Electromagnetic testing (ET) is a nondestructive testing (NDT) method that can be used to detect flaws and structural variances in materials. The most popular form of electromagnetic testing is eddy-current testing (ECT), which traditionally is used to find surface and subsurface flaws in both new and existing metal structures. Applications include inspection of bars, tubes and wire for surface flaws; of welds for cracks; the IDs of heat exchanger tubing, aerospace structures and components; and mechanical components such as bearings and gears.
Electromagnetic testing can also be used to verify that a metallic component has the proper microstructure, including verification of proper alloy content, proper heat treatment (hardness and case depth, for example) and proper physical shape. Verification of the desired end result is important in conventionally heat-treated parts, as well as induction heat-treated parts. ET systems are typically installed in a production line down stream of the heat-treating process.
The electromagnetic structure test is accomplished by placing a component within a varying magnetic field, which is typically created by using an encircling probe consisting of "driver" and "pick-up" windings. The component being tested behaves similarly to a core of a transformer, assisting in the transfer of energy from the primary coil to the secondary coil (Fig.1). Changes in the component's microstructure alter the strength and distribution of the electromagnetic energy in the circuit in predictable relationships.
Probe design and configuration
An electromagnetic testing system consists of an electronic instrument, probes and positioning fixtures. The fixtures are used to hold the component stable relative to the probes. Because ET is very sensitive to variations in physical shape, measuring components in a consistent manner is crucial.
Two ways that testing probes are typically wired are external reference and internal (self) reference (Fig. 2). Effective tests can be arranged either using two external-reference probes, which make measurements by comparing to each other, or by a single internal-referencing probe.
An external-reference coil arrangement provides increased sensitivity over testing with a single probe. In an external-reference configuration, a reference part is placed inside the reference probe and the test part is placed inside a separate test coil. The driver coils of both probes receive the same ac energy input. The pick-up, or sensor, coils in both probes are monitored and compared by means of the test instrument electronics.
Internal-reference, or self-reference, probes offer the advantage of testing using a single probe. In some applications, where a probe may be moved along the length of a part such as a shaft, a single probe may be the only practical solution. The additional winding required in the self-reference probe makes each probe slightly more expensive compared with a single self-referencing probe.
In many cases, critical areas on a component are hardened using induction heating methods. ET coils can be formed into custom probes to test these areas. For example, Fig. 3 shows how coils can be positioned to test bearing races on an automotive spindle.
One additional method to test hardness on large areas is to use a spot probe. A spot probe uses internal-reference coils, which send a varying magnetic field into a localized area (Fig. 4).
Structural changes in materials are indicated by variations in both conductivity and permeability. Permeability changes are of a much higher magnitude compared with the relative conductivity changes that might be detected.
Changes in material structure are often discovered at different testing frequencies using the electromagnetic structure test methodology shown in Fig. 1. If a single known defect is well defined, it is possible to test the component at a single test frequency. Many manufacturers are testing parts using multiple frequencies to ensure that multiple anomaly situations will be discovered. A typical multifrequency test protocol can use four or eight frequencies. Using a broad range of frequencies helps to ensure that most anomalies will be located. A typical eight-frequency test spans from 25 Hz to 25 kHz. Figure 5 shows a testing output for a small steel slug. Differences between this sample and the stored acceptance criteria occur at 25, 180 and 484 Hz.
Using test frequencies below 200 Hz can require that the component being tested be held still during the test. Testing small components such as bearings or small gears can be done in a continuous manner to improve production throughput.
The question most frequently asked about electromagnetic testing is: "Can electromagnetic testing show what is the Rockwell hardness of a part?" Unfortunately, the answer is no. Electromagnetic testing is a qualitative rather than quantitative testing method. It can indicate whether a component being tested has a different microstructure than that of a reference component.
PM part testing case study
One Rockwell hardness correlation test was done on powder-metallurgy (PM) gears used in an automotive fuel pump. In this test, an in-line hardness inspection and sorting system was built to sort 15 distinct star sizes, 10 distinct ring sizes and 3 different lobe shapes per size. Figure 6 shows a sample of the parts tested.
To obtain an appreciation of the changes in hardness of the specimens, several tests were conducted to develop a correlation between a measured Rockwell hardness (HRB) and the ET-estimated hardness. Before the tests were conducted, the hardness of several specimens was measured for use as the known "good" reference specimens. The testing instrument was set up using the reference specimens so an alarm box could be estimated. After establishing the alarm box to show good separation between the hardness highs and lows, sample specimens were passed through an encircling hardness coil.
A total of 380 pieces (38 sets of 10 pieces each) were passed through the hardness coil, and printouts were created for each set. All 380 pieces were then tested using a typical Rockwell hardness tester. The difference between ET hardness estimates and the measured Rockwell hardness readings are shown in Fig. 7. In addition to the initial testing, 22 suspect and failed specimens were tested as well. Rockwell hardness readings, ET readings and microstructure examinations were completed on these 22 pieces, and the results are also included on Fig. 7.
After completing initial testing, a production sampling test was started. During the test, nearly 35,000 pieces were passed through the hardness coil. Random pieces were taken for Rockwell hardness testing to verify the results. The results from this test are shown in Fig. 8. 
Figure 7 shows a good correlation between Rockwell hardness and ET hardness results from initial testing, showing only about 1 to 3 HRB points variation. Both hardness-testing methods were able to detect a low-hardness specimen. The ET hardness values tend to show a much more consistent value compared with Rockwell hardness values at both higher and lower hardnesses. This consistency is more apparent with the sample production testing results (Fig. 8), which shows the ET results at a nearly flat line.
While electromagnetic testing of part microstructure does not provide a quantitative test value, it does have the ability to accurately sort parts based on hardness, case depth and shape. Testing shows that the method can detect a difference of +/-1.0 HRB value under certain conditions. Correlation of ET data with Rockwell hardness test values and destructive testing methods allows manufacturers to easily provide 100% in-line testing at production line speeds. In many cases, the electromagnetic testing is superior to a standard Rockwell hardness test as the percentage of surface area tested by ET is much greater than that of a single hardness test.