The Effect of Current in ...

The Effect of Current in Magnetic Particle Inspection

The types of current used to create the magnetizing effect and the factors
used to determine how much current to use are explained

BY ARTHUR R. LINDGREN


Magnetic particle inspection (MPI) is a nondestructive method for locating surface and near-surface defects in ferromagnetic (iron or steel) parts, by magnetizing the parts and indicating any flaws present using finely divided magnetic particles. Its use is limited to parts having a high material permeability, 100 or more, and therefore cannot be used on nonferrous parts such as copper, brass, aluminum, titanium, austenitic stainless steel, or any of the precious metals.

Today, the method sees action in every type of metalworking plant, either in process control, final inspection, or both. It is used on all types of ferrous castings, forgings, and fabrications during manufacture. Because of its tremendous ability in locating fine shallow surface fatigue cracks, the method is also used in preference to all others for overhaul or maintenance of machinery.

Inspectors should keep in mind the words of Carl Betz, a founding father of Magnaflux Corp. and author of Principles of Magnetic Particle Testing: "The purpose of using magnetic particle inspection is not to find cracks in ferromagnetic parts. Its purpose is to assure the user that a given part is defect-free. The method should detect small, as well as large, defects on all surfaces of the part."

The magnetic particle method makes use of the principle that flux lines flowing within a part, near and tangent to its surface, will be distorted at the location of a discontinuity. A discontinuity is any interruption in the normal physical configuration or composition of the part. Such a distortion causes some of the flux lines to exit and then reenter the part. This phenomenon is called "magnetic flux leakage." The very small north pole on one side of the discontinuity and south pole on the other side are capable of attracting finely divided magnetic particles and form an outline or "indication" of the imperfection. In effect, the particles provide a permeable bridge for the flux lines to travel across.

Types of Magnetizing Current
There are four types of magnetizing currents used in MPI, and there are typical uses for each type. They are as follows:

  • Alternating current

  • Half-wave current

  • Single-phase, full-wave current

  • Three-phase, full-wave current

    It is important that the MPI inspector know the pluses and minuses of each and on what application they find their best use.

    Alternating Current (AC)

    Fig. 1 — Alternating current.
    Figure 1 shows that with AC, the current rises to a peak in one direction, falls to zero, reverses direction, and repeats the cycle in the other direction. The cycle continues 60 times per second. Since this type of current is passing through zero 120 times per second, the magnetic field it generates is not able to penetrate very far beneath the surface of the part. The current and the field generated stay near the skin of the part, no matter how high the value of current used. This is called the "skin effect." Alternating current works very well for locating cracks open to the surface. This includes most fatigue cracks, grinding cracks, heat treat cracks, thermal cracks found in multipass welds, handling cracks in gray iron castings, and all cracks that occur in malleable castings. Under production-like conditions, using the wet method, AC should not be relied on to disclose inclusions more than 0.01 in. (0.25 mm) below the surface, although the field itself might penetrate a bit further.

    Half-Wave Rectified Current (HWDC)

    Fig. 2 — Half-wave rectified (HWDC).
    Half-wave rectified current is a form of direct current since it flows in only one direction — Fig. 2. It is "pulsed" with wide intervals between pulses. The DC provides good penetration and the pulsing makes particle mobility excellent when used with dry magnetic particles. It is great for dry powder multipass weld inspection locating not only thermal cracks, but also other defects like slag inclusions, root cracks, undercuts, and incomplete fusion. An experienced inspector, using dry powder and the "continuous" method, can easily detect flaws 0.25 in. (6.35 mm) below the surface.

    When used with the wet method, HWDC loses its deep penetrating detection qualities since particle mobility now is controlled by bath flow rather than the pulsing. However, it can be relied on to find near-surface defects defined as those just out of range with AC, slightly deeper than 0.01 in. Examples would be a forging lap or a fatigue crack when the surface has been cold worked, closed up, and fused. Near-surface defects also include grinding cracks where the surface has gone into compression. These cannot be located with AC, but may sometimes be located with HWDC. Under production-like conditions, using the wet method, HWDC should not be relied on to disclose subsurface defects more than 0.025 in. (0.65 mm) below the surface.

    Single-Phase, Full-Wave Current (1FWDC)

    Fig. 3 — Single-phase, full-wave rectified (1FWDC).
    In single-phase, full-wave current, rectifier circuitry changes the direction of current flow during the "negative" portion of the AC cycle so that current always flows in one direction — Fig 3. Single-phase, full-wave DC is very close to pure DC. Like HWDC it is also "pulsed," but the pulses are twice as frequent with no interval between. For this reason, particle mobility is only fair. It is seldom used with prods and dry powder since HWDC, is far superior for locating deep seated defects. Penetration is excellent on wet applications. For production work, it can be relied on to locate flaws 0.05 in. (1.3 mm) below the surface and maybe a bit deeper if conditions are ideal.

    Its overall sensitivity, when used with the wet method, is equal to that which can be attained using three-phase, full-wave current (3FWDC). Repeated tests have shown that the slight difference in "ripple" between single and three phase does not affect inspection sensitivity when using the wet method.

    The manufacture of magnetizing equipment producing 1FWDC over 6000 A is impractical due to the heavy primary current required; therefore, all 10,000- and 20,000-A units use 3FWDC current.

    Three-Phase, Full-Wave Current (3FWDC)

    Fig. 4 — Three-phase, full-wave rectified (3FWDC). Three phases of the line are rectified and then combined in one output for use in magnetization (360 pulses per second). This is nearly equivalent to pure DC, about 15% "ripple."
    On three-phase, full-wave units, all phases of the input line circuit are used — Fig. 4. They are each rectified and then combined into one output for use in magnetization. The resulting current output has very good penetration nearly equivalent to pure DC. Sensitivity wise, it compares with 1FWDC current. For production work it can be relied on to locate flaws 0.05 in. (1.3 mm) below the surface and perhaps a bit deeper. Three phase is found in the larger 6000- to 20,000-A units such as those used on heavy aerospace parts.

    Sensitivities
    Table 1 shows estimated sensitivities with regard to defect depth and current type. Sensitivity will vary by product and type of defect. You should use these with discretion until your own product line values are established.


    Table 1 — Sensitivity Estimates According to Defect Depth and Current Type
     
      AC HW FWDC
     
    Field
    Penetration
    Poor Excellent Good
     
    Defect found
    Wet method
    Surface to
    0.01 in.
    (0.25 mm)
    Surface to
    0.025 in.
    (0.65 mm)
    Surface to
    0.05 in.
    (1.3 mm)
     
    Dry method Surface to
    0.01 in.
    (0.25 mm)
    Surface to
    0.25 in.
    (6.35 mm)
    Surface to
    0.05 in.
    (1.3 mm)

    Magnetizing Current Determination
    Many times the inspector must decide how much current to use on a part when there is no sample available containing minimum size flaws to be detected. Guideline formulas are available for selected sizes and shapes of parts, but many times, they can be rather difficult to interpret. I suggest the following procedure to determine an amount of current to start with for a given part.

    Note: If there is a customer specification indicating the amount of current to use, use the following procedures only to confirm that the amount specified is adequate, not to lower it. Let us consider the traditional (formula) rules of current determination for a head shot, a coil shot, and a prod contact shot. We will also review the use of many inspection aids now being used.

    Since a large portion of all magnetic particle inspection is conducted using a wet horizontal unit and black light, we will review single-shot magnetizing first, using either a head shot, coil shot, or both. The factors involved in determining current will also apply to semi-production inspection (same person processes part and inspects it) and production inspection (processing and inspection separate). In this category, think of a 5-lb or less casting, forging, or fabrication used in the aerospace, automotive, or nuclear industry. The amount of current to use on prod contact applications will be discussed separately.

    Current selection for both the head shot and coil shot must take into consideration that other factors that affect locating defects may have deteriorated from their normal condition, namely:

  • Proper bath concentration with no contamination.

  • Adequate black light intensity with minimal visible (white) light.

  • Continuous method being used.

  • Maintaining "balanced fields" when using multidirectional magnetizing.

    Head Shot (Circular Field)

    Fig. 5 — The amount of current for a circular field.
    For the head shot, theory indicates that the target amount of current to use should place the flux density at the knee of the magnetizing curve — Fig. 5. This is the highest point of effective permeability for any given part. The current to attain this value of flux density is 1000 A for each 1 in. (2.5 cm) of outside diameter or diagonal of the part. The same rule applies if a central conductor is used to magnetize the part. Note that the effective range of H (magnetizing force) from d to e is quite wide. Values of B (flux density) vary from one-half the target amount to well over it, from a to c. Beginning in the 1970s, the use of inspection aids dropped this 1000-A target. Many of the more popular specifications now call for 300 to 500 A/in.

    Coil Shot (Longitudinal Field)
    The formula used by most inspectors for determining the approximate amount of current to use when magnetizing a part positioned on the inside surface of a magnetizing coil is

    I =              K        
            L / D X N


    where I is current in amperes, K is 45,000, L is the part length, D is the part diameter, and N is the number of turns in the coil.

    This formula is valid only for parts with an L/D ratio of 2 to 15 and is independent of coil diameter provided the part's cross-sectional area is not greater than one-tenth that of the coil. For parts exceeding this limit, the "constant" changes and the coil diameter is included in the revised formula.

    The formula indicates that the amount of current to use depends entirely on the part's L/D ratio. For example, if a part's length is doubled, the amount of current to use to produce those same indications would be cut in half. To check this theory, a new inspector could take two bolts having the same diameter, 6 in. (15 cm) long, and hold them end to end in the center of the coil. When the coil is energized, use a Hall Effect Meter to measure BT (Gauss — flux density tangent (parallel) to the surface). If one of the bolts is then removed and the coil reenergized, BT on the remaining bolt should drop by 50%.

    Note: The formula tells us that short parts require more force to magnetize them than do longer ones. When magnetizing a part in a coil, magnetic poles appear at the ends of the part. These poles are formed such that they decrease the magnetizing field within the part. This effect is called self-demagnetization. Since the poles are localized to the ends of the part, their effect on magnetizing the part decreases as the part length increases.

    Finally, keep in mind that should the length of a part exceed the coil's inside diameter by more than 50%, the part must be processed more than once to attain best inspection sensitivity. (In examples using a 12-in.- (30-cm-) diameter coil to magnetize a 60-in.- (1.5-m-) long shaft, the shaft should be processed four times. A quick check, using a quantitative quality indicator (QQI), would most likely reduce the number of processings to two.

    Multiple-Pass Welds and Steel Castings
    Where subsurface sensitivity is required, HWDC must be used along with contact prods. Incomplete penetration at the root of the weld, incomplete fusion where the filler metal fails to coalesce with the base metal, subsurface shrink, and inclusions can all be detected within limits. Surface flaws, such as undercut and overlap, surface shrink, and crater cracks, can be detected with either AC or HWDC. Magnetic particle inspection cannot detect porosity.

    To attain subsurface sensitivity, use "the rule of eight." At prod spacing of 8 in. (20.3 cm), use 800-A HWDC to locate a flaw 1/8 in. (3.17 mm) below the surface. To increase sensitivity and locate flaws 1/4 in. (6.35 mm) below the surface, either go to 6-in. (15.2 cm) spacing and 1200 A or 8-in. spacing and 1600 A. Do not apply the powder; blow off any excess amount and then turn off the current. Performing the inspection as the powder is applied attains best sensitivity.

    At high values of current, there exists a dead zone of about 1/2 in. (12.7 mm) around each prod as current enters or leaves the weld. The excessive powder built up around the prods makes less than 6-in. spacing nonproductive. This is the primary reason the inspector must overlap the prod about 1 in. (25.4 mm) as he or she proceeds to inspect long welds.

    The prod must be placed on the weld, never on the base metal being joined. Burn spots on a cast product, such as the welds, are not acceptable if they can be avoided, but, on a wrought product, they are much more likely to be a starting point for metal fatigue. Prod tips must be dressed frequently to avoid arcing.

    Except for crater cracks, all multipass weld defects run with the weld so inspection is required in only one direction.

    If for some reason only surface defects in the weld are to be located, AC magnetizing may be used. Perhaps the inspector is looking only for fatigue cracks. If this is the case, 12-­18 in. (30.5-­45.7 cm) prod spacing may be used. A pie gauge can verify sensitivity.

    Inspection Aids to Determine Which Current to Use
    Pie Gauge. A typical gauge is made of eight, highly permeable, low retentivity steel segments brazed together in the shape of a pie. When held against the surface of the part while it is being magnetized, the magnetic force generates a flux field across one or more of the four simulated cracks in the pie.

    The pie gauge finds its primary use on applications where large parts are being magnetized using clamps and coil wraps with the dry method. Dry magnetic particle indications do stay in place after the force is removed.

    Inspectors on wet applications also use the pie gauge. When a bath containing magnetic particles is flowed over the gauge, good indications form. The indications will stay in place as long as the flux field exists. However, these indications melt away quickly when the current goes off. The indications on the gauge must be viewed under black light while the current is flowing. On complex-shaped parts, processing the part and gauge at the same time is not always easy. When possible, it will furnish reliable information on the direction of flux flow.


    Fig. 6 — Magnetic field during current flow through a part.
    Hall Effect Meter. With the introduction of the Hall Effect probe, the 1000-A figure for a head shot dropped to 500 A and, in some applications, as low as 300 A. The inspector could now measure the exact amount of BTat any point on the part considered critical. Many companies conducted tests on their products. Most agree that Hall Effect readings of 30 to 60 gauss in air on the surface of the part were high enough to disclose defects they considered critical. Inspection supervisors knew these values were conservative and modified them downward for production inspection only when a group of sample test parts containing known typical defects in all critical locations was available. Determining BT at some locations on a part when magnetizing with a coil can sometimes be difficult. The Hall Effect Meter probe coil must be held normal (at an angle of 90 deg) to the surface, fixtured if necessary. Readings taken near the ends of a part can include values of flux density normal to the surface (BN) since it increases rapidly at these locations. For this reason, some firms still do not rely on BT readings taken while magnetizing with a coil.

    Readings of gauss taken by the Hall Effect Meter are a compromise. There is no convenient way to measure the actual amount of flux passing through the defect near the surface of the part, causing a leakage field. We therefore measure the value of BT in the air as close to the part surface as convenient.

    Even this reading is a compromise since we are taking our reading at the centerline of the Hall Effect probe's coil. For a part circularly magnetized, BT in air drops slightly as distance from the part increases — Fig. 6. Conversely, for parts longitudinally magnetized, BT in air increases slightly as distance from the part increases since the self-demagnetizing effect decreases.

    When the use of the Hall Effect meter was first introduced for estimating inspection sensitivity, some firms felt there were too many variables involved, particularly when a coil was being used for magnetizing. Since the advent of the calibrated paste-on defect, the two inspection aids working jointly have proven useful in all industries.


    Fig. 7 — A QQI field indicator. This basic shim satisfies most needs because its circular and crossed bar flaw configuration is suitable for longitudinal and circular fields. The circular flaw is especially useful in balancing multidirectional fields.
    Paste-on Defects — QQIs. For many years, a paste-on defect that worked — one with accurate, reproducible values — was only a dream. The QOI introduced in the 1980s was an answer to that dream — Fig. 7. It is a small steel shim, 0.75 in. (1.9 cm) square, 0.002 in. (50 microns) thick containing a precision etched simulated defect, a circle and cross. It is made of highly permeable, low retentivity steel. When used, it is held against or glued on a part, with the simulated defect face down.

    These QQIs are not the least bit retentive. If they were, they could not be used in multidirectional magnetizing operations. However, at times, a given QQI glued to a part may appear to be retentive by retaining an indication after the magnetizing force has been removed. This happens when the test part is very retentive and contains a longitudinal field. The part itself is magnetizing the QQI. This phenomenon will not occur should the retained field be circular since a circular field has no poles and is entirely contained within the part.

    Sharp cracks, open to the surface, require only one or two gauss of BT, as measured with a Hall Effect meter. Most open seams require the same. If the seam is rolled over and closed at the surface, 5 gauss may be required. Forging laps, which are neither open to the surface nor normal to it, are harder to indicate. They require five or more gauss to provide an indication that may be easily viewed.

    Many inspectors use these QQI shims much the same way as a pie gauge. One of them is glued face down on the end of a thin, flat piece of stiff plastic about 0.75 in. (19 mm) wide and 5 in. (13 cm) long. When held against and processed with the part, the shim indicates the direction and amount of flux flow. Unlike the pie gauge, the particles hold their position after processing. Indications may be viewed any time after processing.


    Fig. 8 — QQI cross-section (not to scale).
    Now take a closer look at the design of the QQI — Fig. 8. The defect depth of 0.0006 in. (15 microns) is the height of the artificial flaw. The defect's depth below the surface is 0.0014 in. (35 microns). The value of the leakage field, which controls the brightness of the indication, will increase the deeper the cut is made in the QQI. There are two reasons for this. First, the number of flux lines intercepted by the artificial flaw increases as the height of the flaw increases. In addition, the number of those flux lines that reach the surface of the QQI increases as the depth of the flaw beneath the surface decreases. The QQI shows a reasonable indication at 5 gauss and a very bright indication at 15 gauss.

    Most aerospace and automotive firms rely on QQIs. The 5 to 15 gauss range is considered both safe and conservative. These companies realize that there are other operating variables involved in MPI processing that may deteriorate, such as bath quality and lighting. Flaws a little bit harder to find are therefore a plus.


    Excerpted with permission from Magnetic Particle Inspection Manual for the Inspector by Arthur R. Lindgren (ronjomac@aol.com). Following his retirement after a 34-year-long career with Magnaflux Corp., Glenview, Ill., Lindgren established L&L Consultants, Inc., Lake Zurich, Ill.