Development of Training Modules for Magnetic Particle Inspection Daigo Kosaka 1,a), David J. Eisenmann 1,b), Darrel Enyart 1,c), Norio Nakagawa 1,d), Chester Lo 2, and David Orman 1,e) 1 Center for Nondestructive Evaluation, Iowa State University, 1915 Scholl Road, Ames, IA 511-342. 2 Western Digital, Van Nuys, CA a) Corresponding author: dkosaka@iastate.edu b) djeisen@iastate.edu c) denyart@iastate.edu d) nakagawa@iastate.edu e) orman@iastate.edu Abstract. Magnetic particle inspection (MPI) is a nondestructive evaluation technique used with ferromagnetic materials. Although the application of this method may appear straightforward, MPI combines the complicated nature of electromagnetics, metallurgical material effects, fluid-particle motion dynamics, and physiological human factors into a single inspection. To fully appreciate industry specifications such as ASTM E-1444, users should develop a basic understanding of the many factors that are involved in MPI. We have developed a series of MPI training modules that are aimed at addressing this requirement. The modules not only offer qualitative explanations, but also show quantitative explanations in terms of measurement and numerical simulation data in many instances. There are five modules in all. Module #1 shows characteristics of waveforms and magnetizing methods. This allows MPI practitioners to make optimum choice of waveform and magnetizing method. Module #2 explains how material properties relate to the magnetic characteristics. Module #3 shows the strength of the excitation field or the flux leakage from a crack and how it compares to the detectability of a crack by MPI. Module #4 shows how specimen status may influence defect detection. Module #5 shows the effects of particle properties on defect detection. BACKGROUND Magnetic particle inspection (MPI) can find small cracks on surfaces of ferromagnetic materials. Although the application of this method appears relatively straightforward, MPI combines a series of complicated techniques into a single inspection. There is an ASTM standard (E-14444) [1] for MPI. The standard explains the method, however, the standard doesn t address the complications inherent to the inspection process. Resolution of many of these complications are handled via training, but often lack quantitative rigor. OBJECTIVES The purpose of this paper is to develop training modules for MPI practitioners. The modules should have following characteristics: The modules should have explanations about the inspection process and details associated with it that MPI practitioners can understand.
The modules should give qualitative general explanations and quantitative solutions with experimental measurements or numerical simulation when possible. TRAINING MODULES We have developed five modules. Table 1 shows contents of the modules. Characteristics of each module are shown below. TABLE 1. Contents of the training modules for MPI. Module number Contents 1 Characterization of current waveforms. Correlation of waveform to bench settings and readout. Correlation of waveform with magnetic field at part surface and with respect to amperage. 2 Magnetization parameters of the part. Measure in situ leakage field around fatigue cracks to identify optimal magnetizing parameters. 3 Detectability of cracks. Field strength or magnetic field and direction at critical locations of parts. 4 Descriptions of Component effects. 5 Bath characteristics. Module #1 MPI equipment provides users with several options of waveforms and magnetizing methods. After completing this module practitioners should be able to make optimal choices of waveform and magnetizing methods based on the inspection requirements. This module shows characteristics of optimal waveforms and magnetizing methods. The module has three following sections. 1. Characterization of current waveforms (AC, HWDC, FWDC). 2. Correlation of waveform to Bench Settings and Readout. 3. Correlation of waveform with the magnetic field at part surface and with respect to amperage. This is the longest of the five modules because it explains the basics of MPI. After completing this module the user should have at least a rudimentary understanding of the technique. For example, characterizations of current waveforms, details of bench readout and experimentally measured current and correlations between detected cracks and magnetization methods have been explained in the module. Figure 1 shows a characterization of current waveform. This figure illustrates the electrical circuitry and the waveform for the FWDC excitation. Module #2 Magnetic material properties are important when crossing the magnetizing method and current value. This module explains how material properties relate to the magnetic characteristics. It also demonstrates magnetic flux leakage which is how Magnetic particles attach to a specimen A knowledge of the correlation between magnetic flux density and field is important for MPI work because it is desirable that the permeability of samples is large enough in MPI. Also, nonlinearity of permeability (hysteresis and high retentivity) is needed in the residual technique. Reliable demagnetization requires larger current than magnetization as the retentivity of the sample must be overcome in each direction as the field reverses direction and steps-down. Therefore characteristic of ferromagnetic materials have been shown in section 1 by using a quotation [2].
In section 2, the determination of the proper magnetizing current levels by using formulas and QQI s [3] is explained. Measured calculated results of situ field leakage are also shown. Because magnetizing current is one of the most important factors it is imperative that MPI practitioners understand the correlation between magnetizing current and the situ leakage field. Figure 2 shows computational representations of DC magnetization. Leakage resulting from a defect drops off very quickly with distance from the surface of the part. Magnetic circuit Inside of MPI equipment Thyristor Secondary side Trigger Transformer Control circuit Primary side Voltage of primary side 1.5 -.5-1 Trigger timing Voltage of secondary side 1.5 -.5-1 Primary side.2.4.6.8.1 Time (sec) Secondary side.2.4.6.8.1 Time (sec) FIGURE 1. Simplified schematic of the electrical circuitry for the FWDC excitation. Two thyrisotors convert the negative part of the AC waveform to positive. The duty cycle can be reduced by the thyristors. The lower plot shows a limited duty cycle waveform, where the RMS current is 5% of that of the full duty cycle FWDC. y x 12 mm Steel 1 mm Crack Current direction 12 mm Flux direction.2 mm ΔB of Y direction (T).1.5 -.5 -.1 1A 2A 4A 1A -2-1 1 2 X position from center (mm) FIGURE 2. These figures show a simulation model and the computations. This figure illustrates a simulation model of DC magnetization. Leakage flux is generated from a position of about ±.4mm from center of a crack. The position of the flux leakage is outside of the edges of crack opening. Therefore the magnetic particle pattern is larger than the crack width. Module #3 This module examines the strength of the excitation field or the flux leakage from a crack and how it compares to the detectability of a crack with MPI. The module has two sections.
1. Detectability of cracks (Head and coil shots). 2. Field strength or magnetic field and direction at critical locations of parts (Head and coil shots). In section 1, the leakage field is shown by using measured and computed results. As shown Fig. 3, Leakage flux is very small. However as shown Fig. 3, an easily discernable magnetic particle pattern is produced by the leakage flux. This shows that the quality of the magnetic particle pattern cannot be determined by just the strength of the leakage flux. In section 2, the measured magnetic field around samples is shown. The field strength varies with location. As a result the location of cracks have a large impact on their detectability. Module #4 This module shows the effects of sample surface conditions on magnetic particle inspection. It includes descriptions of component effects, surface status, and the effect of the shape of the samples. Figure 4 shows photographs of magnetic particle patterns on various surfaces. It illustrates that finding cracks becomes more difficult as brightness of the background approaches the brightness of the indication at the crack. Probability density 4 3 2 1 Without crack With crack 11 12 13 By (mt) FIGURE 3. Situ leakage field in the residual technique and the magnetic particle pattern are shown. Measured situ leakage field in horizontal direction is shown. Difference between the field values was about 1.5mT. Image showing the ease with which you can find a crack indication that has sufficient brightness in a photograph of magnetic particle pattern.
(c) (d) FIGURE 4. Photographs of magnetic particle pattern. Mill scale surface. Belt-sanded surface. (c) Milled surface. (d) Ground surface. Module #5 This module explains magnetic particle density and the effects of contamination of the bath. When the concentration of the particles is too low, the resulting low illumination will cause indications to be missed. When the concentration is too high too bright of a background is produced and indications can blend into the background and be missed. Warm-up of the bath takes time and is often neglected. Particles build up on the surface of the tank and must be scrubbed to obtain proper particle density. Running the pump alone is not enough to clean the surface and stabilize the particle levels. As an example it was found that the particle concentration is still unacceptably low after 2 days of runtime without scrubbing. As shown in Fig. 5, scrubbing all surfaces in the tank is necessary to reach proper particle concentrations. FIGURE 5. Concentration of magnetic particle in the bath.
SUMMARY Five training modules were developed for MPI practitioners. Quantitative explanations by measurement and numerical simulation were used where possible in the modules. It is hoped that these modules will improve understanding the MPI techniques and practices. ACKNOWLEDGMENTS The research team would like to acknowledge the contributions of many of the OEM s and air carriers who have provided advice, guidance and in some cases actual parts for use in the study. We would also like to acknowledge the Federal Aviation Administration, for their support under Grant number 1-G-2. Without their support this work would not have been possible. REFERENCES 1. ASTM International, ASTM E1444 / E1444M - 12, ASTM International, 212. 2. Richard M. Bozorth, Ferromagnetism, IEEE Press, p. 4, 1993. 3. Louis Cartz, Nondestructive Testing, ASM International, p. 157, 1995.