Electromagnetic Testing-ASNT Level III Study Guide ECT

Electromagnetic Testing

Study Guide Eddy Current Testing Revisited My ASNT Level III

Pre-Exam Preparatory Self Study Notes

http://independent.academia.edu/CharlieChong1 http://www.yumpu.com/zh/browse/user/charliechong http://issuu.com/charlieccchong

Fion Zhang at Shanghai

26th April 2015

CHAPTER 1

HISTORICAL BACKGROUND

Belore discussing the principles of eddy current testing, it seems appropriate to discuss brielly facets of magnetism and electromagnetism that serve as the foundation of our study of eddy current testing. In the period from 1775 to

1900, scientific experimenters Coulomb, A Ampere, Faraday, Oersted, Arago, Maxwell, and Kelvin investigated and cataloged most of what is known about magnetism and electromagnetism

Arago discovered that the oscillation of a magnet was rapidly damped when a nonmagnetic conducting disk was placed near the magnet (Figure 1.1).

He also observed that by rotating the disk, the magnet was attracted to the disk. In effect, Arago had introduced a varying magnetic field to the disk

causing eddy currents to allow in the disk producing a magnetic field by the disk that attracted the magnet. Arago's simple model is a basis lor many automobile speedometers used today.

Oersted discovered the presence of a magnetic field around a current-carrying conductor, and he observed a magnetic field developed in a perpendicular plane to the direction of current flow in a wire. Ampere

observed that equal and opposite currents Ilowing in adjacent conductors cancelled this magnetic effect. Ampere's observation is used in differential coil applications and to manufacture noninductive, precision resistors.

Faraday's first experiments investigated induced currents by the relative motion of magnet and a coil (Fig. 1.2)

Faraday's major contribution was the discovery of electromagnetic induction. His work can be summarized by the example shown· in Figure 1.3. Coil A is connected to a battery through a switch S. A second coil a connected to a galvanometer G is nearby. When switch S is closed producing a current in coil A in the direction shown, a momentary current is induced in coil a in a direction(- a) opposite to that in A. If S is now opened, a momeritary current will appear in coil a having the direction of (- b). In each case, current flows in coil a only while the current in coil A is changing.

FARADAY LAW

The electromotive force (voltage) induced in coil a of Figure 1.3 can be expressed as follows:

E = K ∙ N ∙ ∆Ф/∆t

E = Average induced voltage

N = Number of turns of wire in coil B

∆Ф/∆t = Rate of change of magnetic lines of force affecting coil B K = 10-8 _constant

Maxwell produced a two-volume work "A Treatise on Electricity and

Magnetism" first published in 1873, Maxwell not only chronicled most of the work done in electricity and magnetism at that time, but he also developed and published a group of relations known as Maxwell's equations for the elec tromagnetic field. These equations form the base that mathematically

describes most of what is known about electromagnetism today. In 1849 Lord Kelvin applied Bessel.'s equation to solve the elements of an electromagnetic field. The principles of eddy current testing depend on the process of

electromagnetic induction. This process includes a test coil through which a varying or alternating current is passed. A varying current flowing in a test coil produces a varying electromagnetic field about the coil. This field is known as the primary field.

Faraday Law

Increasing current in a coil of wire will generate a counter emf which opposes the current.

Applying the voltage law allows us to see the effect of this emf on the circuit equation. The fact that the emf always opposes the change in current is an example of Lenz's law. The relation of this counter emf to the current is the origin of the concept of inductance. The inductance of a coil follows from Faraday's law.

Since the magnetic field of a solenoid is: B = μNI ∙ (l -1₎ Thus: E = - NA ∙∆B/ ∆t, becomes; E = - N A ∙∆ [μNI (l -1_{)] /} ∆t E = - NAμN ∙(l -1₎ ∙ ∆I/∆t for L = N2_Aμ (l -1₎ E ∝ ∆ Ф/ ∆t (Faraday Law) E = - N ∆Ф/ ∆t Ф = BA B = flux density

A = Area under the influence of B For a fixed area and changing current, Faraday's law becomes: E = - N ∆Ф/ ∆t = -N ∆BA/ ∆t

GENERATION OF EDDY CURRENTS

When an electrically conducting test object is placed in the primary field, an electrical current will be induced in the test object. This c urrent is known as the eddy current. Figure 1.4 is a simple model that illust rates the relations hips of primary and induced (eddy) curre nts. Conductor A represents a portion of a test coil. Conductor B represents a portion of a test object.

Following Lenz's law and indicating the instantaneous direction of primary current Ф_p, a primary field Ф_p is developed about Conductor A. When

Conductor B is brought into the influence of Ф_p, an eddy current l_E is induced in Conductor B. This electrical current lE produces an electromagnetic field Ф_E that opposes the primary electromagnetic field Ф_p. The magnitude of Ф_E is directly proportional to the magnitude of l_E. Characteristic changes in

Conductor B such as conductivity, permeability, or geometry will cause lE to change. When l_E varies, Ф_E also varies. Variations of Ф_E are reflected to

Conductor A by changes in Ф_p. These changes are detected and displayed on some type of readout mechanism that relates these variations to the

characteristic that is of interest. I_p = Primary current

I_E = Eddy current

Ф_p = Primary magnetic flux

FIELD INTENSITY

Ф

Figure 1.5 presents a schematic view of an excited test coil. The

electromagnetic field produced about the unloaded test coil in Figure 1.5 can be described as decreasing in intensity with distance from the coil and also varying across the coil's cross section. The electromagnetic field is most intense near the coil's surface.

Ф_p I_p

The field produced about this coil is directly proportional to the magnitude of applied current, rate of change of current or frequency, and the coil parameters. Coil

parameters include inductance, diameter, length, thickness, number of turns of wire, and core material. To better understand the principles under discussion, we must again look at the instantaneous relationships of current and magnetic flux. The

exciting current is supplied to the coil by an alternating current generator or oscillator. With a primary current lr flowing through the coil, a primary electromagnetic field Ф_p is produced about the coil. When this excited test coil is placed on a conducting test object, eddy currents l_E will be generated in that test object. Figure 1.6 illustrates this concept.

Note the direction of l_p, Ф_p, and the resultant eddy current lE. Although Figure 1.6 shows l_E by directional arrows on the surface of the test object, l_E extends into the test object some distance. Another important observation is that l_E is generated in the same plane in which the coil is wound. Figure 1.7

emphasizes this point with a loop coil surrounding a cylindrical test object (4).

A more precise method of describing the relationships of magnetic flux, voltage, and current is the phase vector diagram or phasor diagrams (4).

Figure 1.8-a. Phasor Diagram of Coil Voltage without Test Object

E = Coil Voltage

E_p = Primary Voltage

E_s = Secondary Voltage = 0 I = Excitation Current

Ф_p = Primary Magnetic Flux

Figure 1.8 shows the effects of a non-ferromagnetic test object on a test coil. Figure 1.8a shows an encircling coil and the resultant phasor diagram for the unloaded coil . The components of phasor diagram 1.8a are as follows:

E = Coil Voltage

E_p = Primary Voltage

E_s = Secondary Voltage = 0 I = Excitation Current

Ф_p = Primary Magnetic Flux

The current (I) and primary magnetic flux (Ф_P) are plotted in phase, and the primary voltage (E_P) is shown separated by 90 electrical degrees. Secondary magnetic flux Фs is plotted at zero because without a test object no

secondary flux exists. Figure 1.8b represents the action of placing a

non-erromagnetic test object into the test coil. The components of phasor diagram 1.8b for a loaded coil are as follows:

E = Coil Voltage

E_p = Primary Voltage E_s = Secondary Voltage E_T = Total Voltage

I = Excitation Current

Ф_p = Primary Magnetic Flux Ф_s = Secondary Magnetic Flux Ф_T = Total Magnetic Flux

Figure 1.8-a. b. Phasor Diagram of Coil Voltage with Test Object E_p + E_s = E_T E_p E_s E_measured = E_T Ф_S Secondary  magnetic flux Ф_T Ф_p Primary magnetic flux non-ferromagnet ic test object Excitation current I Ф_T∠ ≠90º E_T∠ ≠90º

Observing Figure 1.8b we can see by vectorial addition of E_p and E_s we arrive at a new coil voltage (E_T) for the loaded condition. The primary magnetic flux cflp and secondary magnetic flux ells are also combined by vectorial addition to arrive at a new magnetic flux (Ф_T) for the loaded coil. Notice that for the condition of the test object In the test coil, Ф_T is not in phase with the

excitation current I. Also observe that the included angle between the excitation current and the new coil voltage E_p is no longer 90 electrical

Figure 1.8-a. b. Phasor Diagram of Coil Voltage with Test Object E_p + E_s = E_T E_p E_s E_measured = E_T Ф_S Secondary  magnetic flux Ф_T Ф_p Primary magnetic flux non-ferromagnet ic test object Excitation current I

CURRENT DENSITY

The distribution of eddy currents in a test object varies exponentially. The current density in the test object is most dense near the test coil. This exponential current density follows the mathematical rules for a natural expo.nential decay curve (1/e) Usually a natural exponential curve is illustrated by a graph with the ordinate (Y axis) representing magnitude and the abscissa (X axis) representing time or distance.

A common point described on such a graph is the "knee" of the curve. The knee occurs at the 37 percent value on the ordinate axis. This 37 percent point, or

knee, is chosen because changes in X axis values produce significant changes in Y axis values from 100 percent to 37 percent, and below 37 percent changes in X axis values produce less significant changes in Y axis values. Applying this logic to eddy current testing, a term is developed to describe the relationship of current density in the test object. Consider the eddy current generated at the surface of the test object nearest the test coil to be 100 percent of the available current, the point in the test object thickness where this current is diminished to 37 percent is known as the standard depth of penetration (4). Figure 1.9 is a relative eddy

Figure 1.9 - Relative Eddy Current Density

0.37

Current Density at Depth “x”

The current density at any depth can be calculated as follows:

J

= J

e

Where:

J_x = Current density at depth x , amperes per square meter J_o = Current density at surface, amperes per square meter π = 3.1416

f = Frequency in hertz

μ = Magnetic permeability, henries per meter (H∙m-1₎ x = Depth from surface, meters

σ = Electric conductivity, mhos per meter (Siemens∙m-1_?)

The siemens (SI unit symbol: S) is the unit of electric conductance, electric susceptance and electric admittance in the International System of Units (SI). Conductance, susceptance, and admittance are the reciprocals of resistance, reactance, and impedance respectively; hence one siemens is equal to the reciprocal of one ohm, and is also

referred to as the mho. The 14th General Conference on Weights and Measures approved the addition of the siemens as a derived unit in 1971.In English, the same form siemens is used both for the singular and plural

MAGNETIC PERMEABILITY

Magnetic permeability μ is a combination of terms. For nonmagnetic materials:

μ= μ_o = 4π∙ 10-7 _H/m

For magnetic materials: μ = μ_r∙μ_o

Where:

μ_r = Relative permeability, henries per meter (H∙m-1₎

THE STANDARD DEPTH OF PENETRATION

δ

The standard depth of penetration can be calculated as follows:

δ = (πfμσ)

where:

δ = Standard depth of penetration, meters π = 3.1416

f = Frequency in. hertz

μ = Magnetic permeability, H/m

Exercise:

lt should be observed at this point that as frequency, conductivity, or

permeability is increased, the penetration of current into the test object will be decreased. We can use the graph in Figure 1.9 (p. 6) to demonstrate many eddy current characteristics. Using an example of a very thick block of

stainless steel being interrogated with a surface or probe coil operating at a test frequency of 100 kilohertz (kHz), we can determine the standard depth of penetration and observe current densities at other depths. Stainless steel

{300 Series) is non-ferromagnetic. Magnetic permeability μ is 4π∙ 10-7 _H/m

and the conductivity is 0.14∙107 _{mhos per meter for 300 Series stainless steel.}

δ = (πfμσ) -½

δ = (100 x 103 x π x 4 π x 10-7 _{x 0.14 x 10}7₎ -½ _m

Exercise:

Using 1.35 mm as depth “x” from surface a ratio of depth/depth of penetration would be 1. Referring to Figure 1.9, a depth/depth of penetration of 1

indicates a relative eddy current density of 0.37 or 37 percent. What is the relative eddy current density at 3 mm?

Depth “x” equals 3 mm and depth of penetration is 1.35 mm, therefore: 3/1.35 = 2.22δ

Current density = (1/e) 2.22 _{= 0.11 or 11%}

This ratio indicates a relative eddy current density of about 0.1 or 10 percent. With only 10 percent of the available current flowing at a depth of 3 mm,

detectability of variables such as conductivity, permeability, and

discontinuities would be very difficult to detect. The obvious solution for greater detectability at the 3 mm depth is to lower the test frequency. Frequency selection will be covered in detail later in this text.

PHASE/AMPLITUDE AND CURRENT/TIME RELATIONSHIPS

Figure 1.10 reveals another facet of the eddy current. Eddy currents are not generated at the same in stant in time throughout the part. Eddy currents require time to penetrate the test part. Phase and time are analogous; i.e., phase is an electrical term used to describe timing relationships of electrical waveforms.

Phase angle lagging

Figure 1.10 - Eddy Current

Phase Angle Radians Lagging

Phase is usually expressed in either degrees or radians. There are 2π

radians per 360 degrees. Each radian therefore is approximately 57 degrees. Using the surface current phase angle near the test coil as a reference, phase angle current deeper in the test object lags the surface current. The amount of phase lag is determined by:

β = x/δ = x(πfμσ) -½ _{in radian}

where β equals the phase angle lag in radians.

Figure 1.10 should be used as a relative indicator of phase lag. The exact phase relationship for a particular system may be different due to other variables, such as coil parameters and excitation methods. The amount of phase lag for a given part thickness is an important factor when considering resolution. Resolution is the ability to separate variables occurring in the test object; for example, distinguishing two

discontinuities occurring at different depths in the same test object. As an example, let us establish a standard depth of penetration at 1 mm in a 5 mm thick test object. Refer to Figure 1.10 and observe the phase lag of the current at one standard depth of

penetration. Where depth of interest (X) is 1 mm and depth of penetration (δ) is 1 mm, the X/ δ ratio is 1 and the current at depth X lags the surface current by 1 radian.

Projecting this examination, let us observe the phase lag for the entire part thickness. The standard depth of penetration is 1 mm, the part thickness is 5 mm; therefore, the ratio X/δ equals 5. This produces a phase lag of 5 radians or approximately 287 degrees for the part thickness. Having a measurement capability of 1 degree increments, the part thickness could be divided into 287 parts, each part representing 0.017 mm. That would .be considered excellent resolution. There is an obvious limitation. Refer to Figure 1.9 and observe the resultant relative current density with an X/δ ratio of 5. The relative current density is near 0. lt should become apparent that the frequency can be

adjusted to achieve optimum results for a particular variable. These and other variables will be discussed in Section 5 of this study guide.

CHAPTER 1

0.1-1 Generation of eddy currents depends on the principle of: A. wave guide theory.

B. electromagnetic induction.

C. magneto-restrictive forces. D. all of the above.

0.1-2 A secondary field is generated by the test object and is: A. equal and opposite to the primary field.

B. opposite to the primary field, but much smaller.

C. in the same plane as the coil is wound. D. in phase with the primary field.

0.1-3 When a non-ferromagnetic part is placed in the test coil, the coil's voltage:

A. increases.

B. remains constant because this is essential.

C. decreases.

0.1-4 Refer to Figure 1.8b (p. 5): If E_T was produced by the test object being stainless steel, what would the effect be if the test object were copper?

A. E_T would decrease and be at a different angle.

B. E_T would increase and be at a different angle.

C. Because both materials are non-ferromagnetic, no change occurs. D. None of the above.

0.1-5 Eddy currents generated in a test object flow: A. in the same plane as magnetic flux.

B. in the same plane as the coil is wound.

C. 90 degrees to the coil winding plane.

D. Eddy currents have no predictable direction.

0.1-6 The discovery of electromagnetic induction is credited to: A. Arago.

B. Oersted. C. Maxwell.

Figure 1.8-a. b. Phasor Diagram of Coil Voltage with Test Object E_p + E_s = E_T E_p E_s E_measured = E_T Ф_S Secondary  magnetic flux Ф_T Ф_p Primary magnetic flux non-ferromagnet ic test object Excitation current I Ф_T∠ ≠90º E_T∠ ≠90º

Discussion

Subject: Reason out on the following:

0.1-4 Refer to Figure 1.8b (p. 5): If ET was produced by the test object being stainless steel, what would the effect be if the test object were copper?

A. ET would decrease and be at a different angle.

B. ET would increase and be at a different angle.

C. Because both materials are non-ferromagnetic, no change occurs. D. None of the above.

0.1-7 A standard depth of penetration is defined as the point in a test object where the relative eddy current density is reduced to:

A. 25 percent.

B. 37 percent.

C. 50 percent. D. 100 percent.

0.1·8 Refer to Figure 1.9 (p. 6). If one standard depth of penetration was

established at 1 mm in an object 3 mm thick, what is the relative current density on the far surface?

A. 3

B. <0.1

c. 1/3

D. Indeterminate

0:1-9 Refer to Figure 1.10 (p. 8). Using the example in question 1.8, what is the phase difference between the near and far surfaces?

A. Far surface leads near surface by 57 º B. Far surface leads near surface by 171 º

0.1-10 Calculate the standard depth of penetration at 10 kHz in copper; σ = 5.7∙107 _{mhos per meter.}

A. 0.1 mm B. 0.02 mm

C. 0.66 mm

CHAPTER 2

Eddy Current

Eddy Current (EC) testing is based on electromagnetic induction. The technology can be used to detect flaws in conducting materials or to measure the distance between a sensor and a conducting material. The measurement does not require the tested object to be in direct contact

The principle

The basic principle behind standard EC testing involves placing a cylindrical coil, which is carrying an alternating current, close to the test piece. The current in the coil generates a changing magnetic field, which produces eddy currents in the test piece. Variations in the phase and magnitude of these eddy currents are monitored using a second coil (search coil) or by measuring changes to the current flowing in the primary coil (excitation coil).

Image

Variations in the electrical conductivity or magnetic

permeability of the test object or the presence of flaws will change the flow patterns of the eddy currents and there will be a corresponding change in the phase and amplitude of the measured current.

Applications

EC testing can be used to inspect physically complex

shapes and to detect small cracks on or near the surface of a test piece. The inspected surfaces need only minor

TEST COIL ARRANGEMENTS

Test coils can be categorized into three main mechanical groups: probe coils, bobbin coils, and encircling coils.

PROBE COILS

Surface coil, probe coil, flat coil, or pancake coil are all common terms used to describe the same test coil type. Probe coils provide a convenient method of examining the surface of a test object. Figure 2.1 illustrates a typical probe coil used for surface scanning.

Probe coils and probe coil forms can be shaped to fit particular geometries to solve complex inspection problems. As an example, probe coils fabricated in a pencil shape (pencil probe) are used to inspect threaded areas of mounting studs and nuts or serrated areas of turbine wheels and turbine blade

assemblies. Probe coils may be used where high resolution is required by

adding coil shielding. When using a high-resolution probe coil, the test object surface must be carefully scanned to assure complete inspection coverage. This careful scanning is very time consuming. For this reason, probe coil inspections of large test objects are usually limited to critical areas. Probe coils are used extensively in aircraft inspection for crack detection near

fasteners and fastener holes. In the case of fastener holes (bolt holes, rivet holes), the probe coil is spinning while being withdrawn at a uniform rate. This provides a helical scan of the hole using a "spinning probe" technique.

Thread Probe Coils Eddy Current Inspections on RPV bolts and in RPV flat bottom holes

Because of the size of the inspection objects and the inaccessibility of the inside thread a mechanised inspection is necessary. The especially developed bolt inspection tables for the inspection of the thread and shaft regions enable a secured inspection. By the outline guidance of the thread an optimum sensor position is guaranteed, whereas the inspection of the shaft region provides an automatic feed that ensures the complete inspection of the total shaft surface. In case of the flat bottom hole thread inspection an optimum sensor guidance is obtained by a motorised compulsory guidance in the thread. Path sensors allow a detailed eddy current and path record and a resulting well analysable presentation of the C-scan.

ENCIRCLING COILS

Encircling coil, OD coil, and feed-through coil are terms commonty used to describe a coil that surrounds the test object. Figure 2.2 illustrates a typical encircling coil.

Encircling coils are primarily used to inspect tubular and bar-shaped products. The tube or bar is fed through the coil (feed-through) at relatively high speed. The cross section of the test object within the test coil is simultaneously

interrogated. For this reason, circumferential orientation of discontinuities cannot be determined with an encircling coil. The volume of material

examined at one time is greater using an encircling coil than a probe coil;

therefore, the relative sensitivity is lower for an encircling coil. When using an encircling coil, it is important to keep the test object centered in the coil. If the test object is not centered, a uniform continuity response is difficult to

obtained. It is common practice to run the calibration standard several times, each time indexing the artificial discontinuities to a new circumferential

location in the coil. This procedure is used to insure proper response and proper centering.

BOBBIN COILS

Bobbin coil, ID coil, and inside probe are terms that describe coils used to inspect from the inside diameter (ID) or bore of a tubular test object. Bobbin coils are inserted and withdrawn from the tube ID by long, semiflexible shafts or simply blown in with air and retrieved with an attached pull cable. These mechanisms will be described later in the text. Bobbin coil information follows the same basic rules stated for encircling coils. Figure 2.3 illustrates a typical bobbin coil.

Probe coils, encircling coils, and bobbin coils can be additionally classified (5). These additional classifications are determined by how the coils are

electrically connected. The three coil categories are absolute, differential, and hybrid. Figure 2.4 shows various types of absolute and differential coil

ABSOLUTE COILS

An absolute coil makes its measurement without direct reference or comparison to a standard as the measurement is being made (6). Some applications for absolute coil systems would be measurements of conductivity, permeability, dimensions, and

hardness.

DIFFERENTIAL COILS

Differential coils consist of two or more coils electrically connected to oppose each other. Differential coils can be categorized into two types. One is the self-comparison differential, and the other is external reference differential. The self-comparison

differential coil compares one area of a test object to another area on the same test object. A common. use is two coils, connected opposing, so that if both coils are

affected by identical test object conditions, the net output is "0“ or no signal. The self-comparison arrangement is insensitive to test object variables that occur gradually. Variables such as slowly changing wall thickness, diameter, or conductivity are

effectively discriminated against with the self-comparison differential coil. Only when a different condition affects one or the other test coils will an output signal be generated. The coils usually being mechanically and electrically similar allows the arrangement to be very stable during temperature changes. Short discontinuities such as cracks, pits, or other localized discontinuities with abrupt boundaries can be detected readily

The external reference differential coil, as the name implies, is when an

external reference is used to affect one coil while the other coil is affected by the test object. Figure 2.5 illustrates this concept. This system is used to

detect differences between a standard object and test objects. lt is particularly useful for comparative conductivity, permeability, and dimensional

measurements. Obviously in Figure 2.5 it is imperative to normalize the system with one coil affected by the standard object and the other coil

affected by an acceptable test object. The external reference differential coil system is sensitive to all measurable differences between the standard object and test object. For this reason it is often necessary to provide additional

HYBRID COILS

Hybrid coils may or may not be the same size and are not necessarily

adjacent to each other. Common types of the hybrid coil are Driver/Pickup, Through Transmission, or Primary/Secondary coil assemblies. Figure 2.6 shows a typical hybrid arrangement.

A simple hybrid coil consists of an excitation coil and a sensing coil. In the through transmission coil, the excitation coil is on one side of the test object and the sensing coil is on the other. The voltage developed in the sensing coil is a function of the current magnitude and frequency applied to the excitation coil, coil parameters of the exciting and sensing coils, and test object

characteristics. In Figure 2.6 an encircling coil induces circumferential

currents in a cylindrical test object, and the disturbances of these currents are detected by a small probe coil.

ADDITIONAL COIL CHARACTERISTICS

Coil configuration is but one of many factors to consider when setting up test conditions. Other coil characteristics of importance are mechanical, thermal, and electrical stability; sensitivity; resolution; and dimensions. The geometry of the coil is usually dictated by the geometry of the test object, and often sensitivity and resolution are compromised. The relative importance of test coil characteristics depends upon the nature of the test. A blend of theory and experience usually succeeds in selection of proper coil parameters. Coil

Detecting Corrosion in Aluminum with Eddy Current Array Technology

Corrosion is everywhere and aluminum is no exception. Whether used in the petrochemical, the power generation, or the aerospace industry, aluminum is subject to degradation. Without a doubt, there is a real need for a reliable and high-precision non-destructive testing (NDT) method.

In many situations, one must detect and assess the extent of corrosion damage without having direct access to the region of interest. Indeed,

assessing wall loss and pitting on the far side of an aluminum layer is key in a number of situations.

The present document highlights the capabilities of eddy current array (ECA) technology using a particularly interesting application: corrosion detection in the storage tanks of nuclear power plants. The following describes how the technology is used to examine this important asset, which plays a critical role in the safe operation of nuclear plants.

The Challenge

Storage tanks vary in size, shape and material. In the current situation, an aluminum tank with a slightly concave floor (approximately 5.2 m (17 ft) in diameter) was in need of inspection. As in most cases during in-service

inspections, the far side of the aluminum plate was not accessible. This called for a solution capable of scanning through the floor plates in an effort to

detect and characterize corrosion-related defects such as pitting and thinning. An enhanced technique was needed in place of conventional NDT methods such as ultrasonic testing (UT) or single-channel eddy current testing (ECT). In the application herein, the examination was originally performed with UT, which required couplant, a crew of four to five technicians and a significant amount of time because of the small active surface of the transducer

(6.35 mm or 0.25 in.).

Furthermore, a wide ECA probe would need to:

 Adapt to the tank floor’s curvature and other geometric features

 Offer sufficient penetration to scan through thick aluminum (6.35–7.94 mm or

0.250–0.313 in.)

The Solution

ECA technology uses several individual coils, grouped together in one probe. The coils are excited in sequence to eliminate interference from mutual inductance

(something referred to as multiplexing). So doing, the coils work together to scan a wider inspection area than conventional ECT probes, which drastically cuts down on the time required to inspect an entire tank floor.

The absence of couplant, inherent to eddy current testing, is also a natural advantage of the solution over UT. Ectane front The solution developed to answer this challenge consists of three elements — Eddyfi’s EctaneTM, a compact, rugged, battery-operated ECA data acquisition unit; Magnifi®, acquisition and analysis software for graphical display (C-scan), record keeping, and reporting; and, finally, because of the non-linear geometry found in this application, a semi-flexible probe whose active surface could match the tank floor’s geometry.

The ECA probe developed for the application has a flexible active surface 128  mm (5.04  in.) wide adapting to slightly convex or concave geometries. The array features 33 coils, distributed in two rows, and uses multiplexing for enhanced performance.

The coils, 6 mm (0.236 in.) in diameter, are perfectly matched to cover a low-frequency range of 0.6-20 kHz with a central frequency of 5 kHz. This design ensures excellent

A calibration plate was used to validate the probe’s performance. To simulate both localized pitting and plain corrosion, it has a series of flat-bottom holes (FBH) ranging from 1.59 mm (0.063 in.) to 12.7 mm (0.5 in.) in diameter and 10% to 80% of the plate’s thickness.

Thanks to Magnifi, it’s easy to use the phase angle to assess the extent of corrosion, discriminating between near-surface and more distant defects. In addition to the traditional impedance plane, ECA technology offers advanced imaging capabilities. Indeed, Magnifi can generate 2D and 3D C-scans, which proves extremely useful when interpreting signals. Scanning the calibration plate with the ECA probe yielded the following results:

The ECA probe can clearly detect pitting-like indications (down to the

1.59 mm (0.063 in.) FBH at 40% thickness) or thinning-like indications (down to the 12.7 mm (0.5 in.) FBH at 10% thickness). These results proved to be superior to those of the previous examination method.

The solution was deployed on-site and led to the discovery of very degraded tank floor plates. The entire ECA inspection of a typical tank floor was

performed in about a tenth of the time taken with the original UT inspection procedure, and was carried out by a single technician.

The Benefits

The solution developed by Eddyfi to meet the challenge of corrosion detection in nuclear power plant storage tanks has several benefits, useful to other

industries and applications as well:

 Rapid scanning of large regions of interest

 Improved versatility, adapting to curved or irregular surfaces

 High-precision assessment of localized indications (e.g. pitting) and general degradation (e.g. thinning)

 Easier interpretation with C-scan imaging

 Full data recording and archiving capabilities

Eddyfi develops a variety of products, of which the ones presented here are only a few. We have the expertise and flexibility to engineer solutions for the most challenging applications.

Rising to the Ferromagnetic Electromagnetic Testing Challenge

We all rely on carbon steel (CS) welds in our daily lives, whether they are on the structures we use to commute, on the pipelines that carry the fuel we use in our cars, or on the wind turbines that generate the electricity we use to

prepare meals. I think we can agree that we like our CS welds strong and

secure. Hence the need to inspect them for defects thoroughly and effectively. Carbon Steel Welds are Everywhere

Why is that? CS is easy to weld, doesn’t cost too much, and it’s extremely reliable. But. There’s always a but. CS welds are prone to cracking and are sometimes well hidden under layers of paint and coatings used in an effort to preserve assets. The crack defects in CS welds often break their surface and are usually too small for the naked eye to see.

Furthermore, carbon steel is ferromagnetic. This means a high magnetic permeability and little to no penetration of eddy current. We’ve never shied away from a challenge

The Problems with Existing CS Weld Inspection Methods

Conventional methods used to detect cracks in industries relying heavily on carbon steel welds include:

• Penetrant testing (PT)

• Magnetic particle testing (MT)

• ECT pencil probes including ACFM

These methods require extensive and time consuming surface preparation, the

remains of which often end up released in the environment. Which adds to their high dependence on operator skills, somewhat unreliable results, inability to archive

inspection data, and inherently low inspection speeds.

Another method enjoying a degree of success - electromagnetic, this time - is

alternating current field measurement ACFM . This method relies on mathematical models to assess cracks and estimate their depth. However, while it can do what the other techniques can’t, it’s also a slow one that needs, like ECT pencil probes, several scans to cover the entire geometry of the weld while only offering partial data.

Pushing the Limits of Electromagnetic Inspection Technologies

Being so widespread, but not supported properly, CS weld inspection

deserved a better inspection method. To come up with it, we were faced with very interesting technological challenges: How do we scan the entire

geometry of the weld in a single pass to speed up the inspection process? How do we do so without surface preparation? How do we achieve that with reliable positioning and depth information about crack defects? A typical eddy current array (ECA) solution would seem, at first glance, ideally suited to this type of application. It isn’t, however. That’s because typical ECA pancake coil configurations yield signals from which it is difficult to extract depth

information. Furthermore, the presence of liftoff introduces a “drift” of the operation point along this hook, which produces significant phase changes, making depth sizing impossible from a practical standpoint.

Enters TECA – Tangential Eddy Current Array

Through much R&D, we came to the conclusion that using

“ tangential eddy current ” was the most promising avenue towards overcoming these challenges. As mentioned above, conventionally, the axes of pancake coils are

positioned perpendicular to the surface under test. With tangential eddy current, coils are on their sides, with their axes parallel to the surface and the eddy current

generated by the coil flowing parallel to the surface under test, “diving”, so to speak, under it. So how could we use tangential eddy current and leverage the power of an eddy current array? A multiplexed ECA would solve the single-pass problem, as arrays cover a wider area. We analyzed several parameters, including coil

size/impedance/position/configuration, the operating frequency, and the multiplexing pattern (topology), among others, to create an optimal ECA solution. We tested and characterize more than 30 coil configurations over the course of a year of R&D,

coming up with what we felt is the best coil configuration to leverage the power of ECA, striking a balance between coverage, penetration, and resolution. That’s how the

tangential eddy current array (TECA™ ) was born. We were able to observe that TECA generated a relatively flat liftoff signal and defects approximately 90° from the liftoff signal, something that’s not possible using other inspection techniques. The multiplexed eddy current generated by TECA can dive under cracks down to 10 mm (0.4 in). But that doesn’t take care of the geometry issue.

A NEW EDDY CURRENT PROBE - Tangential Eddy Current Array

Lift-off noise is unavoidable so long as the probe picks up the eddy current induced by exciting coil. There the authors have thought of two notions in order to design a new probe that suppresses lift-off noise and detects flaws:

1. One of the methods to eliminate lift-off noise in eddy current testing is to develop a probe picking up the component of eddy current that is generated only by flaws but not by the probe lift-off.

2. Each part of detecting coil windings picks up the parallel component of eddy current to itself.

With the above two notions in mind, the authors have devised a new eddy current surface probe that is composed of a pancake exciting coil and a tangential detecting coil as shown in Figure 1. The circular exciting coil is adopted because it induces eddy current most efficiently. The exciting coil induces axi-symmetric circular eddy current in the test material with no eddy current circulating across the exciting coil circle when there is no flaw in the test material as shown in Figure 2(a). When there is a flaw crossing the circle, some eddy current circulates along the flaw crossing the circle. Since each part of the detecting coil winding picks up the parallel eddy current component to itself, the tangential

detecting coil picks up only the eddy current circulating across the circle as shown in Figure 2(b)-(d). As the new probe scans over a flaw, the detecting coil generates a figure eight signal pattern. If the probe has two tangential detecting coils wound perpendicular to each other, it can detect all flaws in every orientation. The impedance of the exciting coil can also be used to monitor the probe lift-off in order to avoid the probe not detecting flaws in the material.

The new probe is lift-off noise free because the lift-off of the probe from the material does not cause any eddy current to circulate crossing the exciting coil circle. Thus lift-off noise can be eliminated by detecting only the newly generated eddy current by flaws and by not detecting the eddy current induced by the exciting coil when there is no flaw in the test material. The probe is self-nulling because the detecting coil generates a signal only when a flaw causes some eddy current to circulate across the circle.

Since the probe generates minimal lift-off noise, the authors have also thought that the probe lift-off does not influence much to the flaw signal and that the signal phase can be used for evaluating the depth of surface flaws.

Scanning an Entire Weld in One Pass

This was also tricky. The TECA coil design had to be used in such a way as to cover the cap, toe, and heat-affected zone of CS welds, while dynamically adjusting to the weld’s uneven geometry. The challenge lay in bundling the coils in a mechanical package that struck a balance between resolution and sizing capabilities.

After much testing, we designed an ingenious system of independent, spring-loaded fingers that adapt to weld geometries. The individual wedged fingers all incorporate an array of coils, which provides great resolution even at

higher scan speeds, surfing over the uneven geometry of the weld and enabling the a single-pass scans of entire welds.

What About Liftoff?

And you would be right to ask. As I mentioned above, TECA generates a

virtually flat liftoff signal, with crack-like indications approximately 90° relative to this liftoff signal and all the indications featuring the same phase shift.

The software processing Sharck probe data incorporates the equivalent of a three-dimensional depth-to-liftoff-to-vertical-amplitude depth plane that allows compensating for liftoff.

The Final Touch Add to the design a removable high-resolution encoder and you have the final patent-pending Sharck probe capable of positioning cracks, measuring their length, and sizing them as deep as 10 mm (0.4 in), without surface preparation, at up to 200 mm/s (7.9 in/s).

Chapter 2

0.2-1 Differential coils are usually used in: A. bobbin coils.

B. probe coils. C. OD coils.

D. any of the above.

0.2·2 When using a probe coil to scan a test object, ____ _ A. the object must be dry and polished.

B. the object must be scanned carefully to insure inspection coverage.

C. the object must be scanned in circular motions at constant speeds. D. the probe must be moving at all times to get a reading.

0.2·3 A "spinning probe" would most likely be a (an): A. bobbin coil.

B. ID coil. C. OD coil.

0.2·4 A "feed-through" coil is:

A. a coil with primary/secondary windings connected so that the signal is fed through the primary to the secondary.

B. an encircling coil. C. an OD coil.

D. both B and C.

0.2-5 When inspecting a tubular product with an encircling coil, which statement is not true?

A. OD discontinuities can be found.

B. Axial discontinuity locations can be noted.

C. Circumferential discontinuity locations can be noted.

D. ID discontinuities can be found.

0.2·6 An absolute coil measurement is made ____ _ A. by comparing one spot on the test object to another.

B. without reference to or direct comparison with a standard.

Encircling Coil - Defect Detectability

For long defect the self comparison differential coils’ may cancelled each other

0.2·7 When coils in a differential arrangement are affected simultaneously with the same test object variables, the output signal ____ _

A. is directly proportional to the number of variables.

B. is "0" or near-"0."

C. is indirectly proportional to the number of variables. D. is primarily a function of the exciting current.

0.2·8 Which coil type inherently has better thermal stability? A. Bobbin

B. Absolute C. OD

Encircling Coil - Defect Detectability

When coils in a differential arrangement are affected simultaneously with the

0.2-9 A hybrid coil is composed of two or more coils. The coils ____ _ A. must be aligned coplanar to the driver axis:

B. may be of widely different dimensions.

C. must be impedance-matched as closely as possible. D. are very temperature sensitive.

0.2-10 Proper selection of test coil arrangement is determined by: A. shape of test object.

B. resolution required. C. sensitivity required. D. stability.

Absolute and Differential Coils

3. TEST COIL DESIGN

As discussed earlier, test coil design and selection is a blend of theory and experience. Many factors must be considered. These important factors are determined by the inspection requirement for resolution, sensitivity,

impedance, size, stability, and environmental considerations. In order to

better understand coil properties and electrical relationships, a short refresher in alternating current theory is necessary. First, we must examine electrical units-for example, current and its representative symbol I. Current not only suggests electron flow but also the amount. The amount of electrons flowing past a point in a circuit in one second is expressed in amperes; 2π∙1018

electrons passing a point in one second is called 1 ampere.

RESISTANCE

Resistance is an opposition to the flow of electrons and is measured in ohms. Ohm's Jaw is stated by the equation: I = E/R

Where:

R =

ρ

Resistance = Ohms

Specific Resistance

ρ

Length = foot

Thus, the resistance of a 10-foot length of 40 gauge copper wire with a

specific resistance of 10.4 circular-mil-foot at 20ºC would be found as follows: R = 10.4 x 10/ 9.88 = 10.53Ω

In an alternating current circuit containing only resistance, the current and voltage are in phase. In phase means the current and voltage reach their minimum and maximum values, respectively, at the same time. The power dissipated in a resistive circuit appears in the form of heat. For example, electric toasters are equipped with resistance wires that become hot when current flows through them, providing a heat source for toasting bread.

Inductance

Heat generation is an undesirable trait for an eddy current coil. If the 10-foot length of wire used in the previous example was wound into the shape of a

coil, it would exhibit characteristics of alternating current other than resistance. By forming the wire into the shape of a coil, the coil also would have the

property of inductance. The role of inductance is analogous to inertia in mechanics, because inertia is the property of matter that causes a body to oppose any change in its velocity. The unit of inductance is the Henry (H). A coil is said to have the property of inductance when a change in current

through the coil produces a voltage in the coil. More precisely, a circuit in which an electromotive force of one volt is induced when the current is

changing at a rate of one ampere per second will have an inductance of one Henry. The inductance of a multilayer air core coil can be expressed by its physical properties, or coil parameters. Coil parameters such as length,

Figure 3.1 illustrates typical coil dimensions required to calculate coil inductance.

Figure 3.1- Multilayer Coil

r l

An approximation of small, multilayer, air core coil inductance is as follows:

L = 0.8(rN)

∙ (6r + 9l + 10b)

Where:

L = Self-inductance in microhenries (μH) N = Total number of turns

r = Mean radius in inches l = Length of coil in inches

b = Coil depth or thickness in inches

For example, a coil whose dimensions are as follows: r = 0.1 inches

l

As stated earlier, this inductance is analogous to inertia in mechanical

systems in that inductance opposes a change in current as inertia opposes a change in velocity of a body. In alternating current circuits the current is

always changing; therefore inductance is always opposing this change. As the current tries to change, the inductance reacts to oppose that change. This reaction is called inductive reactance in ohm.

The unit of inductive reactance (X_L) Is in ohms. Because the amount of

reactance is a function of the rate of change of current and rate of change can be described as frequency, a formula relating frequency, inductance, and

inductive reactance is:

X

=

ωL = 2πfL

where:

X_L = Inductive reactance in ohms f = Frequency in hertz

For example, using the 32 microhenry coil calculated earlier, operating at 100 kilohertz, its inductive reactance would be found as follows:

X_L = ωL = 2πfL

X_L = 2π ∙ 100 ∙ 103 ∙ 32 ∙ 10-6

XL = 20.106 ohms

Therefore, this coil would present an opposition of 20 ohms to currents with a rate of change of 100 kilohertz due to its reactive component. Unlike a

resistive circuit, the current and voltage of an inductive circuit do not reach their minimum and maximum values at the-same time. In a pure inductive

circuit the voltage leads the current by 90 electrical degrees. This means that when the voltage reaches a maximum value, the current is at "0“

Power is related to current and voltage as follows: P =El where: P = Power in watts E = Volts I = Current in amperes

Notice that in a pure inductive circuit, when the voltage is maximum, the

current is "0"; therefore, the product El = 0. Inductive reactances consume no alternating power where resistive elements consume power and dissipate

power in the form of heat. The opposition to current flow due to the resistive element of the coil and the reactive element of the coil do not occur at the same time; therefore, they cannot be added as scalar quantities. A scalar

quantity is one having only magnitude; i.e., it is a quantity fully described by a number, but which does not involve any concept of direction. Gallons in a

IMPEDANCE

In order to explain the addition of reactance and resistance with a minimum of mathematical calculations we can again use the vector diagram or phasor

diagram to explain this addition (19). A phasor diagram constructed with

Imaginary units on the ordinate. or (Y) axis and real units on the abscissa or (X) axis is shown In Figure 3.2a.

Substituting Inductive reactance (X_L) and resistance (R) we can find the resultant of the vector addition of X_L and R. This resultant vector Z is known as impedance. Impedance is the total opposition to current flow. Further

observation of Figure 3.2b reveals X_L, R, and Z appear to form the sides of a right triangle. The mathematical solution of right triangles states the square of the hypotenuse is equal to the sum of the squares of the other two sldes, or c2 _{= a}2 _{+ b}2 _{, substituting the Z, R & X}

L, the equation becomes

Z2 _{= R}2 _{+ X} L2

Z = √(R2 _{+ X} L2)

Let's try an example. What is the impedance of a coil having an inductance of 100 microhenries and a resistance of 5 ohms and being operated at 200

kilohertz? First we must convert inductance to inductive reactance and then, by vector addition, combine inductive reactance and resistance to obtain the impedance.

Z = [ 52 + (2π∙200 ∙103 ∙100 ∙10-6₎2_]0.5

The term R + jX_M is known as a rectangular notation. As an example, a

resistance of 4 ohms in series with an inductive reactance of 3 ohms could be noted as Z = 4 + j3 ohms. The impedance calculation is then:

Z = √(42₊₃2_{) = 5 ohm}

In coil design it is often helpful to know also the included angle between the resistive component and impedance. A convenient method of notation is the polar form where Tan Ф = X_L/R, where Ф is included angle between

resistance and impedance. In the previous example our impedance magnitude is 5 ohms, but at what angle?

Ф = tan-1 (3/4) = 36.9º

Z = 5∠ 36.9º = |5|_36.9º = 4+j3

Eddy current coils with included impedance angles of 60° to 90° usually make efficient test coils. As the angle between resistance and impedance

approaches 0, the test coil becomes very inefficient with most of its energy being dissipated as heat..

Q or FIGURE OF MERIT

The term used to describe coil efficiency is Q or merit of the coil. The higher the Q or merit of a coil, the more efficiently the coil performs as an inductor. The merit of a coil is mathematically stated as:

Q = X

/R

Where:

X_L = Inductive reactance R = Resistance

For example, a coil having an induct ive reactance of 100 ohms and a resistance of 5 ohms would have a Q of 20.

PERMEABILITY AND SHIELDING EFFECTS

The addition of permeable core materials in certain coil designs dramatically improves the Q factor. Permeable cores are usually constructed of high permeability "powdered iron." Probe coils, for example, are wound on a form that allows a powdered iron rod or slug to be placed in the center of the coil. lt is. common to increase the coil impedance by a factor of 10 by the addition of core materials. This increase in impedance without additional

winding greatly enhances the Q of the coil.

Some core materials are cylinder- or cup-shaped. A common term is cup core. The coil is wound and placed in the cup core. In the case of a probe coil in a cup core, not only is the impedance increased, but the benefit of shielding is also gained. Shielding with a cup core prevents the electromagnetic field from spreading at the sides of the coil. This greatly

reduces the signals produced by edge effect of adjacent members to the test area, such as fasteners on an aircraft wing. Shielding, while improving resolution, usually sacrifices some amount of penetration into the part. Another method of shielding uses high conductivity material, such as copper or aluminum, to suppress high frequency interference from other sources and also to shape the electromagnetic field of the test coil. A copper cup would restrict the electromagnetic field in much the same manner as the "powdered iron cup core" discussed previously. A disadvantage of high conductivity, low or no permeability shielding is that the coil's impedance is reduced when the shielding material is placed around the test coil. The net effect is, of course, that the coil's Q is less than it was when

By Using Core:

Ferromagnetic core

 Q factor increase (increase impedance)

 Shielding effect

 Less penetration

Non-ferromagnetic core

 Q factor decrease (decrease impedance)

 Shielding effect

Probe Shielding

One of the challenges of performing an eddy current inspection is getting sufficient eddy current field strength in the region of interest within the material. Another challenge is keeping the field away from non-relevant features of the test component. The impedance change caused by non-relevant features can complicate the interpretation of the signal. Probe shielding and loading are sometimes used to limit the spread and concentrate the magnetic field of the coil. Of course, if the magnetic field is

concentrated near the coil, the eddy currents will also be concentrated in this area.

Probe Shielding

Probe shielding is used to prevent or reduce the interaction of the probe's magnetic field with non-relevant features in close

proximity of the probe. Shielding could be used to reduce edge effects when testing near dimensional transitions such as a step or an edge. Shielding could also be used to reduce the effects of conductive or magnetic fasteners in the region of testing.

1) Magnetically shielded with ferromagnetic materials

Eddy current probes are most often shielded using magnetic shielding or eddy current shielding. Magnetically shielded probes have their coil surrounded by a ring of ferrite or other material with high permeability and low conductivity. The ferrite creates an area of low magnetic reluctance and the probe's magnetic field is concentrated in this area rather than spreading beyond the shielding. This concentrates the magnetic field into a tighter area around the coil.

2) Eddy current shielding with non-magnetic materials

Eddy current shielding uses a ring of highly conductive but nonmagnetic material, usually copper, to surround the coil. The portion of the coil's magnetic field that cuts across the shielding will generate eddy currents in the shielding material rather than in the non-relevant features outside of the shielded area. The higher the frequency of the current used to drive the probe, the more effective the shielding will be due to the skin effect in the shielding material.

3) Probe Loading with Ferrite Cores vs. Air Cores

Sometimes coils are wound around a ferrite core. Since ferrite is ferromagnetic, the magnetic flux produced by the coil prefers to travel through the ferrite as opposed to the air. Therefore, the ferrite core concentrates the magnetic field near the center of the probe. This, in turn, concentrates the eddy currents near the center of the probe. Probes with ferrite cores tend to be more sensitive than air core probes and less affected by probe wobble and lift-off.

Probe Shielding

Probe Loading with Ferrite Cores vs. Air Cores

Sometimes coils are wound around a ferrite core. Since ferrite is ferromagnetic, the magnetic flux produced by the coil prefers to travel through the ferrite as opposed to the air. Therefore, the ferrite core concentrates the magnetic field near the center of the probe. This, in turn, concentrates the eddy currents near the center of the probe. Probes with ferrite cores tend to be more sensitive than air core probes

Magnetically shielded with ferromagnetic materials or non ferromagnetic materials

Another coil design used for inspection of ferromagnetic materials uses a saturation approach. A predominant variable that prevents eddy current penetration in ferromagnetic material is called permeability. Permeability effects exhibited by the test object can be reduced by means of magnetic saturation. Saturation coils for steels are usually very large and surround the test object and test coil. A steady state current is applied to the saturation coil. When the steel test object is magnetically saturated it may be inspected in the same manner as a non-ferromagnetic material. In the case of mild steel many thousands of gauss are required to produce saturation. In such other

materials as nickel alloys (monel and inconel), the saturation required is much less and can usually be accomplished by incorporating permanent magnets adjacent to the test coil.