Back to Basics
[ click here for the Back to Basics Archive ]
The Use of Eddy Current for Ferritic Weld
Testing in Nuclear Power Plants
NDT grows regularly with the improvement of the classic techniques and their application to solve tough problems. Here is an article that describes the development of an eddy current testing technique and equipment to supplement another classic technique. Sometimes finesse better solves a problem rather than using a bigger hammer - but you must know the basics to do it.
Frank Iddings
Tutorial Projects Editor
INTRODUCTION AND BACKGROUND
This article discusses the development and use of manual eddy current weld testing as a high quality and cost saving NDT method to supplement load and visual testing of nuclear fuel transfer casks in accordance with ANSI N14.6 (American National Standards Institute, 1993). It also discusses the histories of magnetic particle and eddy current weld testing leading up to this application.
The roots of eddy current weld testing as an alternative to magnetic particle testing began in the offshore oil and gas industry in the North Sea (Goldberg, 1998). Concurrently, an effort to perform magnetic particle testing through coating was undertaken in the United States (American Petroleum Institute, 1991; Electric Power Research Institute, 1988; Goldberg, 1985). The high cost of removing marine growth and paint in order to perform underwater magnetic particle testing and topside testing requiring paint removal resulted in research and development projects for cost reducing applications in both magnetic particle and eddy current testing.
Eddy current testing is a proven and accepted method for detecting short indications through paint.
The first breakthrough in performing underwater magnetic particle testing through black oxide - a thin film of 0.1 to 0.15 mm (4 x 10-3 to 6 x 10-3 in.) - came in the early 1980s and resulted in cost reductions of over 300%. A research program found that magnetic particle testing through black oxide is highly reliable and can detect indications as short as 1.5 mm (0.06 in.); magnetic particle testing through black oxide is well accepted.
In 1987, the American Society of Mechanical Engineers (ASME) changed its code for magnetic particle testing to allow performance demonstration for validating testing through coatings (American Society of Mechanical Engineers, 1987). In 1988, the Electric Power Research Institute issued a "First Use Award" for magnetic particle testing through coatings for crane testing at Wolf Creek Nuclear Power Station. Cost savings, due to negating the need to remove paint and restore it, were significant. However, for many applications, magnetic particle testing through coatings for fabrication, repair and in service testing, is not feasible due to detection requirements. Additionally, the majority of painted welds typically have greater than 0.15 to 0.3 mm (6 x 10-3 to 0.012 in.) of coating, the maximum paint thickness typically cited in both research reports and standards for magnetic particle testing through coatings. The EPRI studies, while identifying discontinuity size as an essential variable, never quantified minimum size for detection by magnetic particle testing through coatings.
In the early 1990s, a series of research and development projects (Sea Test Services, 1996) and round robins were performed to determine the use of eddy current as a means to test through coating where magnetic particle testing could not be used, primarily topside applications. These round robins also included the use of computer aided electromagnetic methods, including alternating current field measurement. Initially, the use of one of the first eddy current testing systems designed for weld testing (instrument and probe) showed promise, however its reliability fell short of comparative field results of magnetic particle testing.
A major breakthrough came when CAN Offshore, using a commercial eddy current instrument and weld probe, produced results through paint - on average, 0.4 mm (0.015 in.) - similar to magnetic particle testing on bare metal. It is noted that the majority of eddy current testing research had its roots in Scotland and that the company's operator was Bill Brown, considered the father of modern day reliable eddy current ferritic weld testing. As the eddy current instrument was small and lightweight, it was used with industrial rope access to provide an additional advantage over costly scaffolding (Figure 1).
EDDY CURRENT WELD TESTING: A QUICK PRIMER
Perhaps the easiest way to visualize eddy current weld testing is to compare it to magnetic particle testing. In eddy current weld testing, a small weld probe containing cross wound coils is passed over the weld. Much like magnetic particle testing, an alternating current is induced into the part under test. In magnetic particle testing, the presence of an indication is shown by magnetic particles being attracted to a flux leakage site. In eddy current testing, the leakage site is shown by an electrical perturbation that is displayed on a cathode ray tube screen. In magnetic particle testing, electric discharge machined notches can be used as a means to calibrate as well as to show the system's performance sensitivity. Likewise, in eddy current testing, these notches are used to set sensitivity. The signal produced on the eddy current instrument is set to a certain screen height (similar to in ultrasonic discontinuity detection). To account for paint, previously measured using an eddy current absolute weld probe, plastic shims the same thickness as the area under test are placed over the notch and the gain is adjusted accordingly. The eddy current instrument is set up so that the signal produced from the notch produces a vertical signal and can be processed from normal background noise and signals produced from the toe of the weld and various weld geometries.
The biggest difference between eddy current testing and magnetic particle testing is that in eddy current testing only the area under the weld probe is under test. In magnetic particle testing, a much larger area is tested, with two test directions (longitudinal and transverse) required for full coverage. In eddy current testing, several scans must be taken to cover the full weld, making it inherently slower than magnetic particle testing, although this is negated if paint must be removed then reapplied. Separate scans which test the base metal, the toe of the weld and the weld face are performed. The eddy current inspector scans looking for normal signals produced by the weld toes and weld geometries and for the crack signal (Figure 2).
One other major advantage of eddy current testing, is that it can be used on a wet surface, whereas a rainy day stops magnetic particle testing. Other advantages are that the equipment is lightweight and battery operated. This makes it ideal for rope access application.
The fact that one method is referred to as magnetic particle testing and the other as a type of electromagnetic testing shows the closeness of the two methods.
RESEARCH AND DEVELOPMENT AND FIELD APPLICATION RESULTS
In order to produce a reliability and confidence level for eddy current testing to replace magnetic particle testing, primarily for the topside testing of mobile offshore drilling units, a joint industry project was performed to explore the use of eddy current for ferretic weld testing (Sea Test Services, 1996). The basis of the project was to select a group of specimens that had indications of the type of geometry and size indication required. The primary aim of the project was to examine the in service testing of mobile offshore drilling units. A discontinuity size of 6 mm (0.2 in.) was chosen as the minimum size for detectability, although a small percentage of test specimens had indications as small as 1.6 mm (0.06 in.). Specimens, both actual cutouts of fatigue cracks and manufactured cracks, were also selected on the basis of weld types and connections; the presence of butts, fillets, cruciforms and access holes; indication location, whether in the weld face, toe of the weld or base metal; and indication orientation, whether longitudinal or transverse.
In the study, the results of the eddy current testing compared to those for magnetic particle testing were very promising for the manual eddy current technique, with 87% agreement with magnetic particle testing results. Subsequent eddy current testing results with the test bed had 100% agreement with magnetic particle testing. One of the key factors for reliability was fine tuning the initial recommended practice and using highly qualified eddy current inspectors.
Results of the computer assisted eddy current methods revealed that further development was necessary. The computer assisted electromagnetic methods had results of 62% agreement with the magnetic particle testing results with significant false alarms. Additionally, these electromagnetic methods had constraints in testing indication, which occurred at the ends of weld details, as well as access problems for geometries like "rat holes."
EDDY CURRENT FERRITIC WELD TESTING IN THE NUCLEAR INDUSTRIES
A major nuclear engineering firm was asked by a nuclear operator to explore the use of eddy current testing to replace magnetic particle testing. The nuclear engineering firm contacted an NDT specialist in eddy current and magnetic particle testing with a subspecialty in testing through coatings.The engineering firm's request was unique in that it required a recommended practice for eddy current testing for ferritic welds that validated the ability to detect a 1.6 mm (0.06 in.) discontinuity. Previous eddy current testing recommended practices for in service use had a detectability set at 13 by 0.8 mm (0.5 by 0.03 in.) deep through nonvisual testing, based on blind trial research, although field results had shown that discontinuities as small as 1.6 mm (0.06 in.) length could be detected.
Till now, field and manufactured specimens were used to qualify the method, but there was no calibration standard to validate discontinuity size. A calibration block with an electric discharge machined notch 1.6 mm (0.06 in.) long, 0.5 mm (0.02 in.) deep and 0.15 mm (6 x 10-3 in.) wide was used, as it was already referenced in MIL-STD 271F (United States Department of Defense, 1986). The recommended practice was revised to reflect this sensitivity. Additionally, a calibration block having notches with depths of 0.2 mm (8 ´ 10-3 in.), 0.5 mm (0.02 in.) and 1 mm (0.04 in.) was used to validate other test parameters such as permeability. This recommended practice was then validated in a set of blind trials witnessed and verified by the nuclear power company's quality personnel, its authorized nuclear inspector as well as engineering staff from the nuclear facility.
In addition to qualifying the procedure, eddy current testing personnel were then qualified to the eddy current testing recommended practice by having to detect, in blind trials, a minimum 90% detection on specimens with some 25 or more indications of varying sizes and types in various geometries. The basic eddy current equipment consisted of an off the shelf eddy current machine, an eddy current weld and absolute probe, electric discharge machined calibration blocks, and plastic shims. All materials and test equipment were certified in calibration to NIST standards.
The eddy current testing recommended practice consists of a series of calibration checks including:
- setting discontinuity sensitivity
- checking the symmetry of the cross wound weld probe
- checking the permeability likeness of the steel under test versus the calibration block
- measuring the steel under test coating thickness in order to compensate for gain adjustments on the calibration block.
The only variable that was unknown going into the transfer cask project was the agreement of the calibration block permeability and that of the transfer cask. The calibration blocks were made of a low alloy 4340 steel and AISI 1018 steel. The transfer cask is made from ASTM 588 steel, a weathering steel. No problems were encountered and permeability checks were within acceptable tolerances.
Application
Transfer cask shells are fabricated from ASTM A588 steel and are used for transferring transportable storage canisters containing spent nuclear fuel bundles (Figure 3). Transfer casks are required to be tested on a scheduled basis in accordance with ANSI N14.6.In the past, magnetic particle testing had been done on transfer cask load bearing welds, requiring removal of all paint in those areas. Although magnetic particle testing is ideally suited for detection of surface cracks on ferritic material, magnetic particle testing loses its sensitivity when applied through coating thicknesses greater than 200 to 300 mm (6 x 10-3 to 0.012 in.). For this nuclear application there were two transfer casks as well as a lifting yoke which required NDT. Welds were identified by the nuclear power company's engineering staff and the actual eddy current testing was witnessed by quality personnel (Figure 4).
Benefits and Advantages
Annual testing of special lifting devices such as a transfer cask can be a significant disruption to an operation that normally runs 24 hours a day, every day. This disruption becomes even more severe if the components need to have paint stripped and reapplied for the purpose of testing the load bearing welds. Paint stripping is further exacerbated by the fact that a transfer cask is considered to be contaminated unless a significant effort is expended to prove otherwise. As a result, the nuclear power company incurs significant costs in both time and money to develop the enclosures to contain a contaminated paint stripping operation. Previously, annual testing resulted in out of service time for these components of approximately one month (Cella, 2003).The use of eddy current also precludes having to remove and reapply coatings. In addition, as eddy current has no consumables such as in magnetic particle testing and liquid penetrant testing (particles, penetrant, white contrast paint and so forth) there is no costly clean up of potentially contaminated materials.
Proof Testing
If eddy current testing detects an indication, it is proof tested by magnetic particle testing. The reason is threefold:
- magnetic particle testing provides a visual display of the indication
- magnetic particle testing provides an additional confidence level to the eddy current test result
- if grinding is required, magnetic particle testing is more sensitive to defining the indication as a groove.
CONCLUSION
Eddy current testing is a proven and accepted method for detecting short indications through paint. Its use for the testing of offshore structures is proven in both research and development studies and field practice. Its transference to the testing of nuclear components is valid and the transfer cask eddy current testing project should provide useful for other nuclear operators having similar applications.
ACKNOWLEDGEMENTS
The author would like to acknowledge the following people for their contributions to this article: Fred Cella, of Stone and Webster/Shaw Group; Paul Plante, of Maine Yankee; Helen Goldberg, of Sea Test Services; Dennis Donovan, of Coastal Inspection Services; and Mike Wallace, of Core Technical Services.
REFERENCES
American National Standards Institute, ANSI N14.6, Radioactive Materials - Special Lifting Devices for Shipping Containers Weighing 10 000 Pounds (4500 kg) or More, New York, ANSI, 1993.
American Petroleum Institute, Recommended Practice for Underwater Magnetic Particle In-service Weld Inspection of Offshore Fixed Platforms and Guidelines for Qualification of Inspector Divers (Draft 96-01), Washington, DC, API, 1991.
American Society of Mechanical Engineers, ASME Boiler and Pressure Vessel Code, Section V, Nondestructive Examination, New York, American Society of Mechanical Engineers, 1986 edition, 1987 addenda.
Cella, Fred, "Eddy Current Inspection of Transfer Casks," ASNT Fall Conference, 2003.
Electric Power Research Institute, Reliability of Magnetic Particle Inspection Performed through Coatings, Palo Alto, EPRI, 1988.
Goldberg, L., Reliability of Magnetic Particle Inspection through Coating, Phase II, Houston, Exxon Production Research Company, 1985.
Goldberg, Lawrence O., "Eddy Current Testing, an Emerging NDT Method for Ferritic Weld Inspection," Materials Evaluation, Vol. 56, 1998, pp. 149-152.
Sea Test Services, Ferritic Weld Inspection Using Eddy Current, Joint Industry Study, Merritt Island, Florida, Sea Test Services, 1996.
United States Department of Defense, MIL-STD 271F, 1986.
* 1095 Shady Lane, Merritt Island, FL 32952; (321) 452-5619; fax (321) 453-8777; e-mail <seatest@aol.com>.
Copyright © 2003 by the American Society for Nondestructive Testing, Inc. All rights reserved.
Copyright © 2008 by the American Society for Nondestructive Testing, Inc. ASNT is not responsible for the authenticity or accuracy of information herein. Published opinions and statements do not necessarily reflect the opinion of ASNT. Products or services that are advertised or mentioned do not carry the endorsement or recommendation of ASNT.
IRRSP, NDT Handbook, The NDT Technician and www.asnt.org are trademarks of the American Society for Nondestructive Testing, Inc. ACCP, ASNT, Level III Study Guide, Materials Evaluation, Nondestructive Testing Handbook, Research in Nondestructive Evaluation and RNDE are registered trademarks of the American Society for Nondestructive Testing, Inc.
ASNT exists to create a safer world by promoting the profession and technologies of nondestructive testing.