by Dirk Dusharme

Developing test equipment that can inspect materials or assemblies without destroying or disassembling them is of interest to most industries and critical to many. Aircraft, nuclear facilities and pipeline inspections would prove cost-prohibitive if it weren’t for the array of nondestructive evaluation tools available to those industries.

One key NDE tool is eddy current testing. In commercial use since the 1950s, eddy current testing provides a versatile method for the nondestructive evaluation of conductive materials when access to the material is limited to one side or when it’s necessary to detect flaws beneath the surface. Unlike ultrasonic methods, this technology works even when an air gap exists in the tested material, making it ideal for nondestructive evaluation of laminated metallic materials.

This article discusses some of the newer eddy current technologies and their implications for nondestructive evaluation, particularly when the area under inspection has a large surface or varying thickness.

Eddy current review

If you apply an alternating current to a wire wrapped around a conductor (e.g., the wire in an eddy current probe), an alternating electromagnetic field forms around the conductor. Bring this field into proximity with another conductive material--a pipe, pressure vessel, area of aircraft wing, etc.--and an alternating electrical current will flow in that portion of the material as well. This alternating current causes its own secondary magnetic field that interacts with (i.e., adds vectorially to) the primary field and results in a perturbation to the field around the probe, as illustrated on the right.

As long as there’s no change in the material or its proximity to the probe, the perturbation remains constant and can be measured. If while moving along the material under test, the probe passes over a crack, a region of corrosion, an overstressed area, a hole or some other anomaly, a subtle change in the field occurs. This change can be measured and compared to measurements on a similar known-good material and the results quantified by the operator and the eddy current instrumentation.

Because the response to the magnetic field depends upon the material’s conductivity, magnetic permeability and distance from the probe to the material’s conductive surface, eddy current testers can gather information in addition to the location of flaws or corrosion. For example, they can provide information about the physical properties of the material or the thickness of nonconductive coatings, such as paint. If the material isn’t too thick, eddy currents can also measure thickness.

Changing the frequency of the probe’s excitation current changes the depth to which the probe can penetrate: the lower the frequency, the deeper the measurement. Thus, a low-frequency eddy current instrument can detect subsurface defects or defects on the opposite side of the scanned material. This makes eddy current testing ideal for detecting corrosion or cracks in areas invisible from the surface. Often, measurements are made at multiple frequencies in order to inspect the material at multiple depths.

However, several issues hamper the conventional technology’s productivity when the method is used to inspect large areas of complex parts of varying thickness, such as aircraft frames. First, distinguishing between defects and probe liftoff--or mere structural changes such as plate separation, edges and fasteners--requires multiple-frequency or swept-frequency eddy current equipment, both of which involve more setup time than single-frequency measurement. Second, measuring a sample that has greatly varying thickness requires a different setup for each thickness in order to obtain optimum defect distinction and, for each change in structure, a new known-good specimen must be manufactured for setup purposes. Third, unlike many measurement technologies, conventional eddy current inspection takes comparative rather than absolute measurements and is highly dependent upon users having a good understanding of the type of material under inspection and the faults they expect to find. Therefore, it requires some skill in both setup and interpreting results.

Advances in eddy current technology during the past five years have addressed some of these problems.

Pulsed eddy current

When using conventional eddy current techniques to inspect materials for flaws that might occur at varying depths, inspectors must establish multiple excitation frequencies. In commercial applications, these tests are typically done in two ways: by taking multiple measurements at different frequencies using a single-frequency instrument or by using a multiple-frequency instrument. The former suffers from obvious productivity issues: Before measuring, the user must set up the probe and perform a calibration for each frequency. With a multiple-frequency instrument, the technician can select from a number of frequencies, which the probe will use simultaneously in a single measurement. To a certain extent with either method, users must have some understanding of the material’s properties, the types of flaws expected and the depth at which they expect flaws to occur to select the appropriate frequencies.

A new technology, pulsed (or transient) eddy current, uses a shaped waveform rather than a continuous sine wave to excite the coil and generate an eddy current pulse in the structure. Because a pulse represents the sum of sine waves for a broad band of frequencies, one test pulse can contain all the frequencies needed to perform the tests at different depths, explains Marcus Johnson, associate scientist with the Center for Nondestructive Evaluation at Iowa State University. “The real advantage is that pulsed eddy current is rich in low-frequency components so you’re able to see near-surface and deeper flaws at the same time,” he notes.

Iowa State’s pulsed eddy current probe is excited by a pulse, which generates a broadband eddy current field in the sample. The higher-frequency eddy current components are generated more quickly within the material under test and lead to a rapid, high-amplitude response, as seen by the probe, says Johnson.

If you look at the pulsed eddy current response as a function of time, the waveform’s earlier portion corresponds to near-surface artifacts such as liftoff. The middle portion corresponds to midsample artifacts such as internal cracks or interlayer corrosion, and the tail end of the waveform could indicate overall thickness or show bottom surface artifacts such as corrosion, as illustrated below.

Pulsed eddy currents offer an advantage over multiple-frequency probes because users don’t have to select the frequencies at which to scan, says Robert A. Smith of the United Kingdom-based QinetiQ Ltd.

“The basic difference is that transient eddy current captures data covering all the frequencies in the bandwidth, whereas multiple-frequency eddy current only captures certain specific frequencies,” says Smith. “This means that the optimum frequencies for any part of the structure will always be available from any transient eddy current scan. For a multiple-frequency scan, the optimum frequencies would be determined in advance and will be different for each different structure, requiring several scans over complex structural changes.”

Pulsed eddy current technology offers more information about structure and defects than multiple-frequency eddy currents, adds Jesse Skramstad, president of NDT Solutions Inc., the U.S. distributor of QinetiQ’s TRECSCAN transient eddy current instrument. The probe is also much easier to use, adds Skramstad. “The same probe and simple setup can be used over a wide range of structural thickness, thus speeding up the data acquisition process,” he notes.

This technology offers the following advantages, according to Skramstad:

Because of the broadband characteristic of pulsed eddy current, the full frequency range is captured in one pass.

Large areas of structure with multiple variations in thickness can be scanned without the need for probe or setup changes. The same setup can be used for different structures, which are optimized during post-processing. This eliminates the need for specimens representing the structures encountered, says Smith. “Unlike traditional eddy current measurements, pulsed eddy current doesn’t require the production of multiple known-good specimens for setup or reference purposes,” he says.

Compared to inductive coils, the use of Hall-effect sensors (explained later in the article) as field detectors improves the spatial resolution and detectability of deep defects.

Advanced post-processing analysis tools allow better data analysis by using algorithms for liftoff compensation, edge subtraction (i.e., edge correction), total thickness measurement, plate separation effect elimination, time-slice viewing and time-domain signal processing. All referencing to known-good structure can be accomplished on the scan data itself.

Johnson cautions that with pulsed eddy current there are some signal-to-noise issues to consider that don’t exist with multiple-frequency instruments. In the latter case, the instrument can lock onto the exact frequencies being analyzed, eliminating the noise at all other frequencies.

Eddy current arrays

Scanning eddy current equipment can use an array of individual eddy current probes in order to increase the area that can be scanned in one pass. This has huge implications for industries that require a lot of eddy current measurements, particularly airframe inspection, in which eddy currents are used to examine every fastener on an airframe’s surface looking for minute cracks.

“Right now, every single fastener has to be looked at with a probe using circular symmetry,” explains Ray Rempt, technical fellow with Boeing Phantom Works in Seattle. “It takes a minute or more per fastener if you don’t use an array. You can inspect dozens of fasteners per minute with a scanning array.” Rempt’s work with Boeing involves developing scanning arrays of about an inch wide, enough to span a row of fasteners on a fuselage. He estimates that eddy current inspection accounts for about 80 percent of airframe inspection for aircraft depot operations.

“They cover a larger area in less inspection time,” agrees Marc Grenier, product manager for R/D Tech, a Quebec-based developer of ultrasonic and eddy current instrumentation. “In a single pass, you get a better idea of the structure of the fault. If you try to do this with a pencil probe, it’s very difficult to get an idea of the size of the flaw. With an array, you get a topographic representation of the defect. So you get shape and depth in a single pass.” A pencil probe could collect the same information, but it would take so long to do that the method is usually avoided, says Grenier. One of R/D Tech's array instruments is shown above.

The other advantage of eddy current arrays is that they can be shaped to fit specific geometries--hexagonal, square, flat and even more complex shapes, such as turbine blades. “Instead of using robotics with three, four or five axes to rotate a part around a single probe, you could use a robot with a single axis to simply move a part in and out of an array,” says Grenier.

He stresses that eddy current arrays don’t provide more resolution than would be obtainable with a pencil probe. In fact, when developing arrays for a client, R/D Tech first determines which type of standard probe is required for the measurement task and then creates an array of those specific probe types.

The number of probes in an array also depends on the task, he says. “For manual testing in aerospace, lap joint inspection, looking for corrosion under aluminum or determining cracks around fasteners, we might typically have 32 elements in the array,” explains Grenier. “When you go to tube inspection, our standard X-Probe has 48 elements.” For more complex applications, the company has supplied arrays with up to 200 elements.

Flexible coils

The key component in any eddy current system is the sensor used to measure the magnetic field. Traditionally these have been inductive coils. Although in use for decades, coils do have a few drawbacks. The coils’ physical makeup limits them in resolution and measurement depth, the latter being dependent upon a coil’s insensitivity to low-frequency signals. The coils’ size also limits the size of arrays that can be constructed from them.

To address the latter problem, a team from GE Global Research, GE Aircraft Engines and GE Inspection Technologies has developed a new product: coils constructed on a thin, flexible plastic substrate. Instead of wire wound around a core, these flexible sensors use metal lines (or traces) deposited on a flexible plastic-like material, similar to a flexible printed circuit board. They can be designed with single or multiple coils.

“This technology is conformable; we can shape it to fit any part,” says Mike Bernstein of GE Inspection Technologies. One application involves measuring dovetail slots on jet engines, as illustrated on page 26. Using flexible eddy current sensors in an automated system has allowed GE to reduce the time for this task, which once took 14 hours, down to about 45 minutes.

A concern that immediately comes to mind is the durability of these probes in a production environment. “The arrays have now been in use in our manufacturing operations for several years,” says Bernstein. “The repeatability is incredible and the durability is excellent. Although no probe has an infinite life, the GE array probe is far more cost-effective than single-element applications. These don’t last as long [as conventional probes], but we get a tremendous number of parts inspected per probe.”

The probes are also easily replaced should they wear out, Bernstein notes, and likens the chore to replacing a razor blade.

GE has also developed accompanying integrated systems and software to help deal with typical eddy current measurement issues, such as automation, liftoff and edge effects.

Another flexible and conformable technology has been developed by Jentek Sensors Inc. of Waltham, Massachusetts. The company’s meandering wire magnetometer sensor consists of a primary winding of a specific shape--typically laid out in the shape of a square wave pattern with specific distances between each of the waves. Around this meandering pattern are multiple secondary windings. As with GE’s flexible coil, the MWM is constructed as a thin flexible circuit board.

According to articles published by the company, the sensors have high spatial resolution and reproducibility, allowing their response to be accurately modeled--cutting down on the amount of calibration required. Their thinness allows them to be permanently installed in critical areas for continuous monitoring of fatigue, according to some sources. We were unable to contact Jentek to determine if the product has yet been used in this fashion.

Sensor technologies

Both wound coils and flexible coils measure magnetic flux indirectly; that is, they measure the current created in the coil when its magnetic fields change. This indirect rate-of-change measurement has limitations when trying to probe deeper into a material where only low frequencies can penetrate.

Two new technologies, Hall-effect devices and giant magnetoresistive sensors, address this issue by directly measuring magnetic flux. Both offer better spatial resolution at increasing depth. With a coil you must sacrifice resolution to increase its sensitivity at lower frequencies.

“The primary advantage of the GMR is that it detects the field directly,” explains Rempt. “You’re less dependent on frequency, so you can see deeper better.” Believing them to be more stable and sensitive than Hall-effect devices, Boeing is considering GMR to replace anisotropic magnetoresistive sensors, a similar technology it now uses for its eddy current probes.

Advances in new Hall-effect devices have somewhat addressed Boeing’s concern, argues Johnson. Although he agrees that their sensitivity might be an issue in conventional eddy current measurements, he maintains that it’s less of an issue where pulsed eddy current is concerned because those systems can operate at much higher currents and produce stronger fields within the sensitivity range of Hall-effect probes.

Part of the attractiveness of Hall-effect sensors for Iowa State, QinetiQ and others is that the devices have been mass produced for decades and lend themselves to eddy current sensor applications in terms of cost and functionality.

QinetiQ uses Hall-effect sensors in an array constructed of a single rectangular coil with multiple Hall-effect sensors. “Hall arrays offer additional benefits of good near-surface resolution due to the small Hall-effect sensor size coupled with the deep penetration provided by the large coil, and hence the large magnetic field,” explains Smith, who recently presented such an array for use with transient eddy currents.

Final analysis

Eddy current measurement is a mature technology for the nondestructive evaluation of conductive materials. It can inspect for surface, subsurface and rear surface flaws in thin materials. Recent advances in the technology include pulsed eddy current, eddy current arrays, flexible probes and new probe technologies such as Hall-effect and GMR sensors. All of these advancements in eddy current help improve its use with materials of varying degrees of thickness, increase the speed at which measurements can be taken, allow eddy current probes to reach previously unreachable areas and improve eddy current’s depth of penetration and spatial resolution.

About the author

Dirk Dusharme is Quality Digest’s technology editor.

The author would like to thank the following individuals for their assistance in this article: Marcus Johnson, the Center for Nondestructive Evaluation at Iowa State University; Ray Rempt, Boeing Phantom Works in Seattle; Mike Bernstein, GE Inspection Technologies in Cincinnati; and Marc Grenier, R/D Tech in Quebec.

For extensive and clearly presented information on eddy current and other NDE technologies, visit www.ndt-ed.org, an educational Web site developed by the Collaboration for Nondestructive Testing and supported by academic and industrial sponsors.