Yoke MT: Part I - A 'How-...

 BACK TO BASICS
Yoke MT: Part I —
A 'How-to' Guide for
Inspectors

Understanding the factors that can affect test results can help you
perform better yoke magnetic particle inspections

BY GORDON E. SMITH AND UWE ASCHEMEIER


Doug Taylor (left) of H.C. Nutting Materials Joining Consultants, and Todd Studebaker (right) of H.C. Nutting’s commercial dive team perform yoke MT tests. In the photo at center, under ultraviolet light, MT showed a crack in an elevator component (see the small horizontal line on the left-hand side).

The following is the second of a series that takes us back to the basics of weld inspection. In this second article and the third, which will run in the Summer issue of Inspection Trends, Gordon Smith and Uwe Aschemeier take a comprehensive look at yoke magnetic particle testing. Smith and Aschemeier are working AWS CWIs. Smith is an ACCP NDT Level II and III – UT, a Level III in MT, PT, and RT, and author of Magnetic Particle Testing Classroom Training Book. Aschemeier has been certified in Europe and Canada, and is a CSA Weld Level III inspector. He is also a member of the AWS D1D Subcommittee 4 on Inspection.

Magnetic particle testing (MT) is more than 50 years old as an inspection technology. One of the first applications was in the inspection of welds in ferromagnetic materials such as steel. Before the 1940s, it was found that passing an electric current through the weld zone would allow imaging of welding defects almost invisible to the unaided eye by the application of iron powders that were attracted to the defects. This was so easy and so popular that many welding power supplies had MT prod outputs built in. About the time man first walked on the moon, a giant step was also being made in the magnetic particle inspection world. The electromagnetic magnetic particle inspection device or electromagnetic yoke was starting to take its place. Hardly mentioned in NDE literature of the 1960s, today it is probably the most used form of nondestructive examination of welds other than visual inspection and is a natural companion to visual inspection. It is important for the welding inspector to understand how this technique works, its strengths, weaknesses, and the truths and folklore about it.

This article is intended to supply basic knowledge for the use of magnetic particle examination of welded ferromagnetic materials to detect surface or very near surface cracks, and other unacceptable discontinuities. In principle, this method of magnetic particle inspection involves magnetizing by the use of permanent or electromagnetic yokes the area to be examined and applying ferromagnetic particles to the surface. The particles will form patterns on the surface where cracks and other discontinuities cause distortions in the normal magnetic field. These patterns are usually characteristic of the type of discontinuity that is present.

Relationship with the Welder
The magnetic particle inspector (who should be qualified by an NDT Level III) often works closely with the welder, who is called on by AWS D1.1, Structural Welding Code — Steel, to clean surfaces that are too rough and to chase (remove) discontinuity indications. The welder is often the best source of information about what types of discontinuities are probable and where they are likely to be.

Developing good communication with the welding contractor can result in more effective and reliable testing. A good start is to explain the magnetic particle procedure and the need for particular surface conditions. It is equally helpful if the inspector is informed of the welding process and materials used.

Equipment for Magnetic Particle Inspection
Small portable units are avilable that can be hand-carried to where they are needed. These units may have both AC and half-wave DC capability. They can be used wherever an adequate 115-V AC power source exists. These AC/DC yokes are unique  electromagnets of special design that allow cost-efficient inspection of complex and small parts of small batch sizes.

Surface Preparation
In general, the smoother the surface of the part and the more uniform its color, the more favorable are the conditions for formation and observation of the powder pattern. This applies particularly to inspections being made on horizontal surfaces. For these conditions, the probability of detection (POD) for yokebased magnetic particle testing can reach its maximum value where the smaller indications are repeatedly and uniformly detected by a group of MT inspectors. This POD value decreases as the inspection surface moves away from a horizontal position in front of the inspector. For sloping and vertical surfaces, the dry powder may not be held on a very smooth surface by a weak leakage field. The surface should be clean, dry, and free of grease. The dry particles will stick to wet or oily surfaces, and will not be free to move over the surface to form indications, which could prevent detection of significant discontinuities. A thin film of oil that is sufficient to interfere with the free movement of the powder often remains on surfaces that have been cleaned of grease by wiping with a rag soaked in a petroleum distillate. This thin film can be removed by dusting the surface with chalk or talc, then wiping the surface with a clean, dry cloth. An initial application of the dry magnetic powder, followed by wiping, often will give a surface over which a second application of powder will move readily. Vapor greasing will give a dry oil-free surface. These actions add to the gauge repeatability and reliability (GRR) value of the inspection as applied to a certain part under certain conditions for a group of qualified inspectors.


Fig. 1 — Mike Kreimer uses a portable, battery-powered unit to perform
a magnetic particle inspection.

Any loose paint, rust, or scale must be removed— Fig. 1. If cleaning is accomplished with shot or grit blasting, there is a “peening” effect, especially on softer steels, which may close up or disorganize the surface magnetic domains and reduce the probability of detection of fine surface discontinuities. The effect is more pronounced with shot than with grit, but if these cleaning methods are used, the operator should be aware of the danger of missing very fine cracks. A thin, hard, uniform coating of rust or scale will not usually interfere with the detection of any but the smallest defects. The inspector should be aware of the smallest size defect he/she is to consider, in order to judge whether or not such a coating of rust or scale should be removed.

Paint or plating on the surface of a part has the effect of making a surface defect appear like a subsurface one. The relative thickness of the plating or paint film and the size of the defects sought determine whether or not the coatings should be stripped. The dry method is more effective in producing indications through nonmagnetic coatings than the wet method, but if fine cracks are expected, the surface is to be stripped of the coating if its thickness exceeds 0.005 in. Most cadmium, nickel, or chromium coatings are usually thinner than that, and the plating makes an excellent background for viewing indications. Hot galvanized coatings are thicker and, in general, should be removed before testing unless only gross discontinuities are important. Broken or patchy layers of heavy scale tend to hold powder around the edges of the breaks or patches, and should be removed if they are extensive enough to interfere seriously with the detection of discontinuities.

Inspection Variables
The two basic inspection variables to be considered are the types of magnetization to be used and the type of particles applied. The location of the defects — on the surface of the part or located wholly below the surface — dictates the type of magnetization. Choice of yoke method lies between AC and some form of DC. If the defect is on the surface, either AC or DC may be used and the choice is determined by other considerations. If the defect lies below the surface, AC should not be used as MT yokes do not pass current through the part. The considerations for AC or DC are different, with DC and half-wave DC being used for new fabrications and AC typically being used for in-service parts and machined surfaces.

Inspection
A good light and good eyesight are the principal requirements for observing the presence of indications on the surface of parts. Choosing the best color powder for contrast against the surface is an aid to visibility. On large discontinuities, dry powder buildup is often very heavy, making indications stand out clearly from the surface. For finer cracks the buildup is less, since only the smaller particles are held by the leakage field in this case. For exceedingly fine cracks, some form of the wet method, which is more sensitive to very fine discontinuities, should be used.

In the case of subsurface discontinuities, experience, and skill, or carefully established, controlled practices for repetitive tests are required to secure the best results. The depth below the surface and the size and shape of the discontinuity determines the strength and spread of the leakage field. An experienced inspector will observe the surface as the powder is allowed to drift onto it, and can see faint, but significant, tendencies of the powder to gather. Often indications are seen under these conditions that are no longer visible when more powder has been applied, the excess blown off, and the surface then examined for indications. Standardized techniques with acceptable GRR values for careful and proper application of the powder can provide excellent sensitivity where similar assemblies are repetitively tested.

In the case of alloy steels having high retentivity, indications are held at the defect by the residual field, making the inspection easier. In low-carbon steels, retentivity is low. On these steels, it is important to do the inspection while the magnetizing force is on (continuous method) and the powder is being applied, since indications may not remain in place after the current is turned off. This is particularly true on vertical and overhead surfaces, where gravity plays a part in causing particles to fall away if lightly held. Smaller discontinuities become more important in harder, high-strength steel. Although retentivity is high, great care must be taken in the inspection of these steels, especially for fine cracks open to the surface.

Yoke Break or Lift Test
Forty years ago, a yoke break or lift test was devised to help evaluate the effect of coating type and thickness on the introduction of flux to a ferromagnetic substrate. This simple test relies on the contention that the flux strength produced in a ferromagnetic material may be gauged by the amount of pull produced by the yoke on that material. The test uses a spring scale to pull an energized yoke from a bare test plate or lift a steel plate of known weight and alloy. This is a field test to ensure that the test item is adequately magnetized for inspection. For low-alloy to high-alloy welds, i.e., A36–4340, two separate MT examinations may be required to meet engineering specifications; one test for the low-alloy side and another test for the higher-alloy side. The lower-alloy side MT should always be completed first because the ability of the magnetic media to adequately detect discontinuities may be impaired on the low-alloy side due to higher magnetization requirements for the higher alloy.

Weld Bead Crack Reference Standards

Fig. 2 — The graph shows a decrease in pull-off force with an increase in coating thickness and type.
In an attempt to quantify the thickness of the coating at which sensitivity is diminished, a test was conducted for which four ASTM A-36 plates were prepared, each containing a single shielded metal arc weld bead in which copper ferrite dilution cracking had been induced — Fig. 2. Each plate was then coated with zinc, zinc chromate primer, enamel, or epoxy paints.

In general, detection with metallic coatings more than 0.004 in. thick should be very well qualified, while nonmagnetic coatings (paints) less than 0.0015 in. of a color highly contrasting the magnetic particles may offer increased probability of detection of indications.

Effect of Coating Thickness on Test Results
The data developed in the above experiment show that the introduction of magnetic flux to a ferromagnetic substrate with an alternating current yoke is affected by interposing thicknesses of paint. The extent of this effect may be significant, depending on the type of paint, its magnetic properties, and its thickness.

When welds in existing structures are being tested, they are often coated with paint or rust. One early study showed that the maximum thickness of paint does not significantly reduce the detectability of a particular size discontinuity as much as the type of coating materials and the weld surface profile. Because the type and thickness of coating on an existing structure are generally unknown, the safe thing to do is to remove it in the areas of interest. Although it is possible to make measurements of paint thickness with a portable coating thickness gauge, this does not provide information about the relative magnetic effects of different coatings.

The inspector should know that hot dip galvanization thicknesses may be 300% or more thicker on welds than on the adjacent galvanized plate surfaces. This is due to the increased content of silicon in the weld filler metal attracting more zinc during galvanization than that contained on the surface of the plate. Additionally, structural steel galvanized layers often contain iron trapped in or near the galvanized surface that will produce false MT indications and false coating thickness readings.

Effect of Corrosion on Surface Cracks
In-service cracks are formed with microscopic “tendrils” of metal stretching across the fracture. The iron in ferrite grain form in these tendrils is longitudinally oriented perpendicular to the direction of the crack, i.e., they try to bridge the crack. The stretching/deformation of the cracked steel makes these tendrils easy to magnetize by prealigning their magnetic domains in the ferrite like little bar magnets with their north-south poles bridging across the steel fracture. Even stretching to deform, but not crack the steel will enhance the magnetic attraction of magnetic particle media in these areas. Note that stainless steels, which are generally thought to be nonmagnetic, will develop magnetic areas where deformed enough or even cracked.

Magnetic Particle Field Testing of Structural Welds
Magnetic particle testing is performed on a variety of welds used in bridges, buildings, and other structures. AWS D1.1, Structural Welding Code — Steel, with its nondestructive testing requirement for MT, is often incorporated into construction contracts. Inspectors typically must work under less than ideal conditions, particularly when existing structures are tested. Field welds made to connect fabricated subassemblies can also present significant problems.

Effect of Weld Surface
Most weld surfaces are left in the as-welded condition for service or for testing. It is up to the inspector to decide whether a particular surface is good enough for MT. Dry magnetic particles and half-wave rectified current or alternating current are typically used with electromagnetic yokes.

The mobility of dry powder with the pulsed magnetic fields produced by yokes is essential for effective testing of irregular surfaces. Testing with yokes and fluorescent wet suspensions is increasingly required for welds that are ground smooth, such as those in vessels fabricated to Section VIII of the ASME Boiler and Pressure Vessel Code.

On rough surfaces, particle suspension can be trapped in depressions, creating unwanted background fluorescence. The geometry of the test surface, particularly where fillet welds are used, often requires a yoke with articulating legs. The distance between the yoke poles may have to be changed frequently to fit the available contact areas. Some yokes have adjustable magnetizing current.

Also of note in weld inspection is a surprisingly large difference in the data obtained from single-bead test welds and multiple-bead test welds, especially as it relates to the coating thicknesses that permit yoke magnetization.

It is believed that this difference in test results may be the result of concurrent leakage fields that exist in multiple-bead weld caps and their metal grain structure, but not in the single weld beads. This internal flux leakage in multiple-pass weld caps could produce decreased flux density in leakage fields at the crack sites and in turn could have reduced the ability to attract and hold particles through the coatings. The inspector should know that the inspection POD may be reduced for multiple welds or for repair welds to single weld beads.

Field Strength Adjustments for Tests of Welds

Fig. 3 — ASNT NDT Level III Jack Kirby demonstrates qualified yoke extentions for hidden gear MT.
The magnetic field strength is adjusted to correspond to the required pole spacing and the test material. This is done by observing the test surface for nonrelevant indications and excessive buildup of particles around the poles. With electromagnetic yokes, the current is usually increased to the point where these things occur and is then reduced slightly for the actual test. A magnetic field pie gauge or gauss meter can be easily used if the test is being performed on a horizontal surface with sufficient space for positioning the gauge.

Some yokes do have removable pole extensions to compensate for variations in pole spacing on the surface. These are available as more than one length and are sometimes articulating. Depending on their length and design, the addition or subtraction of yoke pole pieces can reduce the magnetic field strength midway between the poles by a factor of 25 to 50% or increase it by a like amount, possibly creating a “dead zone” due to overmagnetization. It is possible to determine the reduction or increase of field strength caused using a Hall element gauss meter. Many MT inspections do require NDT Level III test design and qualification support. Shown in Fig. 3 is a qualified yoke magnetic particle test. Figure 4 shows the defect found in that test.

Partial penetration in structural welds is common, as are welds joining sections with substantial differences in thickness. These conditions are common causes of nonrelevant indications with half-wave rectified magnetization. This problem may be eliminated by changing to alternating current; because of its lower level of magnetization, magnetic field strength changes with section thickness are minimized.

Although rectified (DC) current is sometimes specified because of its deeper magnetic penetration and its increased ability to detect subsurface discontinuities, alternating current is becoming more widely accepted in the United States. Empirical data have shown that AC yokes are capable of detecting discontinuities not open to the surface to a depth of at least 1 or 2 mm (0.04 to 0.08 in.). With any type of yoke, it is difficult to predict the maximum depth at which a subsurface discontinuity will be detected. Magnetic particle testing folklore says a “skin effect” exists for yoke-based magnetizism. This just plain is not true; the “skin effect” of alternating current only takes place when the magnetization current passes or is induced through the test part, as is in the case of prod-type MT inspection.

Lighting for Weld Inspection

Fig. 4 — Wet fluorescent MT indication detected with remote around-the-corner extensions, actual size is about 0.3 in. long.
It is important to have adequate illumination of the test surface. Even when working outdoors in bright daylight, some welds are covered by shadows including the inspector’s shadow. Some specifications leave determination of adequate visible light intensity to the discretion of the inspector, while others specify a minimum intensity (for example, so many lux at so many inches) from the test surface. The common practice of using a standard flashlight is usually inadequate.

Electric power is required at the test site for the electromagnetic yoke and it can also be used with a portable high-intensity lamp. Instruments such as borescopes and fiberscopes can be used to inspect areas where direct visual access is limited.

When dry fluorescent powders or wet fluorescent materials are used outdoors, it is important to have shielding to minimize visible light. This is at least as critical as providing sufficient ultraviolet light intensity. Additionally, when using UV photodiode flashlights, the use of the correct UV wavelengths and the appropriate viewing/protective glasses is a must.

Selection and Use of Magnetic Media
The choice between dry method and wet method techniques is influenced principally by the following considerations:
  • Type of discontinuity (surface or subsurface). For subsurface discontinuities, dry powder is usually more sensitive.
  • Size of surface discontinuity. Wet method particles are usually better for fine or broad, shallow discontinuities.
  • Convenience. Dry powder with portable equipment is easy to use for tests on-site or in the field. Wet method particles packaged in aerosol spray cans are also effective for field spot tests.

  • Effect of Shape
    The shape of the magnetic particles used has a strong bearing on their behavior in locating defects. When in a magnetic field, the MT particles tend to align themselves along the lines of force. This tendency is much stronger with elongated particles than with more compact or globular shapes because the long shapes develop stronger polarity. This means that MT particles should have different shapes for primary manufacturing welding, where a broad spectrum of MT media shapes may be more desirable for the POD of general welding-type defects and for in-service (primarily fatigue crack detection where elongated particle shapes may enhance the detection of cracks).

    Yoke Procedure
    Surface preparation. Satisfactory results are usually obtained when the surfaces are in the as-welded, as-rolled, as-cast, or as-forged conditions. However, surface preparation by grinding or machining may be necessary where surface irregularities could mask the indications.

    Prior to magnetic particle examination, the surface to be examined and all adjacent areas within at least 1 in. must be dry and free of all dirt, grease, lint, scale, welding slag, spatter, oil,   and other extraneous matter that could interfere with the examination. In accordance with AWS D1.1, this is the responsibility of the welder. It is the authors’ experience that this practice varies widely by industry and geographic location.

    Acceptable weld surface cleaning beyond that of simple slag removal may be accomplished using detergents, organic solvents, descaling solutions, paint removers, vapor degreasing, sand or water blasting, or ultrasonic cleaning methods.

    Magnetization
    A suitable and appropriate means for producing the necessary magnetic flux in the part shall be employed, using one or more of the techniques described in this procedure.

    Examination medium. The finely divided ferromagnetic particles used for the examination typically shall meet the following requirements:
  • Dry particles.Typically, the color of the particles (dry powder) provides adequate contrast with the surface being examined. Additional specific requirements on the use of dry particles are provided in ASME Section V, SE-709, Standard Recommended Practice for Magnetic Particle Examination. Magnetic particle examination typically should not be performed if the surface temperature of the part exceeds 200°F.
  • Wet particles. Typically, the color of the particles provides adequate contrast with the surface being examined. The particles are suspended in a suitable liquid medium in the concentration recommended in ASTM E-709, which contains additional specific requirements on the use of wet particle suspension. Typically, the surface of the part is not to exceed 135°F.

  • It is the responsibility of the NDT Level III to note these conditions and develop and approve alternate procedures as required to meet project, code, and regulatory specifications.

    Magnetization Field Strength
    When it is necessary to verify the adequacy or direction of the magnetizing field, a magnetic particle field indicator should be used by positioning the indicator on the surface to be examined. When using this indicator, a suitable flux or field strength is indicated when a clearly defined line of magnetic particles forms when the particles are applied simultaneously with the magnetic force. When a clearly defined line of particles does  not form, or doesn’t form in the desired direction, the magnetizing technique should be changed or adjusted.

    Note that nonmagnetic shim or yoke leg caps may be used to increase the yoke-to-part spacing and reduce the magnetic flux to inspection requirement levels. This strength-reduction technique is used more frequently on high-alloy steels, nickel, and cobalt materials in well-qualified procedures to meet specific POD and GRR requirements.

    Demagnetization
    When residual magnetism in the part could interfere with subsequent processing or usage, the part is to be demagnetized any time after completion of the examination.

    All ferromagnetic materials will retain some residual magnetism, the strength of which is dependent on the retentivity of the part. Residual magnetism does not affect the part’s mechanical properties. Structural metals like A-36 rarely require demagnetization, but many high-alloy metals may remain magnetized for years.

    In general, demagnetization is accomplished by subjecting the part to a field equal to or greater than that used to magnetize the part, then continuously reversing the field direction while gradually decreasing it to zero. An alternate method is to heat the part beyond its Curie point where the magnetic domain relationship is terminated. Typically, parts that will be heat treated prior to further processing or use need not be demagnetized.

    Effectiveness of the demagnetizing operation can be indicated by the use of an appropriate magnetic field strength meter. However, a part may retain a strong residual magnetic field after having been circular magnetized, and exhibit little or no external evidence of the field. Therefore, the circular magnetization should be conducted before longitudinal magnetization if complete demagnetization is required.

    Field Calibration of Yoke Magnetization Equipment

    Fig. 5 — An H. C. Nutting Co. electromagnetic lift or break test in AC mode
    using a 10-lb weight.

    1. The magnetizing force of yokes may be calibrated by determining their lifting or break power by using a lift weight of known value — Fig. 5.

    2. Each lift weight should be weighed with a calibrated scale from a reputable manufacturer and stenciled with the applicable nominal weight prior to first use. A weight need only be verified again if damaged in a manner that could have caused potential loss of material.

    3. Each weight should be fabricated from A36/1018 steel unless specified by the NDT Level III. All fabricated field test weights should actually weigh at least 5% more than their marked values to ensure meeting designated requirements. The authors have found that local supermarkets have calibrated scales that could be used to weigh test weights.

    Examination
    Direction of magnetization. At least two separate examinations should be performed on each area. During the second examination, the lines of magnetic flux shall be approximately perpendicular to those used during the first examination. A different technique for magnetization may be used for the second examination. This practice increases the POD and GRR values for the test being conducted. You just find more defects.

    Typically for welds this is a XXXXXX type pattern, with the weld being the centerline of the dual magnetization X.
     
    Critical and lifting device weld examination is to be done with the yoke legs collapsed to 3–4 in. or less of spread wherever possible. The same but smaller XXXXXX pattern should be used.

    Examination coverage. All examinations are to be conducted with sufficient overlap to ensure 100% coverage or specification level coverage at the required sensitivity.

    Yoke Technique

    Fig. 6 — Relative sensitivity at depth by magnetization and media type.
    Note: Except for materials 1⁄4 in. or less in thickness, alternating current yokes are superior to direct or permanent magnet yokes of equal lifting power for the detection of surface discontinuities. Sensitivity varies by technique and depth — Fig. 6.

    Structural welds and items being tested for first time postweld acceptance should be inspected with DC or half-wave DC magnetic fields. All in-service inspections and all inspections made upon finished-machined parts should be done only in the AC mode.

    Residual Test Method
    In the residual method, the test object is magnetized, the magnetizing force is stopped, and then the magnetic particles are applied. This method can only be used on materials having sufficient magnetic retention. The residual magnetic field must be strong enough to produce discontinuity leakage fields sufficient for producing visible test indications. As a rule, the residual method is most reliable for detection of surface discontinuities.

    Hard materials with high retention are usually hard to magnetize, so higher than usual magnetizing settings may be necessary to obtain an adequate level of residual magnetism. This difference between hard steels and soft steels is usually not critical if only surface discontinuities are to be detected.

    Either dry or wet method particle application can be used in the residual method. With the wet method, the magnetized test object may be immersed in an agitated bath of suspended magnetic particles or it may be flooded with particle suspension in a curtain spray.

    Continuous Test Method
    When a magnetizing field is applied to a ferromagnetic test object, the magnetic field rises to a maximum. Its value is derived from the magnetic field strength and the magnetic permeability of the test object. When the magnetizing field is removed, the residual magnetic field in the object is always less than the field produced while the magnetizing force was applied. The amount of difference depends on the magnetic properties of the material. For these reasons, the continuous method, for any specific value of magnetization, is always more sensitive than the residual method.

    Continuous magnetization is the only method possible for use on low-carbon steels or iron having little retentivity. It is frequently used with alternating current on these materials because of the excellent mobility of the magnetic particles produced by AC magnetic fields.

    Evaluation of Indications
    All indications are to be evaluated in terms of the acceptance standards of the applicable construction code or engineering specifications. In general, MT specifications concern only the evaluation of the MT Indication not the actual defect itself as such discontinuities on or near the surface are indicated by retention of the examination medium. However, localized surface irregularities due to machining marks or other surface conditions may produce false indications.

    Broad areas of repeatable linear indications on hot-dipped zinc-galvanized welds may be indicative of weld or base material problems that are related to cracking or liquid metal embrittlement conditions.

    Broad areas of particle accumulation, which might mask indications from discontinuities, should be cleaned and reexamined.

    Postinspection Cleaning
    Postinspection cleaning is necessary where magnetic particle materials could interfere with subsequent processing or with service requirements. The purchaser should specify when post-cleaning is needed and the extent required.

    Following are typical postcleaning techniques:
  • The use of compressed air to blow off unwanted dry magnetic particles.
  • Drying of wet particles, then removal by brushing or compressed air.
  • Removal of wet particles by flushing with solvent.

  • Conclusions
    Magnetic particle testing is an effective nondestructive procedure for locating material discontinuities in ferromagnetic objects of all sizes and configurations. It is a flexible technique that can be performed under a variety of conditions, using a broad range of supplementary components.

    Application of the magnetic particle method is deceptively simple — good test results can sometimes be produced with little more than practical experience. Here, in fact, the development of the technique was almost entirely empirical rather than theoretical.


    GORDON E. SMITH (gsmith@hcnutting.com) is Senior Consultant, H. C. Nutting Co., Columbus, Ohio, and UWE ASCHEMEIER is Senior Welding Engineer, H. C. Nutting Co., Cincinatti, Ohio.