Figures 1-3
Figures 4-6

Introduction
This article covers the evaluations performed on steel test ring standards for magnetic particle inspection since 1985. Initially, it was determined that two types of steel test rings were being used; improperly annealed and properly annealed rings. The number of holes that could be detected visually, at selected amperage, varied between the two types of rings.

Attempts were made to prepare a Society of Automotive Engineers Aerospace Material Standard (SAE/AMS) for fabricating the rings. Because of the subjectivity in the number of holes that could be detected by various observers, it was decided to use a more quantitative means for certifying the rings. Evaluations were performed manually to measure the leakage field from each hole at selected amperage settings. A fixture was made to hold the ring so as to obtain repeatable results. The fixture was then automated and dynamic readings were taken from a variety of rings. Acceptance criteria have been established for both the manual and automated techniques and were used to prepare the SAE/AMS standard.

---

Acceptance criteria have been established for both the manual and automated techniques.

---

 

Initial Study
The initial evaluation of a variety of rings revealed a difference in the number of holes detected by five observers using the fluorescent wet continuous technique at amperages from 500 to 3,500 A. Figure 1 shows typical results obtained from five rings (Hagemaier, 1993). Note that the National Institute of Standards and Technology (NIST) ring became magnetically saturated at 1,500 A and ring 87009 became saturated at 2,000 A. Both these rings are considered unacceptable.

 

Manual Flux Leakage Measurement
Each ring was magnetized at increments of 500 A (5003,500 A) and the residual magnetization at each hole was recorded using a gauss meter and transverse field probe. The probe was hand held but in contact with the ring surface. The results were encouraging and a fixture was designed and fabricated to accurately position the gauss meter probe at a constant location and distance above the surface of the steel ring. The rings were magnetized at 500, 1,500, and 2,500 A. Figure 2 shows the results at 1,500 A. Similar results were obtained at 500 and 2,500 A (Hagemaier et al., 1996). The values for the three properly annealed steel test rings are together in a uniform decaying curve. The NIST ring and unannealed steel rings exhibited broad, poorly defined leakage fields and additional flux leakage locations between the normal leakage locations near the holes. This phenomenon was not observed on any properly annealed steel test rings. The leakage readings correlate nicely with the visual observations shown in Figure 1.

 

Automated Flux Leakage Measurement
A table and rotating fixture were designed and fabricated for performing the automated measurements. Each ring was placed on the scanning system table and an internal copper conductor was inserted through the hole in the ring. Two electrical cables were attached to the magnetic particle test machine and clamped near the ends of the central conductor bar (see Figure 3). Liftoff adjustments were made between the Hall detector probe and the ring surface. The ring was then magnetized at the appropriate amperage and the motor was turned on. The motor operated at 600 rpm. The output from the Hall probe was fed through a conditioning amplifier to an oscilloscope and the resulting image was photographed.

The automated results obtained from the nine holes in each ring (Figure 4) can be directly compared to the manual gauss value curves of Figure 2. The sensitivity of the oscilloscope was set at 10 mV per division. The results at 1,000 and 1,500 A were exactly alike. It can be seen from Figure 4 that rings 285-23 and 680-10 yield a very uniform pattern with the peaks along the calibration line and not exceeding the ±5 percent limit lines. Ring 870009 is considered unacceptable from results obtained from the three techniques of measurement (see Figures 1, 2, and 4). The NIST ring yielded such low amplitude signals at 1,000 A and 10 mV per division that the sensitivity of the trace in Figure 4 was increased to 5 mV per division. The low amplitude trace from the NIST ring is also obvious in Figure 2.

 

 

Summary
Evaluations of the steel test rings for magnetic particle inspection have been conducted since 1985, when Kermit Skeie first discovered that the rings were not uniform in magnetic permeability. A survey of users in the USA revealed that the rings fell into five-hole or nine-hole detectability at 2,500 A, full wave direct current (Hagemaier, 1993). Further investigations revealed that the five-hole rings were improperly annealed and became nine hole rings after being properly annealed. Also, analysis of papers by Betz (1967) and Gregory et al. (1972) revealed that they both worked with improperly annealed rings. It was Gregory et al. who first introduced the ring as a standard into the McDonnell Aircraft process specification. Their requirements were transferred to MIL-STD-1949 and ASTM E 1444. Further evaluations of properly annealed rings by the authors showed the number of holes that should be detected, as listed in Table 1. Hence the requirements in ASTM E 1444 should be changed to agree with the author's requirements. However, the SAE/AMS requirements for particle certification was revised in AMS 3040-3046.

The draft of the SAE/AMS standard has been reviewed, revised, and returned to SAE for final review. The requirements of Figure 3 (for manual evaluation) or Figure 4 (ring 680-12 automated evaluation) will be used to qualify newly manufactured rings, or existing rings that have been disqualified by visual testing (Figure 1 and Table 1).

To obtain manual results, graph the results with the hole numbers along the abscissa (horizontal) and the gauss values along the ordinate (vertical). Join each value with a line to generate a smooth curve (Figure 2). After the lot measurement, calculate the average gauss value at each hole for the lot and plot the mean average curve (Figure 5). The mean average value for the lot, at hole 1, is 2.01 mT (20.1 G) and 10 percent of this value is 0.2 mT (2.0 G). The upper and lower acceptance limits for the lot of rings are plotted at ±0.2 mT (2.0 G) from the mean average curve (Figure 5). In Figure 5, rings 2-4 are acceptable and ring 1 is unacceptable.

To obtain automated results, magnetize the ring, start the rotation, set the oscilloscope gain, and adjust the oscilloscope horizontal spacing so that the peak reading from each of the nine holes is uniformly spaced on the cathode ray tube trace. Photograph, or record, the trace from each ring in the batch. Select the ring that gave the most uniform response from the nine holes and the one with the highest amplitude from hole 1. Using a French curve, draw a line through both the positive and negative peaks (Figure 6). Measure the total amplitude of hole 1 and calculate 5 percent of that value. Using the 5 percent value for hole 1, draw duplicate lines above and below the positive and negative median lines (Figure 6).

 

References
AMS 3040B-95, Magnetic Particles, Nonfluorescent Dry Method. Society of Automotive Engineers, Warrendale, PA.

AMS 3041C-96, Magnetic Particles, Nonfluorescent Wet Method, Oil Vehicle, Ready-to-Use. Society of Automotive Engineers, Warrendale, PA.

AMS 3042C-96, Magnetic Particles, Nonfluorescent Wet Method, Dry Powder. Society of Automotive Engineers, Warrendale, PA.

AMS 3043B-96, Magnetic Particles, Nonfluorescent Wet Method, Oil Vehicle, Aerosol Packaged. Society of Automotive Engineers, Warrendale, PA.

AMS 3044D-96, Magnetic Particles, Fluorescent, Wet Method, Dry Powder. Society of Automotive Engineers, Warrendale, PA.

AMS 3045C-96, Magnetic Particles, Fluorescent, Wet Method, Oil Vehicle, Ready-to-Use. Society of Automotive Engineers, Warrendale, PA.

AMS 3046B-89, Magnetic Particles, Fluorescent, Wet Method, Oil Vehicle, Aerosol Packaged. Society of Automotive Engineers, Warrendale, PA.

Bates, B., D. Hagemaier, and J. Petty, "1993 Evaluation of Steel Ring Standards for Magnetic Particle Inspection," Materials Evaluation, Vol. 54, No. 10, Oct. 1996, pp 1207-1209.

Betz, C., Principles of Magnetic Particle Testing, 1967. Magnaflux Corporation, Chicago, IL.

E 1444-94a, Standard Practice for Magnetic Particle Examination, 1996. American Society for Testing and Materials, Conshohocken, PA.

Gregory, C., V. Holmes, and R. Roehrs, "Approaches to Verification and Solution of Magnetic Particle Inspection Problems," Materials Evaluation, Vol. 30, No. 10, Oct. 1972, pp 219-228.

Hagemaier, D., "Evaluation of Steel Ring Standards for Magnetic Particle Inspection," Materials Evaluation, Vol. 51, No. 9, Sep. 1993, pp 1046-1051.

MIL-STD-1949, Method of Magnetic Particle Examination. NPODS, Standardization Documents Order Desk, Philadelphia, PA.

 

* McDonnell Douglas Aerospace, MC C0071-0012, 2401 Wardlow Rd., Long Beach, CA 90807-5309; (310) 593-7304.

 ⁺ Armstrong Research, Huntington Beach, CA.

Copyright © 1997 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]