Capabilities and limitations
July 11, 2006
Although eddy-current testing is king in the tube production industry, flux leakage is worth a look. Using a small test sample, equipment-maker InspecTech found that for 3.50-in. to 4.00-in. carbon steel tubing with wall thickness from 0.056 in. to 0.100 in., the flux leakage method found more defects and turned out fewer false alarms than the eddy-current method did.
Tube producers routinely test weld seams in carbon steel on the tube mill as the product is made. Such testing is a vital part of any well-rounded quality assurance program. Ultrasonic or electromagnetic nondestructive testing (NDT) techniques are acceptable methods under most codes and practices.
Ultrasonic testing is the method of choice, but for thin-walled or small-diameter tubing (wall thickness less than 3 mm or 0.125 inch, or diameter less than 50 mm or 2 in.), ultrasonic testing is not applied easily to online testing. This leaves electromagnetic testing as the most viable option for testing small tubes. Specifically, eddy-current testing has been unchallenged in the tube industry for several decades.
Tube producers have another electromagnetic testing option: flux leakage (or diverted flux).
The standard reference for the NDT industry is "The Non-Destructive Testing Handbook," which now runs to 10 volumes and covers all aspects of NDT. Volume 4, which deals with electromagnetic testing, discusses two factors:
Tube producers want NDT systems that have PD= 1 and PS= 0. This leads to two observations: First, this isn't feasible, and second, the working relationship between PDand PSis influenced by external factors such as operator skill level and the mill environment.
Therefore, selecting an optimum test method for a given environment can be complex, involving not only how effectively a given system will find defects, but also how well the system will be understood and used in the harsh environment of the tube producer's plant.
Simultaneously operating both systems, flux leakage and eddy current, on one tube mill to inspect the same product yielded the differences in the two systems' capabilities.
Principles of Operation. Eddy current (E/C) and flux leakage (F/L) are two types of electromagnetic test.
An E/C test system uses a coil carrying a high-frequency alternating current to create electromagnetic fields in a conductor, causing eddy currents to flow in the conductor. The eddy currents oppose the primary field. The test system uses a probe to detect the eddy currents. If the conductor is uniform in nature, the electrical characteristics are stable. If the conductive medium changes (if it encounters a defect), the electrical equilibrium is disturbed. The result is amplified and presented as an alarm.
It is important to note that carbon steel is a ferromagnetic material, and a direct eddy-current test on a magnetic tube weld is nearly useless because the weld surface generates many meaningless signals. Therefore, it is necessary to apply a strong magnetic field to the steel to render it the same as a nonmagnetic electrical conductor such as copper or aluminum. A magnetic saturator applies the magnetic field.
The distance between the probe and the item under test influences the eddy-current effect, so it is important to minimize any variation in the spacing. Also, the eddy-current field decays in the material, resulting in reduced sensitivity through the thickness of the test piece. Finally, the relative permeability is 1 for nonmagnetic materials, but over 3,000 for unmagnetized steel. This is why the magnetic saturator is needed for measuring anything other than the surface noise in the weld.
The F/L test does not require eliminating the nuisance value of the magnetic permeability of steel. In fact, the F/L method actually uses this property.
If the material is defect-free and homogeneous, the system distributes magnetic flux uniformly throughout the weld zone. Singularities distort the field and cause some stray flux to leak at the surface, where it is detected, amplified, and presented as an alarm.
E/C testing is characterized by a simple relationship between defect depth and signal amplitude. This is not the case for F/L testing. That aside, F/L has a good response to ID singularities, with signal decay proportional to defect depth. Compared with E/C, F/L is easy to calibrate and operate, and it provides good absolute results.
F/L testing also is useful when the parent material is galvanized or aluminized, and the weld zone is remetalized immediately after welding and before testing. In this situation, a thin coating of zinc or aluminum covers the steel to be tested. The conductivity of the zinc or aluminum is about 60 percent on the International Annealed Copper Standard (IACS) scale; the steel's conductivity is around 10 percent IACS. The eddy-current activity is concentrated in the coating, and testing the steel below becomes a second layer problem. The F/L test, on the other hand, responds only to magnetic materials and treats aluminum or zinc coatings as it would respond to air.
An eddy-current test system was installed alongside an existing F/L unit on a tube mill that was producing 3.50-in. and 4.00-in. carbon steel tubing with wall thickness from 0.056 in. to 0.100 in. To eliminate any variations between the two systems, one factory technician supervised the calibration of both systems, using one set of calibration pieces. The data loggers recorded production characteristics and defect indications only when the factory technician was present and supervising both systems. During the course of the test, the data loggers tested 104,700 feet of material.
The trials tested both carbon steel and aluminized product, but the test regimen did not require separate records for each type of product.
The systems automatically marked all defect indications by separate spray paint systems. Quality assurance personnel inspected coupons cut from the painted locations. This study ignored defect clusters at the beginning and end of each coil because tube producers routinely crop these from the finished product.
In most cases, the reason for the indication was visually obvious. When there were no visual clues, the coupon was crush-tested to check for internal defects.
These factor weights were assigned:
The analysis showed that F/L performed well in finding actual defects (categories 1, 2, and 3) and had fewer false alarms (category 4) than did E/C (see Figure 1).
The probability calculations are somewhat crude because of the small sample size and the lack of minute investigation for additional defects that might have been missed by both systems.
If all defects of severity level 1, 2, and 3 are considered rejectable and all category 4 indications are considered spurious, this test generated 27 rejectable indications and 36 spurious indications.
This analysis of PDand PSis superficial at best, and accurately determining these factors in a tube mill would be close to impossible. Nonetheless, the results indicate that the F/L test method has much to offer.
A.C. Richardson is president of InspecTech Analygas Group, 450 Midwest Road, Scarborough, ON M1P 3A9, Canada, 416-757-1179, fax 416-757-8096, www.inspectech.ca. Murray Rose and Rick Northrup, formerly of AK Tube LLC, contributed to this article.