Magnetic Particle Inspection of Tool Steel

What it finds, and how far it reaches

The method magnetizes the part and lets a discontinuity leak flux at the surface, where iron powder settles onto the leak and outlines it. The largest and most important group of discontinuities is the one exposed to the surface, and surface cracks matter most because they do more harm to service life than a buried flaw of the same size. On tool steel the method locates seams, laps, quenching cracks and grinding cracks, and it detects fatigue cracks that open up in service.

A discontinuity is easiest to find when the field runs across it rather than along it, when its depth runs at right angles to the surface, when its opening at the surface is narrow, and when its length is large against its width. Many incipient fatigue cracks and fine grinding cracks are less than 0.025 mm deep with surface openings of perhaps a tenth of that, and those are readily found by the wet method. As a general rule a surface discontinuity whose depth is at least five times its opening at the surface will be detectable, and the gain from a deeper crack fades beyond about 6.4 mm.

Two features of real cracks work against the method and are worth knowing when a report comes back clean. A crack whose faces are pressed together by compressive stress leaves almost no air gap, so it leaks almost no flux and can produce no indication at all, which is common in shallow cracks left by grinding or heat treating. A forging or rolling lap that emerges at an acute angle also leaks weakly, and laps of that kind call for high sensitivity, usually direct current magnetizing with wet fluorescent particles.

Wet particles for cracks, dry particles for rough and buried flaws

Wet particles are the choice for fine discontinuities such as fatigue cracks, and dry particles are the choice for very rough surfaces and for flaws below the surface, where they are usually applied with portable equipment and carried in an air stream. Colour is a practical matter rather than a technical one. Black stands out on light surfaces, red on bright polished surfaces, and fluorescent particles under ultraviolet light give the highest contrast of all.

The fluorescent route has its own numbers. The peak of the useful black light band sits at about 365 nm, and the light at that wavelength is not the harmful short wave ultraviolet that damages skin and eyes. For aircraft-quality fluorescent inspection the minimum intensity at the workpiece surface is 1000 microwatts per square centimetre, and a high-intensity 125 W bulb reaches up to 5000 microwatts per square centimetre at 380 mm. Bath strength matters as much as the lamp, since a weak bath can miss a fine indication entirely and a heavy one buries it in background, and the settling test on 100 mL of agitated bath is the usual way to hold it steady.

Alternating current stops at the surface

Alternating current at 50 or 60 Hz is confined near the surface by the skin effect, so it detects discontinuities that are open to the surface or only a few thousandths of an inch below it. At lower frequencies the skin effect weakens and the lines of force reach deeper. Direct current magnetizes the whole cross section more uniformly, with a straight-line gradient from a maximum at the surface to zero at the centre under circular magnetization. In practice direct current is used on finished parts such as machined and ground shafts, cams and gears, while alternating current is used for fine cracks that actually break the surface and direct current is better for very fine nonmetallic stringers lying just below it.

One further consequence of alternating current is easy to overlook when reading a current setting. The strength of magnetization follows the peak of the sine wave, which is 1.41 times the value shown on an alternating current meter, so an alternating current meter reading is not directly comparable with a direct current one.

The tool steel test that set the depth limit

The clearest published measurement of how deep the method reaches was made on tool steel itself. A ring of unhardened O1 tool steel holding 0.40 percent carbon was made 127 mm outside diameter, 32 mm inside diameter and 22 mm thick, and twelve holes 1.8 mm in diameter were drilled through it parallel to the cylindrical surface at increasing depths below that surface. The centreline depths of the holes ran from 1.8 to 21.3 mm in steps of 1.8 mm. A central conductor, dry particles and continuous magnetization were used, and the current was raised until each hole gave a readable indication.

Threshold currents at which drilled holes in an O1 tool steel ring gave readable magnetic particle indications, with dry particles, a central conductor and continuous magnetization, reproduced from Magnetic Particle Inspection in ASM Handbook, Volume 17. The alternating current column is the one to read first, because it shows the method failing at a depth of 5.33 mm no matter how much current is applied. With wet particles the levels are higher again, and the source gives about 440 A of direct current for hole 1 and about 910 A for hole 3.

The same ring was hardened to 63 HRC and the test repeated, and the cost of that hardness is visible in the current levels. Raising the current does not recover the deep holes, since the limit is set by the current type rather than by the amperage.

Current needed to indicate two depths of drilled hole in the same O1 tool steel ring after hardening to 63 HRC, reproduced from Magnetic Particle Inspection in ASM Handbook, Volume 17. A hard part needs more current still than an annealed one, and at hole 3 the straight direct current figure is about 2700 A.

How the part is magnetized

The current has to be passed in a direction parallel to the discontinuity, so that the field it sets up runs across the crack. A current sent down the length of the part gives a circumferential field that finds longitudinal flaws, and a coil or solenoid around the part gives a longitudinal field that finds transverse flaws. Where the direction of the flaw is not known, both magnetization directions have to be used in turn, and the prods or yoke are repositioned at 90 degrees between the two shots. The published methods and their boundaries are as follows.

General applications, advantages and limitations of the ways a part is magnetized for magnetic particle inspection, condensed from Table 1 of Magnetic Particle Inspection in ASM Handbook, Volume 17. On a long solid part such as a bar or a billet, contacting the ends circularly magnetizes the whole length at once, and on a coil the part should be centred so that the length effectively magnetized in one shot is as large as possible.

Current, and turning the magnetism off again

For circular magnetization the current is set from the diameter of the part. For longitudinal magnetization in a coil it is set from the ampere turn figure, which grows with the length to diameter ratio because the poles at the ends of a short part fight the applied field, an effect that is considerable below a ratio of 10 to 1 and very significant below 3 to 1. A coil around a part magnetizes about 150 to 230 mm either side of itself, so a part longer than roughly 305 to 460 mm has to be moved through the coil or the coil moved along the part.

Current rules for magnetizing a part, reproduced from Magnetic Particle Inspection in ASM Handbook, Volume 17. The diameter is the largest distance between any two points on the outside circumference of the part.

A finished part has to be demagnetized afterwards, or chips will cling to it during later machining. The choice of demagnetizing method turns on the size of the part, its hardness and the rate at which it has to be produced, and an alternating current yoke can serve as the demagnetizer in some cases. The published applicability is as below, where A means applicable and N means not applicable.

Applicability of demagnetizing methods on the basis of part size, metal hardness and production rate, reproduced from Table 2 of Magnetic Particle Inspection in ASM Handbook, Volume 17. A is applicable and N is not applicable. Where the entry reads that a yoke is used for local areas only, the source limits a yoke on a medium-size part to local work.

The part is either held magnetized only while the particles are applied, which is the continuous method, or left with enough residual magnetism to attract the particles afterwards, which is the residual method. Only a metal with sufficient retentivity can be left to the residual method, and the residual field is reliable for surface discontinuities only. The continuous method works on any magnetizable metal and is mandatory for low-carbon steel or iron with little retentivity. Maximum sensitivity on a very fine discontinuity comes from immersing the part in a wet bath, passing the current for a short time during immersion, and leaving the current on while the part is withdrawn and the bath drains.

Grades and stock

Magnetic particle is one method in the set a tool steel order is checked with, and the others sit beside it on the page on choosing a nondestructive test method. How the same methods are applied to bar and billet, including the seam depth a permeability system reaches, is set out on the bar and billet inspection page.

The cracks this method looks for are also the end of a failure sequence that started earlier, and the changes that precede them are described on the decarburization and stock removal page. Where a part has already broken, the marks the crack left on the fracture face are read with the help of the fracture surface features chart, and the wider failure investigation is set out on the tool and die failure analysis page.

Source: ASM Handbook, Volume 17, Nondestructive Evaluation and Quality Control, ASM International, 1989.