ELECTROMAGNETICTESTING I 233
TABLE 1 . Automotive components tested by eddy depth, based on multiple linear regressions of eddy current
current techniques for various characteristics responses at several frequencies against case depth. The
particular frequencies used depend on the location. Gener-
Component Characterisitcs ally, they consist of a combination of one or more low fre-
quencies (5 to 20 Hz), with one or more higher frequencies
Aluminum flat stock MC (100 to 1,000 Hz). The highest frequencies (5 to 10 kHz) are
Alternator shafts CD used for surface hardness measurement.
Axle shafts CD HD
Aluminum castings CD HD Figure 42 shows midshaft case depth as calculated from
Ball joint studs CD HD eddy current measurements versus case depth measured
Bearing races destructively. With perfect correlation, all points would fall
Brake calipers for disk brakes HD on the 45 degree straight line. The data are for a mix of
Brake cylinders HD about 130 production and specially prepared shafts. Root
Camshaft sprockets CD HD mean square error for this measurement is about 0.15 mm
Castings, malleable iron CD HD (0.006 in.) with respect to destructive measurement.
Connecting rods HD
Disk brake rotors FIGURE 40. Test locations on axle shaft (case
Distributor shafts MC depth indicated by dashed line)
Door latch components CD HD
Exhaust valve seats CD ~BUTTON
Fasteners (bolts, nuts)
Flywheels HD II
Forged camshafts HD II ...,_SPLINE
Front wheel hubs CD HD I
Gears of many types CD HD I
HD I
Differential gears OM HD I
Transmission gears CD I
Iron castings OM HD I
King pins OM HD I
Paint on metals CD HD
Parking pawls OM ....,..._ MIDSHAFT
Piston pins FT
Shock absorber tubes HD I
Spring wire OM HD I
Stabilizer bars CD HD MC I
Stabilizer bar stock CD
Starter pinion bar stock CD 'I
Thrust washers
Tin on pistons MC I
Torsion bars MC I
Transmission shafts HD I
Valve seat inserts FT
Water pump shafts OM BEARING
Wheel spindles CD HD
MC
OM HD
OM HD
CD == CRACK DETECTION.
DM CASE DEPTH MEASUREMENT.
FT == FILM THICKNESS.
HD = HARDNESS DETERMINATION.
MC MATERIAL COMPOSITION.
234 I NONDESTRUCTIVETESTING OVERVIEW
FIGURE 41 . Axle shaft test stand equipment Crack and PorosityDetection and
Machined Hole Presence in Master
,., sr [l'l'ING Brake Cylinders
MOTOR Master brake cylinders, cast of aluminum 355 alloy, are
-- TE'iT inspected two at a time by a semiautomatic test system for
three conditions: porosity, cracks and the presence of six
INSTRUMENT machined holes necessary for the part's function. This is
accomplished using a rotary differential dual element eddy
~- COMPUTER current probe excited at 100 kHz. A computer is pro-
grammed ( l) to verify the presence of the holes through
1,,tOTOR the presence and amplitude of signals as shown in Fig. 43
CONTROL and (2) to verify the absence of cracks and porosity by the
, TEST STAND absence of signals between these holes. Porosity, as an
P!ATEN example, is indicated by the signal in addition to the six-
WITH hole signals in Fig. 43. The test cycle of this equipment is as
TEST COJLS follows.
AXLE SHt\FT
IN TEST FIGURE 43. Master brake cylinder cross section
POSIT JON with hole presence and porosity discontinuity
signals
FROM FORD MOTOR COMPANY. REPRINTEDWITH PERMISSION.
FIGURE 42. Plot of axle shaft correlation data CYLINDER
4 (1.6) BORE
SURFACE
3 (I 2) wo ow
!:] :::1
0 :0r
II I
66 €6; 6€;
00
I- V) 2 10 8) 0w ~w> w~> ::J~0 _is' _is'
I (04)
~~ I -' w c30wI f :s._Jll'l.JJ
~ ~0w
s~Qe:'. (!) vi a< ._J a<
vi 0
>Ou-=E Vl '.S CL V)
I II II II
0
LU
0 I 23 4 5 VOLTS PER
0 (0.4) (0.8) (I 2) (I 6) DIVISION
DESTRUCTIVE CASE DEPTH
milimeters (inches)
ELECTROMAGNETICTESTING I 235
1. The operator manually loads a dual fixture with two FIGURE 44. Probe set up for test of plating
parts. thickness
2. The operator initiates the test cycle by pressing dual PROBE IS AN ABSOLUTE
palm buttons. AIR-CORE TYPE
3. Test parts are automatically clamped in place. mPROBE CHANNEL 1
4. Probes, mounted on a common slide base, rotate at
-+--t------+-) -
30 s-1 (1,800 rpm) and are advanced (by 200 mm or --+-a~-)~ - wTEST AREA 25 mm [1 in.) LONG
8 in. stroke) into bores of the two test parts. AT FOUR LOCATIONS
5. The bores of both test parts are scanned simultane- CHANNEL 2
ously.
6. Probes are retracted at the end of the scan cycle and PROBE
parts are automatically undamped at the end of the
retract cycle. TIN PLATING ON PISTON SURFACE
7. The accept or reject visual signal for each part is acti-
vated at the operator's station. DIESEL PISTON
8. The operator manually unloads tested parts.
FIGURE 45. Graph of test responses for tin
The time required for the test cycle is 8 s, not including plate thickness; meter readings indicate
load and unload time. The system is calibrated using a mas- proportion to phase angle of eddy current signal
ter part that has the six required holes and a 0.25 mm
(0.01 in.) deep porosity. This sensitivity level is sufficient to Vl 0.025 mm (0.001 in.)
detect detrimental cracks as well as porosity.
zLVI.lJ
Tin Plate Thicknesson Diesel
Engine Piston u::..::'. O O 19 mm (0.00075 in.)
I
Tin plating of steel pistons for diesel engines is used to
aid lubrication during initial engine operation. The accept- I- 0.013 mm (0.0005 in.)
able thickness is about 0.02 mm (8 x 10-4 in.). These parts
are automatically oriented to inspect four areas rotated -25 +25
90 degrees from the piston pin hole and above and below it.
Specifications for the test include: METER READINGS
1. acceptable thickness: 0.013 mm (0.0005 in.) to Cold Headed Pinion Gear Blank
0.025 mm (0.001 in.); Crack Detection
2. test in four locations; A rotating probe system (see Fig. 46) inspects cold
3. probe test frequency at 70 kHz; headed pinion gear blanks for cracks or seams. Specifica-
4. production rate of 400 per hour; and tions for the test include:
5. part is rotated at 120 rpm during test.
1. rejectable crack depth of0.25 mm (0.01 in.);
Out-of-specification parts are automatically removed from 2. rejectable crack length of 6.3 mm (0.25 in.);
the line. The probe is an absolute air-core type. Test areas 3. inspection rate of 5,100 parts per hour;
are 25 mm (1 in.) long in each of four locations. 4. probe test frequency of 20 kHz; and
5. parts tested at each end on the outside diameter.
Calibration is done by inserting a master with 0.02 mm
(8 x 10-4 in.) thickness of tin plating in the test station and Rejected parts are automatically diverted out of the system.
rotating as shown in Fig. 44. Next, the operator adjusts zero
suppression to meter zero. A 0.013 mm (5 x 10-4 in.) tin
thickness master is inserted and the reject dial is set to
reject this part. Next, a 0.025 mm (0.001 in.) tin thickness
master is inserted to verify that it also rejects. The calibra-
tion is done for both channels of the system (see Fig. 45).
236 I NONDESTRUCTIVE TESTING OVERVIEW
FIGURE 46. Test arrangements for cold headed Parts first enter the machine, rolling from outside diam-
pinion gear blank crack detection eter to outside diameter, and are then tipped over on the flat
face and pushed over a hole. Next, they drop and form a
ROTATING CRACK DETECTING PROBE ( 1.000 rpm) stack of 10parts in the bore of a rotary transformer that con-
tains a crack detecting probe. The bottom half of the part is
DIFFERENTIAL tested for longitudinal cracks, the stack is dropped half a
FERRITE CORES part length and the top half is then inspected.
ONE REVOLUTION Mastering is done by first inserting the cracked reject
master part, notched 0.25 mm (0.01 in.) deep by 6.3 mm
CRACK (0.25 in.) long. The machine is indexed until this part is in
NOISE~ the test location and rejection is verified. The procedure is
repeated with the cracked accept master part, notched
4 VOLTS 0.13 mm (0.005 in.) deep by 6.3 mm (0.25 in.) long, to verify
acceptance.
+ TIME
Hub and Spindle Hardness and
VOLTAGE Case Depth Inspection
FIGURE 4 7. Equipment for hub and spindle Hubs and spindles for front wheel drive vehicles are
hardness and case depth inspection tested for Re 58 minimum hardness and 2 mm (0.08 in.)
minimum case depth at a rate of 1,200 parts per hour using
UPPER TEST PROBE LOWER TEST PROBE a microprocessor based eddy current instrument (see
Fig. 47). A combination of phase and amplitude signals at
500 Hz and a phase signal at 5 kHz are used to perform the
total heat treat inspection in the two inside diameter heat
treated zones on hubs and three shaft locations on spindles.
Hubs enter the machine at random, sliding on their
flanges on a gravity roller track The machine meters them
into the test station, one at a time. When properly staged,
the upper probe comes down into the hub and the lower
probe comes up into the hub. Both probes seat on the
induction heat treated ball tracks (see Fig. 48). The eddy
FIGURE 48. Diagram of system set up for eddy
current tests of hubs
TYPICAL FRONT WHEEL DRIVE
DUAL BALL TRACK HUB
FROM K.J. LAW ENGINEERS, INCORPORATED. REPRINTEDWITH
PERMISSION.
ELECTROMAGNETICTESTING I 237
current instrument then performs a sequential test, first at Typical rejectable conditions include:
the low frequency, both phase and amplitude, then at the
high frequency. The two probes are driven sequentially. The 1. not heat treated (green);
results of each of these six tests are compared with stored 2. heat treated but not quenched or poorly quenched;
data for an accept part and the status of each test is dis- 3. low induction power;
played. Reject parts are automatically diverted. 4. short induction time; and
5. misplaced pattern due to insufficient induction coil
FIGURE49. Diagramof systemconvertedfor
eddy currenttests of spindles insertion.
TEST COIL I KEEPER GROOVE These systems are convertible to test the companion
spindle used with this hub. A three-channel coil assembly is
TEST COIL 2 ----1-- INDUCTION HEAT lowered on to the spindle to test it in three locations as
TEST COIL 3 TREATED ZONE shown in Fig. 49. Rejectable conditions are similar to those
BALL TRACK described above.
PILOT Special masters are prepared representing the types of
failures that can occur in the induction heat treat process.
These are inserted twice a shift to verify that the system
operates properly. Automatic printout data are compared to
previous data to detect coil and probe, machine wear or fail-
ure. The instrument is totally digital and the coils are so pre-
cise that no adjustments are required.
Camshaft Heat Treat Inspection
Because of the irregular shape of camshafts, it is neces-
sary to use comparator techniques for eddy current testing
of hardness and case depth (see Fig. 50). The coils are
FIGURE50. Diagram of comparatorsystemfor eddy currenttests of camshafts
REFERENCE MI\STER PART
LOBES (TYPICAL) REFERENCE COIL
CAMSHAFTJOURNAL (TYPICAL)
COMPARATORSORTING COILS
PRODUCTION PART COIL TRAVEL _._
THREE-ELEMENT AIR CORE TEST COIL
238 I NONDESTRUCTIVE TESTING OVERVIEW
excited at 3 kHz to ensure that the camshaft has been prop- comparator system. Production parts are lifted off a walking
erly heat treated. This machine can test 300 camshafts per beam conveyor and staged automatically in a similar man-
hour. A part with a reject heat treat condition is automati- ner. A reference coil is slowly moved onto the master accept
cally diverted from the system and only acceptable parts are shaft and travels over its full length. Simultaneously, an
put on to the production line for subsequent assembly into identical test coil is moved onto the production part in
engines. exactly the same manner so that both coils are always over
the same lobe number on their respective parts. In this way,
Operation begins when a reference master accept part only variations in heat treat quality will be detected. Geom-
with nominal hardness and case depth is locked into the etry variations along the part will be ignored because they
are being compared to identical geometry on the master
FIGURE 51 . Test equipmentfor camshaftheat part.
treatment inspection;parts enter from back
Calibration is done with identical master parts in both
REFERENCE COIL REFERENCE CAMSHAFT the master reference coil and the test coil. The zero sup-
pression is adjusted to meter zero at center scale. The sec-
TEST ond master part is then removed from the test coil position
CAMSHAFT and the reject master (usually a part that has been heat
treated but not quenched) is inserted to verify rejection.
FROM K.J. LAW ENGINEERS, INCORPORATED. REPRINTEDWITH
PERMISSION Typical rejections include parts:
1. not heat treated (green);
2. heat treated but not quenched;
3. with low induction power;
4. with short heat time;
5. with misplaced pattern (induction coils not fully on to
the part); and
6. in which one lobe not heat treated.
Figure 51 shows a system used for comparator testing of
camshaft heat treatment.
ELECTROMAGNETICTESTING I 239
PART 7
MULTIFREOUENCY TESTING
Requirements for Multifrequency Conventional Remedies for Monofrequency
Testing Problems
Conventional Use of an Eddy Current Signal Improvement in detection and interpretation quality
normally involves filtering, which attenuates interfering sig-
In eddy current testing, discontinuities mus•t be detected nals and improves the signal-to-noise ratio. This technique
assumes that the filters do not act in the same way on two
and the signals must be interpreted. Detection generally types of signal. Thus, abrupt signals with a relatively high
involves an amplitude exceeding a specified threshold value. frequency spectrum may be separated from basically lower
This threshold depends on an assessment by the operator in frequency signals by high pass filtering. However, this does
some cases (when the strip chart recordings show all exces- not solve the problem of indications given by multiple dis-
sive signals) or the threshold may be fixed within automatic continuities. In addition, even when filtering is used, inter-
sorting monitors. Both cases result in a nonrepeatable ference may occur because of the following phenomena.
detection level when signals are superimposed on back-
ground noise, the effect of which may increase or decrease 1. Affecting both the x and y components of a signal, fil-
the relevant signal (see Fig. 52). ters are rarely sufficiently matched to avoid rotation.
Interpretation is frequently based on the phase of the 2. Probe speed variation may modify the filtering result
signal, which provides information concerning the depth of in amplitude and in the ratio between the working
the discontinuity (multiparameter interpretation tech- signal and the spurious signal.
nique). Again, the presence of background noise, vectorially
combined with the relevant signal, completely modifies the These difficulties are solved by the multifrequency tech-
appearance of the signal, changing its phase and making an nique, as described below.
estimate of the signal difficult (Fig. 53). The problem is
even worse in the case of multiple indications. In fact, most Physical Basis of the
simple interpreting techniques assume that the observed Multifrequency Process
signal is the result of a single discontinuity within the area
covered by the probe field. Multiple discontinuities may Linearity Principle
lead to considerable interpretation error. Thus, on a tube, an
external discontinuity and an internal discontinuity, both The multifrequency process is useful for solving multiple
less than 10 percent of the tube wall thickness, can result in signal problems, which include (1) the useful signal plus
the tube being rejected, because the signal phase may indi- background noise or (2) the superimposition of two or more
cate that the tube is close to perforation (Fig. 54). useful indication signals.22•23
FIGURE 52. Influence of background noise on amplitude of signal component
-1-
SIGNAL WITHOUT NOISE BEST CASE WORST CASE
SIGNAL COMBINED WITH BACKGROUND COMPONENT
240 I NONDESTRUCTIVE TESTING OVERVIEW
FIGURE 53. Influence of background noise on The basic assumption, confirmed by practice, is the lin-
signal displayed in impedance plane ear composition of the signals. The signal resulting from two
discontinuities, or from one discontinuity and background
VARIOUS RESULTS OF SIGNAL COMBINED WITH BACKGROUND NOISE noise, is equal to the linear combination (vectorial sum) of
the two signals considered separately (Fig. 55).
The multifrequency process uses a composite signal and
subtracts the undesirable signal to leave only the useful sig-
nal, exactly as if the useful signal had been detected alone
(Fig. 56). Thus, the problem is to produce the signal used
for subtraction.
Production of Subtracted Signal
The subtraction result would be perfect if the undesir-
able signal alone were available. This is sometimes possible,
FIGURE 54. Cathode ray tube display observed when using differential probe on internal and
external discontinuities in tube --
<c:::::: ) +
INTERNAL DISCONTINUITY EXTERNAL DISCONTINUITY RESULTANTSIGNAL
FIGURE SS. Linearity principle -
+
NOISE DISCONTINUITY DISCONTINUITY IN NOISE
FIGURE 56. Subtraction of undesirable noise --
-
SIGNAL WITH NOISE NOISE PURE SIGNAL OF DISCONTINUITY
ELECTROMAGNETIC TESTING I 241
in the case of a perfectly repeatable signal, using computer measurement from another does not give zero; changing the
processing. However, the multifrequency Piocess is far frequency also changes the ratio between the various sig-
more powerful because the signal does not have to be nals. As an example, Fig. 57a shows (1) an undesirable noise
stored, but merely has to be measured. signal resulting from a region close to the inside diameter
probe transducer; (2) a useful signal from further away
Measurement is done with the same probe transducer, (through-wall discontinuity); and (3) a signal resulting from
energized at a frequency resulting in preferential detection a remote external discontinuity. InFig. 57a, a low frequency
of the signal to be eliminated. Unfortunately, the auxiliary separates the phases of the three signals while correctly dif-
frequency not only measures the undesirable phenomenon, ferentiating between them. In Fig. 57b, the use of a higher
but also is sensitive to other phenomena. Subtracting one frequency results in a skin effect and favors the close range
signal, mainly to the detriment of the external discontinuity.
FIGURE 57. Subtractionprincipleused in The phase separation is also wider. Subtraction of the sec-
multifrequencyprocess:(a) low frequency; ond measurement from the first will result in the virtual dis-
(bJ high frequency;(cJ subtraction appearance of the close range noise signal, adjusted
fa) similarly in amplitude and phase in both cases. However, it
will not reproduce the other two signals as detected by the
fb) THROUGH-WALL DISCONTINUITY low frequency probe. These signals are attenuated and the
phase interpretation laws therefore require adjustment
NOISE (Fig. 57c).
fc) THROUGH-WALLDISCONTINUITY Basic Characteristics of the Multifrequency Process
EXTERNAL DISCONTINUITY The previous example shows the advantage of selecting
very different frequencies so that the basic measurement is
disturbed as little as possible (by the auxiliary measurement)
due to attenuation and phase shifting.
However, good subtraction of the undesirable signal
assumes that the frequencies are not too different so that
the signal shape can be retained. In fact, when the two
undesirable responses to the two frequencies have been
adjusted in amplitude and phase, their waveforms are
slightly different, resulting in a subtraction residual, which
should be minimized by adjusting the actual waveform of
the signal.
The signal elimination method gives a measurement effi-
ciency that depends on the difference in type and position
between the useful signal and the undesirable signal. In
fact, a ripple of the same type as the background noise will
be considerably attenuated by the combination whereas a
deep discontinuity will give a signal virtually unaffected by
an elimination of surface noise.
Conversely, a significant advantage of this method is that
it is purely static; the result of the signal combination is
absolutely independent of probe movement speed so that
the result is the same at rest, at low speed or at high speed.
This is a very clear advantage when compared to filter-
ing, which requires strict control of both the speed of the
probe and the size of the discontinuities. With the multifre-
quency method, only discontinuity positions in relation to
the probe must be controlled.
242 I NONDESTRUCTIVE TESTING OVERVIEW
PART 8
MAGNETIC FLUX LEAKAGE TESTING
The technique of magnetic flux leakage (MFL) testing is inspection by wet or dry magnetic particles is often per-
a commonly used inspection method. While some forms of formed, especially if specifications require that only surface
breaking discontinuities be found.
magnetic flux leakage may not be as sophisticated as other
methods of inspection, it is probable that more ferromag- Elongated Parts
netic material is inspected by magnetic flux leakage than by
any other method. The cylindrical symmetry of elongated parts such as wire
rope permits the use of a relatively simple flux loop to mag-
Magnetic flux leakage methods consist primarily of two netize a relatively short section of the rope (shown in
steps. The first is a method of magnetizing the part and the Fig. 59). Encircling sensors are placed at some distance
second is the use of a magnetic flux sensitive detector to from the rope to permit the passage of splices. Such systems
scan flux diverted by discontinuities. (see Fig. 60) might also be used for pumping well sucker
rods and other elongated oilfield parts.
Magnetizing methods have evolved to suit the geometry
of the parts being inspected. The methods include the use After a well is .drilled, the sides of the well are lined
of current applied to the part, conductors that carry current with a relatively thin steel casing material, which is then
through hollow parts, yokes and coils. Many situations exist cemented in. This casing can only be inspected from the
in which current cannot be applied directly to the part, inside surface. The cylindrical geometry of the casing per-
because of the possibility of arc burns. Design considera- mits the flux loop to be easily calculated, so that magnetic
tions for magnetization of parts often require minimizing
the magnetic reluctance of the magnetic circuit, consisting FIGURE58. Magnetizationof shortparts by
of (1) the part, (2) the magnetizing system and (3) air gaps encirclingcoil
that might be present.
COIL
Magnetic flux leakage caused by discontinuities in the
inspected part is detected by a variety of sensors, which
include pickup coils, Hall elements, magnetodiodes, Forster
microprobes (ferroprobes) and magnetic particles. Signals
from probes are processed electronically and presented in a
manner that indicates the presence of the discontinuities.
Discussed below are the basics of systems for the inspec-
tion of many types of ferromagnetic materials, including flat
plate, wire rope and the insides of oil well casing. Next, the
inspection of new oilfield tubular products is discussed.
Types of Parts Inspected by
MagneticFlux Leakage
Short Parts with Little or No Symmetry
For the inspection of these parts, the test object may be
magnetized to saturation by the passage of current through
the part, or by placing it in an encircling coil. If hollow, a
conductor can be passed through the part, and magnetiza-
tion achieved by any of the standard techniques (these
include half wave and full wave rectified alternating current,
pure direct current from battery packs, or pulses from
capacitor discharge systems). Figure 58 illustrates the pas-
sage of a part through a coil. For irregularly shaped parts,
ELECTROMAGNETIC TESTING I 243
FIGURE 59. Setup for magnetic flux leakage saturation of the well casing is achieved. Figure 61 shows a
testing of elongated parts: (a) flux loop for wire typical inspection system (centered in the bore of the well
rope inspection with electromagnet; (b) with
permanent magnets c~sing) that consists of a magnetizing coil, soft iron pole-
pieces and sensors that ride in pads pressing against the
(aJ inside of the casing.
As with inservice well casing, buried pipelines are acces-
sible only from the inside diameter (see Fig. 62). The flux
loop is the same as for the well casing inspection system. In
FIGURE61. Inspection systemfor installed well
casing
(bJ N
FIGURE60. Wire rope inspection system CENTRALIZER
SOFT IRON POLE PIECE
SENSORS
Mt\GNETIZING COIL
SOFT IRON POLE PIECE
CENTRALIZER
EDDY CURRENT
AVERAGE WALL THICKNESS
MEASURING SYSTEM
FROM MAGNETIC DEVELOPMENT CORPORATION. REPRINTED WITH
PERMISSION.
244 I NONDESTRUCTIVE TESTING OVERVIEW
FIGURE 62. Schematic diagram of magnetic flux leakage pipeline inspection system
DRIVE PACKJ\GE FLUX LOOP RECORDER PACKJ\GE
SENSORS
FIGURE 63. Setup for magnetization of cannon this case, a drive mechanism must be provided to propel the
tube by internal conductor method inspection system through the pipeline.
CANNON TUBE Cannon Tubes
I In this elongated part, the presence of rifling affects the
ability to perform a good inspection, especially for disconti-
DIRECT CURRENT nuities that occur in the roots of the rifling. The tube is mag-
POWER SUPPLY netized to saturation (see Fig. 63) and scanned (see Fig. 64)
for the resulting residual induction with Hall elements.
L__J
FIGURE 64. Electronic detection of magnetic leakage associated with cracks in cannon tube:
tangential component of magnetic flux leakage: (a) for smooth bore tube; (bJ for rifled tube
fa) fb) fff}MOTION
8-- RIFLED BORE
-·.,·,·.-.1---~ . I , .,..•• - •·~,-··,'1·~:i
.. . • , •• l .: .- •. .-
ELECTROMAGNETICTESTING I 245
Ball Bearings and Races also act as stress risers, and cracking can originate from
them during subsequent quench-and-temper procedures.
Systems have been built for the magnetization of both Depending on the use to which the material is put, subsur-
steel ball bearings, and their races. One such system uses face discontinuities such as porosity and laminations may
especially fabricated Hall elements as detectors. also be considered detrimental. These types of discontinu-
ities may be acceptable in welds where there are no cyclic
Relatively Flat Surfaces stresses present. They may however give rise to injurious
cracking when such stresses are present.
The inspection of welded regions between flat or curved
plates is often performed using a magnetizing yoke. Sensor In the metal processing industries, grinding especially
systems include coils, Hall elements and magnetic particles. can lead to surface cracking and to some changes in surface
Figure 65 illustrates typical yoke magnetization. metallurgy. Such discontinuities as cracking have tradition-
ally been found by magnetic flux leakage methods, espe-
Threaded Regions of Pipe cially wet magnetic particle inspection.
An area that requires special attention during the inser- Service induced discontinuities include cracks, corrosion
vice inspection of drill pipe is the threaded region of the pin pitting, erosion from turbulent fluid flow or metal-to-metal
and box connections. Common problems that occur in these contact, and stress induced changes in the product metal-
regions include fatigue cracking at the roots of the threads lurgy. In those materials placed in tension and under torque,
and stretching of the thread metal. Automated systems that fatigue cracking is likely to occur. A discontinuity that arises
employ both active and residual magnetic flux leakage tech- from metal-to-metal wear is sucker rod wear in tubing from
niques can be used for detecting such discontinuities (see producing oil wells. Here, the pumping rod can rub against
Fig. 66). the inner surface of the tube, and both the rod and tube
wear thin. In wire rope, the outer strands will break after
Types of Discontinuities Found by wearing thin and inner strands sometimes break at disconti-
Magnetic Flux Leakage nuities that were present when the rope was made. Railroad
rails are subject to cyclic stresses that can cause cracking to
In the metal forming industry, discontinuities commonly originate from otherwise benign internal discontinuities.
found by magnetic flux leakage methods include overlaps,
seams, quench cracks, gouges, rolled-in slugs, and subsur- Loss of metal due to a conducting fluid in the vicinity of
face inclusions. In the case of tubulars, internal mandrel two slightly dissimilar metals is a very common form of cor-
marks (plug scores) can also be considered defects when rosion. The dissimilarity can be quite small, as for example,
they result in remaining wall thicknesses that are below at the heat treated end of a rod or tube. The result is prefer-
some specified minimum. Small marks of the same type can ential corrosion by electrolytic processes, compounded by
erosion from a contained flowing fluid. Such loss mecha-
nisms are common in subterranean pipelines, in installed
well casing, and in refinery and chemical plant tubing.
FIGURE66. Coil magnetizationsystemfor drill
pipe thread inspection;solidstate sensorsscan
threadsin both activeand residualfields
FIGURE65. Y9ke methodfor creationof high
level magneticinduction;flux lines (shown)
shouldbe roughlyperpendicularto suspected
discontinuitydirection
246 I NONDESTRUCTIVE TESTING OVERVIEW
The stretching and cracking of threads is a common magnetic flux leakage techniques during the product's li~~-
problem. For example, when tubing, casing and drill pipe time and the need for the evaluation is mandated by specifi-
are ~verto~qued at the coupling, the threads exist in their cation or by failure in similar parts. Magnetic flux leakage
plastic region. This causes metallurgical changes in the techniques are in common use for this type of inspection.
~etal, and can create regions where stress corrosion crack-
mg (~CC) takes place in highly stressed areas at a faster rate Sensors Used in Magnetic Flux
than m areas of less stress. Couplings between tubes are a Leakage Inspection
g?od example of material that may be highly stressed. Drill
~ipe threads are a good example of such stress causing plas- The Coil
tic deformation and thread root cracking.
The coil is the most commonly used magnetic flux leak-
Effects of Discontinuities age sensor because of the large surface are~ of in~pected
ferromagnetic parts. Coils can be arranged with their faces
I~ so1:1e industries, depending on the use of the product, parallel or perpendicular to the inspected surface (see
certain discontinuities can be tolerated while others cannot. Fig. 67). Coil arrays are used to increase coverage an? the
~any products are now designed so that there is less metal elements of an array might be physically or electronically
~th fewer discontinuities being used for their fabrication. connected to select only discontinuity signals and minimize
Time-to-failure calculations using fracture mechanics allow noise. The signals induced in single coils as they pass
product lifetimes to be predicted, and the product is through leakage fields depend on several factors.
~xpected to have ended its useful life by that time whether it
is defective or not. Usage criteria have also evolved, allowing The distance of the coil face to the inspected surface (h
a used product to remain in the same service or in less criti- in Fig. 67) is the first factor. As h increases, the magnetic
c_al service for long periods of time and so extend the life- flux leakage field decreases and so the coil output electro-
time_ of th~ product. Magnetic flux leakage assessments of motive force decreases. This liftoff effect can also be mea-
the i~tegnty of such used materials are quite common, sured with a Hall element gauss meter. Typical liftoff for coil
m~sdpuecsitan!elys. in the oil and gas production and distribution sensors is in the region of 1 to 3 mm (0.04 to 0.1 in.).
Cr~cks found during testing can be permitted in certain The second factor is the angle at which the discontinuity
~ate~als but not in others. The evolution of codes and spec- field is scanned by the coil. Ideally, the angle between the
ifrcat10ns has permitted, for example, some minor degree of
surfac~ cracking in new ferromagnetic tubing while at the FIGURE 6 7. Parallel and perpendicular coils
same time requiring that similar material with fatigue cracks cutting magnetic flux leakage fields from
~e removed from service. Magnetic flux leakage is used to discontinuity at speed v; for discontinuity fields
~md such cracking in new and used materials. As discussed longer than coil, output of coils is given by
m_ the previous chapter, crack depth is difficult to measure formulas shown
with accuracy by magnetic flux leakage techniques. This is
because the signals from discontinuities found by magnetic y
flux leak~ge are caused by a variety of factors, including the
dept~, width and orientation to the activating field. In view 4
of this, magnetic flux leakage signals should be considered a I
cause.o~ flux leakage that requires further investigation. I
It is important, therefore to determine the service of the x
material before using magnetic flux leakage or any other
n?ndes~ructive testing technique. Cracks and other tight LEGEND
discontmuities are stress risers and can lead to rapid failure
under the wrong circumstances. On the other hand, inclu- EPARALLEL cc (HyL - Hyr)LV
sion~, internal laminations, slag and porosity can be non- ErERPENorcuLAR"" (Hyr - Hy2)Lv
detnmental unless, by corrosion, erosion or some wear L = COIL LENGTH INTO PAGE
mec~an_i~m, they eventually become surface breaking dis-
c~ntmmties. Many materials and structures are fabricated
with w~at is considered a tolerable amount of relatively
small discontinuities. Some of this material is evaluated by
ELECTROMAGNETICTESTING I 247
long axis of the coil and the long axis of the discontinuity In the case of coils oriented with their faces perpendicu-
lar to the inspected surface (Fig. 67), the coil width affects
opening (in the case of a tight crack) should be zero degrees. the final signal because the generated field in the lower coil
Inpractice this cannot alwaysbe arranged, and the coil elec- branch exceeds that generated in the upper coil branch by
tromotive force is cancelled to some degree. This is because an amount that depends on (1) the proximity of the
one end of an elongated coil senses a rising discontinuity branches and (2) the leakage field components at the
field while the other end of the coil, which could be on the respective positions.
other side of the discontinuity, senses a field in the opposite
direction. Such a situation is shown in Fig. 68. The use of nonconducting ferrite as a core increases the
flux through the coil over that from the same discontinuity
The length of the discontinuity is the third factor affect- field scanned with an air core coil. The nonconducting
ing single coil signals. For very short discontinuities, the nature of the ferrite material eliminates the formation of
sensing coil might be longer than the discontinuity. In this eddy currents within the core material as a rapidly changing
case the entire discontinuity field is scanned. When the dis- magnetic flux leakage field is scanned. The nature of the
continuity length exceeds that of the coil length L, the coil flux leakage field also affects single coil signals. The peak
signal will be proportional to L if the discontinuity field is amplitude of coil voltages can be related to certain leakage
uniform over the distance L. In any case the coil electromo- field parameters if a model for the leakage field has been
tive force will be an integral taken along the length of the developed. If the leading and trailing edge fields (HyL, HyT
coil. in Fig. 67) or the upper and lower edge fields (Hy2, Hy1) are
known and if they relate to some discontinuity parameter,
In the case of coils that are oriented with their faces par- then it may be possible to extract this parameter from the
allel to the inspected surface, the leading edge to trailing coil voltage.
edge distance also plays a role in the development of the
output. The general shape for a symmetrical magnetic flux The coil voltage is also affected by two other factors: the
leakage field is shown in Fig. 69a. For optimal sensitivity to number of coil turns and the coil speed over a discontinuity.
tight discontinuity fields, the leading edge to trailing edge The output voltage E is directly proportional to both of
distance should be equal to the liftoff h. these factors. The relationship E = Nd<)>!dt permits the use
of formulas given in Fig. 67.
FIGURE 68. Size and angle at which
discontinuityleakage fields are scanned by The advantage of coils is that they are sturdy and rela-
finite sized coil sensorsaffects signal generated: tively cheap to manufacture. They can also be shaped or
(a) coil smaller than discontinuity(;bJ coil scans wound to suit the inspection task. Their disadvantages are
only end of discontinuity(same situation occurs their size, their velocity dependence and the need to encap-
if discontinuityis smaller than coil); (cJ coil sulate them in nonconducting material with good wear char-
scansdiscontinuityat angle (some signal acteristics.
cancellationcan occur in thissituation)
FIGURE 69. Coil signalshapes: (a) general
fa) fbJ shape of coil signalfrom parallelcoil passing
throughsymmetricaldisconitnuity field;
vB Vt B (bJ generalshape of signalfrom perpendicular
coil passing through same leakage field (signals
(-L-t-- ) -1 ----- L--- 1 are not so symmetricailf discontinuityis angled
to surface, as is case with overlap)
fa) fbJ
fcJ
V B
~1
248 I NONDESTRUCTIVE TESTING OVERVIEW
The Hall Element array of tape recorder heads. Elongated magnetic balloons
also exist for the inspection of the inside surface of tubes.
The Hall element is a solid state sensor that responds in Forster Microprobes
a linear manner to increasing magnetic field intensity pass-
ing through it. The main advantage of the device is the small These devices are small pieces of ferrite that carry one or
size of its active area. Hall devices as small as 0.1 x 0.025 mm more coils. The ferrite is excited by one of the coils at a rel-
(0.004 x 0.01 in.) have reportedly been used in the evalua- ati:ely high ~requency (f), while the signal at a frequency of
tion of bearing races, with a sensitivity of 0.5 V per 2f is taken either from the same coil or from a second coil.
ampere/tesla (0.05 volts per ampere/kilogauss). Those used This signal carries with it the information contained in the
in commercial Hall element gauss meters have larger active stray field from a discontinuity. Typical ferrite core lengths
areas. The second advantage of such devices is that they can are 2 mm (0.08 in.) with diameters of 0.1 mm (0.004 in.). A
be aligned to measure the normal (Hy) or tangential (H) typical excitation frequency would be 140 kHz (see Fig. 75).
component of the flux leakage field trom a discontinuity,
with an amplitude that is not dependent on the speed of the FIGURE 70. Flux leakage pickupsensorsmade
sensor over the discontinuity. from ferrite C-coresand Hall elements (shaded):
(aJ sensor gives unipolaroutput as it passes
The main disadvantage of the Hall device is that, like over discontinuity;(bJ sensor gives bipolar
many other solid state devices, no two have exactly the same output; (cJ Hall elementsare connectedin
sensitivity, and thus when Hall element arrays are used, oppositionso that output is also bipolar
some time must be allotted to the electronic balancing of faJ
the array. Further, because these devices require activation,
additional circuitry is needed, and because they are less fbJ
robust than a coil, they require encapsulation. Such protec-
tion can often lead to a relatively large liftoff between the fcJ
active area of the Hall element and the inspected surface.
FROM US PATENT3,529,236 11970). REPRINTEDWITH PERMISSION.
Recent advances in solid state technology allow Hall ele-
ments to be combined with power supplies and amplifiers in
one chip. Such devices also have good temperature charac-
teristics and have proven useful as inspection sensors at ele-
vated temperatures. Outputs of 75 V/T (7.5 mV per gauss)
are possible with such devices.
Hall elements can also be used with flux concentrators as
shown in Fig. 70.
The Magnetodiode
The magnetodiode is a solid state device, the resistance
of which changes with magnetic field intensity. It consists of
p-zones and n-zones of a semiconductor, separated by a
region of material that has been modified to create a recom-
bination zone (see Fig. 71). Active areas are typically 3.0 x
0.6 x 0.4 mm (0.1 x 0.02 x 0.016 in.), and output signals are
generally larger than for Hall elements, although the
response to field intensity is not so linear for higher fields as
shown in Fig. 72.
Figure 73 shows that frequency response is flat from
direct current field to 3 kHz. Fig. 74 shows that sensitivity is
stable without temperature dependence in the range of -10
to 50 °C.
Magnetic Recording Tape
For the inspection of flat plates and billets, it is possible
to scan the surface with wide strips of magnetic recording
tape. Discontinuity signals are taken from the tape by an
ELECTROMAGNETICTESTING I 249
The Magnetic Particle FIGURE 73. Magnetodiodefrequencyresponse
Magnetic particles are finely ground high permeability
oui -0 Ta= 25 °C
magnetic material that is sometimes dyed for contrast with t§ -3 ....... ._,
the metal surface. This makes the magnetic particle method
a combination of a magnetic flux leakage technique and ~> -6 ' ' ....
-9
FIGURE 71 . Schematic of magnetodiode, I::)-~
showingp-zones and n-zones in semiconductor IQ::-.). -12
materialalong with intrinsic(iJ and 0
recombination(rJ zones
0.02 0.04 0. I 0.2 0.4 I 2 4 IO 20 40
p+ n+
MAGNETIC FIELD FREQUENCY
(kHz)
FIGURE 74. Magnetodiodetemperature
dependence
FIGURE 72. Typical characteristiccurve (voltage AMBIENT TEMPERATURE
V versus magneticfield intensityHJ of 230 A (degrees Celsius)
magnetodiodeshows linearresponseup to
FIGURE 75. Forster type microprobeor
about 40 kA·m-1 (500 OeJ field at 25 -c (77 °FJ Ferroprobe™; sensor excitedat frequencyf;
response taken at frequency2f
VOLTAGE ,.,,,,,,- --
FERRITE
1.6 H (kOe)
'~ 2f
/1.4 I
u
1.2 ~,,. 1.5 2.0
J,/1.0
0.8
0.6
0.5 1.0
/0.4 0.2
,/0.2
,I-2.0 -1.5 -1.0 -o.s 0
I/ 0.4
0.6
J 0.8
1.0
./ ~'/ 1.2
1.4
-J,/' 1.6
250 I NONDESTRUCTIVE TESTING OVERVIEW
visual inspection. Ideal inspection conditions occur when a Where:
fine spray of such particles is intercepted by a flux leakage
field and some of them stick to the field. An advantage over H the magnetic field intensity (A-m-1 );
other forms of magnetic sensor is that the particles have df an element of length (m); and
zero liftoff from the discontinuity field.
I the current in the part (A).
Optimum visual conditions occur when the particles
emit light in that portion of the visible spectrum where the If the part is in the form of a cylindrical bar, the symmetry of
eye is most sensitive (yellow-green). This is particularly the situation allows H to be constant around the circumfer-
effective when all other visible light has been removed. This ence, so the closed integral reduces to:
situation is achieved when low energy ultraviolet light is
used to illuminate the particles and their dye absorbs this 21tRH = I (Eq. 9)
energy and re-emits it in the visible region.
and
Magnetic particle inspection is performed, either in
active or residual field, for a wide variety of parts. It is per- H I
formed as the primary inspection or as a followup inspection 21tR
when discontinuities have been found by other methods.
where R is the radius of the part. A surface field intensity
Typical Magnetic Flux Leakage that creates an acceptable flux leakage field from the mini-
Applications mum sized discontinuity must be used. Such fields are often
created by specifying a certain number of amperes per
Short Parts meter of part outside diameter (or amperes per inch).
For many short parts, the only sensor that can be conve- Example of Transverse Discontinuity Inspection
niently used is the magnetic particle. The part can be
inspected for surface breaking discontinuities during or Because of the demagnetizing effect of the end of a tube,
after it has been magnetized to saturation. For active field automated flux leakage inspection systems do not generally
inspection, the part can be placed in a coil carrying alternat- perform well when scanning for transverse discontinuities at
ing current and sprayed with magnetic particles, or it can be the ends of tubes. This is because the normal component Hy
magnetized to saturation by a direct current coil and of the field outside the tube is large and can obscure discon-
inspected in the resulting residual induction with magnetic tinuity signals. Inspection specifications for such regions
particles. In the latter case, the induction in the part can be often include the requirement of additional longitudinal
measured with a fluxmeter. Wet particles perform better magnetization at the tube ends, and subsequent magnetic
than dry ones because there is less tendency for the wet par- particle inspections during residual induction. This situation
ticles to fur along the field lines that leave the part. These is equivalent to the magnetization and inspection of short
methods will detect transversely oriented, tight discontinu- parts as outlined above.
ities.
The effective permeability of the metal under inspection
The flux leakage field strength from a tight crack is is quite small due to the large demagnetization field created
in the part both by the physical end of the part and by the
ifroughly proportional to the field strength H across the requirement that the flux lines must be continuous (and
must therefore have a relatively short path in the metal).
crack, multiplied by the width Lg of the crack. the inspec- Large values of the magnetizing force at the center of the
tion is performed in residual induction, the value of Hg coil are usually specified. Such values depend on the weight
(which depends on the local value of the demagnetization per unit length of the part because this quantity affects the
field in the part) will vary along the part. Thus the sensitivity LID ratio. Where the part is a tube, the LID ratio is given by
of the method to discontinuities of the same geometry varies the length of metal between poles divided by twice the wall
along the length of the part. thickness of the tube and Lis somewhat indeterminate. As a
rough example, with L = 460 mm (18 in.) and D = 19 mm
For longitudinally oriented discontinuities, the part must (0.75 in.), the LID ratio is 24. Using the McClurg formula
be magnetized circumferentially. If the part is solid, then for effective permeability:
current can be passed through the part, the surface field
intensity being given by Ampere's law:
JH·df = I (Eq. 8) µ 6~ - 5 (Eq.10)
D
ELECTROMAGNETICTESTING I 251
then, theµ value in the above example is 139. For a material to confine the particles to the surface of the inspected part.
This is not the case with dry particles, which have the ten-
with a saturation flux density of 1.8 T (18 kilogauss), the dency to fur along lines of magnetizing force.
magnetizing field intensity H (in kA-m-1) shown in Eq. 11 is
required: In many instances, it may be better to use some other
method for transverse discontinuities, such as ultrasonic or
H B (Eq.11) eddy current testing.
µµo Wire Rope Inspection
1.8 x 47t x 10-7 An interesting example of an elongated steel product
139 inspected by magnetic flux leakage methods is wire rope.
Figure 76 shows the cross sections of typical wire ropes.
10.3 kA·m-1 Such ropes are used in the construction, marine and oil pro-
duction industries, in mining and elevator shafts, and for
or in oersteds: overhead cableways for personnel and raw material trans-
portation. Inspection is performed to detect general wall
H -Bµ loss due to wear and elongation, and the breakage of inter-
nal strands. Typical equipment is shown in Fig. 60 and two
18,000 forms of flux loop are shown in Fig. 59. The type of flux loop
139 used (electromagnet or permanent magnet) depends to
some extent on the accessibility of the rope. Permanent
130 magnets might be used where taking power to an electro-
magnet might cause logistic or safety problems.
is required. An effective way of checking such rough calcula-
tions is to manufacture discontinuities on the inner surface of The general theory for the electromagnetic flux loop is as
the tube and raise the value of H in the magnetizing coil until follows. There are N electromagnet cores each oflength Lm,
the discontinuity is visible during magnetic particle testing. cross sectional area Am and effective relative permeability
When specifications are required for a wide range of tube µm, with two end pieces that have an effective length of LFe
diameters and wall thicknesses, this is the best procedure to and permeability µFe through which a wire rope passes. The
follow. Each case should be treated on its own merits. effective length of the rope in the instrument is LR, its cross
sectional area is AR and its effective relative permeability is
In practice, ends of tubes are inspected for transverse µR· The total magnetic reluctance in the magnetic circuit is:
discontinuities by the following magnetic flux leakage
techniques, both using active fields from an encircling coil.
1. In the technique using a direct current active field, + 2µFe A Fe (Eq.12)
magnetic particles are thrown at the inspected material
while it is maintained at a high level of magnetic induc- +
tion by a direct current field in the coil. This method is
particularly effective for internal cracks. Fatigue cracks where the air gaps have an effective length of LA and an
in drill pipe are often found by this method. effective cross sectional area of AA Further, if there are n
turns on each electromagnet core, each of which carries a
2. In the technique using an alternating current active current of I amperes, then:
field, magnetic particles are thrown at the inspected
material while it lies inside a coil carrying alternating nNI (Eq.13)
current. Using 50 or 60 Hz alternating current, the
penetration of the magnetic field into the material is Where: = the magnetic flux in the circuit;
quite small and the technique is only good for the jH·df = the sum of the individual magnetic reluctances;
detection of outside discontinuities.
and
When inspections for both outer surface and inner sur- = the formal statement of Ampere's law for this
face discontinuities are necessary, it may be best to inspect
first for outer surface discontinuities with an alternating cur- situation.
rent field, then for inner surface discontinuities with a direct
current field. For wet magnetic particle testing, the surface
tension of the fluids that carry the particles is large enough
252 I NONDESTRUCTIVE TESTING OVERVIEW
FIGURE 76. Cross sectional area configurations for various types of wire rope
6 x 19 CLASSIFICATION
•••• 6 x 26 ws
6 x 195 6 x 21 FW 6 x 25 FW
FILLER WIRE WARRINGTON SEALE
SEALE FILLER WIRE
6 x 37 CLASSIFICATION
6 x 31 ws 6 x 36 ws 6 x 37 2 OP 6 x 41 SFW 6 x 49 SWS
2 OPERATIONS
WARRINGTON SEALE WARRINGTON SEALE SEALE FILLER WIRE SEALE WARRINGTON SEALE
FROM WIRE ROPE CORPORATION OF AMERICA. REPRINTEDWITH PERMISSION.
By making suitable estimates of the parameters involved, Internal diameters and metal masses involved in the flux
a reasonably good estimate of the flux in the rope can be loop indicate that some form of active field excitation must
made. Because in this inspection situation, discontinuities be employed. Internal diameters of typical production or
can occur deep inside the rope material, it is essential to transportation tubes range from about 100 mm (4.0 in.) to
maintain the rope at a high value of flux density, 1.6 to 1.8 T about 1.2 m (4 ft).
(16 to 18 kilogauss). Under these conditions, breaks in the
inner regions of the rope will produce magnetic flux leakage If the material is generally horizontal, some form of drive
at the surface of the rope. mechanism is required, and because the inspection device
(commonly called a pig) may move at differing speeds, the
The problem of detecting magnetic flux leakage from magnetic flux leakage sensor should have a response inde-
inner discontinuities is compounded by the need to main- pendent of velocity. For devices that operate vertically, such
tain the sensors at some relatively large distance from the as well casing inspection systems, coil sensors can be used,
rope. This permits splices in the rope to pass through the so long as the tool is pulled from the bottom of the well at a
inspection head. Common sensors include Hall elements constant speed by the workover rig. In both types of instru-
and encircling coils (see Fig. 77). ment, the sensors are mounted in pads that are pressed
against the inner wall of the pipe.
The cross sectional area of the rope can be measured by
sensing changes in the flux loop that occur when the rope FIGURE 77. Full encircling coil for detection of
gets thinner. The air gap becomes larger, and so the value of flux leakage from discontinuities in wire rope;
the field intensity falls. This change can easily be sensed by principle of operation is same as that of parallel
placing Hall elements anywhere within the magnetic circuit. coil in Fig. 67
Internal Casing or Pipeline Inspection
Inspection of inservice well casing or buried pipelines is
often performed by magnetic flux leakage methods. Various
types of wall loss mechanisms occur, including internal and
external pitting, erosion and corrosion caused by contact or
proximity of dissimilar metals.
From the point of view of magnetizing the pipe metal in
the longitudinal direction, the two applications are identical.
ELECTROMAGNETICTESTING I 253
Schematic diagrams for a casing inspection system and a 2. the effect of any second string around the inspected
string (the additional metal contributes to the flux
pipeline pig are shown in Figs. 61 and 62. The flux loop for loop, especially in areas where the two strings touch);
each one is identical and is given by the approximate for- 3. the relatively large current that must be sent down
the wireline to raise the pipe wall to saturation (tem-
mula: .s..µ0NI peratures in deep wells can exceed 500 K (200 °C or
390 °F); and
<I> + '2t g (Eq.14)
µcAc 4. the possibility that the tool will get stuck (this occurs
Ag both downhole and underground, because external
pressures can cause the pipe to buckle).
+ ~
µPAP
Where:
number of turns on the electromagnet; FIGURE78. Typicaldata from internalcasing
the excitation current for the electromagnet; inspectiondevice
magnetic length of the electromagnet's core;
effective relative permeability of the core; CHANNEL 2
cross sectional area of the core;
air gap between poles and pipe; CHANNEL I CHANNEL 3
effective surface area of pole pieces;
magnetic length of the pipe; I 00 0 I 00 2"'scJ 0 75
effective relative permeability of the pipe; and 1,280 (4,200) Naa CIIOILIILIAIRIIAINIJD
cross sectional are of the pipe. CENTRALIZE
11111 I
Because both line pipe and casing are manufactured to INTERNAL 1
outside diameter size, there is a range of inner diameters for CORROSION
each pipe size. Such ranges may be found in specifications. iuwf I COL~
To make the value tg of the air gap as small as possible, soft Cl:::
iron attachments can be screwed to the pole pieces. Typical :J- 111 I
data from a magnetic flux leakage casing inspection system QJ - v,
are shown in Fig. 78. V) COLLAR
aaa
For the pipeline pig, a recorder package is added and 0~.E-e! 1,525 (5,000) COUAR
signals from discontinuities are tape recorded. When tapes
are retrieved and played back, areas of damage are located. CLwLl:::_~Q8.J 11
Pipe welds provide convenient magnetic markers. With the
downhole tool, magnetic flux leakage signals are sent up the ~E .JJ)i!JRNAL
wireline and processed in the logging truck at the wellhead. r.CORROSION
Figure 61 shows a system that also includes a remote field ~
eddy current testing system. 0V) II II111111111
II
A common problem with this and other magnetic flux 1,890 (6,200) "'0,.
leakage equipment is the need to determine whether mag- l -ii111
netic flux leakage signals originate from discontinuities inside aa
or outside of the pipe. Production and transmission compa- COLLAR
nies require this information because it allows them to deter-
mine which form of corrosion control to use. Inspection shoes 1111111
sometimes contain a high frequency eddy current transmit-
ter-receiver system that responds only to inside diameter dis- EXTERNAL
continuities. Thus occurrence of both magnetic flux leakage CORROSION
and eddy current signals indicates an inside diameter discon-
tinuity, whereas occurrence of only a magnetic flux leakage 1111111
signal indicates an outside diameter discontinuity.
11
Problems with this form of inspection include:
II
COLLAR
I II
1. the inability of the magnetic flux leakage system to LEGEND
measure elongated changes in wall, such as might
occur with general erosion; CHANNEL I = AVERAGE OF TWELVE MAGNETIC FLUX LEAKAGE SENSORS
CHANNEL 2 = EDDY CURRENT INTERNAL DISCONTINUITY DISCRIMINATOR
CHANNEL
CHANNEL 3 = LARGEST SIGNAL FROM ONE OF TWELVE MAGNETIC FLUX
LEAKAGE SENSORS
254 I NONDEST'RUCTIVETESTING OVERVIEW
REFERENCES
1. Lang, D.J. "Inspection of Aircraft Surfaces Using 11. Brosky, S.A., et al. "An Eddy Current Test for High
Two Single-Coil Eddy Current Techniques." Materi Strength Seamless Pipe." Association of Iron and
als Evaluation. Vol. 32, No. 4. Columbus, OH: Steel Engineers Convention. Cleveland, OH: Associ-
ation oflron and Steel Engineers (September 1970).
American Society for Nondestructive Testing (April
1974): p 87-92. 12. Mandula, J.M. "Billetscan - A New Eddy Current
2. Grubinskas, R.C. "An Evaluation of Probe Coils with Device for Total Surface Inspection of Square Bil-
Ferrite Cores for Use in Electromagnetic Testing." lets." Proceedings of the 1970 ASNT Fall Conference
Materials Evaluation. Vol. 24, No. 10. Columbus, [Cleveland, OH]. Columbus, OH: American Society
OH: American Society for Nondestructive Testing for Nondestructive Testing (October 1970).
(October 1966): p 557-563.
13. Mester, M.L. "In-Line Inspection of Hot Steel Prod-
3. Rogel, A.P. and J.J. Scalese. "Automatic Eddy Cur- uct." Proceedings of the Thirteenth Mechanical
rent Bolt-Hole Scanning System." Materials Evalua Working and Steel Processing Conference. Warren-
tion. Vol. 41, No. 6. Columbus, OH: American dale, PA: The Metallurgical Society of the American
Institute of Mining, Metallurgical and Petroleum
Society for Nondestructive Testing (June 1983): Engineers (January 1971).
p 839-843.
4. Davis, T.J. "Multifrequency Eddy Current Inspec- 14. Mester, M.L. "Progress in Eddy Current Inspection
tion with Continuous Wave Methods." Materials of Hot CW Pipe." Proceedings of the Fifteenth
Evaluation. Vol. 38, No. 1. Columbus, OH: Ameri- Mechanical Working and Steel Processing Confer
ence. Warrendale, PA: The Metallurgical Society of
can Society for Nondestructive Testing (January the American Institute of Mining, Metallurgical and
1980): p 62-68. Petroleum Engineers (January 1973).
5. Schmidt, TR. "The Remote Field Eddy Current
Inspection Technique." Materials Evaluation. 15. Mester, M.L. United States Patent No. 3,694,735.
Vol. 42, No. 2. Columbus, OH: American Society for WaterCooled EddyCurrent Transducer for Testing
Nondestructive Testing (February 1984): p 225-230. Product While at High Temperature (September
6. Melcher, J.R. and H.H. Woodson. Electromechani 1972).
cal Dynamics. New York, NY: Wiley and Sons
(1968): p 358-583. 16. Stumm, W. "New Developments in the Eddy Cur-
7. Russell, D.E., D.J. Kenway and K.R. Radesh. "New rent Testing of Hot Wire and Hot Tubes." Materials
Developments in Magnetic Tube Inspection." Pro- Evaluation. Vol. 29, No. 7. Columbus, OH: Ameri-
gram, 1986 ASNT Spring Conference. Columbus, can Society for Nondestructive Testing (July 1971):
OH: American Society for Nondestructive Testing p 141.
(1986): p 15.
8. Kikuta, T., J.L. Fisher, S.N. Rowland and W.D. Jolly. 17. Bergstrand, G. and G. Backstrom. "Hot Surface
"Far-Field Eddy Current Model for Carbon Steel Inspection by Eddy Current." Proceedings of Qual
Test3 [Cincinnati, OH]. Columbus, OH: American
Gas Pipes." Journal of Offshore Mechanics and Arc Society for Nondestructive Testing (October 1984).
tic Engineering. Vol. 109 (November 1987).
9. Palanisamy, R. "Finite Element Modeling of 18. Padilla, V.E. and J.W. Parks. "Definition of Fatigue
Remote-Field Eddy Current Nondestructive Testing Crack Geometry by Eddy Current Techniques." Sev
Phenomenon." Proceedings of the Second National enth Symposium on Nondestructive Evaluation of
Seminar on Nondestructive Evaluation of Ferromag Components and Materials in Aerospace, Weapon
netic Materials. Houston, TX: Dresser Atlas (1986). Systems, and Nuclear Applications. San Antonio, TX
10. Jackson, J.F. "Eddy Current Inspection of Smaller (1966).
Diameter, As-Rolled Steel Rounds during Rotary
Straightening." ASNT Fall Conference [Houston, 19. Parks, J.W. "Detection and Definition of Environ-
TX]. Columbus, OH: American Society for Nonde- mental Damage Using Eddy Current Techniques."
structive Testing (September 1976). Proceedings of the American Society of Nondestruc
tive Testing Symposium on Nondestructive Evalua
tion [San Antonio, TX]. Columbus, OH: American
Society for Nondestructive Testing (1973).
ELECTROMAGNETICTESTING I 255
20. Rogel, A.P. and J.J. Scalese. "Automatic Eddy Current 22. Libby, H. Introduction to Electromagnetic Nonde
structive Methods. New York, NY: R. Krieger Pub-
Bolt-Hole Scanning System." Materials Evaluation. lishing ( 1979).
Vol. 41, No. 7. Columbus, OH: American Society for
23. Libby, H. "Multiparameter Eddy Current Con-
21. HNaognedmesatireurc,tivDe.JT. e"sEtidndgy(JuCnuerr1e9n8t 3)I:mpp8e3d9a-n8c4e3. Plane cepts." Research Techniques in Nondestructive Test
ing. New York, NY: Academic Press (1970).
Analysis." Materials Evaluation. Vol. 41, No. 2.
Columbus, OH: American Society for Nondestruc-
tive Testing (February 1983): p 211-218.
256 I NONDESTRUCTIVE TESTING OVERVIEW
BIBLIOGRAPHY
1. Bahr, A.J. Microwave Nondestructive Testing. New 15. Lord, W. and R. Palanisamy. "Development of The-
York, NY: Gordon and Breach (1982). oretical Models for Nondestructive Testing Eddy
Current Phenomena." EddyCurrent Characterize
2. Blitz, J. Electrical and Magnetic Methods of Nonde tion of Materials and Structures. G. Birnbaum and
structive Testing. Bristol, United Kingdom: Adam G. Free, eds. Philadelphia, PA: American Society
Hilger (1991). for Testing and Materials (1981): pp 5-21.
3. Bozarth, R. M. Ferromagnetism. New York, NY: 16. Lord, W., ed. Electromagnetic Methods of Nonde
IEEE Press (1953). structive Testing. Vol. 3. Gordon and Breach Sci-
ence Publishers (1985).
4. Bray, D. and R.K. Stanley. Nondestructive Evalua
tion: A Tool for Design, Manufacturing, and Ser 17. McMaster, R.C. "The Origins of Electromagnetic
vice. New York, NY: McGraw-Hill (1989): p Testing." Materials Evaluation. Vol. 43, No. 8.
.523-553. Columbus, OH: American Society for Nondestruc-
tive Testing (July 1985): p 946-956.
5. Cecco, V.S., G. Van Drunen and F.L. Sharp. Eddy
Current Testing. Columbia, MO: GP Courseware 18. Metals Handbook, ninth edition: Vol. 17, Nonde
(1987). structive Evaluation and Quality Control. Materials
Park, OH: ASM International (1989).
6. Dodd, C.V. Solutions to Electromagnetic Induction
Problem'). Ph.D. dissertation. University of Ten- 19. Nondestructive Testing Handbook, second edition:
nessee (June 1967). Vol. 4, Electromagnetic Testing. Columbus, OH:
American Society for Nondestructive Testing (1986).
7. Forster, F. "Creativity and the Imagination in the
Physics of Non destructive Testing." British 20. Palanisamy, R. "Developments in Eddy Current Non-
destructive Testing." Materials Evaluation. Vol. 49,
Journal of NonDestructive Testing. Vol. 21, No. 3. No. 9. Columbus, OH: American Society for Nonde-
Northampton, United Kingdom: British Institute of structive Testing (September 1991): p 1,158-1,161.
Non-Destructive Testing (May 1979): p 119-124.
8. Hagemaier, D.J. "Eddy Current Standard Depth 21. Palanisamy, R. "Residual Magnetism Effects on the
of Penetration." Materials Evaluation. Vol. 43, Eddy Current Nondestructive Test Signal." Journal
No. 11. Columbus, OH: American Society for of Applied Physics. Vol. 61, No. 8. Woodbury, NY:
Nondestructive Testing (October 1985): p 1,438- American Institute of Physics (April 1987): p 3,205-
1,442, 1,454. 3,207.
9. Hagemaier, D.J. "Nondestructive Testing Develop- 22. Palanisamy, R. Finite Element Modeling of Eddy
ments in the Aircraft Industry." Materials Evalua Current Nondestructive Testing Phenomena. Ph.D.
tion. Vol. 49, No. 12. Columbus, OH: American dissertation. Fort Collins, CO: Colorado State Uni-
Society for Nondestructive Testing (December versity (1980).
1991): p 1,470-1,478.
23. Schmidt, TR. "The Remote Field Eddy Current
10. Hagemaier, D.J. Fundamentals of Eddy Current Inspection Technique." Materials Evaluation.
Testing. Columbus, OH: American Society for Non- Vol. 2, No. 2. Columbus, OH: American Society
destructive Testing (1990). for Nondestructive Testing (February 1984):
p 225-230.
11. Hochschild, R. "Electromagnetic Methods of Test-
ing Metals." Progress in NonDestructive Testing. 24. Stanley, R.K. and G.L. Moake. "Inspecting Oil
Vol. 1. E.G. Stanford et al., eds. New York, NY: Country Tubular Goods Using Capacitor Discharge
Macmillan Company (1959): pp 59-109. Systems." Materials Evaluation. Vol. 41, No. 7.
Columbus, OH: American Society for Nondestruc-
12. Hwang, J.H. Defect Characterization by Magnetic tive Testing (June 1983): p 779-782.
Leakage Fields. Ph.D. dissertation. Fort Collins,
CO: Colorado State University (1975). 25. Stratton, J.A. Electromagnetic Theory. New York,
NY: McGraw Hill Book Company (1941).
13. Libby, H. "Multiparameter Eddy Current Con-
cepts." Research Techniques in Nondestructive Test
ing. New York, NY: Academic Press (1970).
14. Libby, H.L. Introduction to Electromagnetic Non
destructive Test Methods. New York, NY: Wiley-
lnterscience (1971).
8SECTION
MAGNETIC PARTICLE TESTING1
Bernard Boisvert, Dayton, Ohio
258 I NONDESTRUCTIVETESTING OVERVIEW
PART 1
INTRODUCTION
Capabilities and Limitations of already using magnetic particle testing or those preparing
Magnetic Particle Techniques for advanced training in the technique.
Magnetic particle testing can reveal surface discontinu- Principles of Magnetic Particle
ities, including those too small or too tight to be seen with Testing
the unaided eye. Magnetic particle indications form on an
object's surface above a discontinuity and show the location Magnetic particle testing is a nondestructive method of
and approximate size of the discontinuity. Magnetic particle revealing surface and subsurface discontinuities in magneti-
tests can also reveal discontinuities that are slightly below zable materials. It may be applied to raw materials such as
the surface, depending on their size. billets, bars and shapes; during processes such as forming,
machining, heat treating and electroplating; and in testing
There are limits on this ability to locate subsurface dis- for service related discontinuities.
continuities. These are determined by the strength of the
applied field and by the discontinuity's depth, size, type and The testing method is based on the principle that mag-
shape. In some cases, special techniques or equipment can netic flux in a magnetized object is locally distorted by the
improve the test's ability to detect subsurface discontinuities. presence of a discontinuity. This distortion causes some of
the magnetic field to exit and reenter the test object at the
Magnetic particle testing is for ferromagnetic materials discontinuity. This phenomenon is called magneticflux leak
only: it cannot be used on nonmagnetic materials, including age (MFL). Flux leakage is capable of attracting finely
glass, ceramics, plastics or such common metals as alu- divided particles of magnetic materials that in tum form an
minum, magnesium, copper and austenitic stainless steel outline or indication of the discontinuity. The strength or
alloys. curvature of magnetic flux leakage fields is critical in causing
particles to remain held at an indication.
In addition, there are certain positional limitations: a
magnetic field is directional and for best results must be ori- One of the objectives of magnetic particle testing is to
ented perpendicular to the discontinuity. This generally detect discontinuities as early as possible in the processing
requires magnetizing operations in different directions to sequence, thus avoiding the expenditure of effort on materi-
detect discontinuities. Objects with large cross sections may als that will later be rejected. Practically every process, from
require a very high current to generate a magnetic field ade- the original production of metal from its ore to the last fin-
quate for magnetic particle tests. A final limitation is that a ishing operation, may introduce discontinuities. Magnetic
demagnetization procedure is usually required following the particle testing can reveal many of these, preventing flawed
magnetic particle process. components from entering service. Even though magnetic
particle testing may be applied during and between process-
This section provides an overview of the magnetic parti- ing operations, a final test is usually performed to ensure
cle testing process. Topics covered include (1) basic steel that all detrimental discontinuities have been detected. In
and component production and some of the discontinuities welds with a tendency toward delayed cracking, there may
produced; (2) the fundamental theory of magnetism, mag- be a specified time delay between the completion of weld-
netic flux and of magnetic field types; (3) principles of elec- ing and the final test.
trically induced magnetism and magnetizing current;
(4) testing media and processes; and (5) basic principles and The test itself consists of three basic operations: ( 1) estab-
methods of demagnetization. lish a suitable magnetic flux in the test object; (2) apply mag-
netic particles in a dry powder or a liquid suspension; and
Such data can be helpful to managers, supervisors and (3) examine the test object under suitable lighting condi-
personnel outside nondestructive testing who require gen- tions, interpreting and evaluating the test indications.
eral information on the magnetic particle testing process. It
may also be helpful for introductory studies by individuals
MAGNETIC PARTICLE TESTING I 259
PART 2
FABRICATION PROCESSES AND MAGNETIC
PARTICLE TEST APPLICATIONS
The most widely used classification system for magnetic between an ingot and a plate with a width at least twice its
particle inspection considers the origin of discontinuities in thickness.
the stages of fabrication and service. These classes may be
broadly categorized as follows. Inherent Discontinuities
1. Primary production and processing tests are used to This group of discontinuities occurs during the initial
inspect the stages of processing from pouring and melting and refining processes and during solidification
solidification of the ingot to production of basic from the molten state. Such discontinuities are present
shapes, including sheet, bar, pipe, tubing, forgings before rolling or forging is performed to produce intermedi-
and castings. These tests are typically used to locate ate shapes.
two discontinuity subgroups: (a) those formed during
solidification are called inherent discontinuities; and Pipe
(b) those formed during mill reduction are called pri When molten metal is continuously cast or poured into a
mary processing discontinuities.
mold, solidification progresses gradually, starting at the
2. Secondary processing or manufacturing and f obrica sides and progressing upward and inward. There is progres-
tion tests are used to inspect the results of processes sive shrinkage during solidification. For ingots, the last
that convert raw stock into finished components. metal to solidify is at the top and center of the mold.
Forming, machining, welding and heat treating dis- Because of the shrinkage, there is typically insufficient liq-
continuities are detected. uid metal remaining to fill the mold and a depression or cav-
ity is formed.
3. Service tests are widely used for detecting overstress
and fatigue cracking. FIGURE 1 . Crosssectionof ingot showing
shrinkagecavityat top center
Magnetic particle tests are not used to detect corrosion,
deformation or wear, three of the most common service
induced problems.
Basic Ferromagnetic Materials
Production
In the production of ferrous alloys, iron ore is converted
to steel in one or more furnaces where it is melted and
refined and where alloying elements are added. While in
the liquid state, the metal is poured into a mold and allowed
to solidify into a shape typically called an ingot, or continu-
ously cast.
Ingots are quite large and must be formed into more
manageable shapes by hot working through a series of rolls
or mills. These semiflnished shapes are called blooms, ·billets
or slabs, depending on size and shape. A bloom is an inter-
mediate product, rectangular in shape with a cross sectional
area typically larger than 0.02 m2 (36 in.2). A billet can be
round or square with a cross sectional area from 1,600 mm2
to 0.02 m2 (2.5 to 36 in.2). A slab is an intermediate shape
260 I NONDESTRUCTIVE TESTING OVERVIEW
In addition, impurities such as oxides and entrapped Ingot Cracks
gases tend to migrate to the center and top of a mold and
may become embedded in the last portions to solidify. After Contraction of the metal during solidification and cool-
solidification, the upper portion is cut off or cropped and ing of the ingot generates significant surface stresses and
discarded, removing most of the shrinkage cavity and impu- internal stresses that can result in cracking. If the cracks are
rities. However, if the cavity is deeper than normal or if the internal and no air reaches them, they are usually welded
cropping is short, some of the unsound metal will show up shut during rolling and do not result in discontinuities. If
in the intermediate shape as a void called pipe. they are open to the air or otherwise become oxidized, they
will not seal but remain in the finished product.
Figure 1 shows an ingot cross section illustrating the
shrinkage cavity and impurities in the top center. Pipe is During the rolling of an ingot into a billet, oxidized
almost always centered in the semifinished shape and is cracks form long seams. It is common practice to use mag-
undesirable for most purposes. netic particle tests of billets before additional processing.
Such preprocessing tests permit the removal of seams by
Nonmetallic Inclusions grinding, chipping or flame scarfing. If not removed before
rolling or working, seams are further elongated in finished
All steel contains nonmetallic matter that mainly origi- shapes and this may make the final product unsuitable for
nates in deoxidizing materials added to the molten metal many applications.
during the refining operation. These additives are easily oxi-
dized metals such as aluminum, silicon, manganese and oth- Primary Processing Discontinuities
ers. The oxides and sulfides of the metals make up the
majority of nonmetallic inclusions. When finely divided and When steel ingots are worked down to shapes such as
well distributed, these discontinuities are often not objec- billets, slabs and forging blanks, some inherent discontinu-
tionable. ities may remain in the finished product. Inaddition, rolling
or forming operations may themselves introduce other dis-
However, sometimes the additives collect during solidifi- continuities. The primary processes considered here
cation and form large clumps in an ingot. During primary include the hot working and cold working methods of pro-
processing these large clumps are rolled out into long dis- ducing shapes such as plate, bar, rod, wire, tubing and pipe.
continuities called stringers. Inhighly stressed components,
stringers can act as nucleation points for fatigue cracking. In Forging and casting are also included in this category
certain test objects, stringers are acceptable in a limited because they typically require additional machining or other
amount. Government and industry specifications on steel subsequent processing. All the primary processes have the
cleanliness define the amount of inclusions or stringers that potential for introducing discontinuities into clean metal.
may be accepted.
Seams
The addition of lead or sulfur to molten steel is a com-
mon practice for the alloys known as free machining steels. Seams in bar, rod, pipe, wire and tubing are usually
These alloys contain a large number of nonmetallic inclu- objectionable. They originate from ingot cracks and despite
sions that break or chip during machining operations. Mag- preprocessing tests, some cracks can be overlooked or
netic particle tests of free machining alloys often indicate an incompletely removed.
alarming number of discontinuities that are not considered
detrimental in service. Rolling and drawing operations can also produce seams
in the finished product (see Fig. 2). If the reduction on any
Blowholes of the rolling passes is too great, an overfill may then pro-
duce a projection from the billet. This projection can be
As molten steel is poured into an ingot and solidification folded or lapped on subsequent passes, producing a long
commences, there is an evolution of gases. These gases rise deep seam.
through the liquid in the form of bubbles and many escape
or migrate to the cropped portion of the ingot. The reverse also occurs if the shape does not fill the rolls,
resulting in a depression or surface groove. On subsequent
However, some gases can be trapped in the ingot, form- rolling passes, this underfill produces a seam running the
ing the discontinuities known as blowholes. Most blowholes full length of the shape. Seams originating from overfilled
are clean and will weld or fuse shut during primary and sec- rolls usually emerge at an acute angle to the surface. Seams
ondary rolling. Those near the surface may have an oxidized caused by underfilled rolls are likely to be normal to or per-
skin and will not fuse, appearing as seams in the rolled, pendicular to the surface.
forged or extruded product. Oxidized blowholes in the inte-
rior of slabs appear as laminations in plate products.
MAGNETIC PARTICLE TESTING I 261
Seams or die marks can be introduced by defective or Laminations
dirty dies during drawing operations. Such seams are often
fairly shallow and may not be objectionable, especially when Laminations in plate, sheet and strip are formed when
subsequent machining removes the seam. Seams are always blowholes or internal cracks are not fused shut during
objectionable in components that experience repeated or rolling but are flattened and enlarged. Laminations are large
cyclic stresses in service. These seams can initiate fatigue and potentially troublesome areas of horizontal discontinu-
cracks. ity.
FIGURE 2. Formationof seams and laps: Magnetic particle testing detects lamination only when it
(aJ overfillproduces excessmetal squeezed out reaches and breaks the edges of a plate. Laminations that
of rolls; (bJ lap resultswhen projectionis folded are completely internal to the test object typically lie paral-
over and forced back into bars surface during lel to its surface and cannot be detected by magnetic parti-
subsequent pass; (cJ underfillresultswhen cle procedures.
there is not enoughmetal to fill rolls; (dJ seam
in finishedbar occurswhen underfillis Cupping
squeezed tight on subsequent rollingpass
Cupping occurs during drawing or extruding operations
(a) when the interior of the shape does not flow as rapidly as the
surface. The result is a series of internal ruptures that are
serious whenever they occur (see Fig. 3).
Cupping can be detected by the magnetic particle
method only when it is severe and approaches the surface.
Cooling Cracks
Bar stock is hot rolled and then placed on a bed or cool-
ing table and allowed to reach room temperature. During
FIGURE 3. Cuppingformed during drawing or
(b) extruding
(c)
(d)
262 I NONDESTRUCTIVE TESTING OVERVIEW
cooling, thermal stresses may be set up by uneven rates of between the surface and the center. Because of their posi-
temperature change within the material. These stresses can tioning, flakes are not detectable by magnetic particle tech-
be sufficient for generating cracks (see Fig. 4). niques unless machining brings the discontinuity close to
the surface.
Cooling cracks are generally longitudinal but because
they tend to curve around the object shape, they are not Forging Bursts
necessarily straight. Such cracks may be long and often vary
in depth along their length. Magnetic particle indications of When steel is worked at improper temperatures, it can
cooling cracks therefore can vary in intensity (heavier where crack or rupture. Reducing a cross section too rapidly can
the crack is deepest). also cause forging bursts or severe cracking.
ForgingDiscontinuities Forging bursts may be internal or surface anomalies.
When at or near the surface, they can be detected by the
Forgings are produced from an ingot, a billet or forging magnetic particle method. Internal bursts are not generally
blank that is heated to the plastic flow temperature and then detected with magnetic particles unless machining brings
pressed or hammered between dies into the desired shape. them near the surface.
This hot working process can produce a number of disconti-
nuities, some of which are described below. Forging Laps
Flakes During the forging operation, there are several factors
Flakes are internal ruptures that some believe are caused that can cause the surface of the object to fold or lap.
Because this is a surface phenomenon, exposed to air, laps
by cooling too rapidly. Another theory is that flakes are are oxidized and.do not fuse when squeezed into the object
caused by the release of hydrogen gas during cooling. (see Fig. 5).
Flakes usually occur in fairly heavy sections and some Forging laps are difficult to detect by any nondestructive
alloys are more susceptible than others. These ruptures are testing method. They lie at only slight angles to the surface
usually well below the surface, typically more than halfway and may be fairly shallow. Forging laps are almost always
objectionable because they serve as fatigue crack initiation
FIGURE4. Coolingcracksindicatedwith points.
fluorescentmagneticparticles
Flash Line Tears
As the dies close in the final stage of the forging process,
a small amount of metal is extruded between the dies. This
extruded metal is called flash and must be removed by trim-
ming.
If the trimming is not done or not done properly, cracks
or tears can occur along the flash line (see Fig. 6). Flash line
tears are reliably detected by magnetic particle testing.
FIGURE5. Forginglaps in pistonrods
MAGNETIC PARTICLETESTING I 263
Casting Discontinuities WeldmentDiscontinuities
Castings are produced by pouring molten metal into Welding can be considered a localized casting process
molds. The combination of high temperatures, complex that involves the melting of both base and filler metal.
shapes, liquid metal flow and problematic mold materials Welds are subject to the same type of discontinuities as cast-
can cause a number of discontinuities peculiar to castings. ings but on a slightly different scale. In addition, other dis-
Some of these are described below. continuities ·may be formed as a result of improper welding
practices. Some of the discontinuities peculiar to weldments
Cold Shuts are described below.
Cold shuts originate during pouring of the metal when a Lack of Fusion and Lack of Penetration
portion of the molten liquid solidifies before joining with
the remaining liquid. The presence of an oxidized surface, Failure to melt the base metal results in a void between
even though it is liquid or near liquid, prevents fusion when the base and filler materials. This lack of fusion can be
two surfaces meet. This condition can result from splashing, detected by magnetic particle methods if it is close enough
interrupted pouring or the meeting of two streams of metal to the weld surface.
coming from different directions.
With lack of penetration, the root area of the weld is
Cold shuts can be shallow skin effects or can extend inadequately filled. Magnetic particle testing does not gen-
quite deeply into the casting. Shallow cold shuts called scabs erally detect lack of penetration.
can be removed by grinding. Deep cold shuts cannot be
repaired. Heat Affected Zone Cracks and Crater Cracks
Hot Tears and Shrinkage Cracks Cracks in the base metal adjacent to the weld bead can
be caused by the thermal stresses of both melting and cool-
Hot tears are surface cracks that occur during cooling ing. Such cracks are usually parallel to the weld bead. Heat
after the metal has solidified. They are caused by thermal affected zone cracking is easily detected by magnetic parti-
stresses generated during uneven cooling. Hot tears usually cle testing.
originate at abrupt changes in cross section where thin sec-
tions cool more rapidly than adjacent heavier masses. Cracks in the weld bead caused by stresses from solidifi-
cation or uneven cooling are called crater cracks. Cracks
Shrinkage cracks are also surface cracks that occur after caused by solidification usually occur in the final weld pud-
the metal cools. They are caused by the contraction or dle. Cracks caused by uneven cooling occur in the thin por-
reduction in volume that the casting experiences during tion at the junction of two beads. Magnetic particle testing is
solidification. widely used to detect crater cracks.
FIGURE 6. Flash lines and laps in forgings
Manufacturingand Fabrication
Discontinuities
Discontinuities associated with various finishing opera-
tions, e.g., machining, heat treating or grinding are
described below.
Machining Tears
Machining tears occur if a tool bit drags metal from the
surface rather than cutting it. The primary cause of this is
improperly shaped or dull cutting edges on the bit.
Soft or ductile metals such as low carbon steel are more
susceptible to machining tears than harder medium carbon
and high carbon steels. Machining tears are surface disconti-
nuities and are reliably detected by magnetic particle testing.
264 I NONDESTRUCTIVE TESTING OVERVIEW
Heat Treating Cracks FIGURE 7. Grinding cracks indicated: (aJ with
visible magnetic particles; (bJ with fluorescent
When steels are heated and quenched (or otherwise heat magnetic particles
treated) to produce properties for strength or wear, cracking
may occur if the operation is not suited to the material or faJ
the shape of the object. The most common sort of such
cracking is quench cracking, which occurs when the metal is fbJ
heated above the critical transition point and is then rapidly
cooled by immersing it in a cold medium such as water, oil
or air.
Cracks are likely to occur at locations where the object
changes shape from a thin to a thick cross section, at fillets
or notches. The edges of keyways and roots of splines or
threads are also susceptible to quench cracking.
Cracks can also originate if the metal is heated too
rapidly, causing uneven expansion at changes of cross sec-
tion. In addition, rapidly increasing heat can cause cracking
at comers, where heat is absorbed from three surfaces and
is therefore absorbed much more rapidly than by the body
of the object. Comer cracking can also occur during
quenching because of thermal stresses of uneven cooling.
Straightening and Grinding Cracks
The uneven stresses caused by heat treating frequently
result in distortion or warping and the metal forms must be
straightened into their intended shape. If the distortion is
too great or the objects are very hard, cracking can occur
during the straightening operation.
Surface cracks can also occur in hardened objects during
improper grinding operations. Such thermal cracks are ere-
ated by stresses from localized overheating of the surface
under the grinding wheel. Overheating can be caused by
using the wrong grinding wheel, a dull or glazed wheel,
insufficient or poor coolant, feeding too rapidly or cutting
too heavily. Grinding cracks are especially detrimental
because they are perpendicular to the object surface and
have sharp crack tips that propagate under repeated or
cyclic loading (Fig. 7).
Another type of discontinuity that may occur during
grinding is cracking caused by residual stresses. Hardened
objects may retain stresses that are not high enough to cause
cracking. During grinding, localized heating added to
entrapped stresses can cause surface ruptures. The resulting
cracks are usually more severe and extensive than typical
grinding cracks.
Plating, Pickling and Etching Cracks
Hardened surfaces are susceptible to cracking from elec-
troplating, acid pickling or etching processes.
Acid pickling can weaken surface fibers of the metal,
allowinginternal stresses from the quenching operation to be
relieved by crack formation (Fig. 8). Another cracking mech-
anism is the interstitial absorption of hydrogen released by
MAGNETIC PARTICLE TESTING I 265
the acid etching or electrodeposition process (Fig. 9). FIGURE 1 O. Fatiguecrackingin manufactured
components:(a) gear tooth roots;
Absorption of nascent hydrogen adds to the intem~l stress~s (b) automobilecrankshaft;(c) aircraft
component
of the object and subsequently may cause cracking. This
mechanism, called hydrogen embrittlement, can result in (a)
cracking during the etching or plating operation or at some
later time when additional service stresses are applied.
FIGURE8. Treating with acid weakenssurface
metal, allowing releaseof springsinternal
stressesthrough surface cracks
(b)
FIGURE9. Hydrogenor picklingcrackson
steel spring
(c)
266 I NONDESTRUCTIVE TESTING OVERVIEW
Service Discontinuities where surfaces do not visibly separate and magnetic particle
testing is needed to detect and locate the cracking.
Discontinuities also occur from service conditions. Some Fatigue Cracking
discontinuities such as deformation and wear are not
detected by the magnetic particle test, but the technique is Objects subjected to repeated alternating or fluctuating
useful for indicating the discontinuities listed below. stresses above a specific level eventually develop a crack
(Fig. 10). The crack continues to grow until the object frac-
Overstress Cracking tures. The stress level at which fatigue cracks develop is
called the fatigue strength of the material and is well below
All materials have load limits (called ultimate strength). the ultimate strength of the material. There is an inverse
When service stressing exceeds this limit, cracking occurs. relationship between the number of stress applications
Usually the failure is completed by surface fracture of the (cycles) and the stress level necessary to initiate cracking:
object. In this case, the crack is easy to detect and magnetic low cycles and high stress produce the same results as high
particle testing is not required. However, there are instances cycles and low stress.
FIGURE 11 . Fatiguecrack from copper Another factor contributing to fatigue cracking is the
penetrationon journal presence of surface anomalies such as copper penetration
(Fig. 11) sharp radii, nicks and tool marks. These act as
stress risers and lower both the number of cycles and the
stress level needed to initiate cracking. Fatigue cracking
typically occurs at the surface and is reliably detected by
magnetic particle testing.
Corrosion
Magnetic particle procedures are not used to detect sur-
face corrosion or pitting. However, there are secondary dis-
continuities that can be revealed by the magnetic particle
method. When objects are under sustained stress, either
internal or external, and are at the same time exposed to a
corrosive atmosphere, a particular kind of cracking results.
Known as stress corrosion cracking, this discontinuity is eas-
ily detected by magnetic particle testing.
Another occurrence related to corrosion is pitting. Pit-
ting itself does not usually produce magnetic particle indica-
tions (in some applications, sharp edged pits can hold
particles). Pitting can serve as a stress riser and often initi-
ates fatigue cracks. Fatigue cracks originating at corrosion
pits are reliably detected by the magnetic particle method.
MAGNETIC PARTICLE TESTING I 267
PART 3
MAGNETIC FIELD THEORY
Magnetic Domains Figure 13 can be duplicated by placing a sheet of paper
over a bar magnet and sprinkling iron particles on the paper.
Materials that can be magnetized possess atoms that It shows the magnetic field leaving and entering the ends or
group into magnetically saturated regions called magnetic poles of the magnet. This characteristic pattern illustrates
domains. These domains have a positive and negative polar- the term lines offorce used to describe a magnetic flux field.
ity at opposite ends. In macroscopically unmagnetized There are a number of important properties associated with
material, the domains are randomly oriented, usually paral- lines of force.
lel with the crystalline axes of the material, resulting in zero
net magnetization. 1. They form continuous loops that are never broken
but must complete themselves through some path.
When the material is subjected to a magnetic field, the
domains attempt to align themselves parallel with the exter- 2. They do not cross one another.
nal magnetic field. The material then acts as a magnet. Fig- 3. They are considered to have direction, leaving from
ure 12 illustrates the domain alignment in nonmagnetized
and magnetized material. the north pole and traveling through air to the south
pole, where they reenter the magnet and return
Magnetic Poles through the magnet to the north pole.
4. Their density decreases with increasing distance from
A magnet has the property of attracting ferromagnetic the poles.
materials. The ability to attract (or repel) is not uniform over 5. They seek the path of least magnetic resistance or
the surface of a magnet but is concentrated at localized reluctance in completing their loop.
areas called poles. In every magnet, there are two or more
poles with opposite polarities. These poles are attracted to When a bar magnet is broken into two or more pieces,
the Earth's magnetic poles and therefore are called north new magnetic poles are formed. The opposing poles attract
and south poles. one another as shown in Fig. 14.
FIGURE 12. Orientation of magnetic domains: FIGURE 13. Magnetic field surrounding bar
(a) in nonmagnetized material; (b) in magnet
magnetized material
fa)
fb)
FIGURE 1 4. Broken bar magnet illustrating
locations of newly formed magnetic poles
268 I NONDESTRUCTIVE TESTING OVERVIEW
If the center piece in Fig. 14 is reversed so that similar This continued alignment after removal from the exter-
poles are adjacent, the lines of force repel one magnet from nal field is called retentivity and can be an important prop-
the other. If one of the bars is small enough, the lines of erty in some magnetic particle testing procedures.
force can cause it to rotate so that unlike poles are again
adjacent. This illustrates the most basic rule of magnetism: Some examples of ferromagnetic materials are iron,
unlike poles attract and like poles repel. cobalt, nickel and gadolinium.
Types of MagneticMaterials Sourcesof Magnetism
All materials are affected to some degree by magnetic Permanent Magnets
fields. Matter is made up of atoms with a positively charged
nucleus surrounded by a field or cloud of negatively charged Permanent magnets are produced by heat treating spe-
electrons. The electron field is in continual motion, spinning cially formulated alloys in a strong magnetic field. During
around the nucleus. When the material is subjected to a heat treatment, the magnetic domains become aligned and
magnetic field, the electron orbits are distorted to some remain aligned after removal of the external field. Perma-
degree. The amount of this distortion (or the corresponding nent magnets are essential to modem technology; their
change in magnetic characteristics) when subjected to an applications include magnetos, direct current motors, tele-
external magnetic field provides a means of classifying phones, loud speakers and many electric instruments.
materials into three main groups: diamagnetic, paramag-
netic or ferromagnetic. Common examples of permanent magnetic materials
include alloys of aluminum, nickel and cobalt (alnico); cop-
Diamagnetic Materials per, nickel and cobalt (cunico); copper, nickel and iron
(cunife); and cobalt and molybdenum (comol). Supermag-
The term diamagnetic refers to a substance whose mag- net materials such as neodymium-iron-boron and samar-
netic permeability is slightly less than that of air. When a ium-cobalt are now common.
dimagnetic object is placed in a strong magnetic field, the
induced magnetism is in a direction opposite to that of iron. Magnetic Field of the Earth
Diamagnetic materials include mercury, gold, bismuth and
zinc. The planet Earth is itself a huge magnet, with north and
south poles slightly displaced from the Earth's axis. This dis-
Paramagnetic Materials placement results in a slight deviation between geographic
north and magnetic north.
Paramagnetic describes a substance whose permeability
is slightly greater than that of air or unity. When such mate- As a magnet, the Earth is surrounded by magnetic lines
rials are placed in a strong magnetic field, there is a slight of force as shown in Fig. 15. These lines of force make up
alignment of the electron spin in the direction of the mag- what is sometimes called the earth field and they can cause
netic flux flow. This alignment exists only as long as the para- problems in both magnetizing and demagnetizing of ferro-
magnetic material is in the external magnetic field. magnetic test objects. The earth field is weak, on the order
of0.03 mT (0.3 G).
Aluminum, platinum, copper and wood are paramag-
netic materials FIGURE1 5. Magneticfield of Earth
Ferromagnetic Materials / ,_...,
/ / > ,,I,,•1,,-~,-,
Ferromagnetic substances have a permeability that is / I
much greater than that of air. When placed in an external
magnetic field, the magnetic domains align parallel with the I I\
external field and remain aligned for some period of time I
after removal from the field. Ferromagnetism can only be I\
explained via the domain theory. Paramagnetic and diamag- I \
netic materials do not contain such domains. I I
I
I I
1 I
I I
I
I
\ I
I
\
MAGNETIC PARTICLE TESTING I 269
Movement of ferromagnetic objects through the earth objects. When mechanically induced magnetization occurs
field can induce slight magnetization. This is a problem in air- as a result of forming operations, it can be removed by sub-
craft where magnetized components can affect the com- jecting the magnetized object to a routine demagnetization
passes used in navigation. Similarly, demagnetizing can be process.
difficult if certain objects, usually long shafts, are not oriented
in an east-west direction during the demagnetization process. It can be difficult to remove mechanically induced mag-
netization resulting from cold working. Disassembly is usu-
Mechanically Induced Magnetism ally impractical and demagnetization must be accomplished
using portable yokes or cable coils. The operation is compli-
Cold working of some ferromagnetic materials, either by cated when other ferromagnetic components are near the
forming operations or during service, can magnetize the magnetized object: the demagnetizing operation can mag-
netize adjacent objects and a sequence of demagnetizing
operations must then be performed.
270 I NONDESTRUCTIVETESTING OVERVIEW
PART 4
MAGNETIC FLUX AND FLUX LEAKAGE
Circular Magnetic Fields FIGURE 16. Horseshoe magnetillustrating
fundamentalpropertiesof magnetism:
The most familiar type of magnet is the horseshoe shape fa) directionof magneticflux; fb) magneticflux
shown in Fig. 16a. It contains both a north and south pole in air aroundpoles fmovingpoles close
with the lines of magnetic flux leaving the north pole and together raises magneticflux density);fc) fusing
traveling through air to reenter the magnet at the south poles, forminga circularlymagnetizedobject;
pole. Ferromagnetic materials are only attracted and held at fd) discontinuityin circularlymagnetizedobject
or between the poles of a horseshoe magnet. and its resultingflux leakage field
(a)
If the ends of such a magnet are bent so that they are
closer together (Fig. 16b), the poles still exist and the mag- (b)
netic flux still leaves and reenters at the poles. In such a
case, however, the lines of force are closer and more dense. (c)
The number of lines of flux per unit area is called magnetic
flux density and is measured in tesla or gauss.
If the magnetic flux density is high enough, ferromag-
netic particles are strongly attracted and can even bridge the
physical gap between poles that are close enough together.
The area where the flux lines leave the pole, travel through
air and reenter the magnet is Galleda magnetic flux leakage
field.
When the ends of a magnet are bent together and the
poles are fused to form a ring (Fig. 16c), the magnet no
longer attracts or holds ferromagnetic materials (there are
no magnetic poles and no flux leakage field). The magnetic
flux lines still exist but they are completely contained within
the magnet. In this condition, the magnet is said to contain a
circular magnetic field or to be circularly magnetized.
If a crack crosses the magnetic flux lines in a circularly
magnetized object, north and south poles are immediately
created on either side of the discontinuity. This forces a por-
tion of the magnetic flux into the surrounding air, creating a
flux leakage field that attracts magnetic particles (Fig. 16d)
and forms a crack indication.
LongitudinalMagnetization (d)
If a horseshoe magnet is straightened, a bar magnet is
formed with north and south poles (Fig. l 7a). Magnetic flux
flows through the magnet and exits or enters at the poles.
Magnetic poles can be thought of as occurring wherever
field lines enter and leave a magnetized object. Ferromag-
netic materials are attracted only to the poles and such an
object is said to have a longitudinal field or to be longitudi
nally magnetized.
MAGNETIC PARTICLE TESTING I 271
If these magnetic flux lines are interrupted by a disconti- The strength and curvature of the leakage field deter-·
nuity, additional north and south poles are formed on either mines the number of magnetic particles that can be
side of the interruption (Fig. l 7b). Such secondary poles attracted to form a test indication. The greater the leakage
and their associated flux leakage fields can attract magnetic field strength, the denser the indication, so long as the mag-
particles. Even if the discontinuity is a very narrow crack, it netic flux leakage field is highly curved.
can still create magnetic poles (Fig. l 7c) that hold magnetic
materials (the magnetic flux leakage field is still finite). Subsurface Discontinuities
Magnetic Field Strength A slot such as a keyway on the backside of an object cre-
ates new magnetic poles that distort the internal flux flow. If
The strength of a flux leakage field from a discontinuity · the slot is close enough to the surface, some magnetic flux
depends on several factors: (1) the number of magnetic flux lines may be forced to exit and reenter the magnetized
lines; (2) the depth of the discontinuity; and (3) the width of object at the surface. The resulting leakage field can form a
the discontinuity's air gap at the surface (the distance magnetic particle test indication.
between the magnetic poles).
Size and strength of the indication depends on: (1) the
FIGURE1 7. Bar magnetillustrating proximity of the slot to the top surface; (2) the size and ori-
longitudinalmagnetization:(aJ horseshoe entation of the slot; and (3) the intensity and distribution of
magnetstraightenedinto bar magnetwith the magnetic flux field. A similar effect occurs if the discon-
north and southpoles; (bJ bar magnet tinuity is completely internal to the object. Figure 18 is an
containingmachinedslotand corresponding illustration of a keyway on the far side of a bar and Fig. 19
flux leakage field; (cJ crackin longitudinally illustrates a midwall discontinuity.
magnetizedobject,producingpolesthat attract
and hold magneticparticles FIGURE 1 8. Slot or keywayon reverseside of
magnetizedbar
faJ l N~,'',,- ... ,: ,~s l
FIGURE 19. Internal or midwall discontinuity
I II in magnetizedtest object;there may or may
\\ II not be magneticflux leakage, dependingon
~ 1--1 value of flux in object
I\ uI I
\\
II
'...l
S
N
N---~fbJ ~N~IC0PARTICl£S
.---------.5-.:·.:·~:.N.....____.
fcJ ~NETIC PARTICLES
-=--=--~--=--=-- --=-==---=-=-~=---~-=----::::::::-:==-=--~-~-==--==-~=-----==-
272 I NONDESTRUCTIVE TESTING OVERVIEW
Effect of DiscontinuityOrientation FIGURE20. Flux leakage fieldsfrom
discontinuitieswith differentorientations:
The orientation of a discontinuity in a magnetized object (aJ perpendicularto magneticflux; (bJ at
is a major factor in the strength of the magnetic flux leakage 45 degreeangle to magneticflux; (cJ parallel
field that is formed. This applies to both surface and internal to magneticflux
discontinuities. The strongest magnetic flux leakage field is (aJ
formed when the discontinuity is perpendicular to the mag-
netic flux flow. If the discontinuity is not perpendicular, the (bJ
strength of the magnetic flux leakage field is reduced and
disappears entirely when the discontinuity is parallel to the j__ =~_=~=ifcJ ;'.----------------171
magnetic flux flow. -1 -~-;
Figure 20 illustrates the effect of discontinuity orienta-
tion on the strength of the magnetic flux leakage field.
Formation of Indications
When magnetic particles collect at a flux leakage site,
they produce an indication visible to the unaided eye under
the proper lighting conditions. Tighter magnetic flux leak-
age fields create the ideal conditions for particle attraction,
the force on an individual particle being given by the prod-
uct of terms involving the particle shape, the local value of
the field strength and a term involving the local curvature
of the magnetic flux leakage field. Particles then tend to
align end to end in this field. The stronger the ability to
hold particles, the larger the indication.
MAGNETIC PARTICLE TESTING I 273
PART 5
ELECTRICALLYINDUCED MAGNETISM
Circular Magnetization . Inducing Circular Magnetization in a Test Object
When an electric current flows through a conductor such Figure 22 illustrates a method for inducing a circular
as a copper bar or wire, a magnetic field is formed around field using a magnetic particle testing unit. The test object is
the conductor (Fig. 21a). The direction of the magnetic clamped between the contact plates so that electric current
lines of force is always 90 degrees from the direction of cur- passes through it.
rent flow. When the conductor has a uniform shape, the flux
density or number of lines of force per unit area is uniform When tubes are tested by passing a current through
along the length of the conductor and uniformly decreases them, the magnetic flux rises from zero at the inner surface
as the distance from the conductor increases. or axis toward its maximum value at the outside surface. The
inside surface is often equally important when testing for
Because a ferromagnetized object is a large conductor, discontinuities. Because a magnetic field surrounds a con-
electric current flowing through the object forms a circular ductor, it is possible to induce a satisfactory field in the tube
magnetic field. This magnetic field is known as circumferen by inserting a copper bar or some other conductor through
tial magnetization because the magnetic flux lines form the component and passing the current through the bar.
complete loops within the object (Fig. 21b).
This method is called internal conductor magnetization.
A characteristic of circumferential magnetic fields is that Figure 23 indicates a method employed for circular field
the magnetic flux lines form complete loops without mag- inspection of short parts. For longer parts such as steel
netic poles. Because magnetic particles are only attracted to tubes, an insulated metal rod is run along the bore of the
and held where flux lines exit and enter the object surface, tube, exciting it with a high current. The rod should not
indications do not occur unless a discontinuity crosses the touch the inside of the tube and the tube should be placed
flux lines. The resulting accumulation of magnetic particles on wooden planks to isolate it from the ground. These mea-
forms an outline of the discontinuity over its exact location. sures should eliminate arc bums from magnetizing current
or eddy current grounding.
FIGURE21 . Magneticfield generated: Magnetic Field Direction
(a) aroundconductorcarryingelectriccurrent;
(b) around ferromagnetictest object used as The direction of the magnetic lines of force is always at
conductor right angles to the direction of the magnetizing current. An
easy way to determine the direction of the magnetic flux is
faJ MAGNETIC FIELD
FIGURE22. Inducingcircumferentiaml agnetic
~L 1=.(. ,)4~..J ~~II= field in object used as conductor
MAGNETIZING CURRENT
L'"j/J ELECTRIC
CONDUCTOR CURRENT
fbJ MAGNETIC FIELD
=L c%e:Bc~MAGNETIZING CURRENT
L TEST OBJECT
274 I NONDESTRUCTIVETESTINGOVERVIEW
to imagine the conductor held in your right hand with the conductor is oriented in a lengthwise direction by forming
thumb extended in the direction of the electric current flow. the conductor into a coil (Fig. 25). Application of the right
Your clenched fingers then point in the direction of the hand rule shows that the magnetic field at any point within
magnetic flux flow. This is known as the right hand rule the coil is in a lengthwise direction.
(Fig. 24).
When a ferromagnetic object is placed inside a coil car-
LongitudinalMagnetization rying an electric current (Fig. 26), the magnetic flux lines
concentrate themselves in a longitudinal direction. An
Electric current can be used to induce longitudinal fields object that has been longitudinally magnetized is character-
in ferromagnetic materials. The magnetic field around a ized by poles close to each end where the field lines leave
and enter the test object to form continuous loops around
FIGURE23. Inducingcircumferentiaml agnetic the associated current.
field usinginternal conductorfor: fa) tube with
insideand outsidesurfacediscontinuities; When a longitudinally magnetized object contains a
fb) multiplering shapeswith crackingon inside transverse discontinuity, a leakage field is produced that
and outsidesurfaces attracts magnetic particles and forms an indication. Fig-
ure 27 illustrates a typical coil found on magnetic particle
(aJ CONDUCTOR test systems used to locate transverse discontinuities.
FIGURE25. Formationof longitudinal
magneticfield usingcoiledconductor
MAGNETIZING MAGNETIC FIELD
CURRENT
CRACKS CURRENT
fbJ FIGURE26. Testobject containinglongitudinal
magneticfield induced by coil
MAGNETIZING
CURRENT PATH OF PERMANENT OR FLEXIBLE CABLE
MAGNETIZING MAGNETIC LINES OF FORCE
FIGURE24. The right hand rule indicates
directionof magneticflux flow based on CURRENT
directionof magnetizingcurrent
LONGITUDINAL CRACK
(NOT DETECTED)
FORTY-FIVE DEGREE TRANSVERSE
CRACK (DETECTED) CRACK
(DETECTED)
MAGNETIC PARTICLE TESTING I 275
Multidirectional Magnetization FIGURE27. Formationof transverse
discontinuityindicationduring longitudinal
When testing for discontinuities in different directions, it magnetization
is standard practice to perform two tests, one with circular
magnetization and the other with longitudinal. Two or more
fields in different directions can be imposed on an object
sequentially and in rapid succession.
When this is done, magnetic particle indications are
formed when discontinuities are favorably oriented to the
direction of any field. Such indications persist as long as the
rapid alternations of current continue.
276 I NONDESTRUCTIVE TESTING OVERVIEW
PART 6
MAGNETIC PARTICLE TEST SYSTEMS
Stationary Magnetic Particle Test testing unit. The size and weight of power packs prevent
Systems moving them and test objects are accordingly transported
to the test site. The rating or current output of commercial
Wet method horizontal magnetic particle test systems typ- power packs varies widely but is typically from 6 to 20 kA of
ically consist of (1) a high current, low voltage magnetizing magnetizing current.
source; (2) head stock and tail stock for holding test objects
and providing electrical contact for circumferential magneti- The current is applied by cable wraps, formed coils,
zation; (3) a movable coil for longitudinal magnetization; and clamps and prods. Most power pack units incorporate an
(4) a particle suspension tank with an agitation system. adjustable current control, one or two ammeters and an auto-
matic shot duration timer.
The basic components, along with magnetizing control
indicators and ampere meters, are enclosed within a table- Mobile and Portable Testing Units
top structural frame. Systems are available in a large num-
ber of sizes from a 25 mm (1 in.) contact plate opening up to There are many applications where it is not possible to
systems that are 6 m (20 ft) long. The systems provide alter- bring the test object to the magnetic particle system. Mobile
nating current, direct current or a combination of the two, units are one type of equipment that can be transported to
with maximum magnetizing current output from 1 to 10 kA. the test site and still provide relatively high magnetizing cur-
Figure 28 shows a typical stationary or wet horizontal unit. rents (Fig. 29). Traditional mobile units may be considered
small versions of the power pack systems. Some mobile
Power Packs units have a magnetizing current output of 6 kA but most
are limited by size considerations to between 3 and 4 kA.
Power packs are the electrical sources needed to pro- Transportability is also improved by restricting the types of
duce high amperage, low voltage magnetizing current. magnetizing current to alternating current and half-wave
They are used to magnetize test objects such as castings direct current. Magnetizing current is applied to the test
and forgings that are too large to be placed in a stationary object by cable wraps, formed coils, prods and clamps. Oil
field portable magnetizing units can reach 15 kA by capaci-
tor discharge through internal conductors or cable wraps.
FIGURE 28. Typical wet horizontal magnetic FIGURE 29. Typical mobile magnetic particle
particle test system test system
MAGNETIC PARTICLE TESTING I 277
The term portable equipment refers to compact, applied to the coil, a longitudinal magnetic field is created in
lightweight units that can be hand carried to the test site the core and transmitted to the legs. When coupled to a test
(Fig. 30). Some portable units are mounted on wheeled object, a longitudinal magnetic field is generated between
carts to facilitate portability. Like stationary and mobile the poles as shown in Fig. 33.
equipment, portable units come in a variety of sizes, shapes,
weights and amperage outputs. The most common method Yokes are often specified by their lifting ability or the
of applying current with a portable unit is with prods or surface field they create midway between their poles, as
clamps. However, cable wraps and formed coils are also measured with a gauss meter.
used in many applications. Reduced weight and size are
achieved by omitting the step down transformer needed for FIGURE31 . Circularmagneticfield generated
demagnetization. around magnetizingprods
Prods and Yokes FIGURE32. Hand probe or yoke with
transformer
Prods are magnetization accessories that may be used
with stationary, power pack, mobile and portable units.
They typically consist of a pair of copper bars 12 to 20 mm
(0.5 to 0.75 in.) in diameter with handles and connecting
cables. One of the prod handles has a trigger to remotely
activate the magnetizing current from the unit's mainframe.
Prods set up a circular magnetic field that diminishes in
intensity as the distance between prods increases (Fig. 31).
Yokes are often cable connected to a mobile or portable
unit that provides the magnetizing current. A yoke designed
with a self-contained magnetizing source is often called a
hand probe. Hand probes contain small transformers that
generate low voltage and high current (Fig. 32). Yokes usu-
ally contain a magnetizing coil with a core of laminated
transformer iron. Attached to the core are legs that may
either be fixed or articulated. When magnetizing current is
FIGURE30. Portable magneticparticletesting
systemwith fixed distanceprod assembly
FIGURE33. Longitudinalmagneticfield
generatedby yoke
CURRENT
278 I NONDESTRUCTIVETESTING OVERVIEW
PART 7
FERROMAGNETIC MATERIAL
CHARACTERISTICS
Magnetic Flux and Units of B as shown in Fig. 34, the initial curve for an unmagnetized
Measure piece of steel. Starting at point O (zero magnetic field
strength and zero magnetic flux) and increasing H in small.
A magnetic field is made up of flux lines within and sur- increments, the flux density in the material increases quite
rounding a magnetized object or a conductor carrying an rapidly at first, then gradually slows until point A is reached.
electric current. The term magnetic flux is used when refer- At point A, the material becomes magnetically saturated.
ring to all of the lines of flux in a given area. Flux per unit Beyond the saturation point, increases in magnetic field
area is called magnetic flux density (the number of lines of strength do not increase the flux density in the material. In
flux passing transversely through a unit area). diagrams of full hysteresis loops, the curve OA is often
drawn as a dashed line because it occurs only during the ini-
There can be some confusion about the units of measure tial magnetization of an unmagnetized material. It is
used to define these magnetic quantities. The unit of mag- referred to as ,the virgin curve of the material.
netic flux was originally called a maxwell with one maxwell
being one line of flux. The unit of flux density was the gauss When the magnetic field strength is reduced to zero
with one gauss equal to one maxwell per square centimeter. (point B in Fig. 34b), the flux density slowly decreases. It
In 1930, the International Electrotechnical Commission lags the field strength and does not reach zero. The amount
redefined and renamed the gauss as an oersted, or the inten- of flux density remaining in the material (line OB) is called
sity of a magnetic field in which a unit magnetic pole experi- residual magnetism or remanence. The ability of ferromag-
ences a force of one dyne.2 netic materials to retain a certain amount of magnetism is
called retentivity.
In 1960, the International Organization for Standardiza-
tion released ISO 1000: The International System of Units Removal of residual magnetism requires the application
(SI). This document standardizes the metric units for mag- of a magnetic field strength in an opposite or negative direc-
netic flux. Flux intensity is measured using the weber (Wb) tion (Fig. 34c). When the magnetic field strength is first
with one weber equal to 108 lines of flux. The fluxdensity unit reversed and only a small amount is applied, the flux density
is the tesla (T) or one weber per square meter (Wb-m-2); slowly decreases. As additional reverse field strength is
applied, the rate of reduction in flux density (line BC)
1 Wb-m-2 = 1 T = 10,000 gauss. increases until it is almost a straight line (point C) where B
equals zero.
Magnetic Hysteresis
The amount of magnetic field strength necessary to
All ferromagnetic materials have certain magnetic prop- reduce the flux density to zero is called coercive force. Coer-
erties specific to that material. Most of these properties are cive force is a factor in demagnetization and is also very
described by a magnetic hysteresis curve. The data for the important in eddy current testing of ferromagnetic materials.
hysteresis curve are collected by placing a bar of ferromag-
netic material in a coil and applying an alternating current. As the reversed magnetic field strength is increased
By increasing the magnetizing field strength H in small beyond point C, the magnetic flux changes its polarity and
increments and measuring the flux density B at each incre- initially increases quite rapidly. It then gradually slows until
ment, the relationship between magnetic field strength and point D is reached (Fig. 34d). This is the reverse polarity
flux density can be plotted. saturation point and additional magnetic field strength will
not produce an increase in flux density.
The relationship between magnetic field strength and
flux density is not linear for ferromagnetic materials. A spe- When the reversed magnetic field strength is reduced to
cific change in H may produce a smaller or larger change in zero (point E in Fig. 34e) the flux density again lags the
magnetic field strength, leaving residual magnetism in the
material (line OE). The flux densities of the residual mag-
netism from the straight and reversed polarities are equal
(line OB is equal to line OE).
Removal of the reversed polarity residual magnetism
requires application of magnetic field strength in the original
MAGNETIC PARTICLE TESTING I 279
FIGURE 34. Hysteresisdata for unmagnetized steel: (a) virgincurve of hysteresisloop; (b) hysteresis
loop showingresidual magnetism; (c) hysteresisloopshowingcoerciveforce; (d) hysteresisloop showing
reverse saturationpoint; (e) hysteresisloopshowingreverse residual magnetism; (f) complete hysteresis
loop
faJ B+ fdJ 8+
A
A H+
ZERO FLUX DENSITY AND / H+ H- 8-
ZERO MAGNETIC STRENGTH ~ / 8+
H- I A
0 I
I ! ;----. . H+
I
O RESIOUl\t MAGNETCM
REVERSE MAGNETIZATION ~ D 8-
SATURATION POINT 8+
A
8- H+
fbJ 8+ A feJ 8-
H+
RESIDUAL MAGNETISM~ 8 H-
0
H- D
REVERSE
8- MAGNETIZATION
fcJ B+ POINT
c 0 A ff)
8- H+
/H- H-
COERCIVE FORCE D
REVERSE
RESIDUAL POINT
LEGEND
=A SATURATION POINT
8 = MAGNETIC FLUX DENSITY
=H MAGNETIC FIELD STRENGTH
280 I NONDESTRUCTIVETESTINGOVERVIEW
direction. Flux density drops to zero at point F in Fig. 34f The reciprocal of permeability is reluctance, defined as
with the application of coercive force OF. Continuing to the resistance of a material to changes in magnetic field
increase the field strength results in the magnetic polarity strength.
changing back to its original direction. This completes the
hysteresis loop ABCDEF (note that the curve CDEF is a Magnetic properties and hysteresis loops vary widely
mirror image of curve ABCF). between materials and material conditions. They are affected
by temperature, chemical composition, microstructure and
Magnetic Permeability grain size. Figure 36a is a hysteresis loop for hardened steel
and the loop is typical of a material with low permeability,
One of the most important properties of magnetic mate- high reluctance, high retentivity and high residual mag-
rials is permeability. Permeability can be thought of as the netism that requires high coercive force for removal. Fig-
ease with which materials can be magnetized. More specifi- ure 36b is the hysteresis loop for an annealed low carbon
cally, permeability is the ratio between the flux density and steel. It is typical of a material with high permeability, low
the magnetic field strength (B divided by H). Figure 35a is reluctance, low retentivity and low residual magnetism that
the virgin curve of a high permeability material and Fig. 35b requires a low coercive force for removal.
is the curve of a low permeability material.
FIGURE36. Hysteresisloops: (a) hardened
FIGURE35. Magneticpermeabilitycurves: steel hysteresisloop; (bJ annealedlow carbon
(a) high permeabilityvirgin curve; (bJ low steel hysteresisloop
permeabilityvirgin curve
fa)
fa) SATURATION POINT
i:::
/ / vzi
I / ui
i::: I 0
vwzi
I x:3
0 I
LL
I
POSITIVE N1AGNETIC
3 FIELD STRENGTH
.....J
LL
fbJ POSITIVE N1AGNETIC fbJ i:::
i::: FIELD STRENGTH vzwi
zvi
SATURATION POINT 0
w x
0 \
::)
x / .....J
/ LL
::) /
.....J / COERCIVE FORCE
LL
POSITIVE N1AGNETIC POSITIVE N1AGNETIC
FIELD STRENGTH FIELD STRENGTH
MAGNETIC PARTICLE TESTING I 281
PART 8
TYPES OF MAGNETIZING CURRENT
In the very early days of magnetic particle testing, it was Use of Alternating Current in Magnetic Particle
believed that the most desirable current for magnetization Tests
was direct current provided by storage batteries. As knowl-
edge of magnetic particle process expanded and electrical There are three primary advantages to using alternating
circuitry continued to advance, many types of magnetizing current as a magnetizing source. First, the current reversal
currents became available: alternating current, half-wave causes an inductive effect that concentrates the magnetizing
direct current and full-wave direct current. The terms half flux at the object surface (called the skin effect) and provides
wave rectified direct current and fullwave rectified direct enhanced indications of surface discontinuities. This is
current are used for alternating current rectified to produce especially important for inspection of irregularly shaped
half-wave and full-wave direct current. components, such as crankshafts. Magnetic fields produced
by alternating current are also much easier to remove dur-
AlternatingCurrent ing demagnetization. A third advantage is that the pulsing
effect of the flux caused by the current reversals agitates the
Alternating current is useful in many applications particles applied to the test object surface. By increasing
because it is commercially available in voltages ranging from particle mobility, this agitation allows more particles to col-
llO to 440 V. Electrical circuitry to produce alternating lect at flux leakage points and so increases the size and visi-
magnetizing current is simple and relatively inexpensive bility of discontinuity indications.
because it only requires transforming commercial power
into low voltage, high amperage magnetizing current. Concentration of the flux at the test object surface also
can be a disadvantage because most subsurface discontinu-
In the United States and some other countries, alternat- ities are not detected. Another disadvantage is that some
ing current alternates sixty times in a second. Many other specifications do not allow the use of alternating current on
countries have standardized fifty alternations per second. plated components when the coating thickness exceeds
The alternations are called cycles. One hertz (Hz) equals 0.08 mm (0.003 in.). The flux in a test object may not be at
one cycle per second and 60 Hz is sixty cycles per second. peak value, depending on where within the magnetizing
Figure 37 shows the waveform of alternating current. In one cycle the current is turned off. Alternating current is more
cycle, the current flows from zero to a maximum positive effective than direct current on objects with thick nonmetal-
value and then drops back to zero. At zero, it reverses direc- lic coatings.
tion and goes to a maximum negative peak and returns to
zero. The curve is symmetrical with the positive and nega- Half-Wave Direct Current
tive lobes being mirror images.
When single-phase alternating current is passed through
FIGURE 37. Waveform of alternating current a simple rectifier, the reversed flow of current is blocked or
clipped. This produces a series of current pulses that start at
..-- I CYCLE ~ zero, reach a maximum point, drop back to zero and then
pause until the next positive cycle begins. The result is a
~ varying current that flows only in one direction. Figure 38
wzvi shows the waveform for half-wave direct current.
0 Half-wave direct current has penetrating power compa-
rable to single-phase full-wave direct current. Half-wave cur-
:x3 rent has a flux density of zero at the center of a test object
and the density increases until it reaches a maximum at the
LL object surface. The pulsing effect of the rectified wave pro-
duces maximum mobility for the magnetic particles; dry
TIME method tests are enhanced by this effect. Another distinct
advantage of half-wave direct current is the simplicity of its
282 I NONDESTRUCTIVE TESTINGOVERVIEW
electrical components. It can be easily combined with current. The current fluctuation causes a skin effect that is
portable and mobile alternating current equipment for not significant. It is also possible to incorporate switching
weld, construction and casting tests. devices in the circuitry that reverse the current flow. This
permits built-in reversing direct current demagnetization.
One of the disadvantages of half-wave magnetization is Because of its simpler components, the initial cost of single-
the problem in demagnetization: the current does not phase full-wave direct current equipment is much less than
reverse so it cannot be used for demagnetizing. Alternating that of three-phase full-wave equipment.
current can be used to remove some residual magnetism
but the skin effect of alternating current and the deeper One disadvantage of single-phase units is the input
penetration of half-wave direct current cause incomplete power requirement. Single-phase equipment requires 1.73
demagnetization. times more input current than three-phase units. This
becomes very significant at higher magnetizing currents
Full-Wave Direct Current where input values can exceed 600 A.
It is possible for electrical circuitry to not only block (or Three-Phase Full-Wave Direct
rectify) the negative flowing current, but to invert it so that Current
the number of positive pulses is doubled. Figure 39 shows
the waveform of single-phase full-wave rectified alternating Commercial electric power, especially at 220 and 440 V,
current. The resulting current is usually called singlephase is provided as three-phase alternating current with each
fullwave direct current. phase providing part of the total current. Figure 40 shows
the waveform of three-phase alternating current. Three-
Single-phase full-wave direct current has essentially the phase full-wave magnetic particle equipment rectifies all
same penetrating ability as three-phase full-wave direct three alternating current phases and inverts the negative
flow to a positive direction, producing a nearly flat line
FIGURE38. Half-wavedirectcurrentwaveform direct current magnetizing current. Figure 41 shows the
waveform of three-phase full-wave direct current.
+ FIGURE39. Single-phasefull-wavedirect
0 current waveform
ALTERNATING CURRENT INPUT _... I:~zf--
UJ
0:::
0:::
u::)
TIME
FIGURE40. Waveformof three-phase
alternating current
HALF-WAVERECTIFIER ______.. 1/60 2/60
----·---
(\
SECONDS
HALF-WAVE
DIRECT CURRENTOUTPUT
The words you are searching are inside this book. To get more targeted content, please make full-text search by clicking here.
ASNT NDT Handbook Volume 10 OVERVIEW
Discover the best professional documents and content resources in AnyFlip Document Base.
Search
ASNT NDT Handbook Volume 10 OVERVIEW
- 1 - 50
- 51 - 100
- 101 - 150
- 151 - 200
- 201 - 250
- 251 - 300
- 301 - 350
- 351 - 400
- 401 - 450
- 451 - 500
- 501 - 550
- 551 - 600
Pages: