5.3 Demagnetization Procedures and Residual Field Limits

Official MT general topic 6 on the outline administered prior to 15 December 2026 is Demagnetization Procedures. Passing the written exam means you know why a residual field is a problem, how shops take it out, how they prove it is gone, and that the numeric limit lives in the procedure, not in a single ASNT constant.

Why demagnetize

A ferromagnetic part that leaves the magnetic particle booth still magnetized can:

• Hold chips and swarf on subsequent machining, grinding, honing, or assembly
• Disturb instruments — magnetic compasses, some avionics sensors, nearby meters, and some electron-beam or welding equipment
• Interfere with later processing — welding arc blow, uneven coating or plating, chips trapped under paint
• Mask or distort a later MT exam if a leftover field from the first shot fights the next field or makes a residual-method shot meaningless
• Attract debris in service on bearings, hydraulic components, and aerospace mechanisms

Not every part must be demagnetized. Soft, low-retentivity parts that were magnetized only with alternating current may already sit near zero because AC is self-demagnetizing. Many written practices require demagnetization only when residual field would affect the next operation or the customer specification. Hard steels after a heavy direct-current or full-wave shot almost always need a planned demagnetization step.

Do not demagnetize before a residual-method particle application. Residual inspection uses the retained field on purpose. Demagnetize after evaluation (or before a conflicting later shot) when the procedure requires a known low residual.

How residual field is removed

Demagnetization works by cycling the material around a shrinking hysteresis loop until the domains are left disordered. Each reversal is a little weaker than the last. A single opposite pulse that is too strong simply remagnetizes the part the other way.

Decaying AC is the workhorse. The part is drawn slowly through a 50/60 Hz coil, or the coil current is ramped down, so each half-cycle reverses the field at a slightly lower amplitude. Pulling too fast leaves residual. Leaving the part at the coil mouth in a strong AC field and yanking it sideways can remagnetize a local zone. Long bars may need more than one pass, reversing the direction of travel, so both ends see a decaying field.

Reversing DC is for parts alternating current cannot finish: heavy sections, deep residual from a full-wave or DC shot, or materials where the AC skin effect leaves the core magnetized after the surface looks clean. The operator or the bench program reverses polarity and steps the current down to a low value. Skipping steps, or ending on a large pulse, leaves a leftover pole.

Yoke withdraw is the portable version of decaying AC: keep the AC yoke energized and slide it away so the part sees a reversing, diminishing field. A DC yoke pulled off the part can leave residual field. Do not treat DC yoke removal as demagnetization unless the written practice says that specific motion is the qualified method.

Verify with a field indicator or gauss meter

"It looks demagnetized" is not a measurement. Particles falling off the part only tell you the field is no longer holding that powder. They do not give a residual-field number.

• A field indicator (pocket compass-type or a dedicated residual indicator) shows whether a residual field is still steering the needle.
• A Hall-effect probe / gauss meter gives a number, usually in gauss or millitesla. Conversion you should know: 1 mT = 10 gauss, so 0.3 mT = 3 gauss.

Measure at the locations the procedure names — often ends, last contact points, coil-mouth positions, gear teeth, and other geometric poles. A part can be quiet in the middle and still a magnet at a sharp end or a thread. If the next process is machining, measure after the part is out of the bench's stray field, not while it is sitting against a steel table that is itself magnetized.

Probe orientation matters. Follow the meter and procedure: residual checks are often taken with the probe in a stated orientation relative to the surface. A single lucky low reading on the mid-span OD does not clear a shaft that still grabs chips at the headstock end.

Residual limits are procedure-specific

ASNT does not publish one universal residual-field number as the Level II pass/fail criterion. Many shop practices and customer specifications use a typical limit of about 3 gauss (0.3 mT) or less at the surface. Some commodity work allows more. Aerospace, landing-gear, engine, and instrument-near parts are often much tighter and may require a documented survey at named locations.

Teach it this way on the exam:

• If the stem gives a limit, use that limit
• If the stem asks for a commonly cited shop number, ≤3 gauss / 0.3 mT is the usual textbook figure — and you must label it as a typical procedure limit, not an ASNT-published constant
• Tighter than 3 gauss is common when the part will fly, go near a compass, enter a precision coating line, or be welded next

A residual reading is not an amperage setting. Three gauss is not "set the head shot to 3." It is a leftover field after demagnetization, measured with a field indicator or gauss meter.

Special cases the exam likes

Parts that will be welded. Residual field causes arc blow: the arc wanders, undercut appears, and fusion can suffer. Demagnetize weldments, adjoining fittings, and sometimes the fixture before a critical weld, even if the leftover field would have been acceptable for chip control alone. A "3 gauss is always fine" answer is wrong when the next operation is a qualified weld.

Aerospace residual limits. Engine rotating parts, landing gear, actuators, and flight-control hardware often carry customer or prime-contractor residual limits well below the generic 3-gauss shop card. Some travelers require a gauss-meter survey at named stations and a recorded value. Follow the traveler. Do not "shop-limit" an aerospace part because that is what the structural shop uses on plate.

Shipping near compasses and avionics. Large magnetized structures — pipe, plates, assembled gearboxes, spare rotors — can deflect a magnetic compass or upset nearby sensors. Export, aircraft, and some marine rules care about this. Demagnetize and verify before the part is crated next to instruments, loaded near a compass, or installed in an assembly that will be compass-swung.

Subsequent magnetic particle testing. If a residual circular field remains, a later longitudinal shot can produce confusing particle patterns (including magnetic writing). A residual-method exam on the next shift is meaningless if yesterday's field is still in the part. Demagnetize between conflicting shots when the procedure requires a known starting state.

Mixed operations. A part that was prodded in the field, then sent to a wet bench, then finish-machined, may need demagnetization at more than one gate. Each gate uses that procedure's limit. The field exam's "good enough for the weld overlay" is not automatically good enough for the machine shop or the paint line.

Putting the sequence together

A complete Level II mental checklist after the last evaluation:

1. Does the written practice or customer specification require demagnetization for this part and the next operation?
2. Choose decaying AC if the section is within AC reach; choose reversing DC step-down if the residual is deep or the section is thick.
3. Use yoke withdraw only for local AC yoke work, and do not confuse it with pulling off a DC yoke.
4. Verify with a field indicator or gauss meter at the specified points, away from stray bench fields.
5. Compare the reading to this job's residual limit — commonly ≤3 gauss / 0.3 mT in general shop practice, tighter in aerospace, weld-prep, and instrument work.
6. Document as the procedure requires. A demagnetization that is not measured is not finished.

On the general exam, a stem that names chips on a lathe, arc blow on the next weld, a compass in the crate, or a later MT shot is asking why you demagnetize. A stem that names a thick full-wave-DC forging is asking reversing DC, not a single surface AC kiss. A stem that names 3 gauss is asking you to treat that figure as a procedure-typical limit, not as ASNT's one published number for every part.