remote field testing

David E. Russell, Russell NDE Systems, Edmonton,Alberta, Canada

David D. Mackintosh, Queens University, Kingston,Ontario, Canada

Ad A. Shatat, Russell NDE Systems, Incorporated,Edmonton, Alberta, Canada

8C H A P T E R

Remote Field Testing

PART 1. Background

IntroductionRemote field testing is popular because ofits ability to inspect regions not only nearthe probe but also throughout thethickness of the material. Thischaracteristic is especially valuable for thetesting of thick walled ferromagnetictubing, where the technique shows highsensitivity to inside and outside surfacepipe wall heterogeneities. Remote fieldtesting often does not require cleaning ofthe pipe and is not sensitive to internalcoatings. Since 1970, remote field testinghas grown into a mature and recognizednondestructive testing technology. Thischapter introduces remote field testing —its history, applications, strengths andlimitations.

HistoryThe remote field effect was first noted inthe 1940s and was patented by W.R.MacLean in 1951.1 In the late 1950s,Thomas R. Schmidt independentlyrediscovered the technique whiledeveloping a tool for the inspection of oilwell casings. Shell Developmentpurchased the patent rights from MacLeanand had great success with the tool. Atthat time, no electromagnetic techniquesfor examining the casingsnondestructively were available.2 Schmidtspearheaded the development of thetechnique and named it remote field eddycurrent testing to distinguish it fromconventional eddy current testing. Thetechnique as used in industry is nowreferred to as remote field testing. The termminimizes confusion with conventionaleddy current testing and emphasizes themagnetic field interactions exhibited byremote field testing.3 Shell encouraged thecommercialization of the technique bylicensing it to interested parties in the1980s. Several manufacturers immediatelyrecognized the value of remote fieldtesting for the examination of ferrousheat exchanger tubes and beganmanufacturing remote field testequipment.

Over the last 20 years remote fieldtesting has attracted the interest ofresearchers around the world.4 Theresearch was triggered by Schmidt’s 1984

208 Electromagnetic Testing

publication on the usage of acircumferential array of detectors.5W. Lord and others gave remote fieldtesting a firm theoretical basis bypublishing the first in-depth finiteelement study.6 Later Mackintosh andAtherton developed powerful analysistools by recognizing remote field testing’sthrough-transmission character.7 Theimproved understanding of how remotefield testing worked increased its accuracyand acceptance. This resulted intremendous growth in the late 1980s andearly 1990s. Systems were developed usinginternal probes to examine gasdistribution pipelines, oil and gas wellcasings, cast iron and steel water mains,heat exchangers and boilers.Developments since 1990 include thetesting of flat plates (for example, storagetank floors)8 and steel pipes using externalprobes that use a technique similar toremote field testing.

Early remote field testing systems weremuch like the eddy current test systems ofthe 1980s: both used analog circuits,cathode ray tube displays and paper chartrecorders for data storage. Moderninstruments use computers to display andstore data and more advanced systemsalso have automated signal analysisroutines.

PART 2. System Components

FIGURE 2. Electronic components of remote

Probe ConfigurationIn the basic remote field testing probe(Fig. 1), there is one exciter coil and onedetector coil. Both coils are woundcoaxially with respect to the tested tubeand are separated by a distance greaterthan twice the tube diameter. This axialdistance is very characteristic of remotefield testing. If the exciter and detectorwere to be placed close together thedetector would measure only the fieldgenerated by the exciter in its vicinity. Inthat case the setup would basically be astandard eddy current test setup in sendand receive mode.

To observe remote field testing’s uniquethrough-wall transmission effect thedetector needs to be moved away fromthe exciter. The actual separation dependson the application and the probemanufacturer but will always be aminimum of two pipe diameters. It is thisseparation that gives remote field testingits name — the detector measures theelectromagnetic field remote from theexciter. Although the fields have becomevery small at this distance from theexciter they contain information on thefull thickness of the pipe wall.

The dimensions of the coils will varyfrom manufacturer to manufacturer. Thefill factor is the ratio of the effective crosssectional area of the primary internalprobe coil to the cross sectional area ofthe tube interior. Although the fill factorof the coils can be as low as 70 percent itwill usually be similar to the fill factor foreddy current probes: 85 percent or more.A lower fill factor reduces sensitivity tosmall discontinuities but does nototherwise affect the quality of remotefield testing data. The ability to functionwith a low fill factor makes remote field

FIGURE 1. Simple probe for remote fieldtesting.

Tube Energy flow

Corrosion

Detector coilExciter coilProbe

lead

testing especially attractive for pipes withinternal coatings and tight bends.

Remote field testing probes oftencontain arrays of receiver coils. The coilsare connected to a remote field testinginstrument by coaxial cable, where theouter conductor is used to shield theinner conductors from ambient noise. Thecoaxial cable is usually housed in a stiffplastic tube that lets the probe be pushedinto a heat exchanger tube or pipe fordistances up to 30 m (100 ft).

InstrumentationBesides the coils and coaxial cable aremote field testing system contains fourother major components.9

1. An oscillator is used as the signalsource for the exciter coil and as areference for the detector signal.

2. A power amplifier increases the powerlevel from the oscillator signal so thatit can be used to drive the exciter coil.

3. The phase and amplitude detectormeasures the detector coil signal.

4. A microcomputer based storage deviceprocesses and stores the data.

Figure 2 shows how the differentelectronic components interact.

Driving of Remote Field TestProbeThe exciter coil is energized withalternating current at frequencies rangingfrom 50 Hz to 1 kHz for ferrous materials.Higher frequencies are used fornonferrous tubes. The exciter coiltypically carries a current of 0.1 to 1.0 A,

209Remote Field Testing

field test system.

Detectorcoil

Phase andamplitudedetector

Computer

Excitercoil

Poweramplifier

Reference signal

Oscillator

the limitation being probe temperatureand, for some probes, magnetic saturationof the probe core. The detector signaldepends directly on the exciter currentand frequency. Thicker wall penetrationcan be obtained by lowering thefrequency and increasing the excitercurrent. The actual test parameters (suchas test frequency, drive voltage andsample rate) are chosen by taking intoconsideration factors such as probe pullspeed, discontinuity sensitivityrequirements, magnetic permeability µ,tube electrical conductivity σ and tubewall thickness τ.

In general, a lower frequency (up to250 Hz) is used for thick walled and highpermeability pipe. For superior sensitivityor high test speeds, a higher frequency ispreferred. The frequency is chosen as highas possible while minimizing noise andremaining in the remote field zone.Because remote field testing is used invery diverse applications, it is importantto check for the presence of anyelectromagnetic interference noisesources, such as welders, electric motors,power inverters and pumps. Such devicestend to generate noise in the frequencyrange of remote field testing. Modern dayremote field testing equipmentmanufacturers sometimes provide the userwith a noise spectrum, showing theenvironmental noise for a range offrequencies. Some technicians prefer toplace the probe in a thick walled block todetermine the baseline noise at a givenfrequency.

Because remote field testing signals areoften quite small (1 to 10 µV) it isadvisable to avoid using line voltages andtheir harmonics (60 Hz or, in Europe,50 Hz), which can cause interference.

The inspector needs to minimize thebackground noise while keeping in mindthe test factors mentioned above.


PART 3. Detector Signal

Remote Field Energy ZonesFor a remote field probe, there are twodistinct sensing zones with a transitionzone between them (Fig. 3).10 In order,the zones are the direct field zone, thetransition zone and the remote field zone.

As the detector coil distance from theexciter coil is increased, the dominantfield energy changes from direct coupledenergy (between the exciter and detectorcoils, inside the tube) to energy that iscoupled to the detector coil primarily bytransmission through the tube wall.Between these two distinct zones, there isa transition zone where the direct coupledenergy and the indirect coupled energyare comparable in magnitude. Thelocation of the transition zone changeswith frequency, wall thickness,permeability and conductivity.

In an idealized situation (infinitealternating current sheet over aconducting half space), the eddy currentdensity in the case of conventional eddycurrent techniques decays exponentially

FIGURE 3. Zones in remote field testing. Profilesused to indicate direct field region, transition a

Bfie

ld m

agni

tude

, µT

(m

G)

10 (100)

1.0 (10)

0.1 (1.0)

0.01 (0.1)

Indirect energy transm

Exci

ter

coil

Directcoupled

fieldTransition zone

0I diameterdistance

2 d

with depth. This phenomenon, called skineffect, in general limits the application ofconventional eddy current techniques tothe detection of surface or shallowheterogeneities. The remote field eddycurrent technique seemingly violates theskin effect limitation in that it is equallysensitive to inside surface and outsidesurface discontinuities. Lord and otherssimulated the underlying physical processand examined the field distribution tolook for clues to explain this seemingcontradiction.6 A brief summary of theirfindings are reported below.

Finite Element SimulationElectromagnetic induction phenomenaassociated with conventional and remotefield eddy current nondestructivetechniques are both described by thecomplex form of the vector poissonequation:


of B field just inside and outside pipe wall arend remote field zones.10

Amplitude of outside wall

ission path

Detector

Remote field zone

diametersistance

3 diametersdistance

Amplitude of inside wall

FIGURE 5. Flux distribution in pipe as function of time in stepsof 4.1667 ms, which corresponds to 0.5 rad (30 deg)increments of ωt (product of angular frequency ω andtime t) when frequency f = 40 Hz: (a) 0 rad (0 deg);(b) 0.5 rad (30 deg); (c) 1.0 rad (60 deg); (d) 1.5 rad(90 deg); (e) 2.0 rad (120 deg); (f) 2.5 rad (150 deg). Onlytop half of pipe longitudinal section is shown. Flux lines areplotted logarithmically to increase dynamic range of plot.Test conditions and specimen are similar to those in Fig. 4.

(a) (d)

Radi

al d

ista

nce

(pro

por

tiona

l sca

le)

Axial scan direction(distance unit equivalent

to pipe diameter)

Radi

al d

ista

nce

(pro

por

tiona

l sca

le)


to pipe diameter)

(b) (e)



OS

IS

PA

OS

IS

PA

Radi

al d

ista

nce

(pro

por

tiona

l sca

le)

OS

IS

PARadi

al d

ista

nce

(pro

por

tiona

l sca

le)

OS

IS

PA

(1)

where A is magnetic vector potential(volt), J is applied current density (ampereper square meter), µ is magneticpermeability (henry per meter), σ isconductivity (siemens per meter) and ω isangular frequency (radian per second).Lord and others used the finite elementtechnique to study a test geometryconsisting of an axisymmetric excitationand detector coils in a ferromagneticpipe.6

Figure 4 shows experimental andmodel predictions of the detector coilsignal magnitude and phase as a functionof the distance (expressed in multiples ofthe pipe diameter) between the excitationand detector coils. The phasemeasurements are made with reference tothe excitation signal. The magnitude ofthe detector signal decreases rapidly withdistance until a transition region isreached (after about two pipe diameters inlength), after which the rate of decrease ismuch slower. The phase angle, incontrast, changes rapidly in the transitionregion. Some insight into the underlyingphysical process can be gained byexamining the magnetic flux plots.

Figure 5 shows the variation in the fluxdistribution at various points in thesinusoidal excitation cycle. The excitationfrequency is 40 Hz and the plots show theevolution of the magnetic flux pattern asa function of time in steps of 4.1667 ms,which correspond to 0.524 rad (30 deg)increments in ωt. The flux lines areplotted on a logarithmic basis to increasethe dynamic range of the plot. The plotsshow that the flux density decays rapidlyclose to the exciter (near field). The decay

1 2

µωσ∇ = − +A AJ js


Experimental measurement of phase

Finite element predictionof magnitude

Experimental measurementof magnitude

Finite element prediction of phase

FIGURE 4. Experimental and finite element predictions ofsignal magnitude and phase for detector coil for 38.1 mm(1.5 in.) inside diameter ferromagnetic pipe of 5.1 mm(0.2 in.) wall thickness. Test frequency f = 40 Hz; relativepermeability µr = 250; conductivity σ = 143 S·m–1.

Nor

mal

ized

mag

nitu

de(lo

garit

hmic

sca

le)

Axial distance from excitation coil (distance unit equivalent to pipe diameter)

1.00.50

– 0.5–1.0–1.5–2.0–2.5–3.0–3.5– 4.0– 4.5–5.0

3.5 (200)

2.6 (150)

1.7 (100)

0.9 (50)

0

–0.9 (–50)

–1.7 (–100)

–2.6 (–150)

–3.5 (–200)0 1 2 3 4 5 6 7 8 9 10 Legend

IS = inside surfaceOS = outside surfacePA = pipe axis

to pipe diameter) to pipe diameter)

(c) (f)


to pipe diameter)


to pipe diameter)

OS

IS

PARadi

al d

ista

nce

(pro

por

tiona

l sca

le)

Radi

al d

ista

nce

(pro

por

tiona

l sca

le)

OS

IS

PA

FIGURE 7. Equiphase contour plots corresponding to Fig. 8,indicating existence of phase knot (point where phase isundefined) in transition region. Test conditions andspecimen are similar to those in Fig. 4.

Phaseknot

OS

IS

PA0 1 2 3 4

Radi

al d

ista

nce

(pro

por

tiona

l sca

le)

Axial distance from excitation coil(distance unit equivalent to pipe diameter)

Legend

rate is much lower in the remote fieldzone. As an example, at ωt = 1.57 rad(90 deg), the four bands of flux linescontain 90 percent, 9 percent, 0.9 percentand 0.09 percent of the total flux.

Additional insight can be gained byreviewing the equivector magneticpotential contour and equiphase contourplots shown in Figs. 6 and 7. Theequivector magnetic potential contourplot shows the existence of a potentialvalley (that is, a point where the magneticvector potential reaches a minimum) inthe transition region where the root meansquare of the vector magnetic potential iszero. Similarly, Fig. 7 shows the existenceof a phase knot (that is, a point where thephase is undefined) in the transitionregion. The presence of these unusualartifacts indicates that the energy flows inan unusual pattern.

Figure 8 shows model predictions ofthe real component of the poyntingvector, which shows the magnitude anddirection of the energy flow at each pointin space. The plot shows that the fieldpattern in the transition region arises as aconsequence of the interaction betweentwo energy streams. The outward boundenergy stream interacts with the inwardlydirected energy stream in the transitionregion, giving rise to the potential valleyand phase knot. The presence of potentialvalley leads directly to the magnitude plotshown in Fig. 4. Because the energy flowloop includes the regions near the outerand inner walls of the pipe and becausethe detector coil output is a function ofthe field that it is immersed in, the

FIGURE 6. Equivector magnetic contour plots showingexistence of potential valley (point where magnetic vectorpotential reaches minimum) in transition region. Testconditions and specimen are similar to those in Fig. 4.

Potentialvalley

OS

IS

PA0 1 2 3 4

Radi

al d

ista

nce

(pro

por

tiona

l sca

le)

Axial distance from excitation coil(distance unit equivalent to pipe diameter)

LegendIS = inside surface

OS = outside surfacePA = pipe axis

technique is sensitive to both insidesurface and outside surfacediscontinuities. In other words, the fieldin the vicinity of the detector coil is aconsequence of the superposition of thefields generated by the energy flows nearthe inner and outer walls of the pipe.Consequently, the method is equallysensitive to inside surface and outsidesurface discontinuities.

A detector placed in the remote fieldzone will provide information from twopipe wall transitions: one at the exciterand one at the detector. The totalattenuation (in decibel) and phase change


IS = inside surfaceOS = outside surfacePA = pipe axis

FIGURE 8. Finite element model predictions of realcomponent of poynting vector, showing magnitude anddirection of energy flow pattern in region.

Radi

al d

irect

ion

(arb

itrar

y un

it)

Longitudinal direction(arbitrary unit)

(in degrees) is the sum of the attenuationand phase change at both locations.Schmidt5 discovered that the attenuationand phase change could be derivedapproximately by using the skin effectequation for depth of penetration:

(2)

and:

(3)

where f is frequency (hertz), δ is standarddepth of penetration (meter), µ0 ismagnetic permeability of free space(ratio), µr is relative magnetic permeability(ratio), σ is tube electrical conductivity(siemens per meter) and τ is tube wallthickness (meter).

When the probe is in the nominal tubewall setting, the signal from the remotefield testing detector coil is used as thereference voltage against which changescaused by wall thinning and otherdiscontinuities are measured. Theattenuation and phase rotation angles ofremote field test signals are governed byrules like those for eddy current signals.Equation 1 shows that, if the frequency,conductivity or relative permeability goesup, then the standard depth ofpenetration into the material goes down— and vice versa. Therefore, a relativelyhigh test frequency (500 Hz to 1 kHz)would be chosen to test thin materials —for example, a wall less than 1 mm (0.04in.) — or to test materials that have a lowpermeability such as magnetic stainlesssteel or low electrical conductivity such asnickel. Conversely, if the tube wall is thick— for example, greater than 3 mm (0.12in.) — or has a high relative permeability(for example, seamless carbon steel), thena lower frequency would be moreappropriate (for example, 50 to 200 Hz).The quantity that sets the standard depthof penetration is the product of thefrequency, permeability and conductivity.Instead of resolving these factorsseparately, it is often more convenient todetermine their product.

To analyze remote field discontinuitysignals using the skin effect theory, thewall thickness must be expressed instandard depths of penetration, where astandard depth of penetration isequivalent to a radian (rad) and 1 rad =180·π–1 ≅ 57.3 degrees. Consider a casewhere the nominal wall thickness of thepipe is five standard depths of penetrationand wall loss at one point is fullycircumferential and 1.5 standard depths ofpenetration deep. When diffusingthrough the pipe wall, the signal will be

Phase lag r= µτ π µ σf 0

δπ σ

=µ µ

1

0f r


delayed by five standard depths ofpenetration or 5 rad (286 degrees)through nominal thickness and by 3.5 rad(201 degrees) at the circumferentialdiscontinuity. As the detector coil movesfrom nominal pipe into the area of metalloss, its phase lag will decrease by 1.5 rad(86 degrees), which makes the phaseincrease by 1.5 rad (86 degrees), a positivechange at metal loss. A similar analysiscan be applied to the amplitude. Innominal pipe, the signal is attenuated bya factor of exp (5) = 148 whereas at thediscontinuity location the amplitude isreduced by a factor of exp (3.5) = 33. Thedetector coil output will therefore increasein amplitude by a factor of exp (1.5) = 4.5.

In general, the relatively simple skineffect equation describes remote fieldtesting behavior very well. However, whena discontinuity’s depth is such that theremaining wall thickness is less than onestandard depth of penetration, the remotefield response will actually start to deviatefrom the skin effect equation.7 Thedeviations are especially noticeable inamplitude. In those cases, moresophisticated techniques such as thosedescribed elsewhere, in this volume’schapter on modeling, are required.

PART 4. Selection of Remote Field Testing

Features of Remote FieldTesting11

Remote field testing can be used for allconventional carbon steel materialspecifications, diameters and wallthicknesses. It is therefore used in manydifferent types of heat exchangers,including fossil fuel boilers (especially inwater wall and generator bank tubes),black liquor recovery boilers, shell andtube exchangers and air fin coolers.

Remote field testing operates atrelatively low frequencies. Typicalfrequencies are in the range of 40 to500 Hz.

The test speed for carbon steel tubes isabout 150 mm·s–1 (30 ft·min–1). Atwo-person crew can examine from 200 to500 tubes measuring 9 m (30 ft) long inan 8 h shift, depending on access, setuptime, number of discontinuitiesencountered and other factors.

Remote field testing is a noncontacttechnique, so the probes have minimalfriction with the pipe wall and require nocouplant.

SensitivityThe accuracy for remote field testing inthe straight part of the tubes is about10 percent of wall thickness for generalwall loss. The accuracy is generally less(20 percent of wall) for highly localizeddiscontinuities and in bends or nearexternal conducting objects because of thechanges in magnetic properties of thetube in the bend area and because ofshielding effects of external objects.

TABLE 1. Techniques used for nondestructive t

Electrom_____________________Eddy Current _____________________

Characteristic Conventional Satura

Skilled technicians required yes yesSpeed (m·s–1) 0.9 to 2.0 0.61Wall loss identifieda yes yesMinimum probe fill factor 0.8 0.7Ferromagnetic tube test no sligh

a. Inside diameter versus outside diameter.b. Ultrasonic testing requires seals to center the probe and

Remote field testing is also equallysensitive to inside and outside surfacediscontinuities but usually cannotdiscriminate between them without thehelp of near field coils. Remote fieldtesting is relatively insensitive to scale andmagnetic debris.

A large fill factor is not required forremote field testing and centralization isnot critical (as it is with ultrasonic, eddycurrent and flux leakage testing).However, a small fill factor will result indecreased sensitivity to smalldiscontinuities.

Other Test TechniquesTo fully appreciate the strengths andweaknesses of remote field testing, it isuseful to compare it to other techniquesused in industry. Included here are eddycurrent testing, saturation eddy currenttesting and magnetic flux leakage testing.

Although eddy current testing ispredominantly used for the testing ofnonferromagnetic tubing, it can be usedto test slightly ferromagnetic materialssuch as nickel copper alloy by using itwith a direct current biasing field strongenough to magnetically saturate the pipewall. This saturation eddy current techniqueis unsuited for thick walled, highlymagnetic material, such as carbon steel;however, various methods have beenfound useful for carbon steel tubes andpipes. These include the magnetic fluxleakage technique and ultrasonic testing.Table 1 provides a quick comparison ofdifferent nondestructive test methods forheat exchanger testing.


esting of heat exchanger tubes.

agnetic Techniques________________________________Techniques Magnetic Ultrasonic__________________tion Remote Field Flux Leakage Testing

yes yes yes0.61 0.61 0.04no yes yes0.7 0.8 noteb

t yes yes yes

retain fluid coupling.

Eddy Current Testing versusRemote Field TestingIn typical eddy current testinginstruments, the impedance of theinspection coil is measured. Usually thecoil is part of a bridge circuit thatbecomes unbalanced as the coil passesover a change in material thickness,permeability or conductivity.Discontinuities are characterized and sizedby the phase rotation and attenuation ofthe signal as compared to a referencestandard. The test coil in eddy currenttesting can be an energized coil or it canbe a passive coil that receives its energyfrom a separate energized coil in closeproximity (send and receiveconfiguration). Common coilconfigurations are absolute or differentialcoils; axial or radial coils; and bobbin orpancake coils.

Remote field testing has manysimilarities to eddy current testing butthere are also major differences.

1. In remote field testing, the exciter coilis always separated from the receivercoil or coils by at least two tubediameters. As such, remote field testcoils are always in a send and receiveconfiguration.

2. In remote field testing, the energyfrom the exciter coil passes throughthe tube wall twice, once whenleaving the exciter and again whenpassing back through the wall at thedetector.

3. The sensitivity to discontinuities onthe outside of a tube is reduced in theeddy current technique whereasremote field testing maintains almostequal sensitivity to discontinuitieseither inside or outside the tube.

4. Remote field test systems measure thephase and amplitude of a signal. Eddycurrent test systems may measure thesame quantities in send and receiveconfigurations or may measure theimpedance of the test coil.

5. Eddy current technique probes aresensitive to changes in the proximityof the test coil to the tube surface.This change is known as probe wobble— as the probe passes through thetube, it can be pushed to one side ofcenter by internal scale or dents. Evenif the tube is clean, the eddy currenttesting probe can wobble unless it iscentered with mechanical guides.

6. Remote field probes are relativelyinsensitive to probe wobble and areforgiving if the probe is undersized orpushed to one side of the tube.


7. Because of the much lower testfrequencies used for remote fieldtesting in steel (and because themeasurement of phase usually requiresat least one time period of theexcitation signal), remote field probesmust be moved more slowly thaneddy current probes. The remote fieldprobe must be near the smallestdiscontinuity required to be detected for atleast one cycle in order to detect thediscontinuity.

8. Absolute coils for both remote fieldand conventional eddy currenttechniques are both sensitive totemperature variations over the lengthof the tube.

9. In steel tubes, remote field testing ismore sensitive to circumferentialcracks that interrupt the lines ofmagnetic flux. Eddy current bobbinprobes are more sensitive to axialcracks in tubes, which interrupt theeddy currents.

10. Because of its so calledthrough-transmission nature, remotefield testing can examine thickermaterials than eddy currenttechniques can.

11. Because it is commonly used for steeland cast iron, remote field testing isgenerally carried out at a lowerfrequency than eddy currenttechniques are.

Magnetic Flux Leakage versusRemote Field Testing With a magnetic flux leakage probe, theenergy source is an axially alignedmagnetic field (as with a remote fieldtesting probe). Discontinuities aredetected by the magnetic flux leakageprobe as some of the magnetic flux linesleak out of the pipe and are detected bypassive coils or sensors passing though theleakage field. This arrangement is similarto that for remote field testing except thatthe remote field test probe generates analternating current field whereas magneticflux leakage testing uses a direct currentfield. When pickup coils are used, theremote field testing probe itself does notneed to be moving to measure the wallthickness, because the excitation isalternating.

The remote field eddy current signal islikely to offer additional informationbecause both phase and amplitude of thesignal can be analyzed. In contrast, onlythe amplitude of the magnetic fluxleakage signal is available. Consequently,remote field testing provides two pieces ofinformation, usually enough to permitcalculation of tube wall thickness.

PART 5. Signal Analysis

Data Presentation

Strip ChartRemote field test data are recorded incomputer memory or hard drive andphase amplitude diagrams (voltage planes)are displayed on instrument monitors innear real time as the test progresses. Theraw data from the detector are storedeither in phase amplitude format or asin-phase and quadrature components. Thedata can be recalled for display, analysisand reporting purposes after the testprocess is completed.

A strip chart displays coordinates fromthe phase amplitude diagram (forexample, an x,y display, a phase display or

FIGURE 9. Strip chart recordings: (a) phase and log amplitudesignals for absolute probe; (b) x,y voltage signals;(c) differential signals; (d) mixed frequency signals.

(a) (b) (c) (d)

a log amplitude display) as a functioneither of time or of the axial distancealong the length of the tube. In the stripchart of Fig. 9, which shows phase andlog amplitude for an absolute coil,deflections to the left represent metal lossand, on the right, wall thickening as inthe case of tight fitting support plates orbaffle plates. Phase and log amplitude arethe preferred quantities for the absolutecoil strip chart display because they areboth linear indicators of overall wallthickness (as opposed to the in-phase andquadrature components, which make amore suitable display for differentialcoils). To display phase and amplitude onthe same strip chart, the amplitude ismultiplied by the factor for conversionfrom radian to degree (180·π–1).

Complex Plane DisplaysThe voltage plane and x,y displays providemaps of the detector coil output in polarcoordinates (Figs. 10 and 11). On polar


FIGURE 10. Phase amplitude diagram: voltage planerecording.

Reference curve of thickness to amplitude

Zerovoltage

point

0,0Nominalreference point

1,0

FIGURE 12. Strip charts and phase amplitude diagrams showindications from short discontinuity.

displays, signals are drawn as vectorpoints with the angle representing thephase and the radius representing theamplitude. Remote field testing signals onthe voltage plane or x,y display are scaledand rotated to a convenient position forviewing. This manipulation makes iteasier to measure and recognize deviationsfrom the nominal position. With absolutedetectors the signals can be scaled androtated about the origin to place thenominal signal at (1,0) in rectangularcoordinates. In other words the signal fornominal tube is placed at 1 V normalizedat zero degrees. With differential detectorsthe signal is often shifted to place it at theorigin (0,0) even though the signal isusually not actually zero.

Besides the detector trace, the voltageplane has a number of static components:the origin, the exponential skin depthreference curve and the X and Y axes. Thereference curve is a feature unique toremote field testing and is very helpfulwhen identifying and sizing anomalies.The curve starts at 0,0 (that is, zerovoltage at origin) and follows anexponential spiral path as the overall wallthickness of a tube is decreased. It istheoretically possible to place thicknessvalues for full circumferentialdiscontinuities directly on this curve;however, the values would be different forshort discontinuities versus longdiscontinuities, as explained below.Although axial distance information isnot displayed on a voltage plane, itremains a very powerful tool for sizingdiscontinuities. Strip charts are useful


FIGURE 11. Phase amplitude diagram: x,y plane recording.

X axis

Y axis

Signal anglewith respect

to X axis

Zero voltage point

0,0

x,y display

because they usually show distanceinformation.

As mentioned before, the strip chartrecords what the detector coils senses asthe probe is pulled through the tubebeing tested whereas the voltage planeshows selected discontinuity signatures atany point along the strip chart. As aresult, the strip chart displays and x,yvoltage plane displays are often put sideby side. This presentation lets thetechnician use the complex planes toidentify and size discontinuities whilesimultaneously using the strip chart tofind the axial location of thediscontinuity. Combining both displaysaccelerates discontinuity reporting.

Signatures of Short andLong DiscontinuitiesIn the remote field technique,discontinuity indications are recognizedby their shape and size. As each of theexciter and detector coils in the basicprobe passes a discontinuity, there is achange in the voltage of the receiver coil.Therefore, if the discontinuity is shorterthan the distance from exciter to detector,there will be two distinct signals — one asthe exciter passes the discontinuity and asecond one as the detector passes it(Fig. 12). When the physical sizes of theexciter and detector coils are about equal,the two strip chart responses will be

Absolute signals Differential signals

Phase amplitudeand log signals

Differentialsignal

Voltage plane display

Phase amplitude diagrams

x,y plane display

FIGURE 13. Voltage plane signals from long discontinuityshow direction of wall thickness changes.

Absolute signals Differential signals

Strip chart signals Phase amplitude diagrams

comparable. However, in some cases, theexciter coil design is optimized forgenerating the electromagnetic fieldwhereas the detector dimensions areoptimized for resolution and sensitivity.As a result, the exciter coil may be longerthan the detector.

The voltage plane trace in Fig. 12corresponds to the gated area in the stripcharts. The double response is alsoexhibited on the voltage plane by twodistinct traces moving from the normalwall point (0,0 for differential and 1,0 forabsolute coils) vertically to theexponential skin depth spiral. The factthat the voltage plane traces terminate onthe reference spiral shows that thediscontinuity must extend fully aroundthe entire circumference of the pipe. Theangle of the traces with respect to thehorizontal axis permits calculation of thedepth of the discontinuity.

The X,Y display in Fig. 12 ischaracteristic of a differential detectorsetup. A differential detector coil consistsof two identical coils displaced axiallyfrom each other and wound in oppositedirections. Differential coils are excellentedge detectors and can neutralize to someextent the double response characteristicof remote field testing. As such,differential detectors improve remote fieldtesting’s sensitivity to smalldiscontinuities. Differential detectors areless sensitive to tapered or smooth wallloss. The origin for the differential coiltrace is chosen at 0,0 because undernominal pipe wall the two halves of thedifferential detector cancel each other out.

As soon as one half moves underneatha discontinuity, the detector circuit willbecome unbalanced and a resultant signalis measurable. When the second half ofthe detector moves underneath the samediscontinuity, it will produce a similartrace in the opposite direction. The twoopposite responses from a differentialdetector as it moves underneath a pipeanomaly is very characteristic and is oftenreferred to as a differential kick. Thedifferential kicks are clearly visible on thesecond strip chart in Fig. 12. The angle ofthe differential kick on the phaseamplitude diagram can also be used toestimate the depth of the discontinuity.

Eddy current test technicians will oftenprefer to rotate the differential signal froma through hole to 40 degrees so that itlooks like a familiar signal on an eddycurrent impedance plane. However, this iswhere the similarity ends. In remote fieldtesting, signals from increasing wall loss(outside diameter or inside diameter)always rotate counter clockwise. Eddycurrent test signals from inside diameterdiscontinuities rotate clockwise andoutside diameter discontinuities rotate

counter clockwise as they increase indepth.

With the absolute coil, if thediscontinuity is longer than the spacingfrom exciter to detector, the signal phaseand the log of the signal amplitude willboth double in size, resulting in a headand shoulders signal (Fig. 13). This signal isthe result of the overlap of the exciter anddetector responses. The correspondingtrace on the voltage plane is called a dogleg because of the bend in the indicationwith machined discontinuities. Noticethat the discontinuity is much harder tocharacterize with differential detectorsbecause of its long and gradual nature.

Voltage Plane Polar PlotDisplayIn remote field testing, the materialvolumes at the exciter and again at thedetector are both interrogatedsimultaneously. As a result, the sensitivityto discontinuities located on the inside ofthe tube is about the same as thesensitivity to external discontinuities;however, there is no way to tell themapart unless a separate, high frequencycoil is added in the direct field (whichwould be sensitive mainly to internaldiscontinuities). On the voltage planepolar plot display, thickness decreases(due to metal loss) rotate the signalcounter clockwise either along or insidethe reference curve (see Figs. 14 and 15).

The shape and orientation of adiscontinuity will affect the remote fieldtesting indication. Two smalldiscontinuities with the same amount of


FIGURE 15. Voltage plane signals indicate discontinuity depthand discontinuity volume.

volume loss but different depths willshow different indications. In general,deeper discontinuities will show up aslarger, more pronounced signals unlessthey are extremely small in volume.

The orientation of the discontinuity isalso of importance. From eddy currenttesting, it is intuitively expected thatdiscontinuities aligned in the axialdirection show larger responses thandiscontinuities oriented circumferentially:an axially aligned discontinuity interferesmore with the circumferentially flowingeddy currents. In remote field testing, theopposite is true: discontinuities orientedin the circumferential direction are morepronounced.

This difference exists because themagnetic field interaction exhibited byremote field testing makes it behavesomewhat similar to magnetic fluxleakage testing. A circumferentialdiscontinuity forms a large interruptionfor the axially aligned magnetic fluxwhereas a thin axial discontinuity barelyinfluences the flux. The disturbance in themagnetic field caused by thecircumferential discontinuity results in alarger remote field test signal response,which can then be used to estimate thecircumferential extent of thediscontinuity.

If the tube wall is locally thinned onone side of the tube, a line can be drawnfrom the nominal point through thesignal tip; the line will point toward thespot on the reference curve thatrepresents the same reduced wallthickness if the thinning were all aroundthe circumference. By measuring thephase angle of the signal the remainingwall thickness can be estimated. By


FIGURE 14. Voltage plane indication of longdiscontinuity, made with absolute anddifferential coils. Overall thickness of tubedecreases evenly from point A to point B.

Thickness decrease

Thickness increase

A

B

Full wall is normalizedat point A, wherex = 1 and y = 0.

Qua

drat

ure

com

pon

ent

(rel

ativ

e sc

ale)

In-phase component(relative scale)

calculating the ratio of the signal heightfrom a one-sided discontinuity to a signalfrom a circumferentially thinned tubewall, the circumferential extent can alsobe calculated. Phase angle, wall thicknessand circumferential extent can becritically important when characterizingdiscontinuities.

By observing the signal shape, phaseangle and relative size on the voltageplane many discontinuities can becharacterized and sized for depth andcircumferential extent. The axial length ofa discontinuity can also be measured byrecording the data on a strip chart as theprobe is pulled through the tube. Virtuallyall remote field test instruments displaythe data as strip charts and voltage planes.Many instruments have automatic depthsizing and reporting software.

Common Signatures inRemote Field TestingSignals from tubes can be categorized asfollows: (1) general wall loss (with longlength and extensive circumferentialextent), (2) long one-sided discontinuities,(3) short circumferential discontinuitiesand (4) small volume, localdiscontinuities.

General Wall LossDiscontinuities that are longer than theprobe are classified as longdiscontinuities. These discontinuities, if

1

4

Legend1. Increasing depth.2. Increasing depth and increasing volume.3. Increasing depth.4. Increasing volume.

2

3

Qua

drat

ure

com

pon

ent

(rel

ativ

e sc

ale)

In-phase component(relative scale)

they affect the entire circumference of thetube evenly, will produce signals on thevoltage plane that closely follow thereference curve. Technicians identifydiscontinuity length from the strip chartrecording and decide if the discontinuityis long or short. On the strip chart, thephase and log amplitude signals will trackin synchronization with each other if thediscontinuity is perfectly circumferential.

The phase and log amplitude signals oflong discontinuities are roughly doublethose from short discontinuities of similardepth, so it is important that thetechnician recognize the characteristics ofeach signal type. Figure 16 shows generalwall loss as well as signals from shortdiscontinuities.

Long One-Sided DiscontinuitiesDiscontinuities that occur on one side ofthe tube and are longer than the distancefrom the exciter coil to the detector coilare classified as long one-sideddiscontinuities. Such discontinuities willproduce signals inside the reference curve(that is, left of the curve) that may followa track that is parallel to the curve andthat may have an included loop that doesnot return all the way to the nominalpoint at 1,0. These discontinuities oftentaper at each end back to nominalthickness. On the strip chart recording,the phase and log amplitude signals ofone-sided discontinuities do not track

FIGURE 16. Examples of general wall loss plus local pitting(chelant corrosion, inlet erosion and general pittingcorrosion).

Phasetrace

Log amplitude trace

Signals from smallvolume, localdiscontinuities

Tube supportplate signal

General wall losssignals followreference curve

Strip chart signals Phase amplitude diagram

together. The phase trace will generallydeviate further from the baseline positionthan the log amplitude trace. Examplesare steam impingement erosion and tubeto tube fretting (midspan erosion).Carbon dioxide corrosion can also be longand one-sided.

Short CircumferentialDiscontinuitiesDiscontinuities that are shorter than theprobe (that is, shorter than the distancefrom exciter to detector coil) and are all ofthe way around the tube are classified asshort circumferential discontinuities(Fig. 17). These discontinuities produce atypical double signal on both voltageplane and strip chart. The phase and logamplitude strip chart traces will tracksynchronously if the discontinuity isperfectly circumferential. Examples arebaffle wear and condensate grooving.

Small Volume, LocalDiscontinuitiesDiscontinuities that are primarily on oneside of the tube and have limited axialand circumferential length are classified assmall volume, local discontinuities.Examples are pitting and cracking. Thesediscontinuities produce signals that arevery small in amplitude and are left of thereference curve on the voltage plane. Thesignals are sharp and will repeat as each


FIGURE 17. Short, circular discontinuity caused by condensategrooving next to support plate.

Nominal tube support signal

Support signal withcondensate groove

on each side

Strip chart signals Phase amplitude diagram

coil passes the discontinuity. Differentialdetector coils are preferred for detectionof small volume discontinuities becausethey produce a signal twice as large asthat from an absolute probe.

Reference StandardsAll nondestructive test methods usereference standards to comparediscontinuity signals with those fromknown machined discontinuities. Remotefield testing is no different: its referencestandards are tubes with artificialdiscontinuities for calibration. However,remote field testing requires referencestandards for each variation in tubediameter, wall thickness (tube gage),conductivity and permeability. In eachreference standard, referencediscontinuities must be machined toclosely simulate the discontinuitiesexpected in the tube or pipe beingexamined. ASTM E 2096-00 mentions twopossible reference tube styles.11 Customersmay specify different discontinuity typesif they expect to encounter discontinuitiesother than the suggested types.

When ordering the manufacture of areference tube it is important to providethe following specifications.

1. Specify the tube material.2. Specify the tube diameter and gage.3. Specify the tube manufacturing

technique, whether seamless orelectric resistance welded.

4. Describe any heat treating that thetubes to be tested have undergone.

5. Space the discontinuities at least fourtube diameters apart and at least fourtube diameters from each end.

6. Require the reference discontinuitiesto be machined with a series of smallcuts using sharp machine tools andcopious coolant. This machiningprevents local heating that can changethe permeability and conductivity ofthe tube material.

7. Specify the tolerances on thediscontinuity depths and ask for themto be machined according to theactual, measured, tube wall — not thenominal wall thickness that the tubeis specified to be. Tubes aremanufactured to tolerances of, forexample, +12 percent and –10 percent.

8. Specify that all edges of discontinuitiesare to be radiused to avoid edgesignals.

9. Specify how the tube is to beidentified with a permanent label,listing the discontinuity depths,reference serial number, material andwall thickness.

10. Specify whether the tube is to beprotected with a coat of paint.


In an ideal world the reference tubewould be made from the same material asthose in the tube bundle to be tested;however, this is usually not possible.Differences between the reference tubeand the tubes tested will introduceinaccuracies in data analysis. Consider thefact that the tube wall thickness can varyby +12 percent to –10 percent (seamlesstubes can vary by this much from oneside of the tube to the other). Thechemical makeup of the tube can havesimilar tolerances. Add to this themachining tolerances; differences in heattreatment and the magnetic history of thereference tube and the tube to beexamined and these combined parameterscan lead to substantial differences whencomparing the signals from the two tubes.An informed inspector can compensatefor most of these differences if thatinspector can identify them.

Effects of Probe SpeedThe ability of remote field testing todetect and quantify discontinuities relieson the quality of the signal received. Theprobe must be in the vicinity of thediscontinuity for at least one excitationcycle in order to detect it. Ideally theprobe senses a discontinuity for severalcycles so that any noise signals can beaveraged out of the data.

If a test frequency of 100 Hz is beingused, there are 100 opportunities persecond to measure the signal. If thesmallest discontinuity that is required tobe detected has a length of 3 mm(0.12 in.), then the fastest speed possibleis 300 mm·s–1 (12 in.·s–1). To build in asafety margin, the test speed should be nofaster than half this speed. The probe pullspeed should be slow enough so that thedigital sample rate allows the field profilenear the probe to be accurately recorded.

It is equally important that the speedof testing be as constant as possible.Sudden changes in speed can result inanomalous signals.

Tube Support PlatesTube support plates are normally made ofsteel and are common in heat exchangers.These plates have the effect of absorbingmagnetic energy from the probe. In thevicinity of a tube support plate, theremote field test signal tends toconcentrate in the plate, creating amomentary increase in signal amplitudefollowed by a substantial decrease whenthe exciter and detector coils are onopposite sides of the plate. Luckily, ifthere is any wall loss caused by corrosionor erosion at, or under, the tube support

FIGURE 19. Remote field testing signals at tube support plate:before and after frequency mixing.

Normal tubesupport plate

signal

Flaw next to tube support plate

Before frequency mixing After frequency mixing

plate, there is usually a residual remotefield test signal that can be used to detectthe loss.

Figure 18 shows a remote field testsupport plate response on the voltageplane. On the voltage plane, the supportplate causes a characteristic whale shapedsignal. The nose of the whale correspondsto the increase in amplitude just beforethe detector or exciter goes beneath thesupport plate. This increased amplitude iscaused by the electromagnetic wavetaking the path of least reluctance (that is,through the support plate), thus causing alocal increase in the field intensity.12 Afterthe energy has diffused through the airgap, it suffers heavy attenuation in themetal. If the tube support plate is thickand in tight contact with the tube, theremay be no measurable signal left until theprobe has completely passed the plate.The tail of the whale corresponds to thesituation when transmitter and receiverare located on opposite sides of thesupport plate. The tail is shorter when thegap between pipe and plate is larger orwhen the plate is relatively thin. The noseof the whale usually occurs at around 1 to5 mm (0.04 to 0.2 in.) from the edge ofthe support plate.

Support plates signatures aresometimes modeled for calibrationpurposes by metal rings with the samethickness as the desired support plate. Therings can be too small, causing the

FIGURE 18. Typical tube support plate signal on remote fieldtesting voltage plane display.

Qua

drat

ure

com

pon

ent

(nor

mal

ized

)

1.0

0.5

0

–0.5

Skin depth attenuation spiral

A

BCD

In-phase component (normalized)

LegendA. Both transmitter and receiver are located under nominal pipe wall.B. Amplitude increases as receiver approaches plate.C. Amplitude decreases rapidly as receiver moves beneath plate.D. Receiver and transmitter are on different sides of plate.

–0.25 0 0.25 0.50 0.75 1.00 1.25 1.50

electromagnetic wave to wrap itselfaround the outside of the ring rather thansqueezing through the gap between tubeand ring. Proper calibration rings shouldhave diameters of at least 140 mm(5.5 in.).

Mixed SignalsAn effective way to cancel the tubesupport plate signal is to mix the signalsfrom two frequencies in order to enhancedetection of discontinuities under or nearthe plate. Figure 19 shows signals fromdiscontinuities near a tube support plateboth before and after mixing. In remotefield testing, the tube support plateindication is only suppressed and notremoved. There is still less accuracy andsensitivity near a tube support plate.


PART 6. Field Operation

FIGURE 20. Array probe in use on 75 mm (3 in.) pipeline.

Test Considerations12-14

The purpose of a typical remote field testis to determine the depth and type ofmetal loss in each tube. A typical boilerinspection will have the followingconsiderations and steps.

1. A technician may decide that cleaningis required to allow passage of theprobe. In most cases, boiler tubes donot require cleaning before a remotefield test.

2. The boiler operator generally providesdrawings of the boiler showing thelocations of soot blowers, accessopenings, test ports and other areas ofinterest. A tube numbering system isestablished.

3. If the tubes are 6 m (20 ft) or less inlength, the examination can beconducted from the mud drum. If thetubes are longer, it is generallyadvisable to remove the steam druminternals and do the examinationfrom there.

4. It may be decided to test all the tubesor only selected tubes in suspectedtrouble areas. If damage on the bendsis suspected, it is necessary to examinemany tubes in the same row so thatbend signatures can be compared.

5. During data acquisition, an assistanthandles the probe while thetechnician operates the instrument.All data are stored.

6. When the data are being analyzed, thetechnician pays special attention tothe location of soot blowers, to thedirection of gas flow on the hot side,to types of water treatment and to thehistory of previous failures. Knowledgeof these factors helps analysis bygiving insight into types ofdiscontinuities that might be found.

7. A field report is written on site at thecompletion of the job. In this way, theboiler operator can make immediatedecisions concerning tube plugging,repairs or replacement. A computerreport can be created that may includea color coded tube sheet map of theboiler.


Special ProbesTo improve sensitivity of the remote fieldtechnique to discontinuities close to atube support plate or tube sheet, probescan contain detector coils on each side ofan exciter coil. Alternatively, two exciterscan be used with one or two detectorsplaced between them. The disadvantageof these probes is their increased lengthand more complex data.

Pipes with diameters larger than25 mm (1 in.) often require more thanone detector coil to improve detection oflocal discontinuities. Array coils are usedto segment the detector section toimprove small volume discontinuitysizing. The probe in Fig. 20 is being usedto examine a 75 mm (3 in.) diameterpipeline used to transport gas from thewellhead. Figure 21 shows the displaynarrowed to array probe signals ofinterest.

External probes are similar in design toremote field testing probes; however,because external probes are sensitive toliftoff, probe travel must be carefullycontrolled to eliminate false signals.

Array probes require special software todisplay and analyze the large quantities ofmultichannel data. It is possible forsoftware to analyze remote field test dataautomatically. In all cases of automaticdata analysis the success of the analysis

FIGURE 21. Array probe signals.

Signals from 14circumferential spot

sensors

Mappedindications

Channels ofinterest

(channels 3 to 6)

Phase amplitudediagram of

channels 3 to 6

depends on the criteria that the operatoruses to train the software to recognizesignals of interest. It is also veryimportant to have a skilled technicianreview a percentage of the calls made byautomatic analysis software to verifyaccuracy. Only one selected channel at atime can be displayed on a voltage planedisplay.

SummaryRemote field testing is a versatile testtechnique that can be used effectively totest steel tubes, pipes and plates. It canalso be used to test thick walled,nonferrous tubes with equal sensitivity tointernal and external discontinuities.


1. MacLean, W.R. Apparatus forMagnetically Measuring Thickness ofFerrous Pipe. United States Patent2 573 799 (1951).

2. Schmidt, T.R. “History of theRemote-Field Eddy Current InspectionTechnique.” Materials Evaluation.Vol. 47, No. 1. Columbus, OH:American Society for NondestructiveTesting (January 1989): p 14, 17-18,20-22.

3. Atherton, D.L. and W.M. Czura.“Finite Element Calculations for EddyCurrent Interactions with CollinearSlots.” Materials Evaluation. Vol. 52,No. 1. Columbus, OH: AmericanSociety for Nondestructive Testing(January 1994): p 96-100.

4. Hoshikawa, H., K. Koyama, J. Koidoand Y. Ishibashi. “Characteristics ofRemote-Field Eddy CurrentTechnique.” Materials Evaluation.Vol. 47, No. 1. Columbus, OH:American Society for NondestructiveTesting (January 1989): p 93-97.

5. Schmidt, T.R. “The Remote Field EddyCurrent Inspection Technique.”Materials Evaluation. Vol. 42, No. 2.Columbus, OH: American Society forNondestructive Testing (February1984): p 225-230.

6. Lord, W., Y.-S. Sun, S.S. Udpa andS. Nath. “A Finite Element Study ofthe Remote-Field Eddy CurrentPhenomenon.” IEEE Transactions onMagnetics. Vol. 24. New York, NY:Institute of Electrical and ElectronicsEngineers (January 1988): p 435-438.

7. Mackintosh, D.D., D.L. Atherton andP.A. Puhach. “Through-TransmissionEquations for Remote-Field EddyCurrent Inspection of Small-BoreFerromagnetic Tubes.” MaterialsEvaluation. Vol. 51, No. 6. Columbus,OH: American Society forNondestructive Testing (June 1993):p 744-748.

8. Sun, Y.-S., L. Udpa, S. Udpa, W. Lord,S. Nath, S.K. Lua and K.H. Ng.“A Novel Remote-Field Eddy CurrentTechnique for Inspection of ThickWalled Aluminum Plates.” MaterialsEvaluation. Vol. 56, No. 1. Columbus,OH: American Society forNondestructive Testing (January 1998):p 94-97.

9. Kilgore, R.J. and S. Ramachandran.“Remote Field Eddy Current Testing ofSmall-Diameter Carbon Steel Tubes.”Materials Evaluation. Vol. 47, No. 1.Columbus, OH: American Society forNondestructive Testing (January 1989):p 32-36.

10. Atherton, D.L., D.D. Macintosh,S.P. Sullivan, J.M.S. Dubois andT.R. Schmidt. “Remote Field EddyCurrent Signal Representation.”Materials Evaluation. Vol. 51, No. 7.Columbus, OH: American Society forNondestructive Testing (July 1993):p 782-789.

11. ASTM E 2096-00, Standard Practice forIn Situ Examination of FerromagneticHeat-Exchanger Tubes Using Remote FieldTesting. West Conshohocken, PA:ASTM International (2000).

12. Shatat, A. and D.L. Atherton. “RemoteField Eddy Current Inspection ofSupport Plate Fretting Wear.” MaterialsEvaluation. Vol. 55, No. 3. Columbus,OH: American Society forNondestructive Testing (March 1997):p 361-366.

13. Smith, H. and D.D. Mackintosh.“Remote Field Eddy CurrentExamination of Boiler Tubes.”Proceedings of EPRI Workshop:Electromagnetic NDE Applications in theElectric Power Industry [Charlotte, NC,August 1995]. Palo Alto, CA: ElectricPower Research Institute (1995).

14. Mackintosh, D.D., D.L. Atherton,T.R. Schmidt and D.E. Russell.“Remote Field Eddy Current forExamination of Ferromagnetic Tubes.”Materials Evaluation. Vol. 54, No. 6.Columbus, OH: American Society forNondestructive Testing (March 1996):p 652-657.


References

remote field testing

Documents

byremote field testing

historythe remote field

remotefield testing

value of remote fieldtesting

years remote fieldtesting

themagnetic field interactions

russell nde systems

technique bylicensing