions with the same Pc value have the same operating point on the
normalized impedance graph. If, for example, test article resistivity measurements
were required (for checking consistency of alloying elements for instance), the best
accuracy would be achieved by operating neu the knee of the curve where there is
good discrimination against lift-off. (Equation 5.4 does not include skin depth effects,
which may be an overriding consideration).

To operate at the knee position in Figure 5.15 a probe diameter and frequency
combination are selected such that p c ~ 10. The value of Pc in equation 5.4 is
given in SI units; we can use the following version using more familiar units.
-4 -2
P a: 7. 9 x lOr f Ip (5.5)
c

where r is the mean radius, mm

f is frequency, Hz
p is electrical resistivity, micohm-centimetre
( ~r 2 1 for nonferromagnetic material)

It should be noted that the characteristic parameter, Pc' must be used in conjunction
with Figure 5.15 (obtained analytically); it cannot be used to obtain Figure 5.15.

5.7 DEFINITION OF "PHASE" TERMINOLOGY

This section attemps to clarify the concept of phase. The voltage/impedance graphs,
presented in Section 5.5, are used as a link between impedance diagrams and the
display on an eddy current instrument monitor.

In eddy current work the most confusing and often incorrectly used term is PH ASE.
Part of the problem arises because of the existence of two eddy current methods, coil
impedance and send-receive. In this section an attempt is made to clarify some of the
muJ tirle uses of the word.

Figure 5.16 shows the impedance of a probe touching test material. The two axes
represent the quadrature components, V Land V R' of voltage across a coil. In the
absence of real numbers, the axes can also be considered as the normalized
parameters wL/wLoand RL/WLo'

The following list summarizes uses of the term PHASE. One or more of these are
often used without adequate explanation because the term will have a colloquial
meaning.
wL
1. e l ' O. 1 - Arc tan R • angle between total voltage vector and res is tive
vol tage vector. L

NOTE: An impedance bridge measures amplitude of the impedance vector

Z and the angle 0 where the resistance includes Rnc This vector could
therefore not be shown on Figure 5.16. (It is shown on the impedance
diagram in Figure 5.8(a».

Change in phase of normalized resultant voltage vector as probe is moved

over a defect.
 - 73 -

Phase between secondary voltage (induced voltage) and primary voltage

(exci tation voltage). Send-recei ve instruments measure secondary vol tage.

Change in phase of secondary voltage as probe is moved over a defect.

This is approximately the phase measured by some send-receive eddy
current instruments without X- Y outputs.

Phase between the voltage signals obtained from LIFT-OFF and a crack or
void. It is related to PHASE LAG B explained below. (8 J is about double
the phase lag.)
83 is used to estimate defect depth during ET.

6. B. PHASE LAG (not shown in Figure 5.16) of eddy currents below the surface·
relati ve to those at the surface. It was deri ved in the eddy current density
equation Chapter 2, i.e. B - x /6 for semi-infinite plates, where x is
the distance below the surface and B is in radians.

Many eddy current instruments have a PHASE knob by which the entire
impedance voltage plane display can be rotated. It is common practice to
rotate the display to make LIFT-OFF horizontal. (On an eddy current
instrument display, absolute orientation of inductive and resistive axes
may be unknown).

Phase between inductive voltage and current in a ci rcui tj 0 - 90 0

5
5.8 SELECTION OF TEST FREQUENCY

5.8.1 Inspecting for Defects

The first question that must be answered before proceeding with an inspection is: For
wha t type of defects is the inspection being done? 1£ the defects are cracks: What is
the smallest defect that must be detected? Are the cracks surface or subsurface?
Are they likely to be laminar cracks or normal to the test surface? A single general
inspection procedure to verify the absence of any and all types of defects often has
little merit. Inspections often require two ot more test frequencies and/or different
probes to accurately identify defects.

Test frequency can be selected without knowledge of the characteristic parameter,

Pc .or the operating point on the impedance graph. It should be chosen for good
discrimination between defects and other variables. The most troublesome variable is
LIFT-OFF variations, so separation of defects from lift-off is the foremost
consideration.

Only the skin depth equation has to be used,

<5 - 50~ • DUD

(2.l3a)

A test frequency where <5 is about equal to the expected defect depth provides good
phase separation between lift-off and defect signals. Figure 5.17 illustrates the
display on an eddy current instrument monitor as a probe passes over surface and
 - 74 -

CO IL

~jOfF
SURFACE CRAC !<
SUBSURFACE
VO 10 (A )
SUBSURFACE
O/ff/ij~/ va 10 (B )
INCREASING
LIFT-OFF
,\ SUBSURFACE
\ VO I D (A)
SUBSURFACE x .y DEFECT SIGNALS
SURFACE VOID (B)
CRACK

(a) (b)

FIGURE 5.17
Typical Response Signals for Two Types of Defects

subsurface defects. Test frequency is such that 6 equals depth of deepest defect, and
instrument controls are selected such that a signal from lift-off is horizontal. Note
the difference in signal amplitude and angle relative to lift-off of subsurface voids A
and B. This results from skin depth attenuation and phase lag.

If, during inspection, a signal indicating a defect is observed, test frequency may be
altered to verify whether the sigp.al represents a real defect or the effect of another
variable. This discussion is expanded in the next chapter under Signal Analysis.

j.8.2 Measuring Resistivity

Resistivity can be measured at small localized areas or by sampling a larger volume

of a test article to determine bulk resistivity. The volume of material interrogated
depends on probe size and test frequency. For bulk measurements a large probe would
be used and a low frequency to maximize penetration. The skin depth equation is
again used to estimate depth of penetration at the test frequency.

Electrical resistivity measurement is a comparative technique; reference samples of

known resistivity must be used for calibration. Variables that affect the accuracy of
resistivity measurement are lift-off, temperature, and changes in the flow of eddy
currents in test articles not related to electrical resistivity (such as cracks, thickness
and surface geometry).

For best discrimination between resistivity and other variables the operating point on
an impedance graph should be considered. Figure 5.12 illustrated the effect of test
frequency on normalized probe impedance. At the top of the graph the angle,
between lift-off variations and the resistivity curve, is small. Moving down the curve
 - 75 -

the angle, separating the two variables, incre a _~ s towards the knee with no
appreciable change beyond that. However, small lift-off variations, at the bottom of
the curve, produce a lar ge impedance change. The best operating point is somewhere
between t he two extremes, near the knee of the impeda nce curve.

,I ~EfEREHCE SAl/fH
UIP£DANC£ POINT
-"" " " ' 55 I' n - COl

"-
,IIiPEOANCE PO INT
OF UHIl.HOIM_
INCREAS ING ''-l
LIFT ~FF
!- ~
--r 52,. n -em REFERENCE
SAW PLE
V
"
-- ,
u
'~2I'n-CI!I
:::>
o
z

(b ) EOOl CURRENT IHSTRUWEMT IIOHITOR DISPLAY

RESI STANCE

\» I WP[ OANC[ GR APH RESISTiVITY [frEeT

FIGURE 5.18
Resistivity Measurement and the Impedance Graph

Figure 5.18 shows the method of manipulating test conditions to best deal with lift-
off. Figure 5.l8(a) shows the resistivity impedance curve with a frequency and probe
selected to operate near the knee. Figure 5.18(b) is an enlarged section of the curve
rotated so lift-off signals are approximately horizontal. This is the view on an eddy
current instrument monitor.

Next consider temperature effects. First, test article resistivity will be a function of
temperature so test sample and standards should be at uniform temperature. A
greater potential error is in probe wire resistance, RDC. The coil wire resistance is a
part of the probe impedance circuit, so variations in temperature which affect coil
resistance will appear as an impedance change. For greatest accuracy, the inductive
reactance, XL, should be large compared to coil wire resistance; XL / ~c > 50
is desirable.

Obviously this condi tion is not easily satisfied at low tes t frequencies where inductive
reactance is low. One solution is to use a large diameter probe cupped in ferrite. The
large diameter and ferrite c up will both increase XLI ROC

Another solution is to use a Send-Receive instrument. Such an instrument has a high

input impedance, sensing only voltage changes in the receive coil. Coil wire
theretore has no effect.

Consider next the effect of changes in eddy current path not related to electrical
resistivity. If the test is supposed to be a measurement of electrical resistivity,
thickness should not influence the signal. The skin depth equation rnust again be used.
Test article thickness should be equal to or greater than three skin depths, t > 3 <5

t > 3 x 50 ~ mID
~f

or
> 22500 Hz
p
2
t

where t is thickness, p is resistivity in microhm-centimetres, and f is frequency

Other sources of signals are edge effects and surface geometry. When the test
article's edge is within the probe's magnetic field, an increase in resistance to eddy
current flow will be detected. Edge effect can be reduced by probe design, such as a
ferrite cupped probe, or by increasing tes t frequency.

If the surface of the test article is contoured, the magnetic flux coupling will differ
from that of a flat surface and a correction factor may be required.

Cracks or voids are usually less of a problem. The signal from a crack will be very
localized whereas resistivity variations are usually more gradual. The best procedure
to determine if a localized signal is from a change in resistivity is to rescan with a
smaller probe at higher and lower frequency (at least three times and one third the
test frequency). The angle between the signals from lift-off and resistivity should
vary only slightly whereas the angle between lift-off and defect signals will increase
with frequency.

An example of resistivity variations in a zirconium alloy, due to a change in oxygen

concentration, is shown in Figure 5.19.
 - 77 -

TEST ARTICLE WIDTH

( 35 em)

-
Y r-
I LO -
VOLTS

~
r-

r-
I

X, VOL TS '"
a.. W
-'

-
w U
.-
'" '"
~
Z
co:
0:
co:
(a) X-Y DISPLAY OF COIL IMPEDANCE FROM co: (.)
>- v> ....
CHANGE IN ELECTRICAL RESISTIVITY >-
z
Z
.... Z
V>
u.J
UJ ::II W ....
z W UJ
o a: ..
a.. u >-
::II -
'" :z w
0 - ex>
(.) ~
o
~ >

f
(b) WODIFIED C-SCAN DISPLAYING Y-COMPONENT
OF COIL IMPEDANCE VECTOR FROM A CHANGE
IN ELECTRICAL RESISTIVITY

AGURE 5.19
Eddy Current Signals from a Change in Electrical Resistivity on the
Surface of a Zr-Nb Test Article. Test Frequency = 300 kHz.

5.&.3 Measuring Thickness

Test frequency should be chosen so 'lift-off' and 'change in thickness' signals are
separated by a 90 0 phase angle, see Figure 5.20(a). This frequency can be calculated
using the skin depth equation. A reasonable approximation for thin sections is
obtained when

t/6 - 0.8 (5.6)

which converts to

f - 1. 6 p/ t 2 • kH z (5.7a)
 4. ..J ... ~ '-"1"-' ~I I-J' ....... -

is relative permeability (~ - 1 for nonferromagnet ic materia!).

r

In testing thick material, this equation can similarly be used to choose a test
frequency to separate lift- off and subsurface defect signals by 90°. Equation 5.7(a)
can be used by replacing t with X,

f - 1. 6 p / x 2 kHz (5.7)

where x is depth of subsurface defect.

~mlAStlG I--
THICKNESS

INCREA IN~
LlFl'()fF

....
u ./
z

-...
oC
0-
I I ,/
u
....
or;
t
I ' 0 .3 - BALANCE POINT -
FOR NOMINAL
-INCREASING -

~
- TNt"jSS TCTESSI-
0-
L> I = 0.5
i!: 1
~
(b) EOOT CURRENT INSTRUMENT MON ITOR DISPlAY

RESI STAllCE

(I) IIIPf:OANCE GRAPH · RESISTIYllY AND THICKNESS EFFECT

FIGURE 5.20
Thickness Measurement and the Impedance Graph

Conventional thickness measurement is to display the lift-off signal horizontal (along

the X axis) and use the vertical signal (along the Y axis) to measure thickness, see
Figure 5.20(b). If the signal on the instrument monitor is set to move from right to
left as the probe is moved away from the test article, a vertical movement up or
down denotes decreasing and increasing thickness respecti vely.

5.8.4 Measuring Thickness of a Non-Conducting Layer on a Conductor

An insulating layer will not conduct eddy c urrent s so measurement of its thickness is
essentially a lift-off measurement (provided it is non-ferromagnetic), i.e. the
distance between the coil and test article. At high test frequency a small variation in
 - 79 -

lift-off produces a large change in probe impedance as shown in the impedance graph
of Figure 5.9.

To minimize the signal frorn variations in the base material , the test should therefore
be done at the highest practical frequency. The maximum frequency would be limited
by probe-to-instrument impedance matching, cable resonanc e problems and cable
noise.

The measurement is a comparative technique so standard reference thicknesses must

be used for calibration.

5. 8.5 Measuring Thickness of a Conducting Layer on a Conductor

Measurement of the thickness of a conducting layer on a conducting test article can

be done provided there is a difference in electrical resistivity ( 60) between the
two. The measurement is essentially the same as the thickness measurement
described in Section 5.&.3. There is one important difference; varIables in the base
plate, in addition to the variables in the layer, will affect the signal.

Figure 5.2l(a) shows a computer simulation of a layer thickness measurement. The

model shows the magnitude and direction of variables when attempting to measure a
layer (clad l), nominally 0.75 mm thick,with resistivity P = 3 ~n.cm on a base
(clad 2) with resistivity 5 ~Il. em. The plot is part of a normalized impedance graph.
In addition to material property variables, the parameter of space (gap) between the
layers is shown as well as the effect of an increase in test coil temperature. At
10 kHz, t /6 is 0.& and, as predicted, the angle separating signals from lift-off and
layer (clad 1) thickness is about 90°. Unfortunately, so are the signals from test coil
temperature, gap, and resistivity of the base (clad 3). Some of these parameters can
be discriminated against at higher and/or lower test frequencies.

EDOY CU&&£NT IMPEDANCE PLINE

t"~
r 20
%

Am ; i
1li
. 110
. .~~ Ej
.... \

111 '~. •
:: • + 10%
::1 /
.,' . - 40%
• - 50%

I '-~.'/ _ .11 _
.• n - 20% /. p ....., ... ,
+ 50% -..:.::: ,.c.:'" _.. .
. Js (.J\ .. .~ .. ,J
.2C_"-"",-(,--01,--)_ __
o •• c·
.114 ...
~ ' · IES I SI I Wll , l

. It 2

RANGE or YA& 118 LES SHOI N IN COMP UIO R PLOTS

.no
Ut(OU(Ntl • 10 U ti
(0'
. • J•

. 0;20 .0 ; ' 0 . 0;60 ~5' 0 C6 00 pOO 06.0 . (1;60 .0"0 "'00

IO..,AUlED IIESIST.\JIi:C£ ~ •
(.)

FIGURE 5.21
Computer Simulation of a Multi-Layer Sample
 - 80 -

5.9 PROBE-CABLE RESONANCE

Probe-cable resonance must be considered when operating at high test frequencies

and/or using long signal cables, e.g., frequencies greater than 100 kHz and ca bles
longer than 30 m. Most general purpose eddy current instruments cannot operate at
or close to resonance.

Probe-cable resonance can be modelled as shown in figure 4.5. In simple terms,

resonance occurs when inductive reactance of the coil equals capacitance reactance
of the cable, i.e. when
wL EO l/wC

where w is angular frequency, in radians/second. L is coil inductance in henries and

C is total cable capacitance in farads.

Transforming this equation and substituting w-2nf shows resonance occurs when
frequency is

fr - 1/27T Ii:C (4.6a)

This approach is sufficiently accurate for most practical applications. A more

rigorous approach to resonance is presented in Section 4.3.

Resonance is apparent when a probe and cable combination, which balances at a low
frequency, will not balance as frequency is increased. At the approach of resonance,
the balance lines on the eddy current storage monitor will not converge to a null.
The two blancing (X and R) controls will produce nearly parallel lines rather than the
normal perpendicular traces, on the storage monitor. A number of steps can be taken
to avoid resonance:

1. Operate at a test frequency below resonance, such that f is less them O. 8f r'
2. Select a probe with lower inductance. (Since f r is proportional to 1/ IT':'
inductance must be decreased by a factor of four to double resonant frequency).
3. Reduce cable length or use a cable with lower capacitance per unit length (such
as multi-coax cables). This will raise the resonance frequency since capacitance
is proportional to cable length and f r is proportional to 1/ ff.
4. Operate at a test frequency above resonance, such that f is greater than 1. 2 f r'
However, above resonance the sensitivity of all eddy current instru:nents
decreases rapidly with increasing frequency because capacitive reactance
(X c -1/ we) decreases, and current short circuits across the cable, rather
than passing through the coil.
 - 81 -

5.10 SUMMARY

Test probes induce eddy currents and also sense the distortion of their flow caused by
de fe c ts. Surface probes contain a coil mounted with its axis perpendicul a r to the test
specimen. Bec ause it ind uces eddy c urrents to flow in a cir cul a r path it can be used
to se nse all defects independent of orientation, as long as they have a component
perpendicular to the surfa ce. It cannot be used to detect laminar defects.

For good sensitivity to short defec ts, a small probe should be used; probe diameter
should be a pproximately equal or less than the expected length. Sensi tivi ty to short
subsurface defects decreases drastically with depth, even a 'thin' 5 mm sample is
considered very thick for eddy current testing.

The analysis of eddy current signals is the most important and unfortunately the most
difficult task in a successful inspection. A thorough understanding of impedance
graphs is essential to manipulate test conditions to minimize undesirable test
variables. The characteristic parameter for surface probes is used to locate the
operating point on the impedance diagram. It is given by
4
l' - 7.9 x 10- r2 f/p (5.5)
c
r
where is mean coil radius, mm; f is test frequency, Hz; and p is electrical
resistivity, microhm-centimeters.

The cri terion for defect detection with impedance plane instruments is phase
discrimination between lift-off noise and defect signals. Test frequency is chosen
such that 'lift-off' and 'change in wall thickness' signals are separated by a 90 ° phase
) angle. This can be derived from the following equation:
../

')
f - 1.6 p/t- , kHz (5.7)

where t is sample thickness, mm.

If inspection is performed at high test frequencies and/or with long cables, it is

desirable to operate below probe-cable resonance frequency. This is normally
achieved by using a probe of sufficiently low inductance.

To optimize test results, the inspector has control over probe size and test frequency.
In choosing probe diameter the following must be considered:

(a) operating point on impedance diagram

(b) probe inductance and resistance
(c) sensing area
(d) sensitivity to defect length
(e) sensitivity to defect depth
(f) sensitivity to lift-off
(g) sensitivity changes across coil diameter (zero at centre)
(h) sensitivity changes with ferrite core or cup.

Choice of t est frequency depends on:

(a) depth of penetration

(b) phase lag
 - 82 -

(c) operating point on impedance diagram

(d) inductive reactance
(e) probe-cable resonance

5.11 WORKED EXAMPLES

;5.11.1 Effective Probe Diamete r

PROBLEM: Determine sensing diameter of a 5 mm probe when

(a) testing 316 stainless steel (p - 72 microhm-centimetres) at
2 MHz,
and
(b) testing brass ( P - 6.2 microhm-centimetres) at 10 kHz.

SOLUTION:

(2.13a)

(b) <5 - 50~6'i .. 1.25 mm

10

Deff - Dc + 40 - 5.0 + 5.0 - 10 mm

Characteristic Parameter

PROBLEM: If an available probe had coil dimensions of 10 mm outer diameter

and 4 mm inner diameter, determine the best frequency for
resistivity measurements of a zirconium alloy (P - 50 microhm-
centimetres).

SOLUTION: The best frequency for resistivity meaSLJre'11ents is when the

operating point is at the knee location on the impedance diagram.
This occurs when the characteristic parameter P c-1 O. Using
equation 5.5,
Pc • 7.9 x 10- 4 (10.0: 4.0) 2 f/50 - 10

therefore, f = 50 kHz.

(This calculation places no emphasis on skin depth effect, which may

be an overriding consideration).
 L

L
L

L
J

,
I-

I-
I
i
 - 83 -
J

CHAPTER 6 - SURFACE PROBE SIGNAL ANALYSIS

6.1 INTRODUCTION

Manufacturing and preventive ;-naintenance inspection of "flat" components with

surface probes is one of the oldest and most important applications of eddy current
testing. Manufacturing inspection of small steel components for defects and hardness
is alonost exclusively performed by eddy current methods. For safety reasons and
preventive maintenance (savings on replacement costs and downtime) inspection of
aircraft components for cracks and heat treatment effects has been performed since
commercial aircraft first went into service. Eddy current testing is one of the most
effective NDT methods for the above applications because it doesn't need couplants,
it is fast, and 100% volumetric inspection is often possible.

This chapter describes how to maximize signal-to-noise by proper choice of test

frequency and minimizing "lift-off" noise. Emphasis is given to signal analysis and
how to recognize and discriminate between defect signals and false indications. An
attempt is made throughout this chapter to illustrate discussions with real or
simulated eddy current signals.

6.2 EDDY CURRENT SIGNAL CHARACTERISTICS

6.2.1 Defect Signal Amplitude

A defect, which disrupts eddy current flow, changes test coil impedance as the coil is
scanned past a defect. This condition is showm pictorially in Figure 6.1 which
portrays eddy currents induced by a surface probe in a defective plate. Eddy currents
flow in closed loops as illustrated in Figure 6.l(a). When a defect interferes with the
normal path, current is forced to flow around or under it or is interrupted completely.
The increased distance of the distorted path increases the resistance to current just
as a long length of wire has more resistance than a short length.

Eddy currents always take the path of least resistance; if a defect is very deep but
short, current will flow around the ends; conversely, if a defect is very long
(compared to the coil diameter) but shallow, the current will flow underneath. In
summary, defect length and depth (and width to some degree) increase resistance to
eddy current flow and this, in turn, changes coil impedance. (The effect of defect
size on flow resistance in tube testing is derived in Section 8.2.1).

In terms of the equivalent coil circuit of a resistor in parallel with an inductor and its
associated semi-circular impedance diagram (Section 3.5), a defect moves the
operating point up the impedance diagram. Increasing resistance in a test article
changes both probe inductance and resistance.

In the preceding discussion the defect was considered to disrupt the surface currents
closest to the coil. Consider the difference between surface and subsurface defects.
When a surface probe is placed over a deep crack of infinite length, the surface
currents must pass underneath the defect if they are to form a closed loop, see
Figure 6.2(a). This is not the case with subsurface defects as shown in Figure 6.2(b).
Although the void in this picture is not as far from the surface as the bottom of the
crack, the void may not be detected. Eddy currents concentrate near the surface of a
conductor, and therefore, tests are more sensitive to surface defects than internal
defects.
 - 84 -

SURFACE COIL

EDOT CURRENTS

TEST PLATE
kg
~
~~~ r "'""':'" "'"
f-
~ EoOT CURRENT DISTORTION
AT CRACK

EOOT CURRENT (D) EDDY CURRENTS TAKE THE PATH OF LEAST RESISTANCE
DISTORTION
UNDER OR AIIOUNO A DEFECT

(~) EDDY CURRENTS FLO' IN CLOSED PATHS. A DEFECT

INTERFERES 11TH THE NORMAL PATH

FIGURE 6.1
Eddy Currents in a Defective Plate

The skin depth equation helps in the understanding of this phenomenon. In Chapter 2
it was shown that current density decreased with distance from the surface in the
following proportions:
- 6396 of the current flows in a layer equivalent in thickness to the skin depth, 6 ,
- 87% flows in a layer equivalent to two skin depths, 2 6 ,
- 9596 flows in a layer equivalent to three skin depths, 3 6 •

SURfACE COIL

tua
YOlO

....
-......
c..>

....

. ...-
""
::>

co ...
.,
........-
---...
L> ...

.... co
C>
( 3) EDDY CURRENT flOW UNDER • CRAC~ ( b) EDDy CURRENT flOt AROUND A
SU8SURFACE VOID

FIGURE 6.2
Eddy Current Flow in the Presence of (a) Surface and (b) Subsurface Defect
 - 85 -

Since only 5% of the current flows at depths greater than the 3:5 there is no
practical way to detect a subsurface defect at this distance from the surface. But in
the case of a long surface defect 3 a or greater in equivalent depth, most of the
current is flowing under the defect. Surface cracks wili be detected and depth can be
estimated even if eddy current penetration is a small fraction of the defect depth.
Once eddy currents are generated in a metal surface, they w ill follow the contour of
a crack because a potential is set-up about the crack.

6.2.2 Defect Signal Phase

From the above description one cannot predict a defect signal in detail, only its
relative amplitude and direction on the impedance diagram. A more complete
explanation requires inclusion of phase lag. Consider the cr'oss section of a surface
probe as shown in Figure 6.3(a). This pictorial view shows the distribution of magnetic
field magnitude and phase around a coil as derived by Dodd(2). The solid lines are
contours of constant magnetic field strength; the dashed lines represent constant
phase. Since the magnetic field and induced eddy currents have approximately the
same phase, the dashed lines will also represent the phase (B) of the eddy currents.
Ampli tude drops off exponentially with distance and eddy current flow increasingly
lags in phase (relative to eddy currents adjacent to the coil) both with depth and with
axial distance from the coil. Skin depth effect occurs in both radial and axial
directions.

Figure 6.3(a) permits an approximate derivation of eddy current signals for the
shallow surface, subsurface and deep surface defects illustrated. One needs to
establish a reference phase direction as starting point; the LIFT-OFF direction is
convenient and can be defined as the signal resulting from increasing the space
between the coil and test article, starting from the point when the space is minimum.

The signal or effect of defects can be imagined as the absence of eddy currents which
were flowing in the area before the defect existed at this location. As the defects
approach the coil from positions 0 to 5 in Figure 6.3(a), the signal on the eddy current
storage monitor moves from point 0 to 5, tracing the curves illustrated in Figure
6.3(b). This proc.edure is reasonably straight forward for shallow surface and
subsurface defects since they are localized and only intersect one phase and
amplitude contour at any given position. For the deep defect one has to divide the
defect into sections and determine weighted average values for amplitude and phase
a t each position.

The shallow surface defect in Figure 6.3(b) has a large component in the lif t-off
direction; primarily its approach signal makes it distinguishable from lift-off. As
defect depth increases, signals rotate clockwise due to increasing phase angle. The
angle indicated in Figure 6.3(b) is not the value calculated from the phase lag
equation,

8 .. x/a (2.14 )

where B is phase lag (radians), x is distance of defect below the surface (mm) and a is
skin depth (mm).

The angle between lift-off and defect signals is about 2 B • Although probably not
strictly true, one can imagine defect phase angle as the sum of a lag from the coil to
the defect and the same lag back to the coil.
j
- 86 -

o. O~

(o~sr,.r

PH'SE 181

. - - -_\- ---

SH'LLO' DEFECT

SUBSURfACE OUECT _ _ _..J s

OHler POSl r (ON IIIII (0 )

. ~

DEEP ~UEO

SUBSURFACE
DHEcr

SHALLOW DHECT ~_ •

~~
s
L 1r r - OFF

o
(b)

FIGURE 6.3
Derivation of Eddy Current Signal Appearance for Three Types of Defects
 - 87 -

The foregoing discussion assumes that the defect is a total barrier to the flow of
current. Although this assumption is valid for most cracks or discontinui ties t some
cracks are partial conductors. Fatigue cracks, formed when the test article is under a
tensile stress, can become tightly closed when stress is released. The result is that
some fraction of eddy currents could be conducted across the crack interface and the
magnitude of the coil impedance change due to the defect will be less. The phase lag
argument is still valid; a deep crack will still be dis tinguishable fro:n a shaJIow crack
by the shape of the eddy current signal, but the sensitivity to such a crack will be
reduced because of smaller arnpli tude.

6.3 EFFECT OF MATERIAL VARIA nONS AND DEFECTS IN A FINITE THICKNESS

For each test, one must decide on the test frequency to use and on the phase setting.
The conventional way of setting phase on an eddy current instrument is to display the
"lift-off" signal horizontally (on the X-axis) with the impedance point moving from
right-to-Ieft as the probe is raised. All material variables will then display an eddy
current signal at an angle clockwise to the lift-off signal.

7 mm

n I
1 IX1 I><J
WflVT/ZZZZ2ZzzzZ/J
~
f
,
t

I
LI FT -OFF 1.5 mm
p = 72 }J- n· em 2.0 mm
fLr = 1.0

Probe Response to Various Test Parameters at Three Frequencies

Discrimination between defects and other variables is accomplished through pattern

recognition and varying test frequency. Figure 6.ft displays the change in coil
impedance loci for various parameters at different test frequencies. The electrical
resistivity (6p) signal angle, relative to lift-off, increases only slightly as frequency
 \
,"" "-
\: +:-;= 0 0 ~?-==t
,,
\ "-
"-
2 ... 1,5 MO 61

\
',..
0, ,
\~
,-'
wl
\~,
, \
p , 12 ,. a· em
\
\ ~II = +25 eu
\
, \ 6' • -25%
\
\ ll~ = +25%
0.25 ... '. ,. , 1.0
\
, .
0.'
'<y.
\ 6,.' 6,.

: 6.
200 kHz

0.1

(a)
1.0
"-
,\ , 1I FT ·OFF
1\ "
1\ '
0.9 \ "

,
1\
, \\
.... ,
, ,6"~6
,.
6'

'\ \ 0.25 .......... • . 10 kHz

0.1 ,
\

\.-
\
\
\
2 ..
6.
I:;' \
1-:' \
0.1
,~
,-, \
\
\
.. l \ \
.. l, \ \
\ \ 12 ... n· co>
\ \ "
6, .25%
0.6
\ '\ 61 -25%
\
\
\
, \ ,
6,.' +25%
,.
, 0.25 _ \6,. :6,. 1.0
0.5
\, \t,.. 6'
'

\ 50 kHz
I
0.25 ...
o.f \t>r{.,,"
.",(' 200 kH Z
.' 61

0.1 0.1 o. ) o.• 0.5

( b)

FIGURE 6.5
Computer Simulation of Probe Response to Various Test Parameters
 - 89 -

is increased, whereas a change in plate thickness ( (j t) signal angle continually

increases with frequency. The angle, between the signal from lift-off and plate
thickness change, equals about twice the phase lag across the plate thickness. The
signal from a change in magnetic permeability (6 u) of the plate is approximately
90° to the lift-off signal at low frequency and decreases only slightly with increasing
frequency.

Figure 6.5(a) illustrates a computer simulation of coil response to various test

parameters. The simulation is based on the same probe and test sample used in the
previous figure. Comparison of these two figures reveals computer simulation gives
very realistic results.

Note at 50 kHz the increase in magnetic permeability signal (6 \.l) is to the right of
the electrical resistivity signal for the 7 mm probe. Forthe 25 mm probe at 50 kHz it
is to the left of the 6 p signal. As the operating point moves down the impedance
curve with increasing probe diameter, a resistivity signal rotates CW relative to a
permeability signal. Note also that the permeability signal is not perfectly parallel to
the inductive reactance axis. This is due to the skin depth and phase lag changing
with permeability, rotating the signal CWo

During general inspection for all parameters in a thin plate test frequency is normally
chosen such that 'lift-off' and 'change in plate thickness' signals are separated by 90°
on the impedance plane. This frequency is empirically derived by setting ratio
between plate thickness and skin depth equal to approximately 0.8,
tic 0.8 (5.6)

Substituting in equation 2.13 yields

2 (5.7)
f = 1.6 p/t kHz
where p is electrical resistivity (microhm-centimetres), and t is plate thickness
(mm).

This frequency has been proven in practice on various conductivity samples and
various probe diameters. The 90° phase angle increases only slightly with increasing
probe diameter, see Figure 6.5(b). All defect signals (from surface or subsurface
defects) will fall inside this 90° band. Shallow defects, cracks or pits, on the opposite
side of the plate will produce a signal whose angle approaches that of wall thickness,
i.e 90°. Shallow defects on the surface nearest the probe will produce a signal whose
angle is close to that of lift-off.

The two methods of discriminating between defects and other variables, pattern
recognition and varying test frequency, complement each other. Consider signal
pattern behaviour due to nominal wall thickness and resistivity variations. These
variables normally change gradually along a sample. Whereas cracks, pits, and
subsurface voids or inclusions exhibit a step change. Discrimination between these
variables is enhanced by analyzing their behaviour at different test frequencies, as
shown in Figures 6.4 and 6.5. An extremely important point to remember is that all
defects will fall between the 'lift-off' signal angle and the 'decrease-in-wall-
thickness' signal angle regardless of frequency. (For practical applications this
statement is valid; however, the signal from a shallow defect with length greater
than a probe diameter may dip slightly below the lift-off signal).
 - 90 -

COIL

r' '1 CALIBRATION

~~0 / CRAC~
LI fT -OFF " " REAL
1 CRACKS
T~ I Ip ~
/0.5 mlfl 2
mm
SAMPLE : p = 50 J-Ln • em
J-Lr = 1.00

CRACK
CRACK
\ 2 mm DEEP NOTCH

--J~
LIFT-OFF LJ FT ·OFF
0.5 mm DEEP NOTCH 0 . 5 mm DEEP NOTCH

FREQUE~CY = 50 kHz FREQUENCY = 300 kHz

FIGURE 6.6
x- Y Display of Coil Impedance Vector from Calibration Grooves
and a Real Crack. Estimated Depth = 1.3 mm.

6.4 COIL IMPEDANCE CHANGES WITH DEFECTS

• 6.4.1 Surface Defect Measurement

Figure 6.6 illustrates the method used to predict depth of surface defects. Pattern
recognition is used where coil impedance response from the defect is compared with
calibration defects. To estimate defect depth by pattern recognition, the real and
calibration defect signals must be comparable in amplitude. This can be achieved by
changing the gain of the display (normally by decreasing the calibration defect
signals). Defect depth is estimated by interpolation.

Ampli tude of defect signals is not a reliable parameter ior estimating defect depth.
Amplitude is affected by length and the degree of contact across the two interfaces
(e.g., crack closure). Whereas t he coil impedance locus (the X- Y display of coil
impedance) depends mainly on the integrated response with depth of the eddy current
phase lag.

6.4.2 Subsurface Defect Mea sure ment

Signals from subsurface defec t s, Figure 6.tO(b), have a n average phase a ngle relative
to lift-off of approximately 2 B where 8 is the phase lag of the eddy currents at
 - 91 -

depth x. This signal is similar to a change in wall thickness signal and its phase was
denoted t>y 0 3 in Figure 5.16.

6.5 COIL IMPEDANCE CHANGES WITH OTHER VARIABLES

6.5.1 Ferromagnetic Indications

In eddy current testing the test coil is sensitive to many test parameters. One
variable that often causes problems is magnetic permeability. At common test
frequencies one can easily mistake a signal due to increased permeability
(ferromagnetic indication) for a serious defect. The following discussion briefly
outlines the problem and shows how one can differentiate between defects and
ferromagnetic indications.

It is generally recognized that magnetic saturation is required for eddy current

testing of ferromagnetic alloys. Conversely, saturation is not usually employed when
testing "non-magnetic" alloys such as austenitic stainless steels and nickel base
alloys. Unfortunately, these alloys and any alloys containing iron, nickel or cobalt
can display variations in magnetic permeability. This is caused by the strong
dependence of magnetic properties on metallurgical variables such as composition,
grain size, thermal processing, cold work, contamination and segregation.

The following are examples of ferromagnetic indications in nominally nonmagnetic

alloys which have been encountered:

Ferromagnetism associated with manufacturing defects in Inconel 600

extrusions (possibly from chromium depletion at the surface).
Ferromagnetism associated with EDM calibration grooves in Type 304 stainless
steel.
Permeability variations occuring in austenitic stainless steel castings probably
due to segregation (or possibly contamination).
Ferromagnetic inclusions in zirconium alloys resulting from pick-up during
forming.
Magnetite ( Fe 3 0 4) deposits on heat exchanger tubes due to steel corrosion
somewhere else in the cooling system.

The first two types of defects would have made defect depth predictions seriously
inaccurate, and the last three types of ferromagnetic indications could have bee;')
mistaken for defects such as cracks or pitting.

Some of the anomalous ferromagnetic indications listed above could be suppressed by

saturating the test area with a permanent magnet possessing a flux density of a few
kilogauss. If saturation is not possible (or incomplete) there is another way to
determine if an indication is due to a defect or a magnetic effect. The method
involves retesting at a much lower frequency. It is illustrated in Figure 6.7 for the
case of a surface probe passing over defects and a ferromagnetic inclusion.
 - 92 -

A t typical test frequencies 000-500 kHz) there is little phase separat ion between the
signal from defects and magnetic inclusions. As test frequency is reduced, the
operating point moves up to the impedance curve and defect signals rotate as shown.

1.00
lIFi {Iff

~
W 1
Y

,
/
fERROlllCHfllC
INCLUSION FERROlllCHEllC
INCLUSION

.95

tRACKS
.10
PROBE Oil = 7 nun L.O
0 .5 /11111 DEE P
.15 SAIIPLE P = 50,. 0 · em
HOICH
...AI~
1 )
", = I 101HZ

......, . 10
2 """ DEE P

.-..
z

-...
'"'
' -'
or:
.15
~""
c LIFT oOH
....
-
.10 0 .5 """OEEP
-"
00 1Hz
lc
z .65
1200 I

I
2 mm DEEP
I

~
.60
I
I
.55 l O.
O~ __ ~ __ ~ __ ~ ____ ~ __ ~
0.5 mm DEEP
0.05 0.10 0.15 0.20 0.25
5001HZ
NORIIALI ZED RES I STANCE R,

FIGURE 6.7
Coil Impedance/Voltage Display at Three Frequencies

The important point to note is that relative to lift-off, defect signals rotate CCW
whereas the magnetic inclusion signal rotates CW and approaches 90 0 at low
frequency (approximately 10 kHz or lower for the above probe and sample). On the
impedance diagram of Figure 6.7 the direction of the ferromagnetic signal would not
vary appreciably with frequency; increased permeability primarily increases coil
inductance.

When a magnetic inclusion is not on the surface - if it is subsurface or on the opposi te

side of a thin test plate - there is the added complication that the angle of the signal
will be rotated relative to the angle of a ferromagnetic indication on the surface
adjacent to the coil. This arises from phase lag across the plate thickness. The
previous approach of retesting at reduced frequency will also serve to distinguish
between defects and magnetic inclusions. If the phase of the signal from the
indication increases to 90 0 relative to 'lift-off', it is a ferromagnetic anomaly; if it
decreases to nearly 0 0 , it is a defect.
 - 93 -

To summarize:
(a) Many nominally "non-magnetic" alloys can exhibit ferromagnetic properties and
almost any alloy can pick-up magnetic inclusions or contamination during
manufacture or service.
(b) At normal eddy current test frequencies magnetic indications will often appear
similar to defects
(c) Magnetic indications can be distinguished from defects by retesting at a
reduced test frequency.

6.5.2 Electrical Resistivity

Electrical resistivity is a rnaterial parameter which, unlike a defect, usually varies

over significant area. However, if it is localized, and the eddy current signal is small,
it could be be mistaken for a small defect. The best means of distinguishing the two
is to rescan with a smaller probe a the same test frequency, at three times the test
frequency, and at one third the test frequency. Unlike a defect signal, the angle
between resistivity and lift-off changes little with frequency. See impedance graph in
Figure 5.9.

As with the detection of any signal source, resistivity is affected by skin depth. At
high frequency, when skin depth is small, there will be greater sensitivity to surface
resistivity variations. At lower test frequency, eddy currents penetrate deeper into
the material so the measurement will represent a larger volume.

6.5.3 Signals from Changes in Sample Surface Geometry

Abrupt changes in surface curvature result in eddy current signals as probes traverse
them. It causes changes in coupling creating a large lift-off signal and the curvature
also changes eddy current flow distribution creating an effective resistance change,
yielding a signal at an angle to the lift-off direction. The combined effect may be a
complicated signal, as shown in Figure 6.8. The appearance of this type of signal will
not change significantly when rescanned at higher and lower test frequency.

Such signals can be difficult to analyze because they depend on how well the probe
follows complicated surface curvatures. Basically the direction of the impedance
change obeys the following rules when using surface probes:

decreasing radius of curvature on an external surface, e.g., ridge, produces a

change in the direction of increasing resistivity,
decreasing radius of curvature of an internal surface, e.g., groove, produces a
change in the direction of decreasing resistivity.

Figure 6.8(a} illustrates the signal as a probe traverses a shallow groove (decrease in
surface radius) on the internal surface of a 100 mm tube. Figure 6.8(b) shows the
signal as a probe traverses a flat (increase in surface radius). The test was done with
a 9 mm diameter probe at a test frequency of 300 kHz.
 - 9lt -

1 VOLT

-----
c::::::::: cr: - 1 VOLT
•
(a) WIDE SHALLOW GROOVE (b) LOCAL'FLAT SPOT

FIGURE 6.8
X-Y Display of Surface Coil Impedance for Internal Surface Variations
in a 100 mm Diameter Tube

6.6 CALIBRA nON DEFECTS

Analysis of eddy current signals is, for the most part, a comparative technique.
Calibration standards are necessary for comparing signal amplitude and phase (shape)
of unknown defects to known calibration defects. Calibration signals are also used for
standardizing instrument settings, i.e., sensitivity and phase rotation.

Existing national specifications and standards only supply broad guidelines in choice
of test parameters. They cannot be used to establish reliable ET procedures for most
inspections. Figure 6.9 shows a calibration plate proposed by the authors for general
application. The effect of the following can be established using this plate:

1. Varying Electrical Resistivity

2. Varying Thickness
3. Surface Geometry (Curvature)
4. Defect Length for Constant Depth
5. Defect Depth for Constant Length
6. Increasing Subsurface Defect Size for Constant Defect Depth
7. Increasing Distance of Subsurface Defects from the Surface with Constant
Defect Size
8. Varying Thickness of a Non-conducting Layer (lift-off)
9. Varying Thickness of a Conducting Layer
10. Ferromagnetic Inclusions
 - 95 -

I I I J J I I

~ON 'CONouellNe COPPER CHROIi I Ull

LATER LATER PLATE

0.2 ... I. 0 "'" 0.1 ...

D. I .... D.S ... . ,5 '"'"

-

0.05 ...
~I .01 ....

(bl BACK 510£

211ft

1.5 1M

0.7 mI.

..
0.12 0.25 0.5
DEPTl1.
1.0
~II
2.0 •• 0
..
u •

I:

.-..
I I w.
I. 2. 4. 10
.~

LENGTH . Oft
25
0=2"'" CONSTANT OEPTH = 0.5 .,...

0=0 . 5 mil
>0
1=0 1M
COPPER •
IRON
0
YOlO
RI : 50 ow R1 = SO

(il FRONT SIDE

FIGURE 6.9
Calibration Standard
 - 96-

More than one calibration plate would be required to cover a complete range of
materials. A group of three would normally suffice, comprising base materials:
aluminum alloy, p-=4 un. em; bronze, p - 2 5 ~n . C In ; and Type 316 stainless steel,
p --74 ~n. em.

Figure 6.1O(a) illustrates eddy current signals obtained with an absolute surface probe
from some of the calibration block defects. Figure 6.10(b) illustrates signals from the
same defects using a differential surface probe, similar to that in Figure 5.2(c).

0.5 mm DEEP
4 mm DEEP 4 mm OEEP

mm DEEP

0.5 mm DEEP

LIFT-OFF

SURFACE DEFECTS

0.7 mm DEEP
0.7 mm OEEP

1.5 mm OEEP
1 .5 mm DEEP
I
-i
LIFT·OFF

SUBSURFACE DEFECTS

(a) (b)

FIGURE 6.10
Eddy Current Signals With (a) Absolute and (b) Differential Surface Probes
 - 97-

6.7 SUMMARY

Defect signal amplitude is a function of defect length, depth and closure (if a crack).
Signal phase is primarily a function of defect depth. For volumetric Inspection of thin
material the following test frequency should be used:

f - 1.6 p/t 2 t kHz (5.7)

where p is electrical resistIvity, microhm-centimetre, and t is wall thickness, mm.

At this frequency there is good discrimination between defects and lift-off signals
but not between defects and ferromagnetic signals. Magnetic indications can be
distInguished from defects by retesting at reduced frequency. Defect signals rotate
CCW (a'pproaching 0°) whereas ferromagnetic signals rotate CW (approaching 90°)
relative to lift-off signals.

There are few national standards governing eddy current inspections with surface
probes. For effective inspection, a calibration block should simulate the test piece
and contain appropriate surface and substrate defects along with ferromagnetIc
inclusions. Basic knowledge of phase lag and impedance diagrams is also required for
reliable analysis of eddy current indiCations.
- - -- ----
 - 98 -

CHAPTER 7 - TESTING OF TUBES AND CYLINDRICAL COMPONENTS

.1 INTRODUCTION

Tubes or rods up to about 50 mm diameter can be inspected for defects with

encircling coils. Defect sensitivity in larger diameter components decreases because
the inspected volume increases while defect "volume" remains the sarne for a given
defect. For larger diameters, surface probes should be used to obtain higher defect
sensi ti vi ty, see Chapter 5.

The components can be in the form of wire, bars or tubes and round, square,
rectangular or hexagonal in shape, as long as appropriate coil shapes are used.
Inspection is fast and efficient since an encircling coil samples the complete
circumference of the component, allowing 100% inspection in one pass.

Defect detectability depends on disruption of eddy current flow. Therefore, the best
probe is the one which induces highest possible eddy current density in the region of
material to be inspected, and perpendicular to the defect.

When planning an inspection, the following questions must first be answered:

For what type of defects is the inspection to be performed?
If cracks are expected, do they have directional properties?
Does the material or components in close proximity have ferromagnetic
properties?

Once these questions have been answered one can decide on suitable probe design,
test frequency and calibration standards. With the proper procedures one can
discriminate between defect signals and false indications as well as determine depth
once a defect is located. These procedures are based on a knowledge of impedance
diagrams and phase lag.

7.2 PROBES FOR TUBES AND CYLINDRICAL COMPONENTS

1. 2• 1 Probe Types

Four common probe types for testing round materials are illustrated in Figure 7.1: (b)
and (d) are differential probes, (a) and (c) show absolute probes. Each type contains
two separate coils to satisfy AC bridge circuit requirements, which is the typical
mode of operation of most eddy current instruments, see Chapter 4. These bridges
require matching coils on two separate legs of the bridge to balance, thus perrnitting
amplification of the small impedance differences between the two coils. If the two
coils are placed side-by-side, both equally sensing the test material, the probe is
"differential". If one coil senses the test article, the other acting only as a reference,
the probe is absolute.

Figure 7.l(a) and (c) show effective designs for absolute probes; the piggy-back
reference coil is separated from the test article by the test coil and therefore
couples only slightly to the test article (fill factor < <1) •
 - 99 -

CENTERING OI SC S

REF ERENCE COI L

GUIOES
ItSI COIL
RE FE .~ENCE CO I L

(AI (NCIRCLING PROBE . ABSOLUTE (el INTERNAL PROSE • • SSOLVTE

(P IGGY-SACK REF ERENCE) (PIGGY-SACK REfERENCE)

(D) INTERNAL PROBE . OIFFERENTIAL

( S) ENCIRCLING PROSE . DIFFERENT IAL

FIGURE 7.1
Tube Probe Types

Coil Size

The best compromise between resolution and signal amplitude is obtained when coil
length and thickness equal defect depth. See Figure 7.2 for a labelled diagram of a
probe cross section.

As a general guideline for tube inspection, coil length and depth should approximately
equal wall thickness. However, to improve coupling, a rectangular cross section with
thickness reduced to one-half the length can be used. For greater sensitivity to small
near surface defects, coil length and thickness can both be reduced further.
Unfortunately this will result in a decrease in sensitivity to external (far surface)
defects.

Coil spacing, in differential probes, should approximately equal defect depth or wall
thickness for general inspections.
 - 100 -

COIL SPACING

CO I L TH I CKNESS
+--f---+-_. - - D (AVERAGE COIL DIAMETER)
TUBE-COIL
CLEARANCE
7ZZfZZZZZ/ZZZZZZZZZZZZ/

FIGURE 7.2
Probe Coil Nomenclature

For increased sensitivity to near surface defects, spacing can be reduced at the
expense of a reduction in sensitivity with distance from the coil.

Probe-to-tube clearance or gap should be as small as possible. In most internal tube

inspections, a gap equal to half the wall thickness is common. A larger gap (smaller
fill-factor or coupling) results in a small decrease in near surface defect resolution
and a large decrease in signal ampll tude for all types of defects.

~ . __ L Comparing Differential and Absolute Probes

Absolute probes with a fixed reference coil are essential to basic understanding. They
enable study of all physical properties of a test article by plotting characteristic
impedance loci.

When an absolute coil signal is plotted as a function of distance (as the probe travels
along a tube axis) dimensional variations and discontinuities can be separated. See the
example of Figure 7.3(b). The signal is a. function of effective cross-sectional area of
eddy current flow, i.e., wall thickness in the case of tubes, and can be analyzed like a
surface roughness trace with the extra advantage that subsurface flaws can be
sensed.

In tube testing with an internal coil, absolute probe signals from defects and supports
are simple and undistortedj signals from multiple defects and defects under support
plates are often vectorially additive.
 - 101 -

Differential probes have two active coils usually wound in opposition (although they
could be wound in addition with similar results). When the two coils are over a flaw-
free area of test sample, there is no differential signal developed between the coils
since they are both inspecting identical material. However, when first one and then
the other of the two coils passes over a flaw, a differential signal is produced. They
have the advantage of being insensitive to slowly varying properties such as gradual
dimensional variations and temperature: the signals from two adjacent sections of a
test article continuously cancel. Probe wobble signals are also reduced with this
probe type. However, there are disadvantages; the signals may be difficult to
interpret, even to the extent of being misleading. Defect signals under support plates
can be extremely complicated. The signal fro~ a defect is displayed twice: once as
the first coil approaches the defect and again for the second coil. The two signals
froi11 a mirror image and the signal direction from the first coil must be noted. If a
flaw is longer than the spacing between the two coils only the leading and trailing
edges will be detected due to signal cancellation when both coils sense the flaw
equally.

:, :, suPPORr PLATE POSITION

:Z;;;%~//J~ I~ SECT ION THROUGH TUBE

SHO"NG CORRODED ARE'

OIFfERENTIAl COilS ABSOLUTE COIL

(a)

A, ~ .,-. . ...... . I RACE "TH ABSOLUTE

PROBE

___--::====~=_ .." CO" I' COO,,,,,,

___VU~V_~~
10 J

01 STANCE
IRACE .,IH DIFFERENTIAL

I PROBE

--,,-,t---- "'ll lOSS T COvPONENI

FIGURE 7.3
Eddy Current Y -Channel Recordings from a Brass Heat Exdlanger Tube

OD = 26.9 mm~ t=I.1mm, f90 = 21 kHz

 - 102 -

An even more serious situation occurs with differential probes when the ends of a
flaw vary gradually; the defect may not be observed at all. A.n example of this is
shown in Figure 7.3; this brass heat exchanger tube suffered general corrosion as well
a s localized corrosion on either side of a support plate. The gradual upward trendof
the Y -DISTANCE recording in Figure 7.3(b) shows the pronounced grooves at A and B
a re superimposed on an area of general wall thinning in the vicinity of the support
pia teo Note the response of a differential probe to the same defect in Figure 7.3(c).
The differential probe senses the localized grooves but the Y-DlST ANCE recording
shows no indication of the gradual wall thinning which was apparent in Figure 7.3(b).
Table 7.1 compares advantages and disadvantages of the two probe types.

TABLE 7.1 1 ~
COMPARISON OF ABSOLUTE AND DIFFERENTIAL PROBES

ADV ANT AGES: DISADV ANT AGES:

ABSOLUTE PROBES

- respond to both sudden and gradual - prone to drift from temperature

changes in properties and dimensions instability
- combined signals are usually easy to - more sensitive to probe wobble than
separate (simple interpretation) a differential probe
- show total length of defects

DIFFERENTIAL PROBES

- not sensitive to gradual changes in - not sensitive to gradual changes (may

properties or dimensions miss long gradual defects entirely)
- immune to drift from temperature - will only detect ends of long defects
changes - may yield signals difficult to
- less sensitive to probe wobble than an interpret
absolute probe
I
r.2.3 Directional Properties

When inspecting for defects, it is essential that flow of eddy currents be as

perpendicular as possible to defects to obtain maximum response. If eddy currents
flow parallel to a defect there will be little distortion of the eddy currents and hence
Little change in probe impedance.

The eddy current flow characteristics of circumferential internal or external probes

are listed and illustrated in Figure 7.4.
 - 103 -

COl L EOO' CURR ENTS

EOOT CURRENTS fLO. IN ClO SEO PAlHS .

..
EOO' CURRfNT fLO,1 P'R'llfL TO [DO' CURRENI fLO' OIWINIS>< [ S TO
lIWITEO TO CONOUCIING WAT[RIAL COIL "NOINGS . HOI SfNliTIYE ZERO IT TH[ CEHTRE Of A SOLID ROO
TO PIIRE 1I CI RCU.f EREN II iL CR in S NO I[NIITIVIl! iT CEHTRE

fOOT
EOOT CURRENTS
CO IL
CURREN TS

fOOT CURRENT FLO'S PARAllEL E:lOT CURRENTS CONCENIRATE NEAR THE

TO TUBE SURflCE NOT S[NSITIY[ SURfIC[ CLOSE TO THE CO I L OE PTH
TO Li.,NAR SEPiRilIONS . Of P!:NEIRAT!ON I I CONIROLLEO BY fEST FREOUENcY

FIGURE 7.4
Directional Properties of Eddy Currents in Cylindrical Test Articles

In addition to considerations of eddy current flow direction the following are

important:

Magnetic flux is not bounded by the tube wall but will induce eddy currents in
adjacent conducting material, e.g. tube support plates in heat exchangers.
Eddy current coils are sensi tive to ferromagnetic material introduced into a
coil's magnetic field. The ferromagnetic material need not be an electrical
conductor nor need it form a closed path for eddy currents.
Eddy currents coils are sensitive to all material variations that affect
conductivity or permeability.

7.2.4 Probe Inductance

The equations quoted in Section 5.2.3 to calculate inductance for surface probes are
also used to calculate inductance of probes for testing tubes and cylinders. The
important aspect of inductance is that probe impedance, which is a function of
inductance, must be compatible with eddy current instrument and signal cables,

Zprobe = ~R2 + x~
where XL - 2 1T f L when f is in hertz and L in henries and R is coil wire
resistance in ohms.
t.

- 104 -

TABLE 7.2 ENCIRCLING OR INTERNAL COIL IMPEDANCE

Do = 8.9 mm D
0
= 12.7 mm Do = 15.9 mm D
0
= 19.1 mm D
0
= 22.2 mm Wire Size

L = 6.1 lIH L=ll \JH L = 15 liB L = 20 \JH L = 25 \JH 31 Awe

N = 25 n
(0.23 mm)
R =0.3 n R = 0.4 R = 0.5 n R = 0.6 n R = 0.7 n

L = 23 L = 42 L = 59 L = 77 L = 96 34 AWe
N = 49 (0.16mm)
co
R =1 R = 1.5 R=2 R=2 R=3

L = 64 L = 110 L = 160 L = 210 L = 260 37 Awe

N = 81 (0.11 mm)
R =3 R=5 R=6 R=8 R=9

L = 200 L = 360 L = 510 L = 660 L = 830 39 Awe

N = 144 (0.089 mm)
R=9 R = 14 R = 18 R = 22 R = 26

L = 490 L = 880 L = 1.24 mH L = 1.62 mH L = 2.02 mH 41 Awe

N = 225 (0.071 mm)
R = 24 R = 35 R = 45 R = 55 R = 64
 - I 05 -

~ost eddy current instruments will operate over a fairly broad range of probe
impedance without a substantial reduction in signal-to-noise ratio or signal
amplitude. An instrument input impedance of 100 ohms is typical, although a probe
impedance between 20 and 200 ohms is normally acceptable, unless the test
frequency is too close to probe-cable resonance frequency, see Section 7.2.5. Exact
probe inductance calculations are therefore not essential.

To facilitate impedance calculations Table 7.2 has been prepared. This table lists coil
inductance and resistance (with probe in air) for various diameters and wire sizes
while keeping coil cross section constant at 1.2 mm x 1.2 mm. (These dimensions are
fairly typical of tube wall thickness in heat exchangers). With the aid of this table,
and knowledge that inductance is proportional to the square of number of turns and
the square of mean coil diameter (L a NZ0 2 ), one can usually make a reasonable
estimate of wire size and number of turns for a particular probe.

7.2.5 Probe-Cable Resonance

Probe-cable resonance must be considered when operating at high test frequencies

and/or using long signal cables, e.g. frequencies over 100 kHz or cables longer than
30 m. Most general purpose eddy current instruments cannot operate at or close to
resonance.

Probe-cable resonance can be modelled as shown in Figure 4.5. In simple terms,

resonance occurs when inductive reactance of the coil equals capacitive reactance of
the cable, i.e. when

wL • l/wC

where w is angular frequency, radians/second

L is coil inductance, henries
C is total cable capaci tance, farads

Transposing this equation and substituting w c 2n f shows resonance occurs when

frequency is

f - 1/2n ILC
r
This approach is sufficiently accurate for most practical applications. A more
rigorous approach to resonance is presented in Section 4.3.

Resonance is apparent when a probe and cable combination, which balances at a low
frequency, will not balance as frequency is increased. At the approach of resonance,
the balance lines on the eddy current storage monitor will not converge to a null. The
two balancing (X and R) controls will produce nearly parallel lines, rather than the
normal perpendicular traces, on the storage monitor. A number of steps can be taken
to avoid resonance:

1. Operate at a test frequency below resonance, such that f t est is less than 0.8 f r .
2. Select a probe wi th lower inductance. (Since f r is proportional to 1/ IL
inductance must be decreased a factor of four to double the resonant
frequency).
 - 106 -

3. Reduce cable length or use a cable with lower capacitance per uni t length (such
as multi-coax cables). This will raise the resonance frequency since capacitance
is proportional to cable length and f r is proportional to 1 I rc .
4.. Operate at a test frequency above resonance, such that f t: est is greater than
1.2 fro
However, above resonance the sensitivl ty of all eddy current instruments
decreases rapidly with increasing frequency because capaci tive reactance
(X c = l/wC) decreases, and current short circuits across the cable rather
than passing through the coil.

7.3 IMPEDANCE PLANE DIAGRAMS

Eddy current probe for testing cylindrical components differ mechanically from those
for plate testing, but coil impedance can be treated similarly for both test coil
configurations. The impedance display treatment introduced in Chapter 5 applies for
internal and external circumferential coils with the following changes:

1) Lift-off becomes "fill-factor". Fill factor is a measure of coupling between the

coil and test object. In general, it is the fraction of magnetic field that crosses
the test object; for a long coil, this is the fraction of the test coil area filled
with test material. Fill-factor, n (eta), is the ratio

(7.la)

for an encircling coil,

and -2 2 (7.lb)
n '" D ID.1
for a bobbin type internal coil,

where Qo is cylinder diameter

D is average coil diameter
and Di is tube internal diameter

Fill-factor is always a quantity less than or equal to one (n < 1.0) For a coil
inside a tube the impedance change due to decreasing n is thesame as an increase in
Di (with constant wall thickness). For a coil around a tube or cylinder, decreasing n
is the same as decreasing Do.

ii) Probe diameter in plate testing is replaced by ~ube or cylinder diameter in ET of

cylindrical components. They have a similar effect on the operating point on
the impedance diagram.

Figure 7.5 summarizes the effect of test and material variables on a simple
semicircular impedance diagram. Note the similarity of changes in resistivity, test
frequency, diameter and fill-factor with the surface probe results of Figures 5.9 to
5.13.
 - 107 -

O[CRfASlljG
FILL·FACTOR (T)

O.B c;;;;;w;=r;;"l

....'-' --_.
z
~
.......
u

'"
....
>-
0.6 \
IHCREASI~G
RESISTIVI Tr <p)
7T::
.~HI~
.• All lUBE

::;
=>
0
~

...
0 O.c

/ INCREASING
~

...
-'
FREQUENCY (') and
'"
tr
0
z
DIAI/ETER (0)

0.2
:rJ' 1.0

0.2 o.c 0.6

NORI/ALIZEO RESISTANCE

FIGURE 7.5
Simplified Impedance Diagram of a Long Coil Around a Non-magnetic Thin-wall
Tube Showing Effect of Test and Material Variables

Impedance diagrams presented in the literature are often only strictly valid for long
coils (much longer than material thickness), coil lengths for inspection are normally
only a fraction of material diameter. Decreasing coil length has an effect similar to
decreasing fill-factor, it causes the impedance diagram to be smaller than expected
(but similar in shape) from coil and test material geometry. Following sections will
present impedance diagrams for tubes and solid cylinders. For simplicity a fill-factor
of lJllity will be used.

7.3.1 Solid Cylinders

The impedance diagram for a solid cylinder (diameter, Do) inside a long coil is
shown in Figure 7.6. As in Figure 7.5 an increase in test frequency or diameter moves
the operating point (the point on the impedance diagram that specifies the
normalized inductive reactance and resistance of the test coil) down the curve while
an increase in resistivity moves it up the curve. This diagram applies to both wires
and round bars.
 - 108 -

o;-Q
L . , COil

~ IHCREA SING RESISTIVITY

DECREASING fILL-FICTOR
-
u
:::>
o
z

...
,.
c
/IHCREISING FREQUENCY
ac
'"z

NORWILIIEO RESISTANCE

FIGURE 7.6
Impedance Diagram for a Solid Cylinder

The shape of impedance diagrams for cylinders differ markedly from a semicircle,
particularly at higher test frequencies. The shape difference is due to skin effect and
phase lag, factors which were not included in arriving at the semicircular shape in
Chapter 3. At high test frequencies the curve approaches the X and Y axes at 45°.

In testing cylinders with an encircling coil it should be recognized that sensitivity to

defects at the centre of bar or wire is zero, regardless of test frequency. The reason
for this is illustrated schematically in Figure 7.7 which shows plots of eddy current
densi ty across a cylinder. Defects have to disrupt eddy current flow in order to affect
probe impedance. It is apparent from Figure 7.7 that eddy current density is always
zero at the centre of a cylinder resulting in no sensitivity to defects.
 IDa',
LOW FREQUEHCY 8 > ~
4

o
IHTERWEOIArE FREOUENCY S , ~
4

HI GH FREOUEHCY
-'
c
:II
a::
o
z
o
RAOIUS

FIGURE 7.7
Schematic of Eddy Current Distribution in a Cylinder
Surrounded by an Encircling Coil

, 7.3.1.1 Sensitivity in Centre of a Cylinder

It was stated in the previous section that eddy current density in the centre of a
cylinder is zero and hence there is no sensitivity to defects. The relationship of
current flow with depth into a cylinder is derived (very approximately) below, for the
case of no skin depth attenuation and long coils. From Faraday'S Law,
V
5
- -N ~
dt

The magnetic flux densi ty, B, is approximately constant inside a long coil, hence

¢ - BA
_ (B)(lTr 2 )
where r is radial distance from centre of cylinder;
 I;
=4
- 110 -

therefore 7
V
s - -Nnr 2 dt
dB

-,
or 2
V a r
6

Resistance to flow of current is proportional to path length and resistivity and

inversely proportional to cross-sectional area, Ac'

2nrP 2nrp
-= -A
- • unit length x unit depth
c

or R a: r
s

Since I
6 - V
2"
6
by Ohm's Law

and Z

therefore,
-h 2
6
+ (WL)2 - R
s at low test frequency and no skin depth effect,

V
s
I -= a:
6 R
6

or a: r
16

Therefore, eddy current flow is proportional to radial distance from centre of a

cylinder. Hence no current flows at the centre (at r=O) and there is no sensitivity to
defects.

A
 - III -

7.3.2 Tubes

The impedance diagram for an extrernely thin-wall tube with either an internal or
external circumferential coil is a semicircle. This shape is only obtained when wall
thickness, t, is much less than skin depth (t < < <5 ), i.e. skin effect and phase lag are
negligible. This situation will rarely be encountered in practice, especially at
intermediate and high test frequencies, but the concept is useful since it defines one
of the coil impedance limits.

With an external coil the other limit is defined by the impedance curve for a solid
cylinder (maximum possible wall thickness). The impedance diagram for any tube
tested with an external coil, hence, has to lie between the two broken curves in
Figure 7.8, for example the solid line applies to' a tube with internal diameter 80% of

-<:::

" '\ ."

.... \\
-
c
u
u
z

....
c
DC
o
\ \\
\
I
.. ,
\\
....>
-
' -'
:::>
o
z
~t--7r--- CYLINOEA (OJ = 0)

TUBE (0, IDa = 0 .)

-....
o
THIN tALL (OJ::::Oa)

./
/;;r-- DECAEASING WALL THICKNESS

- /'
./

NORilALIZEO RESISTANCE

FIGURE 7.8
Impedance Diagram for a Tube with Encircling Coil Showing Effect
of Decreasing Wall Thickness

the outside diameter i.e., D i / Do· 0.8. Tubes with D i / Do greater than 0.8
would lie to the right of the solid line. The dotted lines in Figure 7.8 trace the shift in
operating point as wall thickness decreases (D 0 constant, D i increasing). Note the
spiral shape of the wall thickness locus. The thick wall end of the curve deviates fro~n
a semicircle locus.

This is attributed to phase lag across the tube wall and forms the basis for eddy
current signal analysis which will be treated in detail in Chapter 8.
 - 112-

Figure 7.8 also illustrates the dependence of the terms "thick-wall" and "thin-wall" on
test frequency. Near the top of the diagram (low frequency) a tube with D i / Do - 0.8
qualifies as thin wall, there is no phase lag across the tube wall, t < < 6. Near the
bottom (high frequency) the same tube becomes thick- wall because thickness
becomes much greater than skin depth, for eddy current purposes the tube now
appears as a solid cylinder.

When a tube is tested with an internal circumferent ial coil the impedance diagram
for a thin-wall tube remains semicircular but that for a thick-wall tube differs
markedly from a solid cylinder; compare Figures 7.8 and 7.9. The impedance locus for

"-
"- ,,
\
\
... \

-.........
u
z \
THIC~ IALL TU8E ' (Oj«Oo)
u

'"'
-
u
=>
o
z
TUBE (O,ltl o : 0 .9)

;...L~__ TlIIH IALl (OJ::::O.)

/J •
.:)
#. /
....; - DECREASING fALL TlfIC~ESS

...... /
/
/
./
.,/

NORWALIZED RESISTANCE

FIGURE 7.9
Impedance Diagram for a Tube With Internal Coil Showing Effect
of Decreasing Wall Thickness

any given tube will again fall between the dashed curves at intermediate frequencies
and approach the thin-wall curve at low frequency and the thick-wall curve at high
frequency as shown for tubes with D i / Do - 0.8 and 0.9. As in the previous
figure, a change in wall thickness produces a coil impedance change along the dotted
lines tracing a spiral shaped curve. Again, this departure from a semicircle is
attributed to phase lag across the tube wall.

. . 3.3 Characteristic Frequency for Tubes

Section 5.6 described how the Characteristic Parameter P c· r2w~a, introduced

by Deeds and Dodd, enables presentation of the effects of changes in 1", w , j.l and a on
a single impedance diagram. This alJowed test coil impedance to be specified in terms
of a single quantity rather than four independent variables. One could use this
 - 113 -

parameter in testing cylinders and tubes. However, most eddy current Ii terature
refers to a similar variable, the characteristic or limit frequency, f g usually
attributed to Forster. It differs from p because probe radius, . r. is replaced
with tube or cylinder dimensions. C

By definition, f g is the frequency for which the Bessel function solution, to

·V1axwell's magnetic field equations for a finite test object, equals one. (Bessel
functions are similar to, but more complex than trigonometric sine and cosine
functions). For a solid cylinder or thick-wall tube tested with an encircling coil,
fg .. 5.070
02 • kHz
llr 0 (7.2a)

wi th P in microhm-centimetres and Do in millimetres.

For a thick-wall tube with an internal coil,

5.07p (7.2b)
f g - , kHz
2
llrOi

For a thin-wall tube with internal or external circumferential coils,

f
lC S.07p , kHz
g lJrOit (7.2e)

The ratio f / f g defines the operating point on impedance diagrams. For non-
magnetic materials (lJ r "1), frequency ratio for cylinders and thick-wall tubes
tested with external coils is given by
2
flf • fO o 15.07p (7.3a)
g

where f is test frequency in kilohertz.

For a thick-wall tube tested with an internal coil,

(7.3b)

For thin-wall tubes tested with internal or external coils,

(7.3c)
 - lIt.: -

I a

t; os
~ lHIC~ -Ull TUBE ( INTERNAL COIl)
u
c fl l a ' fOjl / 507p
~

'"
~

->
~
0 .6
u SOLID CYLINOER cEITE~AL COIL)
=>
III a ' 10 0 1 : 5 DIp
0
~

--
0
~

THIN -'All TUBE

---' o•
i
0
(INT£RHAl , £ITERNAl COilS)
fila' ID , t1501p
z

o2

0 .2 o• 0 .6

NORMALIZED RESISTANCE

FIGURE 7.10
Impedance Diagrams for Tubes and Rods with Long Coils and Unity
Fill-factor Showing Variations of flf AJong Impedance Loci
g

Figure 7.10 shows impedance diagrams for thin-wall tubes, solid cylinders and thick-
walJ tubes with values of f / f g (from 0 to infinity) on the curves. The impedance
plots are both different in shape and have drastically different f / f g ratios. For
example, at the "kneel! in the curves a thin-wall tube has f / f g -1, for a cylinder
f / f g- 6 and a thick-wall tube has f / t g- 4 • These differences originate in the
defining equations which contain Do 2 , nl and nit. To find the operating point
on an impedance diagram using frequency ratio one has to know the geometry (tube
or cylinder). For tubes which do not satisfy the conditions for either thin or thick
wall, calculations of f / f g is not possible except near the top and bottom of
impedance diagrams where curves for intermediate wall tubes converge with the
thin-and thick-wall curves, respectively.

In addition to defining operating point, frequency ratio can also be used for
extrapolation or scale modelling using the similarity condition. This condition stat.es
if two objects. have the same f / f r. then eddy curre~t distribution is identical in
each. Hence If test frequency f 1 meets test requirements for article 1\10. 1, one
can calculate f 2 for article No.2 from the following:

For cy linders,
 - 115 -

for thin-'oVall tubes,

and for thick-wall tubes (internal inspection),

7.3.4 Computer Generated Impedance Diagrams

As indicated in the previous section, exact analytical solutions (Bessel function

solutions) for impedance loci of test coils around or inside tubes are only possible for
limiting cases. These solutions have the additional drawback that they are only
strictly true for long coils. An alternative was made available by C.V. Dodd and his
co-workers (~) at Oak Ridge National Laboratories. They developed computer
programs to calculate coil impedance. These are valid for all coil lengths, internal
and external coils and all tube wall thicknesses. Such computer programs permit
paper experiments to define operating point as well at the effect of variations in coil
size and shape, resistivity, wall thickness and test frequency .

Figure 7.11 is an example of computer generated impedance display for a short

internal coil in an Inconel 600 tube at various test frequencies. Fill-factor and the
effects of small changes in resistivity (60) t wall thickness (6 t) and magnetic
permeability (6 ~) were examined at each frequency. Note the similarity with the
impedance plots of Figure 6.5 obtained for a surface probe. The angular (phase)
separation between fill-factor, 60 t 6 t and lllJ provides the basis for eddy
current signal analysis which will be treated in Chapter 8•
• . 00
..........
-'-" lI' 25 "HI

1I'~6~
0 . 11
1I~ \
\
\
,,
,,\

,,
~ 0 .'2 ,
,
=
.
lUI[ ' D." 2.I _
~
~
,• 0, .l: 10 '1 ...
I
I 1 s I l' Nt
I
I
, I' 100 J#- n· C'"
& 0 II
• COIL 0. , . , _

-..
I
co I D. : 5_' _
~
,
I
L ' , 1_
f
~ 0 . ... 1 A. ' · 21<
~p : ·1~ ·..
0,. ': ·25~

6~ ' ·1'
250 '"'

0 . '0

0 . 001 o 01 o '2
MOlllfl ll /[O RES' S' h CE

FIGURE 7.11
Computer Simulation of Probe Response to Various Test Parameters
 - 116-

'.4 CHOICE OF TEST FREQUENCY

Test frequency is often the only variable over which the inspector has appreciable
control. Material properties and geometry are normally fixed and probe choice is
often dictated by test material geometry and probe availabiiity. Choice of a suitable
test frequency depends on the type of inspection. Testing for diameter variations
normally requires maximum response to fill-factor which occurs at high frequencies.
Testing for defects requires penetration to possible defect locations; surface defects
can be detected at higher frequencies than subsurface defects. Maximum penetration
requires a low frequency which still permits clear discrimination between signals
from harmless variations in material properties and serious defects. The above
factors show choice of test frequency is usually a compromise.

'.4.1 Test frequency for Solid Cylinders

As discussed in Section 7.3.1, the sensitivity at the centre of a cylinder, with an

encircling coil, is zero at all test frequencies. Therefore, there is no advantage in
using a very low test frequency to increase penetration.

'v1aximum test sensitivity is obtained when the impedance diagram operating point is
near the knee of the curve. This condition occurs when f If g - 6. At this point
balanced sensitivity to defects, resistivity and dimensions is obtained. At this test
frequency, DolO ~ 3.5. Increasing the frequency ratio f Ifg to 15 or 20
improves discrimination between surface defects and fill-factor variations (probe
wobble), at the expense of reduced sensitivity to subsurface defects. Maximum
sensitivity to diameter variations is obtained at higher test frequencies, f/ f g - 100
or more.

A frequency ratio lower than 6 will result in a decrease in phase lag and therefore
less phase discrimination between defects and f111 factor. To distinguish between
ferromagnetic variations (or inclusions) and defects, the operating point should be on
top quadrant of the impedance diagram. A frequency ratio of approximately two
( f I f g = 2) would achieve this.
I
:.4.2 Test Frequency for Tubes

When inspecting tubes for defects, the criterion to satisfy is (a) phase discrimination
between defect signals and other indications and (b) good phase separation between
internal and external defect signals. A test frequency, proven in practice on many
types and sizes of tubes, is the frequency f 90 which yields 90° phase separation
between fill-factor variations (and internal defect signals) and external defect
signals. The frequertcy f 90 is empirically derived from the ratio between thickness
,:lnd skin depth, slightly larger than one,

t/o = 1.1

and converts to
2
30/t kilohertz (7.4)

where 0 is resistivity in microhm-centimetres and t is tube wall thickness in milli-

metres. This equation is valid for both internal and external coil inspection and is
roughly independent of tube diameter. At f 90 ,there is good sensi ti vi ty to internal
 - 117 -

and exterlial defects and little sensitivity to magnetite deposit and ferromagnetic
support pIa tes.

The characterisitc frequency ratio f If g cannot be used to satisfy the criterion of

phase discrimination, because the f g equation is not a function of phase lag. It
would also be wrong to use it for defect detection because it is a function of tube
diameter. The latter would require different test frequencies for different diameter
tubes to keep f I f g constant.

If one desires to distinguish ferromagnetic signals from other indications, the

operating point should be on the top quadrant of the impedance diagram for thin-wall
tubing, Figure 7.W. This point is located by calculating the test frequency to make
the characteristic frequency ratio equal to or less than 0.5 (f / f g 2. o. 5) •

Inspection Standards and Specifications

A number of industrial codes cover eddy current tube inspection. The various AST,'.A
specifications are E-215 (aluminum alloys), E-243 (copper and copper alloys), E-426 .
(stainless steels) and E-571 (nickel alloys). None of the ASTM standards specify test
frequencies, they sometimes present normal ranges such as 1 to 125 kHz for
aluminum alloys. Such numbers are of little use in deciding on a suitable test
frequency for a particular test. The ASME ooilerand Pressure Vessel Code, Section
V, Article 8 (1980) specifies test frequency in terms of the angle between through-
wall and external defect indications from a calibration tube. The procedure specified
will normally yield a frequency higher than f90 ,perhaps as high as 2 f 90

;\ 1ost calibration tubes consist of drilled holes of various diameters and/or various
depths from the external surface. Some calibration tubes have EDM (electric
discharge machining) notches in the circumferential and axial directions and on both
internal and external surfaces.

7.5 PROBES FOR DETECTING CIRCUMFERENTIAL CRACKS

A conventional internal circumferential (bobbin) probe induces a flow of eddy

currents parallel to the coil windings and therefore circumferential in direction
(Figure 7.4). To sense a defect, coil impedance must change; this will occur only if
the eddy current flow path is dis turbed. Circumferential defects parallel to this
current, which present no area perpendicular to this path, will therefore not be
sensed.

(a ~ (b)

FIGURE 7.12
(a) Probe No.1 - MUlti-pancake Coil Probe
(b) Probe No.2 - Zig-zag Coil Probe
 - 118 -

To detect circumferential defects the coil must induce currents at an angle to the
cracks. Two possible types of probes are (a) surface probes and (b) zig-zag probes.
j-igure 7.12 shows examples of such probes. The surface probe induces currents in a
circular pattern whereas the zig-zag probe induces currents to follow the 30 0 coil
angle. The probes shown in Figure 7.12 are differential. [n the surface probe
configuration a multi-coil array is used; the four surface coils in edch row are
connected in series and the two rows are connected differentially. A single absolute
surface coil can also be used, provided the probe maintains contact with the tube
surface by spring force or other means (otherwise lift-off noise would be intolerable).
See Figure 7.13 for the cross section of a typical spring-loaded internal probe for
tube testing.

C ABL E TUBE TE S T
corL PLASTIC
CONN£( TOR ULL

HFEHNCE SPRING
COIL

FIGURE 7.13
Spring Loaded Internal Surface Probe for Tube Inspections

A single surface probe is unquestionably the easiest to use; signal analysis is discussed
in Chapter 6. The main disadvantage is the partial circumferential coverage;
multiple passes or helical scanning are necessary for 100% coverage. Another
disadvantage of the surface probe configuration (single or multiple) is the loss of
sensitivity with distance from the coil. If surface coils are small, as will be the case
for most tube inspections, the reduction in sensitivity with distance from the surface
will be greater than with circumferential coils, see Section 5.3.1. The sensitivity to
s:Tlall localized defects originating from the outside surface could be as much as to
times lower than the sensitivity to internal defects. A zig-zag coil has less
attenuation to outside defects, it falls into the circumferential class in this respect.
Neither zig-zag nor surface coil probes will give uniform sensitivity around their
circumference. There will be peaks of maximum and minimum sensitivity depending
on the angle between eddy current path and defect orientation. This can best be
visualized by considering a short circumferential crack passing over the coils: there
will be areas, such as at the peaks of the zig-zag, where eddy current flow is almost
parallel to the crack, resulting in poor sensi tivi ty.

Figure 7.14 shows examples of signal response to real circumferential fatigue cracks
wi th the probes discussed above.
 - 119 -

(a) YUl T I -PANCAKE

CO I L PROBE
~~r-----""Y
1/
CR~C K S

l\ •
x

(b) ZIG -ZAG

CO Il PROBE •
~~-------------

(c) BOBBIN
CO I L PROBE
.. - y

FIGURE 7.14
Eddy Current Scans of Circumferential Cracks in Inconel Tubing
(Signal Amplitude Normalized to a 1.6 mm Diameter
Through Hole). f = 400 kHz.

7.6 SUMMARY

Test coils induce eddy currents and also sense the distortion of their flow caused by
defects. Encircling or bobbin probes have test coi1(s) mounted with their axes parallel
to the tube or rod axis. Since the coils are wound circumferentially the induced eddy
currents 3.Iso flow circumferentially. They cannot be used to detect circumferential
cracks, laminar defects, nor defects in the center of a rod.

As a general guideline for tube inspection, probe coil length, depth, and spacing (i f
differential) should approximately equal wall thickness.
 - - - 0 - " - ' _ .......... _1 ... '-_ II'-UI.

exchanger inspection. However, for short localized defects, differential probes (two
test coils side-by-side) are normally preferred.

Analysis of eddy current signals is the most important and unfortunately the most
dif ficul t task in a successful inspection. A thorough understanding of impedance
diagrams and effect of phase lag is needed to manipulate test conditions to minimize
undesirable test variables. The Characteristic Frequency for tube inspection is used
to locate the operating point on the impedance diagram. It is given by

f :a 5.07p/Dt kHz (7.2c)

g
where p is electrical resistivity and 0 is tube internal diameter (for bobbin probe)
and external diameter (for encircling probe); t is tube wall thickness.

One needs to know the operating point on the impedance diagram to determine
effects of fill-factor, electrical resistivity, and magnetic permeability. The optimum
sensitivity to fill-factor is near the bottom of the impedance diagram, in the middle
for electrical resistivity and at the top for magnetic permeability.

When inspecting tubes for defects, criteria to satisfy are (a) phase discrimination
between defect signals and other indications and (b) good phase separation between
internal and external defect signals. For general purpose testing the frequency given
by
kHz
(7.4)
is used where t is wall thickness in mm. This frequency yields 90° phase separation
between internal and external defect signals and little sensi ti vi ty to magnetic
deposits and ferromagnetic support plates.

Special probes are needed to inspect for circumferential cracks or defects close to
tubesheets. Single, spring loaded, surface probes are effective.

? WORKED EXAMPLES

7.7.1 PROBLEM:Calculate frequency to operate at the knee location of the impedance

diagram for a cylinder 5 mm in diameter and electrical resistivity
p • 10 microhm-centimetres.

SOLUTION:
• 6
(7.3a)

therefore f E 6 x 5.07 x 10
52
• 12 kHz

----------------
 - 121 -

7.7.2 (a)
PROBLEM: Calculate the test frequency to inspect Inconel 600 tubing with
D 1 = 10. 2 mm • t :z 1. 1 mm and p E98 Jnjcrohm-
centimetres.

SOLUTION: Best test resul ts are obtained when there is sufficient phase
separation between internal and external defect signals. A phase
separation of 90 0 allows good discrimination between the two and
reasonable defect depth estimates. To achieve 90 0 phase separation,
the test frequency is determined by

(7.4)

(derivedfrom t/6 "1.1)

3 x 98
= 245 kHz
(1.1)2

Therefore 245 kHz is the required frequency.

7.7.2 (b)
PROBLEM: Determine the approximate operating point on the impedance
diagram, for problem (a).

SOLUTION: Since t /6 .. 1. 1 this tube cannot be considered thick or thin

walJed. Therefore, neither equation 7.2{b) nor 7.2{c) is strictly
valid. However, for t/6 > 1, equation 7.2(c) for thick-wall tubing
will yield an approximate solution.
2
fffg .. fD /S.07 P (7.3c)
i
2
.. 245 x (10.2) f5.07x 98
- 51. 3

This would place the operating point on the lower quadrant (much
lower than the knee location) of the thick-wall curve of Figure 7.10.

7.7.2 (c)
PROBLEM: Calculate a test frequency for the above tube suitable for
discriminating between ferromagnetic inclusions and defects, when
testing with an internal probe.

SOLUTION: The operating point should be on the top quadrant of the impedance
diagram for thin-wall tubing, Figure 7.10. This point is located by
calculating the test frequency to make the ratio of Forster's
characteristic frequency equal to or less than 0.5.
 - 0.5
therefore
f - (0.S)(S.07P)/D 1 t

• 0.5 x 5.07 x 98/10.2 x 1.1 - 22 kHz

Therefore, at 22 kHz (9% of f 90 ). it should be possible to
discriminate between defects and ferromagnetic indications.

- .--- -~
 - [23-

CHAPTER 8 - TUBE TESTING - SIGNAL ANALYSIS

8 .1 INTRODUC TION

a nd :n- se:- vi c e inspe ctio n of t ubes is one of the most importan t

\\2:1'..;£ac t '.Jr~ ng
a pplic 3tions o f ed dy c urrent testi ng. For in -service inspection of small-bore tubing in
pa rt icular , eddy c urr ent is by far the most frequentl y used method. Access is usually
li: n ited to t ube en ds which makes other NDT techniques diff icult or impossible to
apply.

This chapter emphasi zes in-service testing of tubes using internal probes. This
approach is taken because testing of solid cylinders and tubes with external coils
(manufacturing inspection) is generally less complicated. If the reader understands in-
service inspection he should encounter no problem applying similar principles to other
test si tuations·.

Reasons for the appearance of impedance plane eddy current signals are presented
first. Repetition from previous chapters is intentional, it was desired to keep this
chapter as independent as possible without excessive cross-referencing. Discussion of
s imple defect indications is followed by superimposed signals which are frequently
encountered during in-service inspection such as defects at baffle plates and
tubeshee t s . .A. section dealing with surface probe internal tube inspection is included,
difficult test situations have been resolved with this technique. Signals which could
be mistaken for real defects (anomalous indications) are the subject of another
section. The chapter concludes with a discussion of multi frequency testing, including
its advantages and limitations.

An 3.ttempt is made throughout this chapter to illustrate discussion with real or

simulated eddy current defect signals.

8.2 EDDY CURRENT SIGNALS

8.2.1 Defect Signal Characteristics

. 'V
,
ei.o ~, 6
A defect, which ~is.rupt·s eddy current flow, changes test coil impedance as the coil is
scanned past the detect. A non-rigorous derivation of this effect can be obtained
using Figure 8.1 which portrays eddy currents induced in a tube with either an
internal or external coil. Consider a unit length of tube as being the secondary
winding of a transformer (similar to treatment in Chapter 3). The resistance of a
conductor of length i ,cross-sectional area A and resisti vi ty P is

R = ~"I r· / A , 0 ;! :~ ~;

Wit hou t a defect, resistance around thi s t ub e is

\
\• R0 - 2rrrp/t (8.1a)

In troduct ion of a long defect, of dept h h, which constri c ts eddy current flow over the
d ist an ce 60 (in radians), increases total resistance to

R = 2lirO/t + 6eh~p/t(t-h) (8.1b)

 - 124-

or R - Ro (defect free resistance) + t,R (duetoderect).

FIGURE 8.1
Schematic Illustration of Eddy Current Distribution
Around a Defect in a Tube

A short defect will also increase resistance but by a smaller llR since current can
flow both under and around it. Note that it is width of affected zone, 118 , rather
than actual defect width which determines effect of the defect on resistance. In
summary, the above argument illustrates that defect length, depth and width (to
some extent) all increase resistance to current flow and hence defect signal
ampli tude.

In terms of the equivalent coil circuit of a resistor in parallel with an inductor and its
associated semicircular impedance diagram (Chapter 3), a defect moves the operating
point up the impedance diagram. Increasing resistance in a specimen changes both
probe inductance and resistance.

The above discussion does not predict a defect signal in detail, only its approximate
amplitude and direction on the impedance diagram. A more complete explanation
requires inclusion of phase lag. Consider an absolute coil around a cylindrical sample
as in Figure 8.2(a). (The treatment for a differential coil would be similar but more
complicated because the twin coil configuration generates two mirror image signals
and cross-coupling between the two coils causes further complications). Figure 8.2(a)
shows the distribution of magnetic field amplitude and phase around a coil as derived
by Dodd (~). The solid lines are contours of constant magnetic field strength; the
dashed lines are constant phase. Since magnetic field and induced eddy currents have
about the same phase, the dashed lines also represent the phase of the eddy
 - 126-
o 0)

,OOS''''' "lI! S
\

0·----- __\..1... __ _ -------

(0 )
tl£F(CT PIlSIfIO •

. ~

DEEP DEFECT

SUBSURFACE

(b)

FIGURE 8.2
'Derivation of Eddy Current Signal Appearance for Three Types of Oefects

With this phase setting and at f 90 an 00 defect shows wall loss (+- Y) in a tube
without a change in fill-factor as in Figure 8.3(b). An ID defect consists of wall loss
(+ Y component) as well as a large fill-factor (-X component) because of decreased
coil/tube coupling. The through-wall defect (hole) signal contains elements of both IO
: and 00 defects and hence yields a signal which falls between the two. Note that all
defect signals must fall between decreasing fill-factor and OD defect signals. -
 - 125-

cJrrents. Similar diagrams could be derived for coils inside or around tubes.
,- \rnplitude drops off exponentially with distance and eddy current flowincreasingly
lags in phase (relative to eddy currents adjacent to the coil) both with depth and with
axial distance from the coil. Skin depth effect occurs in both radial and axial
di ,ections~

Figure 8.2(3.) permits derivation of eddy current signals for the surface, subsurface
and deep defects illustrated. One needs to establish a reference phase direction as
starting point, the fill-factor direction is convenient and can be defined as the signal
resulting from a very shallow surface defect which only decreases coupling without
changing phase lag distribution. Hence choosing the phase contour which just touches
the surface under the coil as the 0 0 contour fixes fill-factor direction as in Figure
8.2(b). The signal or effect of defects can be imagined as the absence .of eddy
currents which were flowing in the area before the defect existed at this location. On
moving the coil (or defects past the coil) from positions 0 to 5 in Figure 8.2(a), one
observes the change in amplitude and phase sketched in Figure 8.2(b). This procedure
is reasonably straight forward for the surface and subsurface defects since they are
localized and only intersect one phase and amplitude contour at any given position.
For the deep defect, one has to divide the defect into sections and determine
weighted average values for amplitude and phase at each position.

The surface defect in Figure 8.2(b) has a large fill-factor component, primarily its
approach signal makes it distinguishable from fill-factor. As defect depth increases,
signals rotate clockwise due to increasing phase angle.

The angle between fill-factor and defect signals in Figure 8.2(b) is about 2 6 ,where
B .. x/5. Although probably not strictly true, one can imagine defect signal
phase angle as the sum of a lag of Bfrom the coil to the defect and the same lag back
to the coil.

8.2.2 Effect of Test Frequency

We can now combine Figure 8.2 results with impedance diagrams from Chapter 7 to
illustrate the effect of test frequency on defect signal appearance. Figure 8.3(a)
shows part of Figure 7.9, the impedance diagram for a tube with DilDo - 0.8
tested with a short internal coil. The dotted lines trace the impedance change with
decreasing Do, An external defect (00 defect) in a tube is essentially a decrease in
Do with D i held constant, therefore the dotted lines trace the change In impedance
as a coil is scanned past an 00 defect. Note the similarity between the subsurface
defect in Figure 8.2(b) and the OD defect at 2 f 90 in Figure 8.3(a). The display is
normally rotated counter-clockwise to make a signal from fill-factor approximately
horizontal. This is achieved by rotating the phase control knob on the eddy current
instrument.
 - 127-

, "- OEC~E lS I",

I ~\ "- fiL i. - f ACi e R
; 1\ ,
I \ \ "-
I \\ .... , ------o........r--
J! \\ \\
~ ~. f ' A
----',.-.~--

: I \ \
~ ! ~ \
:
>
\ \
\ \
OfU£CSI "G
\ \ f il L -fAC TOR
\ \ 10
\ \
\
\
00 OEfECT

NORIO\Ql\ Z(O RES I STANt (

FIGURE 8.3(a)
Relation Between Impedance Diagram and Defect Signal Appearance
+Y

.- ,
00 DEFECT
THROUGH-WALL -x +x
DEFECT
I0 DEFECT - - -__
-y
DECREASING FILL FACTOR
FIGURE 8.3 (b)
Defect Signal Appearance at f90

Figures 8. 3(a) and 8.4 show what happens to defect signals with changing test
--
frequency. Reduced frequency results in rotation of defect signals towards the fill-
factor direction. At very low frequencies (Jess than f 90 /4 ) signals from
different t ypes of defects become difficul t to distinguish due to small phase angle
separation.

Increasing test freque nc y increases phase separation between ID and 00 defect

signals as predicted by phase lag. At f 90 the 10 and 00 defect signals are separated
by about 90° with low sensitivity to tube supports and external deposits. At higher
test freque\lcies, 2 f90 and above, higher sensitivity to probe wobble and dents is
obtained and the increased angular separation of defect signals makes it difficult to
discriminate between 00 defects and probe wobble or fill-factor variations, see
Figure 8.4(c).
 - 123-

C. D. OF c·: T

H.) L E

10

PROBE
PROBE WOBBLE
wGBBLE

HOlf

10

PROBE
wOBBLE

(e) 2 t 90

FIGURE 8.4
Appearance of Calibration Defect Signals at Different Test Frequencies

,. 2.3 Calibration Tubes and Simple Defects

Beth manufacturing and in-service inspection require calibration tubes with artificial
defects for initial instrument set-up and subsequent signal analysis and
interpretation. These tubes should be identical in material and size to tubes to be
tested. Minimum calibration requirements include ID, 00 and through-wall defects
(see also the ASTM and ASME codes cited in Section 7.4.2). For in-service inspection,
expected signal sources such as baffle plates, magnetite deposits and dents are useful
and often essential for reliable signal analysis. Figure 8.5 shows typical signals, at
f 90 t from a calibration tube suitable for in-service heat exchanger inspection. Both
absolute and differential probe signals are shown. The 90° phase separation between
ID and OD defects also exists for differential probes. Note the similarity with the
signals derived in the previous section.
 - 129-

P' OBE

OUTSIDE

"OLE

INSI DE
~ E C' E 'S I N G
<ILL " C IO~

SUPP OR T
PLATE

FIGURE 8.5
Eddy Current Signals from a Typical Calibration Tube.
Test Frequency f90 = 250 kHz.

Qualitative reasons for the appearance of 10, 00 and through-wall defects were
presented in Section 8.2.2. The other signals in Figure 8.5 can be explained in a
similar fashion. Magnetite is ferromagnetic non-conductor, its signal is due to its high
permeability. As indicated in Figure 7.11 increasing permeability of tube material
yields a signal which falls between 00 and through-wall defects. The magnetite
signal in Figure 8.5(b) is essentially such a signal rotated about 90° clockwise because
of phase lag across the tube wall. A dent places tube material in close proximity to
the coil resulting in improved coupling (increased fill-factor) and hence yields a signal
opposi te to decreasing fill-factor. Probe wobble yields a signal very close to the f iU-
factor direction because radial displacement of the coil reduces the coupling to the
tube. The reason for baffle plate signal appearance is due to a combination of
factors. For carbon steel baffles, the effects of high magnetic permeability and
intermediate resistivit y partiall y can c el res ulting in small signal amplitude. Phase lag
across the tube wall rotates this signal clockw ise.
 - 130-

O.2S . ..
DENT
CARBON STEEL
SUPPORT

--
'-------+

FIGURE 8.6
Appearance of Quadrature Components on a Chart
Recording for a Calibration Tube

l eddy current tube testing one normally records the quadrature components
\Vertical, Y; horizontal, X) of coil impedance on a two-channel strip chart recorder as
shown in Figure 8.6. With phase adjusted as shown, any real defect will exhibit a Y
component. The X-channel information is required for detailed signal analysis to
decide type and depth of defects which can only be performed reliably through phase
analysis. Accurate phase analysis can be done on-line by monitoring the signals on an
eddy current instrument storage monitor. Alternatively an X-Y recorder or similar
device permits hard-copy storage of quadrature signalS.

A flaw indication on an X- Y monitor is normally a curved locus; it does not have a

simple and unique phase angle. If an absolute probe is used the significant angle to
measure is the tangent angle at the defect signal tip, see Figure 8.7(b). If a
differential probe is used, the phase ang~e is the slope of the straight line joining the
end points of the "figure-8" signal, see Figure 8.7 (c). Figure 8.7(a) illustrates the
change in phase angle wi th defect depth. This curve should be used only as a guide
since defect signal phase angle can change with defect and probe geometry.
 - 131 -

0.0. DEFECT
I .0. 0.0.
DEr-ECT S DEFECTS THROUGH !
100 ~OLE I
8~~ , t
I

~
90

80
/
--l
/ 1.0. DEfECT
----'
c:::::::
3
70 /
L.!-
a
bO / (b) ABSOl UTE
~
/ THROU GH
z 50
UJ
LJ
/ 8~lE 0.0. DEfECT
='.w 40 /
.:L
30
/
::r::
~
0..
LU 20
c:::l
1.0. DEFECT
10
0
0 10 40 50 60 70 80 90 (e) DIFFERENTIAL
\ 20 30

SIGNAL PATTERN PHASE ANGLE (8), DEGREES

( a)

FIGURE 8.7
Eddy Current Phase Angle/Defect Depth Calibration Curve at f90

When an eddy current signal source is located it is often useful to retest at other
frequencies to confirm a defect exists and/or to improve depth estimate. Defect
depth is estimated from signal pattern recognition and verified by comparison with
calibration defect signals at various test frequencies. Normally, frequencies of one-
half and twice f 90 are sufficient. However, to check for magnetic deposits or
indusions a frequency of one-tenth f 90 or less may be required (see Sections 7.4.2
and 8.3.1). Figure 8.4 shows effect of changes in frequency on calibration signals.
Increasing test frequency increases phase separation between 1D and OD defects as
predicted by phase lag. It also increases sensitivity to probe wobble and dents but
lowers sensitivity to tube supports and external deposits. One might question the
validity of comparing machined holes and grooves in calibration tubes with real
defects to estimate type and depth. The following examples justify this approach.
uniform with localized severe pitting. An absolute lnternal prODe Wd~ u~eu LV VULCJil
signals from artificial defects and three of the rocali zed pits. The phase angle of the
first two corrosion indications shows they are 00 defects, comparison with the
calibration defect led to a depth estimate of 25 to 50%. Independent mechanical
measurement found deepest penetration to be 50% for both defects. The third defect
has a noticeably different phase angle from the first two. It approaches the angle for
a through-wall hole, hence its depth was estimated to be 50 to 75% (actual
measurement yielded 75%).

1.0' C1m
25% 00
ECCENTRIC
CALfB R.'\ nON HOLE GROOVE
DEFECTS
10% ID '-......1
ECCE;'1T_RIC~
GRCOY::

CORROSION ~
D"'US )

FIGURE 8.8
External Corrosion in a Copper Tube

(Do = 15.9 mm, t = 1.0 mm, f90 = 5.3 kHz)

An example of stress corrosion cracking (SeC) in Type 316 stainless steel, from a
heavy water plant heat exchanger, is shown in Figure 8.9. The crack extends nearly
half way around the tube. Phase angle of the crack signal shows it extends through
the tube wall. Since the eddy currents flow parallel to coil windings, circum-
ferentially, the large crack signal is due entirely to the component of the crack along
the tube axis. The intergranular, branching nature of sec generally permits their
detection. Since a defect must disrupt eddy current flow to be detectable, if
circumferential cracks are suspected, fatigue cracks for example, special probes are
required, see Section 7.5 and 8.2.5.
 - 133-

30 40 60 80 90
~Jd,I~"~~~hlhl~~~~illiliWil~~~~~

3.2 m"1
HOLE - - -
1.6n~
HOLE
J 50"; OD
CONCE/HR Ie
GROOVE

CAll BRA Tl ON
DEt::ECTS

FIGURE 8.9
Stress Corrosion Cracking in Type 316 Stainless Steel Tubing

(Do = 19.1 mm, t = 1.8 mm, f90 = 68 kHz)

8.2.4 Vectorial Addition and Defects at Baffle Plates

During in-service inspection of tubes in heat exchangers, tube supports (baffle plates)
are frequently defect prone regions. Inspection for defects at baffles is possible
because eddy current signals are often vectorially additive. This permits analysis of
superimposed signals; the signals can be (mentally or graphically) subtracted from the
total indication with resultant separated signals appearing similar to calibration
defects. Vectorial addition provides the basis for multifrequency eddy current testing
(Section 8.4).

Figure 8.10 illustrates how signals from a steel baffle plate and an external groove
are added to obtain a superimposed indication. The difference between the end points
of the baffle plate and baffle and groove signals equals the indication obtained from
the groove by itself.
 - 134 -

CM8')~1
5 TEEL
BAFfLE

FIGURE 8.10
Vectorial Addition of Eddy Current Signals

Figure 8.11(a) shows a section of stainless steel tube removed from a power plant
heat exchanger with part of the carbon steel support plate still in place. The support
shows considerable corrosion; originally there was about 0.25 mm clearance between
the tube and the hole in the plate. Corrosion products have completely filled the gap
leading to crevice corrosion evident in Figure 8.1 I(b) which is a similar tube with the
plate removed. Calibration signals are presented in Figure 8.1l(c). The eddy current
signal from the baffle plate region of Figure 8.Il(a) is shown in Figure 8.Il(d). This
seemingly simple signal is actually quite complex. The upward component is due to
external pitting similar to that in Figure 8.II(b). The presence of a support plate
should result in -X, -Y signal components; in fact a + X deflection is observed. This is
the result of denting of the tube. Denting is circumferential constriction of tubes
due to compressive stresses exerted by baffle plate corrosion products such as
magnetite. The presence of magnetite can also contribute to signal distortion
particularly at low test frequencies. Tube denting is of concern because, in addition
to complicating eddy current signal analysis, it can lead to further tube damage such
as stress corrosion cracking or thermal fatigue because tubes are no longer free to
expand and contract during thermal cycling.
 - 135 -

(a) ( b)

00 r.RDD'/E

I D (,'100'1£

DENT

(c ) ("'L1BRArION (d)

FIGURE 8.11
Corrosion and Denting Under a Steel Baffle Plate

(Do = 15.9 mm, t = 1.25 mm, f90 = 80 kHz)

Another example of defects near a carbon steel tube support is shown in Figure 8.12.
These were obtained from a brass, thermal power plant condenser tube which
suffered erosion/corrosion on either side of supports. This is the same tube as in
Figure 7.3. Defect signals from the baffle plate vicinity are so large the support
signal is obscured. The main point of this example is the advantage of using phase
angle, rather than amplitude, to judge defect severity. Defect B with both
differential and absolute probes has a phase angle approaching that of a through-wall
hole, i.e., it probably extends at least 75% through the wall. Defect A on the other
hand is vertical and hence is probably no deeper than 50% even though it exhibits
greater amplitude than B.
 - lJr; -

DEfE CT
SIGNALS

00
r.qcc'l (
no

)
r.iiQCV E

g
\ ~ MM
5: : :L 1 f ",m
H'):" ~
B.),': L,. E.
')
~
~

(D)

DiFfEREN t iAl
ABSOLUTE

FIGURE 8.12
Quadrature Eddy Current Signals from the Brass Tube in figure 7.3

To this point we have only considered ferromagnetic tube supports, carbon steel is
the material used in most heat exchangers. With magnetic baffle plates vectorial
addition appears to apply for all types of defects. Unfortunately deteriorating water
quality, denting problems and longer service life requirements have made it necessary
to construct some heat exchangers with non-ferromagnetic support plates. Vectorial
addition of eddy current signals involving nonmagnetic supports is generally not valid.
Several factors contribute to this situation, nonmagnetic supports yield much larger
signals than magnetic supports. The larger signal from nonmagnetic baffle plates
effectively reduces signal-to-noise making small defects more difficult to detect.

Possibly the most diffiCult defects to detect under non-magnetic supports are those
of the same width as the plate, e.g., fretting wear from tube vibration.

F igure 8.13(a) illustrates such a si tuati on, a brass baffle pla te with a copper-n ickel
tube containig simulated 50 % deep fretting wear. The same defect with a ma gneti c
baffle plate is shown in figure 8.13(b) for comparison.

Proble ms in detecting defects at non- magne t ic supports cannot be overcome by

emplo ying a multifrequenc y eddy current te chnique. The multi frequency approach
 - 137-

relies on vectorial addition being valid (Section 8.4). Sensitivity can be improved by
employir'5 special probes as will be shown in Section 8.2.6.

!I
~ ,- !";' XiMUM G,l.P
Ii
:\
l;~ . ~

O
'1\ . 'l
/. :% '
~ . ~
/j, - V/
''/////0
/,' //

:, . 'l
~

CONTACT 00 GROOVE

00 I;ROovE

BAF"LE WITH (';ROovE

MAXIMUM r;AP +
BA"C:LE
BRASSBAFFLE
IN CONTACT--..::!

BRASS
BAFFLE
( a)
(b)

FIGURE 8.13
Wear Under (a) Non-Ferromagnetic and (b) Ferromagnetic Baffle Plates

8.2.5 Tube Inspection at Tubesheets

Heat exchanger tubesheets are usually made of carbon steel, eddy current response
should therefore appear sim ilar to a baffle signal. In addition, a large fill-factor (tube
expansion) signal is also obtained as a resul t of tubes being rolled into tubesheets.
Rolling eliminates corrosion prone crevices and also helps hold tubes in the tubesheet.
With carbon steel tubesheets, expansion usually yields the largest signal component,
the tubesheet only contributes appreciably at test frequencies below f 90 •
Figure 8.llt shows tube configuration at a tubesheet and typical eddy current signals.

Occasionally one may encounter a tubesheet clad with a corrosion resistant alloy such
as stainless steel or Incone!. If the cladding is non-magnetic the same complications
arise as with non-magnetic baffle plates (Section 8.2.4). Fortunately, most tubesheets
are only dad on the primary side (near tube ends) where service related defects
rarely occur.
 - 138 -

: ;
srUL
IUBESHEE.T

H E £L
ruB( ; ... [!.T

~,
2 ' to f .0

FIGURE 8.14
Schematic of Tube Geometry at Rolled Joint in Tubesheet
and Associated Eddy Current Signals

The end of the rolled joint at the inboard edge of a tubesheet is a defect prone area
because of high residual and service stresses and also because deposits tend to
accumulate at this location which can lead to corrosion. Eddy current indications
with bobbin-type probes from defects in this region can be difficult to interpret
because of excessive signal distortion from tube expansion. Sensitivity may be
improved by employing a spring loaded surface probe as discussed in next section.

~ '2.6 Testing Tubes with Internal Surface Probes

During in-service inspection of tubes, situations arise where conventional

circumferential probes (both differential and absolute) prove inadequate. The case of
circumferential cracks was treated in Section 7.5. Surface probe designs have also
been found to yield improved test results in the case of defects at non-magnetic
baffle plates and at heat exchange r tubesheets.

Surface probes have sever al advantages over bobbin-type probes. They can be made
much smaller than tube diameter and hence sample a smaller volume of tube
periphery, this provides inherently greater sensitivity to small defects. Spring loading
of a surface probe against the tube wall eliminates much of the fill-factor (lift-off)
distortion caused by tube expansion in tubesheets. The main drawback to surface
probe tube testing is that a number of scans have to be made for complete
 - 139-

circumferential coverage. Conventional probes sample the entire tube in a single

scan.

E .~O CC INCON EL 60 0
ROL LE D JOI NT TUSE ~ALL

-==-~~- 7 0 ~ r:O
t t t DefECT
C B A

TUBE SHEET
EXPAtlSiON/TUB E~H'ET A'
$lG.'lAL ~' lTHOUT -DtFECT
B'C: •
c---------------~+ "' C'

H rrH DEFECT AT
E~D OF ~0LLED J01NT

CONVENTIONAL SURFACE
PROBE PROSE

FIGURE 8.15
Comparison of Eddy Current Test Results in Heat Exchanger Tubesheet
Region with Conventional and Surface Probes

(Do = 12.5 mm, t = 1.2 mm, f90 = 200 kHz)

Figure 8.15 illustrates surface probe testing at the tubesheet region of a power plant
steam generator. It compares signals, from what is believed to be OD corrosion
damage at the end of the rolled joint, obtained with conventional and surface probes.
The reason for the characteristic AlBIC' surface probe signal is as follows. As the
probe is withdrawn from the tube (direction of arrow) it encounters the start of the
expanded area. Failure of the probe to follow this contour exactly results in an
increasing lift-off signal, A'B I, superimposed on the impedance change, AIC', due to
the presence of the tubesheet. Both defect signals were obtained from the same
tube, note the considerable improvement in sensitivity obtained with the surface
probe. This tube was in fact leaking.
 - 140-

()

50% 0 0
ECC ENT R!C
GR OO VE

BRASS BAFFLE

50% GR OOV E

BAFFLE
(NO GAP )

FIGURE 8.16
Internal Surface Probe Testing for Fretting Wear under a
Non-Magnetic Baffle Plate. (Compare with Figure 8.13 Results)

A second example of improved sensitivity with an internal surface probe involves

fretting wear under non-magnetic baffle plates. Figure 8.16 shows results. Compare
with Figure 8.13(a) which shows test results for the same defect obtained with an
internal circumferential probe. With no gap, the 50% groove was barely detectable
with a conventional probe, while Figure 8.16 shows this defect is easily detected with
a surface probe.

8.3 ANOMALOUS EDDY CURRENT SIGNALS

Some eddy current signals can be mistaken for defe c t indic a tions; these are called
false or anomalous signals. The y a rise because of the high sensitivity of eddy currents
to many variables and demonstrate the need for tho rough analysis before concluding
that ever y eddy current signal re presents a defe c t. The followin g examples illustrate
more common ones which have been encount e r e d in practice.
 - 141 -

8.3.1 Ferromagnetic Inclu s ions and Deposits

Materials with relative magnetic permeability greater than 1.0 affect eddy current
response drastically. Skin depth and prob e inductance are both affected by
permeabili tYi perme a bili ty valu e s of 50 to several hundred are typical.

Before citing specifi c examples c onsider the general approach to identify ing si gnals
fr o m magnetic m a ter ials. Suc h signal s can be dist inguished from real defects by
reducing test frequency to move the operating point near the top of the impedance
diagram. Figure 8.17 illustrates the procedure where 1, 2 and 3 represent
ferromagnetic material on the inside, in the tube wall and on the outside
respectively. It may be difficult to achieve a sufficiently high operating point with
some instruments and probes when testing low resistivity, large diameter tubes.
However, if a low enough frequency is achieved, real defect indications will fall
nearly parallel to fill-factor whereas high permeabili ty indications are nearly
perpendicular to fill-factor. At 240 kHz (f 90) in Figure 8.17, 1 and 2 could easily
have been mistaken for ID defects. There is no confusion at 10 kHz since it is known
that all defect indications must fall between fill-factor and an 00 defect signal. The
following two examples demonstrate the procedure to discriminate false defect
(ferromagnetic) indications.
HRROI/AGNEr I C
AN Ollll1 ES

oa J a
T ~ f I . 110m

:
GROOVE GROOVE

9\ : 1~7U
(

ABSOLUTE PROBE
/
G:J = :
INCONEL 600 ruBE

"- - 20
'\.
\
~ 1"3· \
\
...
u
\
- SO
...z
-
.......
u

a:
. BS
\
I
OECR EAS ING FIL L FAC TOR

I
.......
Q

I
-
.:
. 80
'7 0 .6 ~
0 i 100
'"
co:
Q
Z
/ 1. 0. 0 .0.
. 75

0~
/

o;c~Q)
O'----'----'--_-L-_
o 0 . 05 0 . 10 O. IS

NOR~ILl l E O RES ISTINCE . ~

.., L .

FIGURE 8.17
Coil Impedance Display at Two Test Frequencies
 - 142 -

Ferromagnet ic inclusions are occasionally encountered during eddy current testing of

non-magnetic materials. These arise from chips or filings from steel tooling and
handling equipment which are embedded during manufacture. The surface of
nominally non-magnetic stainless steels and nickel-base alloys can also become
magnetic as a resul t of cold working or through alloy depletion from oxidation or
corrOSIon.

0.0. DEFECT I • D. DEFECT

250 kH z FERROMllGNET IC -
INCLUSION
\ \ FERROMAGNET I C
0.0. INCLUSION
~.o.
50 kHz

INCLUSION

--=-===I.:::D:::'~ I NCl U51 orl

_0_._D••

. '0 kHz

FIGURE 8.18
Defect and Magnetic Inclusion Signals Obtained from a New Inconel
600 Tube (Do = 13 mm, t = 1.1 mm) with an Absolute External Coil.
f90 = 250 kHz

Though one might consider a magnetic inclusion a defect, there are several reasons
why it is important to identify the origin of an indication. Even very small, perhaps
insignificant, magnetic inclusions can yield sizeable eddy current signals because of
the extreme scnsi ti vi ty to magnetic permeabili ty. A second reason to determine
defect origin is so measures can be taken to minimize further damage; magnetic
inclusions are nearly always manufacturing defects. Figure 8.18 shows the signal from
a magnetic inclusion in new Inconel 600 tubing at various test frequencies. These
results were obtained with an external encircling probe; this explains the reversal in
appearance of TO and 00 defects from previous examples. The magnetic inclusion
yields a signal whose angular separation from the fill-factor direction increases as
test frequency is reduced. The response of real defects is just opposite.
 - 143-

10.0. DE f ECT

--B~.-
I .0. GErE CT ~

I N7ERNIIL
/
250 :" Hz
M:' GN ETi TE

MAGNET I TE
7
--~
ID . .... ~

MAGNETliE

50 kHz

~
MAGNETITE
1.0. 0.0.
__
----------
10 kHz

FIGURE 8.19
Defect and Magnetite Signals from an Inconel 600 Tube
(Do = 13 mm, t = 1.1 mm) Obtained with an Absolute
Internal Probe. f90 = 250 kHz)

Figure 8.19 shows eddy current response to magnetite (Fe 3 0 4 ) deposits inside an
Inconel 600 tube at various test frequencies. As in the previous example, the
existence of ferromagnetic material is verified by lowering test frequency; magnetite
signals rotate dockwise whereas defect signals rotate counter-dockwise. One could
easily mistake the magnetite signals for real defects at 250 kHz and 50 kHz.
Reducing test frequency can also be used to verify the presence of magneti te on the
outside of a tube. This approach has been used to measure the height of sludge
deposits (containing magnetite) above tubesheets during in-service inspection of
vertical heat exchangers.

Figure 8.20 shows the eddy current signals from a Monel 400 steam generator tube
with external wall thinning near a tube support. The tube wasinspected with an
absolute saturation probe and the signals recorded with wall thinning giving a
vertically upward signal. At 50 kHz the vertical component of the complex signal is
from wall thinning and the horizontal signal is primarily from magnetic deposit. At
200 kH z (2 f 90) the vertical component is again from wall thinning but the
horizontal signal is primarily from an increase in tube magnetic permeability because
of incomplete magnetic saturation under the carbon steel tube support. At 400 kHz
eddy currents just barely penetrate through the wall. In this case the signal is
primarily from tube magnetic permeability variations.
 - 144.-

0.0. GROOVE

DENT

BAFFLE f3 = 400 kH z
BAFFLE PLATE
PLATE

'2: 200 kHz BAFFL E

PLATE
f 1'= 50 kHz
MAGNET I TE
CALI BRA TI ON TUBE
S I GHALS
-_....

, MAGNETITE
f2 = 200 kH z fJ = 400 kHz
fl =50 kHz

ACTUAL DEFECT SIGNAL

FIGURE 8.20
Eddy Current Signals from Monel 400 Tube at Baffle Plate Location.

(f90 = 100 kHz)

L3.2 Conducting Deposits

The most probable conducting deposit which may be encountered during in-service
tube testing is copper. Copper taken into solution in one part of a cooling circuit,
from brass tubes for example, can re-deposit at another location at the expense of a
less noble metal such as iron. An example is shown in Figure 8.21 which is a copper-
alloy tube from an air conditioner heat exchanger. Copper deposits occur near tube
supports, maximum thickness was 0.05 mm. Even such a thin deposit yields a large
eddy current signal since copper is a good conductor. Figure 8.21 shows response from
both absolute and differential internal probes. The absolute probe gave eddy current
 - IL!5-

signals with no -+- Y component, clearly indicating the non-defect nature of the
anomaly.

The differential probe signal is not nearly as clear and illustrates another limitation
of differential probes. Comparison of the deposit indication with calibration defects
could easily lead one to conclude the presence of an OD defect; particularly if the
eddy current resul ts were compressed on X and Y channel recordings as is often the
case during in-service inspection. With a differential probe, one has to observe defect
sense (arrows) to distinguish between deposit signals and those from real defects.

Copper Deposits

I I I

30 40 50 60 10 BO
I I I I I

l .J I1O 00

i
G=?OO"E

I 6 ,'1"\

HOI_E

s rEE~
B~CI:"L£

A%OlUTE DIFF~?ENT IAL

FIGURE 8.21
Eddy Current Indications from Copper Deposits on a Copper Alloy Tube

(Do = 19 mm, t = 1.1 mm, £90 = 57 kHz)

Figure 8.22 shows simulated copper deposit signals at different test frequencies.
There is a noticeable change in phase angle with increasing deposit thickness as well
as test frequency. At frequencies above f90 there exists a possibility that deposits
could be mistaken for 10 defects, even with an absolute probe. The procedure for in-
service inspection of nuclear power plant boilers specified by ASME (.!l) leads to test
 frequencies between f 90 and 2 f 9 0 . This appeC.rs to be a weakness in the code
'.vhich may lead to revision if copper deposits prove :nore common as boilers age.
Inspection of Figure 8.22 reveals that clearer discrimination between copper and
defects is achieved at f 90 /2 rather than at f 90 . Optimum test frequency for
copper coated tubes appears to be the frequency which just leaves signals below the
horizontal fill-factor direction.

... - 7':,<>:' 00 ECCE>ITRfC GRCOV;:

B - 10% ID CONCENTRIC GROOVE
C - 0.13 mm THICK COPPER AROU ,'I O T U'3:::
0- 0.05 r.1m THICK COPPER AROUND TUElE

FIGURE 8.22
Eddy Current Signals Obtained with an Internal Circumferential Probe
from Simulated Copper Deposits on Tubes

. )~ MUL TIFREQUENCY EDDY CURRENT TESTING

4.t Background

Successful in-service Eddy Current inspection relies on eddy current probes that can
sense defects and an analysis of eddy current signals. Both aspects are equally
important. While scanning each tube, eddy current signals are obtained from baffle
plates, magnetite deposits, dents, tubesheets, tube expansion, etc. and maybe
defects. One must, therefore, discriminate between defects and insignificant signals
and even more important, estimate defect severity when it occurs together with
other signal sources. It would be much easier if the data could be processed to
contain only defect signals; Multifreguency ET can do this.

In multifrequency testing, two or more sinusoidal signals of different frequencies are

fed simultaneously to a single eddy current probe. Gain and phase of the output signal
from each frequency can be separately controlled.
 - 14-7-

THROUGH BAFFLE
WALL HOLE PLATE
10 GROOVE 0 0 GROOVE I ~ "AGuE TI TE 1.3 mm
\ \ II DE.'d ~ PO"
r---,C==~====-~===='C'=========~~==~:~·~==~~~--r~~1
L_~=~~D~~I~=-=-=-=-=-,,=d=======::=;:;.===::;;::::;;===:16d
:::=:=:. 1\5 mm

TUBE ~
CALIBRATION

I O.

I.O.~

MAGNETITE ~
-~=-- DENT

PLATE

f 1 = 20 k.H z f 2 =IOOkHZ f J =500kHZ

(a) ( b) ( c)

FIGURE 8.23
Internal Probe Response to Various Test Parameters.
f 90 = 130 kHz.

f 1 = 20
(a)
1\
f 2 = 100 kHz
(b)
~
f J =500
( c)
kHz

FIGURE 8.24
Eddy Current Signal at Baffle Plate Position in Tube of Figure 8.11
f90 = 130 kHz.
 - 148-

These signals can then be combined to eliminate unwanted signals and leave only the
defect signa l. This m ethod is only effective if a defect signal differs
characteristic2lty from unwanted signals and if signals are vector ial ty addi ti ve. The
first con d i ~ion makes detection of internat defec ts , in the presence of internal
Y2rlatlons, impossib le. The second requirement makes the method ineffective for
deteCT lon of fretting wear under non-ferromagn e tic baffle plates (Section 8.2.4). As a
consequence of combining signats from three different frequencies, defect signal
amplitude decreases and instrument no ise increases.

Eddy current penetration and phase lag are a function of frequency; increasing test
frequency reduces penetration and increases phase lag. Since an eddy current signal is
a function of current densi ty and phase lag, it is possible to change the response to
various signal sources by changing test frequency.

If one simulates a heat exchanger tube with defects, deposits, dents and support ·
plates, one obtains the following results:
(a) at high frequencies, only internal defects and dents are detectable, Figure
8.23(c).
(b) at intermediate frequencies, all features are detectable and there is phase
discrimination between internal and external defect signals (because of phase
lag across the wall) and other signals, Figure 8.23(b).
(c) at low frequencies, baffle plates and magnetite deposits yield predominant
signals with little phase separation between internal and external defect
signals, Figure 8.23(a).

With this background in mind, one can decide which combination of frequencies
should be used to eliminate extraneous (unwanted) signals. The following two
examples illustrate these effects.

For the dented tube example described in Section 8.2.3 (Figure 8.11), the extraneous
signals making up the composite signal at f = 100 kHz can be determined by re-
inspecting the tube at higher and lower test frequencies. If the signals from the
actual defect in Figure 8.24- are compared with the corresponding calibration signals
in Figure 8.23, one can see at 500 kHz the signal is primarily from a dent while that
at 20 kHz contains a large baffle plate signal component.
,
!6 , ,2 Multifrequency Testing of Dented Tubes

With a single frequency eddy current inspection, tube supports and dents tend to mask
signals from tube damage under tube supports. This makes detection and estimation
of severity difficult and time-consuming. In the remaining section we show how
multifrequency simplifies the inspection of the dented tube described previously.

Figure 8.25 illustrates the tube stripping sequence; one or more signals are removed
by each mixing of two frequencies. By proper manipulation of the signals from the
two lower frequencies, baffle plate and magnetite deposit signals can be eliminated.
Howe ve r, the resul tant eddy current signal is still distorted by the 'dent ing' signal.
Again, by combining this resultant signal with the signal from a higher test
frequency, the dent signal can also be eliminated. The tube now looks bare. ff a
defect existed under the baffle plate, it would be very easy to detect, the resultant
signal contains on ly information from the 00 corrosion. This process of unwanted
signal elimination is like solving three simultaneous equations with three unknowns
 - 149-

and solving for the paramete r Xl - defect.

( I (I
,
')

20 ,H ~ 500 ;"H l

.,
FIGURE 8.25
Tube Stripping Sequence by MuItifrequency

As shown in Figure 8.23, the signal at each baffle plate is a composite signal
comprising a baffle plate, magnetite deposit (or baffle plate corrosion products), dent
and defect signal. Figure 8.26 illustrates elimination of baffle plate and magnetite
signals. The probe is moved back-and-forth under the baffle pIa te and the signaJ is
monitored on the storage scope in the chopping mode, where both frequency signals
are displayed simultaneously.
 - 150-

Bar <l E
PLA TE

'I ~ITH PHASE

" ~Il" vE.l'(Al AIrs
COMPRESSED e, 0.74 ROTAT lOr; or lQ'

" : f1 - 'I /
\
~--...--~

~ RES I OUAl "ACNET I Tf SIGNAL

FIGURE 8.26
Suppression of Baffle Plate and Magnetite Signals

·re f 2 signal is first rotated to match the f 1 signal orientation. Then f 1

umplitude is changed to match, as nearly as possible, thef Isignal size. In this case,
this method by itself doesn't work. However, by decreasing the vertical component of
the f 1 baffle plate signal, one obtains a good match. On substracting the signal,
! through an electronic mixer (C 1) ,the signals from the baffle plate and the
magnetite deposit both nearly disappear. A small residual signal remains due to
different approach signals at the two test frequencies, indicated in Figure 8.26 by the
two open circles. Although the baffle plate signals are identical, the two points do
not coincide; the baffle plate is sensed earlier at the lower test frequency. This
residual signal is insignificant for this application though it can become quite serious
when testing for small cracks under non-ferromagnetic baffle plates.
 - 151 -

DOlT

~ f 3 500 kHz

~ENT

C~=Cl-f3
\ I

T
, RESIDUAL DENT SIGNAL

FIGURE 8.27
Suppression of Dent Signal

.27 illustrates how one can eliminate the 'denting' signal from the resultant
f 2-f 1) signal. This is achieved by first matching the phase and amplitude
of the 1 and f 3 'dent' signals and then using a second mixing module (c 2) for
subtra tlon.

Figure .28 traces the above sequence for two defective tubes, and shows the eddy
current signal becoming simpler to analyze with each step. On comparing defective
tube si nals with those from a calibration tube, one observes the f 2 defect signal is
distort d by the baffle plate, dent and/or magnetite deposit. The C1 signal is only
distort d from the dent signal, and c 2 is a dear signal indicating outside diameter
(00) pi s approximately 5096 deep. Even an inexperienced inspector could analyse
these r suIts.
 - [52-

o0
I O.
lo a 0 0 100 . 0 0
I OC-.'\. I
-~~
10 " . Cl~ 1 8RHI OH Tuer

10 ~

DENT

DEFECTIVE
TUBE NO 1

c,

DEfECTivE
TUBE NO . 2

c,

FIGURE 8.28
Multi requency Eddy Current Signals from Defective Tube

When using ulti-frequency to eliminate "internal diameter (ID) noise", such as

signals from yelic internal diameter variations ("pilger noise or die chatter"), dents
and probe wo ble, the signal amplitude from internal defects is drastically reduced.
However, sig al amplitude from external defects is not altered significantly.
Multlfrequen y is more effective for external defect detection than for detection of
internal defe ts in tubes. .

SUMMARY

Defect signal amplitude is a function of its axial and circumferential extent as well
as depth. Def ct signal phase is primarily a function of depth. For general purpose
volumetric in pection of heat exchanger tubes, a sui table test frequency is
2
f 9 0 ,.. 3 p / t • kH z (7.4)
where r is electrical resistivi ty and t is wall thickness.

Inspection at f 9 0 allows defect depth to be estimated on the basis of signal phase.

To discrimin te between defects and ferromagnetic deposits a lower test frequency
should be use; normally 10 or 20% of f 90.
 - 153 -

Signal response from most significant service induced defects is usually comparable
in am litude to that from a 1.6 mm diameter through hole. Stress corrosion cracking,
gener 1 corrosion and fretting wear give large signals whereas pitting corrosion and
fatigu cracks give small signals.

Testi g for fretting wear under non-ferromagnetic support plates is difficult and
unrelt ble with bobbin type probes, because defect and support plate signais are not
vecto ially additive. A surface type probe should be used.

Multi requency equipment can be used to eliminate unwanted components from

compl x signals such as support plates and probe wobble. This greatly simplifies
signal analysis.
 - 15 u -

CHAPTER 9 - METALLURGICAL PROPERTIES AND TESTING

FERROMAGNETIC MA TERIALS

'.1 INTRODUCTION

One can find numerous references in NuT publications dealing with eddy current
measuremeln t of material properties, such as chemical composition, hardness,
strength, oprrosion damage, degree of cold work and extent of beth carburization and
decarburizqtion. In fact, none of these properties and material conditions are
measured ~rectly. Eddy current testing is sensi tive to material properties through
their effect of resisti vi ty and magnetic permeability. As such, eddy currents only
provide indkect measurement of material properties and ca're must be taken to
ensure tha ~ some unforseen material variation does not lead to false conclusions. Two
precaution~ will help avoid false test results:

(a) a sou~d basic understanding of ET as outlined in previous chapters

(b) use of sui table standards for any particular test; the condi tion of such standards
shoul~ be verified by independent methods, e.g., hardness tests, tensile tests.

A complet~ treatment of materials property evaluation by eddy current testing is

beyond the lscope of this manual. The basics are cov p ' ~d and a few examples
presented. i

:;. 2 ELECTRIC~L CONDUCTIVITY

;.2.1 Factors Affecting Resistivity

All matericj.ls possess intrinsic resistance to electron flow (current) which is termed
resistivity ( P ,microhm-centimetres). The resistance of a conductor is given by

R = O'1./A ohms

where £. ~s length (cm) and A is cross-sectional area (cm 2 ) • Resistivtty values for
various ma ~ erials are listed in Table 9.1.

Conductivi ~ y ( cr. siemens/metre)* is the ease with which electrons can move
through a rpateria1. It is the reciprocal of resistivity. In eddy current testing,
conductivih is frequently given as a percentage of the International Annealed Copper
Standard (% lACS). In this system conductivity of pure, annealed copper at 20°C is
set to 100$ and conductivity of other materials is given as a percentage of copper.
Conductivi ~ y of a material can be calculated from its resistivity,

% I AC S - 172/ P

*C o nvers !1on: G c 10 8 /0, S/m or mho/m

 - 155-

Incre~sing temperature normally increases resistivity (decreases conductivity) as

shown l in Figure 9.1. Over a limited temperature range the variation is usually linear
accor ~ ing to the rei a tion
!
p - p (1 + auT)
o

TABLE 9.1

E~ECTR1CAL RESISTIVITY OF COMMON CONDUCTORS AT 20 0

e

M/\ TERrAL RESISTIVITY CONDUCTIVITY CONDUCTIVITY

(lJ.n.cm) (siemens/m) (% lACS)
7
Silver 1.6 6.14 x 10 105
Sapper 1.7 5.81 100
~old i 2.4 4.10 70
Aluminum
i~:~- T6 (AI AlIo )
l' 2.8
5.3
3.55
1.89 .
61
32
5.9 1.70 29
Magnesium 4.6 2.17 37
Admiralty Brass ! 7.0 1.43 24
Iron ! 9.7 1.03 18
Phosphor Bronze I 16 0.63 11
Lead , 20.6 0.49 8.4
70 Cu-30 Ni I 37.4 0.27 4.5
Monel I 48.2 0.21 3.6
Zirconium ' 50 0.20 3.4
Titanium I 54.8 0.18 3. 1
304 SST I
70 0.14 2.5
Zircaloy-2 72 0.14 2.4
Inconel 600 98 0.10 1.7
Hastel10y X I 115 0.087 1.5
Waspaloy I 123 0.081 1.4
fi-6/\ 1-4V I 172 0.058 1.0
!

wher P is resistivity at temperature T (·C). Po is resistivity at a reference

rature To' a (0 C -1) is thermal coefficient of resistivity and 6 T is the
rature difference (T-T 0) • For common metals and alloys values of Cl range
from ess than 0.00 I to over 0.01, 0.004 is fairly typical.

Allo)ng normally increases resistivity. Figure 9.2 shows even small alloy additions to
alumi~um can increase resistivity appreciably. The conductivity of binary Cu-Ni
 - 156-

60~
...
E
. 50 Q "0.04

.
C
u 40 PLATI~UI.I
E Q .. 0.004
~
0
~

u
E

~
~

>
~
(f)

Vi
\oJ
a::

200 400 600 800 1000 1200 1400

TEMPERATURE (-Kl

FIGURE 9.1
Effect of Trmperture on the Resistivity of Copper, Platinum and Titanium

.
'"
..E
.,uc
E
&.
o
~

u
E

1-0 2.0 3.0 4.0 5.0 6.0

ALLOY CONTENT (mass %)

FIGURE 9.2
Effect of 'Alloying Elements on the Electrical Resistivity of Aluminum.
 - 157 -

alloy ~ is shown in Figure 9.3. The dependence of c onductivity on c o mposition pro vides
o ne bf:is is fo r e dd y curr e nt sor ting o f mi xed all oys . O xygen impurit y in zirconium a nd
t it a niium all oys changes r esistiv it y consider ab ly . Figure 5.19 showed a no n- un ifo r m
oxygEtn di strib uti on in a zir co nium - ni obium det ec t e d by e ddy curr ent testin g .

' 100 t
80

COPPER/NICKEL ALLOYS
60

40

l 2D
I

o 20 40 60 80 100
MASS % COPPER
100 80 60 itO 20 a
MASS % NICKEL

FIGURE 9.3
VariatiOin in Electrical Conductivity of Nickel-Copper Alloys with Composition
!j

Cold 't'0rk increases resistivity through introduction of lattice defects in metals. At

normc+l temperatures, cold work has a relatively small effect on conductivity (<10 %)
and c4n usually be ignored. The degree of cold work in some austeni tic stainless
steels I can be determined by E T, this is possible because cold work makes them
ferr0rfagnetic, not because of a resistivity change.
i

9.2.2 Mater ~ al Sorting by Resistivity

This i~ normally an eddy current surface probe method. Two instrument types are
commpnly used. Impedance display instruments offer a comparative method as
treateld in Section 5.8.2; the lift-off curves for unknown materials are compared wi th
those bf known standards and the resi s tivity of the unknown is estimated b y
inter p~ lation. Meter readout instruments are also available with built-in "lift-off"
co m p ~ n s ation which are calibrated directly in % U\CS. Both types of instruments
r e quir ~ care on the part of the opera tor to insure meaningful resul ts. Effects wh ich
can cq ntribute to erroneous resul ts follow (for more details see Section 5.8.2):
 - 158-

(a) too ~ ow a test frequenc y can make material thickness appear si milar to
resi ~tivity changes.
(b) sam~le curvature affects coil coupling and hence its response (edge and other
georJIetry effects have a similar response).
(c) t oo nigh a test f,equency couIq sense alloy changes at the surface of oxidi zed or
corrbded materials.
(d) conqucting and nonconducting coatings affect test coil impedance.
(e) amb~ ent temperature var ia tions result in changes in sample resistivity and test
coil !resistance.

The abov~ potential error sources can largely be overcome through use of suitable
standards 'which duplicate ma terials to be tested.

i
i·

--- ---

/'
/'
,/ ---- -
//CONDUCTIVITY
/

I 10 100 1000
TIME AT TEMPERATURE (h)

j FIGURE 9.4
Var~ation of Mechanical Properties and Conductivity in 7075-T6
Aluminum EX]X)Sed at 205°C

An exam~le of eddy current testing to determine heat treatement state of an

alum inunjl alloy is shown in Figure 9.lt. These resul ts are from Pellegrini (lQ) who
indicatesl that the technique can be used to judge the fitness of overheated material
for furth~r service. A similar approach has been used to assess heat treat condition of
titanium la11oys.
i
 - 159 -

MAGN~nC PROPERTIES

For eddlY current purposes one can classify materials as ferromagnetic (magnetic) or
non-ferr oma gnetic (nonmagnet ic). Diamagne tic and paramagneti c materials can ::,e
co ns id ered nonmagnetic. Ferr o mag ne tism has its o r ig in in a quantum mechanics
effec t , It he "excha nge inter act ion". It occ urs in the elements iron, cobalt, nickel and
som e o~ the rare earth metal s. These elements have partially filled d and f electron
shel ls . fA ll oyi ng w i th el e men t s wh ich ha ve a higher elec tron to atom ratio fills these
d a nd f phells and ma kes t he resu lt ing alloys less magnetic, e.g., copper added to
nickel ( ~\onel) and chromium added to iron (st a inless steel).

The ma~n feature separating magnetic from nonmagnetic materials is magnetic

permeapility,. u which is a measure of a material's intrinsic ability to conduct
magnet~c flux. It is defined as the induced magnetic flux density, B, divided by
externa~ magnetic field intensity (magnetizing force), H,
i
! U = B/H

For air ~nd nonmagnetic materials u is a constant,

, -7
Uo = 4Tr X 10 webers/ampere-metre

when B ~s in tes1as* (T) or webers/metre

2
and H is in ampere/metre (Aim).

Sim pi ifika tion resul ts if one uses relative permeabili ty, which is defined as
I
li u r = u / U 0 (dimensionless)
i
Relativ¢ permeability has the same value in all magnetic systems of units. For
magnetilc materials j..l r can be very large, whereas for nonmagnetic material
Ur - 11. 0 •
!
9.3.1 Magneti!c Hysteresis
I

When af aterial is magnetized in a coil, the magnetic field intensity, H, is

proportional to coil current. If alternating current is applied to a magnetizing coil a
o-H 100 results as shown in Figure 9.5. As H increases from zero for the first time,
B incre ses along the DC curve, path No.1. When H decreases, B also decreases but
along p~ th No.2. The difference between paths 1 and 2 is termed hysteresis. When H
has fall ~ n to zero and residual flux densi ty remains in the material, B r' called
retentivlty or reSidu,a l flux density. On decreasing H further (reverse or negative
current) flux densi ty decreases to zero at He which is the coercive magnetic
intenSitt or coercive force. Decreasing H stili further drives the curve to point s 1·
Additio al AC cycles will retrace the loop. At point S 2 the material is saturated,
from S to S 3 the B-H curve is linear with slope ~ 0 • Flux density at
saturati n depends on material; carbon steel satura t es at a bout B = 2 tesla wherea s
Mon el 4pO saturates at about 0 .3 tesla.

*Convertsion: 1 t e s 1 a ,. 10 4 gauss; 1 Aim = 0.0125 66 oersted.

i
 - 160-

Br
s
~
He
--~----~~~~~---------H

FIGURE 9.5
Hysteresis (or ~H) Loop

o ':1.2 Permeabilit

current inspection of ferromagnetic materials several kinds of permeability

play an mportant role. Normal permeability, ~ r, is a measure of a material's
ability 0 conduct magnetic flux; it is an important factor when determining the ease
with wh ch a magnetic material can be saturated.

permeability of concern in ET is incremental or recoil permeability, U t:. .It

d as

B is the change in flux density which accompanies a change in rnagnetizing

force, H, created for example by an eddy current coil's alternating current. An
increm ntal t:. H can be superimposed at any point on a DC magnetization curve as
illustra ed in Figure 9.6.
 - 161 -

I 6 [

!
:
I
1. 4 r-

1.2 1
I
1
1
,

:z
I

-- 'T
a

w
:::>
a

-
:z

w
0.8
fLt'~8/0H

....z 0.6
'-' /'
....
X

./
..-
IRON

sao
~AGNErrZI~G FORCE 1.4 1m I

FIGURE 9.6
DC Magnetization Curve and Recoil Permeability for Iron

At H=O we have the relative initial permeability, lJ i . In a magnetic material

wi thout I a biasing DC magnetic field, the normal permeabili ty is equal to the
Increm1ntal permeabillty,

l.1 r- il.1 L
- l.1
lA
i

In eddY~Urrent testing, test coil inductance and depth of penetration are influenced
by incr mental permeabiE ty not normal permeability. However, throughout this
report i is assumed that the eddy current test is performed without DC bias and with
a low m~gnetiZing force (low alternating coil current). In this case, l.1 r - l.1/). , and
for sim llfication purposes ~ r is .u.sed in the skin depth .and inductance equations
and imp dance diagrams; lJ r is used throughout the manual to denote incremental
permea ility (l.1 ) unless otherwise stated. In addition, all permeabilities
6
describ1d hereafter in the text of the manual are relative permeabilitles and are
therefo~e dimensionless.
i

When ai increasing DC magnetizing field is applied, a nonlinear B-H relationship

results 4s shown in Figure 9.7. The incremental permeability continuously decreases
until sa1uration is achieved. At saturation ~ 6 -1. O. The normal permeability,
instead'i first increases to a maximum value and then decreases gradually, see
Figure 1-7; at saturation it can stW be very large.
 - 162-

. 1

::r
60
....>-
-
-'
I I)

...x
~

'"
....
C>.

...
>
....
~

...
-'

'"

FIGURE 9.7
Magneti tion Curve, Incremental Permeability and Normal Permeability
for a 3Re60 Tube Sample

9.3.3
I
Factors
!

netic materials do not have unique magnetization curves but depend

n factors such as

th rmal processing history,

m chanical processing his tory,
ch mica! composition,
int mal stresses,
te perature (if close to Curie temperature).

The fall wing examples illustrate the effect of some of these variables.

Figure 9 8 shows B-H curves, at room temperature, for three supposedly identical
Monel 4 0 tube samples. The differences are attributed to variations in nickel/copper
content ithin the normal alloy specification range.

Figure 9.9 shows variation of magnetic permeability with cold work in Type 300
series st in less steels (2). In these "nonmagnetic" austenitic steels a ferromagnetic
martens te phase forms during cold working increasing the magnetic permeability. In
contrast, most normally ferromagnetic materials exhibit a decrease in permeability
 - 163-

as a res It of cold work. The 300 series stainless steels can also become
ferrom gnetic as a result of welding, a magnetic delta ferrite phase forms during
solidifi ation.

"~"'L .;)0 rUBE,

'~O. "I ~ TIC a G. S. «
----- _ - - - - - -

~t~/~;_;-,-;_~_-_=',----,-_-"=____
I
O"l 0 ..l....---J._ _

O. / o. 4-00 800 -1200- 2000 24-00

/ ",/ .....-
O. 2 ~.::;::::--
/../
.....-
L..--.-- ./
/
:5 o. /
/ ruBE 1-2\0

V)
UJ
/
f-- I
I
/
/
/
a.lt ------------------------
-! ___________________ -_-_- ..
/ - -.. -:;~' !

:~~----
II I
400 - 800-1200---------

/ r: :E- -. -l-~~ - - - .. ~ .

/~~~~~ , /-rUBE I' 252

200 400 600 800 1000 1200 1400 1600

H AMPERES/METRE

FIGURE 9.8
Magnetization Curves for Various Monel 400 Samples

.- . ~
 >-
=
-l
c::l
-<
L:.J
:::!
~
UJ
Co. 10
u
;::
uJ
Z
l)
-<
:::!
uJ
>
;::
<
...J
UJ
~ 310
1.0 316

20 '10 60 80 100

f COLO 10RI:

FIGURE 9.9
Variation of Relative Permeability with Cold Reduction
for Various Austenitic Stainless Steels (~)

1. 5,---------r---------,--------,---------,

6 MPa NO STRESS

I. a

...,-..
u

..
z
:z:
O. 5

AIIN(ALLE O IRON

25 50 7S 100
~A(NfTlZING FORCE I ~ 'm I

FIGURE 9.10
Effect of Elastic Strain on the Magnetization of Iron (2.)
 - 165 -

figcm~ 3.10 shows changes in S-H curves for iron with internal stress. Note that these
stre sSi levels are purel y elastic, well below the yield s trength. The changes in B-H
(2nd per ;nea.bility) are due to magnetostrict icn.

The cpove exa rnpies illustrate the inherent v2.ria~i!ity of B-H and hence permeability
of fe~ro'l1agnetic materials. Incremental permeability affects an eddy current co il's
inductance as well as depth of eddy current penetration into a material. The large
v2.riaqions in per :neability shown above make conventional eddy current testing for
defec~s in magnetic materiais very difficult if not impossible.

The best solution to eddy current testing of a magnetic material for defects is to
bring it to a condition where U 6. :a 1. 0 • A few slightly magnetic materials can be
heated above their Curie temperature to make the:n nonmagnetic. Monel 400 heated
to between 50° and 70°C has been tested in this manner. Most materials have too
high a Curie temperature to be tested by this approach. The only other way to
decre~se U 6. to unity is by magnetic saturation. This topic is treated in a
subse41uen t section.
;

'7.4 TES-qNG MAGNETIC MATERIALS

9.4.1 SimDI ~fied Impedance Diagrams

A qU1·lit.~ltiVe understanding of the effect of permeability on coil impedance can also
be ob ained by the equivalent circuit and its associated semicircular impedance
diagr m treatment of Section 3.5. Coil inductance is a function of magnetic flux
through it; flux increases in the presence of a magnetic material. For a cylinder
surro~nded by an encircling coil, coil inductance is proportional to both the cylinder's
permfability and its cross-sectional area,
. 2
! L 0: U D
I pro
i
wher~ Lp is primary coil (probe) inductance, U r = U 6. is the cylinder's
incre ental permeability and Do its diameter. An increase in permeability or
diam ter will increase coil inductance. By a similar treatment to that presented in
Chap er 3, one can generate the impedance diagrams of Figure 9.11. Figure 9.11 (a) is
obtai ed by plotting the encircling coil impedance normalized to the inductive
react nce in air. It illustrates the effect of permeability and cylinder diameter. As
perm ability or cylinder diameter increases (with constant coil diameter) coil
impe ance increases drastically. (This explains the good response to ferromagnetic
inclu ions and deposi ts discussed in Sections 6.5.1 and 8.3.1). There is no phase
separ tion and hence no discrimination between variations in permeability and
cylin er diameter. However, there is about 90° phase separation and hence excellent
discr'mination between variations in permeability and resistivity.
 - 166 -

wL wL wL
wLo wLp wLo

1.0

R4/ w Lo
:
(0) C ~ L1NDER (b) CYLINDER (c) PLATE

, FIGURE 9.11
Simplif~ed Impedance Diagrams for Ferromagnetic Cylinders and Plates
:

i
!
I
Figure 9.111(b) is obtained by plotting the encircling coil impedance normalized to its
inductiv~ reactance with the ferromagnetic cylinder inside the coil. This figure
indicatesl the effect of permeability and cylinder diameter on operation point
loca tion. i An increase in both permeability and cylinder diameter moves the operating
point DolWN the impedance curve (for constant fill factor).
i

Surface ~robe inductance also depends on test sample permeability (L p is

proportiqnal to lJ r ) • An increase in permeability moves the operating point UP
the impe ~ance locus as shown in Figure 9.1l(c). However, unlike curves for a cylinder
where th~ semicircle increases drastically in size, the curve for a surface probe
increase$ only a small amount as previously shown in Figure 5.10. This results from
much les~ efficient coupling with surface probes as compared to encircling coils. A
surface ~ robe with a ferrite core (or cup) coil permi ts better magn e tic coupling
(decreasfd magnetic reluctance) and hence yields a larger impedance diagram than a
similar ~ir core coil. An additional obs ervation can be made fro :n Fi g ure 9.11 (C);
magneti<l: permeability has the same effe ct as electrica l resisti vi ty a nd henc e thes e
t wo pa r ~ me te r s cannot be sepa rate d when using a surface probe .
 - 167-

70
In ST AI NLE SS 5 TttL

....u
60 l I ~Hl

....Z
-
.......
u

0<

....>-
-
U
::l
40

Cl

-
Z

a
)0
....
'"
...-'x 20
""0
%

10

10 20 )0

NORMALIZED RESISTANCE

FIGURE 9.12
rimental Normalized Impedance Diagrams for Three Types 329
Stainless Steel Samples Tested with a Long Encircling Coil

' 9.~.2

Figu e 9.12 shows experimental impedance curves for three different Type 329
stain ess steel samples tested with long encircling coils. These curves differ markedly
from a semicircle at the lower section of the impedance diagrarh~ where the curve
appr aches the Y-axis at 45° rather than 90°. These curves are nearly identical in
shap to that presented in Figure 7.6 for a nonmagnetic cylinder. But, while the
nonm gnetic curve intersects the reactance axis (Y-axis) at La, the Figure 9.12
curv s intersects this axis at their respective ~ r values. Magnetic saturation of
these samples would reduce them to a common curve intersecting the axis at 1.0.
This igure is another example of typical permeability variations which may be
enco ntered in supposedly "identical" samples.
 - 168-

I / INCRE!5IN( P RoeE
I [JJI<Elt~
~

I/
-
u
z
1/

-
u
..,
'"
w
fI
/'
#/
INCREASING
FnCUE~CT

>

u
=>
w
z
1.0 .'
Cl
..,
....
-'
•
% INCREASING
cr
Cl PERIHAEILITT
i :z

I INCHASING
I 1lE51SlIVl TY

Ii
I

'r
I
NORMALIZED RESISTANCE
iI
I
FIGURE 9.13
I pedance Diagram for Ferromagnetic Material Showing
, Effect of Material and Test Parameters
I
I
!
Figure 9. 3 shows an actual surface probe impedance diagram for magnetic material.
The shape differs appreciably from a semicircle. Most test variables have a similar
effect on he impedance diagram as for surface probes on nonmagnetic material
(Section 5.5). To measure magnetic permeability in the presence of lift-off noise,
probe dia eter and test frequency should be chosen to operate in region A.

Eddy curr nt inspection of magnetic materials for defects is difficult or -impossible

because 0 random permeability variation as discussed in Section 9.3.3. In addition
there are skin depth limitations. Without saturation, the initial permeability can
range fro 50 to over 500. Since depth of penetration is inversely proportional to the
square ro t of permeability and test frequency,

6 ex: 1/1'f""W
r

to obtainrqua1
i
penetration requires a reduction in frequency by the same factor of
50 to ove 500~ Unfortunately, lowering frequency moves the operating point to
Region B n Figure 9.13 where there is poor signal separation between lift-off,
permeabi 1ty and resistiv ity as well as reduced sensitivity to defects.
i
 I

I - 169-

I
3efo~e leaving Figure 9.13 consider the characteristic parameter, r:2wlJrO
(Sec ion 5.6). Figure 9. i 3 shows the parameter is not generally valid for
ferr magnetic materials. It indicates an increase in t.! rshould ;nove the operating
point down the impedance curve like increasing frequency or probe diameter. In
praClice exactly t.he opposite occurs. The characteristic parameter should only be
used for finding ooerating point of surface probes on nonmagnetic materials.

9.4.3 \-1aterial Sartin b Ma netic Permeability

Deta led treatment of this topic is beyond the scope of this manual. This section is
esse tially a warning.

properties of magnetic materials aifect permeabili ty as discussed in Section

Eddy current testing has been used to sort mixed alloys as well as measur~ment
dness, decarburization, carburization, degree of cold work, strength,
ducti ility, etc. A standard, ASTM E566-76, offers broad guidelines on this eddy
curre t application.

V1ean ngful results with such testing requires at least the following:
understanding of the variables affecting a material's electrical and magnetic
properties
a sound knowledge of eddy current testing
adequate standard samples verified by destructive examination or other
independent methods.

9.4.4 Testi for Defects in Ma netic Materials

Previ us sections explained why saturation is required to suppress effects of usually

harmless permeability variations which could be mistaken for, or obscure, defect
signals. We only consider testing of cylindrical materials'; similar techniques can, at
least n theory, be applied to surface probe testing.

Manu acturing inspection of rods, wires and tubes is accomplished fairly simply by
exter aI, water cooled magnetizing coils through which the material is passed. ASTM
stand rd E309 covers such testing. In-service inspection again presents the most
diffic It situation due to access and space limitations.

9.14 compares V-channel eddy current signals from a Monel 400 tube at f90
t and with magnetic saturation. Saturation results in good defect detection.
Perm ability variation due to cold work and internal stresses at a slight bend in the
tube re completely suppressed by saturation. This tube was saturated by
superimposing the AC eddy current signal on DC magnetization power. Saturation of
Monel 400 is also achieved by incorporating permanent magnets in the probe (~).
 - 170-

SU PPORT
O.D. PLATE FLAT PITS
1"
-.;.>l;~.-..:\:o:=::~=·= CJ= = = = it:========/=:::L7---J1 ~~l~ ~R AT ION
DEFECT HOl E

L-I

..,.-,..~ EDDY C URRE NT

TEST WITHOUT
, SATURAT ION
SLIGHT BEND
IN TUBE

EDDY CURRENT
TEST WITH
MAGNETI C
\....JI_.II----"- SAT U RAT ION
(10 X ABOVE GAIN)

FIGURE 9.14
Eddy Current Signals from a High Magnetic Permeability Monel 400 Tube.
Test Frequency = 50 kHz
 - III -

Satuf2.ion with DC magnetization is limit::d by coil heating. Heat dissipation is

propor ional to current squared and coil wire resistance (P OI.J e r = 12 R) • To
increase magnetization (H is proportional to I) pulse saturation is used. The saturation
curren (DC) is switched on-and-off at regular interVi3.ls thereby reducing the heating
effect. The test current (AC) is superimposed on the sa turation current and the eddy
curren signal is sampled only at maximum saturation. One commercial instrument,
operating on this principle, is currently available. Testing speed is a function of pulse
ra te, i general it is much slower than conventional testing.

If mag~etic saturation at defects is not complete, an eddy current test becomes a

test forr permeability, not eddy current testing as described in previous chapters. This
can be understood from Figure 9.15 which illustrates the change in eddy current
signals from calibration defects in a magnetic stainless steel tube as degree of
saturation is increased. The eddy current signals were obtained with an absolute
bobbin type probe. Since defect signal amplitude decreases as saturation is
approa hed, instrument gain was doubled for the 20 and ItO ampere saturation results.
Magne ization was achieved with an external, water cooled coil; 10 amperes produced
about 2.8 x 10 4 Aim or 350 oersteds. Figure 9.15 shows one has to be saturated
well pa t the knee in the magnetization curve (over 20 amperes) before eddy current
defect ignals appear normal, like those from nonmagnetic materials.

son for the charateristic eddy current signals from partially saturated tubing
is mor dearly apparent in the eddy current impedance display of Figure 9.16 which
include impedance response as magnetization level increases. This figure shows, at
partial saturation (less than 10 amperes), defect signals consist nearly entirely of
increas ng and decreasing permeability. The initial increasing permeability signal
compo ent is attributed to less saturation on either side of machined calibration
defect while the decreasing permeability component is due to more intense
saturat on in the reduced tube-wall region at defects.

Similar results ere obtained with internal saturation using DC magnetization or

perma ent magnets. A single rare-earth permanent magnet was found to be
equival nt to about 5 amperes of an external magnetizing current for this tube size
materi 1.
 - 172-

(XTERN.l.l. MAGNET IZING COIL

THROUGH HOl.E I
~= 1=:;=:== , z'"
c:========~1 : I-INTERNAL ABSOLUTE PROBE

IDGR~VET I ~, I . 2 I

00 GRoovE

A B
I \

ABC 0
1 • \ '\

A PROBE weesl
_-----C
---
--- -- --
B THROUGH He -;-
AI
C 00 GROOVE
o 1.0 GROOV!
10 15 20 25 30 35 40
MAGNETIZING CURRENT (A 1

FIGURE 9.15
Eddy Cur ent Signals from E-Brite 26-1 Tube With Increasing Saturation,
(f90 = 100 kHz at Complete Saturation)

Eddy curr nt testing at partial saturation may seem attractive since defect
sensitivit is very high, it may in fact develop into a useful 'JDT technique. However,
there are rawbacks; lJ 6 is greater than one and is variable. This means eddy
current netration is not defined and conventional phase analysis is impossible.
Testing tu es for defects at magnetic suppor:ts could be a very questionable
procedure since large permeability signals would be encountered which could be
mistaken or or obscure defects. Even the best available saturation methods still
encounter problems in detecting defects at steel baffle plates in some Monel 400
tubes whi h are only slightly magnetic.

Eddy curr nt testing at partial saturation only measures permeability in a thin

surface 1 er adjacent to the test coil. This classifies the technique with NOT
methods ch as magnetic particle inspection and leakage flux testing. Leakage flux
testing re ponds to the distortion of magnetic flux at defects in a magnetized
material sing pickup coils or Hall effect sensors. Partial saturation ET with surface
 - 173 -

prob s has an advantage over encircling (or internal) probes in the ability to separate
perm ability from lift-off variations (Figure 9.13).

THROUGH HOLE EjERNAL MAGNETIZING COIL

~!772-t nst7?7'zz / 2 7 Z ( 7,7 Z 7771

c:===========tC=J-INTERNAL .ABSOLUTE PROBE

i
t
l,?W
ZU
_Z
(f')<l
.\\
BCD VZ'71l'12
ID GROOVE'
Z Z z;:Z Z 7

/
00 GROOVE
~z 7 Z ? ? ? % Z Z? , ,

<If-
WU c 0

L-------~, f
cr:::l
uO
ZZ
I
I

I
~--~/
1\
\\
BALANCE paiN
IN\AIR
I \
I \
\ \
\ \FI L -
\ \ FA
I \
I \
\
\
\
\
INCREASIN
FLUX DEN I T Y _
(DECREASI G
PERMEABI ITY)
\
\
\ 8
MAGNE
CURREN
IZI~G
(AM\S) \

\
\
I
10
\ABC
\\\ D

/"
\ A - PROBE W08BLE
\
\
, 8
C
-
.
THROUGH HOLE
O. D. GROOVE
20
30/~

4
_____________
-\
ABC 0
D - I. D. GROOVE

FIGURE 9.16
Edd Current Signals from E-Brite 26-1 Tube with Increasing Saturation,
f90 = 100 kHz
 ' . I

- 174 -

:;n exampl of the dangers of ET ferromagnetic materials at partial saturation is

illustrated i Figure 9.17. It shows eddy current signals from calibration defects in a.
3Re60 heat exchanger tube tested with a differential probe. (3Re60 requires a flux
density of 0.6T for complete saturation). Calibra tion defects yield signals
which chan in phase with increasing depth leading to the conclusion one may have a
viable test echnique. However, elastic deflection of the tube at a support plate gives
change of rmeabili ty signals nearly identical to serious (50% and 75%) defects. This
is due to m gnetostriction: changes in magnetic properties due to elastic stress such
as shown in Figure 9.10.

PARTIAl SATURATION PROBE

_________1 .;'

HOl E JR. 60

\
TUBE
lC ~ ~ 5
EHFL E

Cb

\
l.' , 7 ... 6 rIO' 5 "" ~ am J 10111 2 ""' 0

FIGURE 9.17
Eddy C t Signals from 3Re60 Tube With Partial Saturation for Various
evels of Elastic Stress. Test Frequency f90 = 230 kHz.
 - 175-

T:,e problem of Figure 9.17 was overcome with a multimagnet probe similar to that
d veloped for'v1onel 400 tubing (~). This eliminated the false defect signals at tube
s ppor"LS and made these heat exchangers inspectabte by conventional ET techniques.
[t was fortunate these particular heat exchangers had nonmagnetic, Type 304
st inless steel, support plates. This permits tube saturation in the vicinity of
s pports. If the supports had been magnetic they would have provided a low
r luctance al ternative path to the saturation field lea'iing the tube only partially
s turated. Nonmagnetic support materials improve inspectability of ferromagnetic
t· bes e'/en though fretting wear may be difficul t to detect wi th a conventional
b bbin-type probe as discussed in Section 8.2.4.

9.5 S~MMARY

dy current testing can be used to measure electrical resistivity and magnetic

rmeability. This parameter, in some cases, can be correlated to a material's
c emical composi tion, hardness, heat treatment, etc. and therefore provide an
in irect measurement of material properties. Material sorting by electrical
r sistivity can be done with general purpose eddy current instruments or with special
in truments with meter output calibrated in % lACS. Care must be taken to obtain
r liable results. Material sorting by magnetic permeability is not simple. It requires a
s und knowledge of magnetic properties and eddy current testing. Most of the
cfmmercial equipment make use of hysteresis distortion and the method is empirical.
It is more reliable to use general purpose eddy current equipment to roughly measure
m gnetic permeabili ty and then correlate to material property. .

T1sting ferromagnetic materials for surface defects is possible but often unreliable.
If Imaterial can be magnetically saturated, it appears as non-ferromagnetic material
toj the eddy currents. Testing at partial saturation results in good sensitivity to
dElfects and to ferromagnetic anomalies but can result in false indications. It is
p ssible to magnetically saturate some ferromagnetic tube alloys in unsupported tube
sections, but nearly impossible under ferromagnetic baffle plates.

gnetic permeability affects the following:

depth of penetration
probe inductance
operating point on impedance diagram
characteristic defect signal is no longer dependent on phase lag
drastically decreases signal-to-noise ratio.

9.6 RKED EXAMPLES

9.6.1 Convert resistivity of 5.5 microhm-centimetres to conductivity in %

Il\CS.

%IACS = 172/ p

= 172/5.5 = 3 1. 3 %
 - 176-

9.b.L Pure annealed iron under a magnetizing force, H, of 40 Aim results

in a magnetic flux density, B, of O.OnT. Determine magnetic
permeability and relative permeability.

SOLUTI
',1. K B/H - 0.028/40 - 7.0 x 10 -'~ henry/m

7
- U/U - 7.0 x 10-4/4n x 10- c 557 (dimensionless)
o
E: Tesla and ampere/metre are the preferred metric units for magnetic
flu density and magnetic field strength respectively but gauss and oersted
(no metric uni ts) are still often used. To complete problem 9.6.2 in the gauss,
ted system requires the following calculations:

0.028 T x 104. gauss = 280 gauss

B =
IT

4-0 A/m x 0.012566 oersted = 0.503 oersted

H = I AIm

= B/H = 280/0.503 = 557 gauss/oersted

= U/ u
o
= 557/1.0 = 557 (dimensionless)

9.6.3 PROBLEt Calculate standard eddy current depth of penetration in carbon steel
at a test frequency of 10 kHz (a) without saturation and (b) with
complete saturation. P ,. 15 microhm-centimetres, U :a 300
i
SOLUTIO N: ,. ,.
U~ Wr Wi
(a) Fro Equation 2.13(a)

<5 = 50 /P/fUt:,

~
1
= 50 104. : 300

= 0.11 mm

(b) = 1.0 at saturation

W~

<5 = 50 ~ 10 41 : 1.0

= 1.94 mm
 - 189 -

Lift-od - 39-43, 78
Limit Flrequency - 112-117
Magnet c Field - 6, 7
\1agnet c Flux - 7-10
·Magnet c Flux Density - 7, 159
'v\agnet c Permeability - 11, 12,66,67,89,91,92, 141, 159-165
~v1agnet c Saturation - 159-161, 169-175
Magnet zing Force - 159, 161
'v1odula ion Analysis - 46
Noise - 32, 34,37,46,81,152,182
Non-ferromagnetic - 10, 91, 142
Null Sa ance (Bridge Balance) - 32, 33
Oersted - 6, 8
Ohm's I w - 8, 16, 56, 110
Operati g Point - 25-29,71,92,113-115,124,141
Oscillat r - 5,31,39
Pararne er - 60, 183
Per for ance Standard - 183
Permea ility (Magnetic) - 11, 13, 66, 67, 89, 91, 92, 141, 159-165
Phase - 72, 73
Phase L g - 2, 14-17, 73, 85
Phasor 20
Primary Circuit - 8, 24
Primary Field - 183
Probe - 51-57, 98-106
Receive Coil - 6, 23, 62, 75
Referen e Coil - 32, 52, 53, 99
Reslsta ce - 18, 26-29, 123-125
ResistivIty - 12, 16, 66, 67, 74, 93, 154-158
Resona ce - 35, 36, 80, 105
Saturati n (Magnetic) - 159-161, 169-175
Second a y Field - 10, 184
Seconda y Voltage - 73
Send-R ceive - 6, 23, 31, 41-44,75
Sensing oil - 6, 23
Signal - 184
Signal-t -Noise Ratio - 58, 184
Similari y Condition (Law) - 71, 114
Sinusoid 1 - 5, 11
Skin De th - 13, 14, 16, 117
Skin Eff ct - 11
Speed 0 Response - 48
Standar Depth of Penetration - 12-17
Surface Probe - 51-55
Test Co 1 - 52, 53
Vector - 20
Voltage 8,9, 20, 31
Voltrnet -6
~ - 177 -

CHAPTER 10 - SUPPORTING INFORMATION

I
~
10.1 NOM~NCLATURE
QUALITY UNIT
SYMBOL QUANTITY NAME SI UNIT SYMBOL

2
A Cross-Sectional area metre 2 In
r Radius metre m
I Length metre rn
t Thickness metre In
w Width metre In
D Diameter metre In
B
2 2
Magnetic flux densi ty weber/metre or tesla Wb/m , T
C Capacitance farad F
f Test frequency hertz Hz
f90 Optimum tube testing frequency hertz Hz
fg Characteristics or Limit
frequency hertz Hz
fr Resonant frequency hertz Hz
H Magnetic field intensi ty amperes/metre or A/m
(Magnetizing force) lenze
I Current ampere A
2 A/m2
J Current density amperes/metre
L Self Inductance henry H
N Number of turns (Windings) dimensionless
Pc Characteristic Parameter dimensionless
)1
R Resistance ohm
RL Resistive load ohm n
V Electric potential volt V
x Depth below the surface metre In
XL Inductive Reactance ohm n
Xc Capacitive Reactance ohm Q
Z Impedance ohm Q
6 Standard Depth of Penetration metre In
\.l Per meabili ty henry/metre H/m
p Resistivity microhm-centimetre \.In·cm
0 Conducti vi ty siemens/metre S/m
cf> Magnetic flux weber Wb
TJ Fill Factor dimensionless
B Phase Lag radian rad
w Angular frequency radians/ second rad/s
0 Angle between Z &:: R degree 0
-----~--~.- -~-----------------------
i

" -178-
,
,,
).2 DEFINITI9NS

This sectio!n lists the most common terms covered in the manual. For each term, the
symbol, tht SI un its and the section where the topi c is covered is given, followed by
the definit on.

A bsol ute probe

I

See ~ections 5.2 and 7.2.

/\ pr t be having a single sensing coil.

AJternatin~ current
I
- I AC • amperes; see Chapters 2 and 3.
- A cu rrent flow changing in amplitude and direction with time.

Anomaly i
I
I

See tections 6.5 and 8.3.

An u expected, unclassified eddy current signal.
A fa se defect indication.

Bridge I
See ection 4.2.1.
Elec rical circuit incorporating four impedance arms.

Calibratio standard

A te t standard used to estimate defect size and set-up instrument.

I

Capacitiv, reactance

xc' o! ms; see Section 3.2.

- The pposition to changes in alternating voltage.

Character stic parameter

!

~~woj.J , dimensionless, see Section 5.6.

It al ows test coil operating point to be specified in terms of a single quantity
rath r than four independent variables.
:
Characterjistic or limit freq>..Jency
I
- f g' hrrtz, see Section 7.3.3.
I

Characte~istic frequency ratio

!

f If g ~i dimensionless, see Section 7.3.3.

It al ows the test coil operating point to be specified in terms of a single
qua tity rather than four independent variables.
I
 - 179-

Circumferential coil

see encircli ng and internal probes.

Conouctivity

a (sigma), siemens/metre; see Sections 2.4 and 9.2.

Measure of the abili ty of a material to conduct current (alternating or direct
current).

Conductor

Material capable of carrying electrical current.

Coupling

The coil's magnetic field couples to the test sample.

The change in probe impedance is directly proportional to probe-sample
coupling.

Current

I, amperes, see Section 3.3.

Flow of electrons.

Depth of penetration (standard)

<5 (delta), millimetres; see Section 2.4.

The depth at which the eddy current density has decreased to lie or 36.&% of
the surface density.
Also referred to as skin depth.

Defect

A flaw or discontinuity that reduces a material's integrity or load carrying

capacity -may involve a loss of material.

Differential probe

see Sections 5.2 and 7.2.

A probe having two sensing coils located side-by-side.

Direct current

- IDe , amperes; see Section 3.3.

A current flow that is constant in amplitude and direction with time~

Discontinuity

A defect.
 - 180-

Eddy currents

see Chapter 2 and Sections 5.2.2 and 7.2.3.

A closed loop alternating current flow induced in a conductor by a varying
magnetic field.

Eddy current method

An electromagnetic NOT method based on the process of inducing electrical

currents into a conductive materi2.1 and observing the interaction between the
currents and the material. In France it is known as the 'Foucault currents'
method.

Edge effect

see Section 5.8.2.

Signal obtained when a surface probe approaches the sampJe's edge.

Effective depth of penetration

see Section 2.4.

Depth at which eddy current density drops off to 5% of the surface density.

End effect

see Section 5.8.2.

Signal obtained when an internal or encircling probe approaches the end of a
tube or rod (similar to edge effect).

Encircling probe (Coil)

see Section 7.2.

Also referred to as a feed-through coil.
A probe which completely surrounds test material; can be absolute or
differential.

Feed-through coil

see encircling probe.

Ferrite

Ferromagnetic oxide material.

Used for cores in high frequency transformers.

Ferromagnetic

see Section 9.3.

A material with a relative magnetic permeability greater than 1.0.

Fill-factor

n (eta), dimensionless; see Section 7.3.

It isa measure of coupling between the coil and test object.
Fraction of the test coil area filled by the test specimen.
 - 181 -

Flaw

A defect.

Foucault currents method

In France the Eddy Current Method is known as the 'Foucault currents' method.

Frequency

f, hertz, see Section 2.lt.

Number of cycles of alternating current per second.

Frequency (angular)

w (omega), radians/second; see Section 3.2.

Angular veIoel ty, where w - 2 'IT f.

Hysteresis

See Section 9.3.1.

Magnetization curve.

lACS

International Annealed Copper Standard, see Section 9.2.

- 0 % lACS - Conducti vity as a percentage of that of pure copper.
IACS
Inductance

L, henries, see Section 3.2.

Ratio of the total magnetic flux-linkage in a coil to the current flowing through
the coil.

Impedance

Z, ohms, see Section 3.2.

The total opposition in an electrical circuit to flow of altenating current.
Represents the combination of those electrical properties that affect the flow
of current through the circuit.

Impedance method

Eddy current method which monitors the change in probe impedance; both phase
and amplitude.

Inducti ve reactance

- XL. ohms, see Section 3.2.

The opposi tidn to a change in alternating current flow.
 - 182-

Inductor

A coil.

Internal probe (Coil)

see Chapters 7 and 8.

A probe for testing tubes (or holes) from the inside. The coiJ(s) is
circumferentially wound on a bobbin.

Lift-off

L.O., mm, see Sections 5.5 and 5.8.4.

Distance between the coil of a surface probe and sample.
It is a measure of coupling between probe and sample.

Magnetic flux

cp ,webers, see Section 9.3.

Magnetizing force

H, amperes/metre, see Section 9.3.2.

Magnetic field intensity.

Magnetic flux density

B, testa, see Section 9.3.

Modulation analysis

An instrumentation method which separates desirable from undesirable

frequency signals from the modulating envelope of the carrier frequency signal.
Test sample must move at constant speed.

Noise

Any undesired signal that obscures the signal of interest .

It might be electrical noise or a signal from specimen dimensional or property
variations.

Null balance

see Section 4.2.1.

Ohm's law

Electromotive force across a circui t is equal to the current flowing through the
circuit multiplied by the total impedance of the circuit.

Operating point

see Sections 3.5, 5.6 and 7.3.3.

Point on the impedance diagram that specifies the normalized inductive
reactance and resistance of a coil.
 - 183-

Oscillator

The electronic unit in an eddy current instrument that generates alternating

probe exci tation current.

Parameter

A ma terial property or instrument variable.

Performance standard

Also referred to as Reference Standard.

A test standard used to qualify and calibrate a test system for a particular test.

Permeability (Magnetic)

lJ (mu), henry/metre; see Sections 2.4- and 9.3. or lJ r' dimensionless, relative
magnetic permeabili ty.

Ratio between flux density, B, and magnetizing force, H. Permeability

describes the intrinsic willingness of a material to conduct magnetic flux lines.

Phase lag

B (beta), radians or degrees; see Section 2.4-.

A lag in phase (or time) between the sinusoidal currents flowing at the surface
and those below the surface.

Phasor

see Section 3.3.

A vector describing sinusoidal signals; it has both amplitude and phase.

Primary field

The magnetic field surrounding the coil due to the current flowing through it.

Probe

Eddy current transducer.

Reference coil

Coil which enables bridge balancing in absolute probes. Its impedance is close to
test coil impedance but does not couple to test material.

Resonance

See Sections 4.3, 5.9 and 7.2.5.

A circuit having an inductor and capacitor connected in series or parallel.
When inductive reactance equals capacitive reactance the circuit is tuned or in
resonance.
 - l81t-

Resistance

R, ohms, see Section 3.2.

The opposition to the flow of electrical current.
Applies to DC and AC.

Resistivity

p ,microhm-centimetre; see Sections 2.1t and 9.2.

Reciprocal of conductivity (p -1 /a ) •

Saturation (Magnetic)

A condition where incremental magnetic permeability of a ferromagnetic

material becomes 1.0.

Secondary field

The magnetic field produced by induced eddy currents.

Send-receive

See Sections 3.1t, 1t.5 and 5.1t.

The variations in the test object which affect current flow within the test
object can be detected by observing their effect upon the voltage developed
across a secondary receive coil.

Signal

A change in eddy current instrument output voltage; it has amplitude and phase.

Signal-to-noise ratio

Ratio between defect signal ampli tude and that from non-relevant indications.
Minimurry acceptable ratio is 3: 1.

Skin depth

See depth of penetration.

Skin effect

See Section 2.4.

A phenomenon where induced eddy currents are restricted to the surface of a
test sample. Increasing test frequency reduces penetration.

Surface probe

See Chapters 5 and 6.

A probe for testing surfaces, whiCh has a finite coverage. The coil is usually
pancake in shape.
 - 185-

Test coil

Coil coupled to test material. It senses geometric, electric and magnetic

changes in test material. .

Voltage
'.

V, volts, see Section 3.3.

Electric potential or driving force for current.
Output signal from an eddy current instrument.

Voltmeter

The instrument used to measure voltage.

Vector

see Section 3.3.

A quantity having amplitude (magnitude) and direction. Normally represented as
a line whose length represents the quantity's magnitude and whose angular
pOSition, the phase (relative to some reference).
 - 136-

.3 ABBREVIA nONS FOR NONDESTRUCTIVE TESTING TERMS

The following abbreviations of nondestructive testing terms are used where required
in nondestructive testing standards published by CGSB:

Terminology Abbreviation

Nondestructive testing NDT

Eddy current testing ET

Hardness testing H

Leak testing LT

Magnetic particle testing MT

Penetrant testing PT

Radiographic testing RT

Ultr asonic testing UT

Visual testing VT

References:

Recommended Practice SNT - TC-IA: Personnel Qualification and Certification in

Nondestructi ve Testing

American Society for Nondestructive Testing - 1980 Edition

Symbols for Welding and Nondestructive Testing

American Welding Society
 - 187-

10.4 REFERENCES

-. 1. H.S. Jackson, "Introduction to Ele c tric Circuits", 2nd ed ition, Prentice-Hall

Inc., Englewood Cliffs, New Jersey (1965),

2. C.V. Dodd, "The Use of Computer-Modelling for Eddy Current Testing",

Research Techniques in NDT, Vol. III, edited by R.S. Sharpe, Academ ic Press
Ltd., London, p. 429-4 79 (1977).

3. H.L. Libby, "Introduction to Electromagnetic Nondestructive Test Methods",

Wiley-Interscience, New York (1971).

4. "Nondestructive Testing Handbook", Vol. II, edited by R.C. McMaster, Ronald

Press, New York, p. 36.1-42.74 (1963).

5. "Eddy Current Testing, Classroom Training Handbook", General

Dynamics/Convair Division, San Diego, California (1979). CT-6-5 Second
Edition.

6. W.J. McGonnagle, "Nondestructive Testing", 2nd edition, Gordon and Breach,

New York, p. 346-390 (1961).

7. F .R. Bareham, "Choice of Frequency for Eddy Current Tube Testing", British J.
Applied Physics, 11, p. 218-222 (1960).

8. V.S. Cecco, "Design and Specifications of a High Saturation Absolute Eddy

Current Probe with Internal Reference", Materials Evaluation, 37, p. 51-58
(1979). -

9. J. Stanley, "Electrical and Mangnetic Properties of Metals", American Society

for Metals, Metals Park, Ohio (1963).

10. H.V. Pellegrini, "Assessing Heat Damage in Aluminum Alloys with an Eddy
Current Testing Technique", Metals Progress, 117, p. 60-63 (1980).

11. ASME Boiler and Pressure Vessel Code, Section V, Article 8, Appendix 1, "Eddy
Current Examination Method for Installed Non-Ferromagnetic Steam Generator
Heat Exchanger Tubing" (1978).

12. "Nondestructive Inspection and Quality Control", Metals Handbook, Vol. 11, 8th
edition, American Society for Metals, Metals Park, Ohio, p. 75-92 (1976).

13. R. Hochschild, "Electromagnetic Methods of Testing Metals", Progress in

Nondestructive Testing, Vol. 1, MacMillan Co., New York, p. 59-109 (1959).
 - 188 -

10.5 INDEX

Absolute Probe - 51, 98-102

Alternating Current - 8, 16,20-22
Anomaly - 91, 140
oridge - 32-36
Bridge Balance - 32-36
Calibration Standard - 94-96, 117
Capacitive Reactance - 19
Characteristic or Limit Frequency - 112-117,120
Characteristic Frequency Ratio - 112-117, 121
Characteristic Parameter - 51, 69-71, 81, 82, 112
Circumferential Coil - 98, 102, 117
Conductivity - 11, 154-157
Coupling - 23, 27, 51, 100, 106
Current - 5-10, 20-22
Defect - 51, 60,61, 73, 83-90,94,123-138,169
Depth of Penetration (Standard) ,.. 12-17, 74
Differential Probe - 53, 54., 98-102
Direct Current - 20, 21
Discontinuity - 179
Eddy Currents - 5-17, 55, 102, 103, 123
Eddy Current Method (Testing) - 1, 18, 51,83,91, 123,154
Edge Effect - 76
Effective Depth of Penetration - 14.
Encircling Probe (Coil) - 98, 106, 109, 113, 142
End Effect - 180
Excitation Coil - 6, 11, 41, 62
Faraday's Law - 8, 16, 45, 55, 64, 109
Faraday, M. - 2
Farad - 19
F eed-Through Coil - 180
Ferrite - 37, 180
Ferromagnetic - 10,91, 159
Fill-Factor - 27, 106-108, 141
Flaw - 181
Forster - 2, 113
Foucault Currents Method - 181
Frequency - 5,8, 12, 16,67, 113, 116, 120, 121
Frequency (Angular) - 8, 181
Frequency Response - 48
Hall Detector - 6, 31, 42, 172
Henry - 18
Hysteresis (3-H curve) - 159-163
lACS - 154-157
Impedance - 8, 9, 19, 23, 30
Impedance Diagrams - 23-29
Impedance Method - 23, 31
Inductance - 18, 57, 58, 103, 105
Inductive Reactance - 19, 25, 64, 167
Inductor - 18
Internal Probe (Coil) - 99
Lenz's Law - 9, 22