What both methods can and cannot do
Magnetic particle and dye penetrant are both surface methods. They find cracks that break the surface, and in the case of magnetic particle, discontinuities a short distance beneath it. Neither one sees into the body of a weld or a casting. If the question is whether there is lack of fusion halfway through a 25 mm butt weld, no amount of penetrant will answer it — that is an ultrasonic or radiographic job.
Within that limit they are the most sensitive, fastest and cheapest tools available. A fatigue crack in a lifting lug or a monorail flange will be found by magnetic particle in minutes, at a size no visual inspection would reliably pick up.
Magnetic particle testing
The part is magnetised. Where a discontinuity interrupts the flow of magnetic flux, some of that flux leaks out at the surface and forms two tiny opposing poles either side of the defect. Ferrous particles applied to the surface are attracted to that leakage and pile up along it, producing an indication many times wider than the crack itself. That amplification is the whole trick — it turns a crack far too fine to see into a visible line.
Practical points that decide whether it works:
- Ferromagnetic materials only. Carbon steel, low-alloy steel, cast iron and most structural steels, yes. Austenitic (300 series) stainless, aluminium, copper, brass and titanium, no — they cannot be magnetised, and the method simply does not function.
- Two directions, always. A crack lying parallel to the flux lines barely disturbs them and will not show. Every area has to be magnetised twice, roughly 90° apart. Skipping this is the most common way a real defect is missed.
- Yoke technique for site work. An AC electromagnetic yoke placed across the area is portable, needs no electrical contact with the part, and leaves no arc strikes. Prods pass current through the component and can burn it, which is why they are usually prohibited on in-service plant.
- Sub-surface reach. With the right technique, magnetic particle will reveal discontinuities lying a couple of millimetres under the surface. Penetrant will not find those at all.
- Coating tolerance. A thin, sound non-magnetic coating — paint, plating — reduces sensitivity but does not stop the method. Thick or blistered coating does.
- Wet fluorescent or dry visible. Wet fluorescent particles under UV-A are the more sensitive option but need darkness, which is often impractical on a façade or a structure in daylight. Dry visible or colour-contrast particles are the workhorse for site inspection at height.
Carried out in Australia to AS 1171, with acceptance to the applicable fabrication or in-service standard.
Dye penetrant testing
A low-viscosity penetrant is applied to a clean surface and left to dwell — usually 10 to 30 minutes — while capillary action draws it into anything open to the surface. Excess penetrant is removed carefully, then a developer is applied, which acts like blotting paper and pulls the trapped penetrant back out. It bleeds into the developer coating and spreads, again producing an indication much larger than the defect.
What governs whether it works:
- Material is irrelevant, porosity is not. Any non-porous material can be tested — the method does not care whether it is magnetic. Porous materials cannot, because the whole surface holds penetrant.
- The defect must be open and clean. This is the real constraint. Paint, plating, oil, water or blasting media in the crack will block the penetrant. So will smeared metal: grinding, machining, shot blasting and wire brushing all drag material across a crack and close its mouth. A ground weld repair that has just been dressed may need etching before it can be honestly penetrant tested.
- Surface finish. A rough as-welded surface holds penetrant everywhere and produces a background that masks genuine indications. Magnetic particle copes with the same surface far better.
- Temperature. Standard penetrants are qualified for roughly 10 °C to 50 °C. Hot steel in the sun in a Queensland summer will exceed that, and the penetrant dries in place; the inspection either waits or uses a qualified high-temperature procedure.
- Colour contrast or fluorescent. Red-on-white colour contrast is read in daylight and is the practical choice on site. Fluorescent penetrant is more sensitive but needs UV and darkness.
Carried out to AS 2062, with acceptance to the applicable standard.
Choosing between them
| Magnetic particle (MT) | Dye penetrant (PT) | |
|---|---|---|
| Materials | Ferromagnetic only | Any non-porous material |
| Finds | Surface and slightly sub-surface | Surface-breaking only |
| Rough surface | Tolerant | Poor — high background |
| Thin coating | Works with reduced sensitivity | Does not work |
| Time per area | Minutes | 30–45 minutes with dwell |
| Orientation | Two directions required | Not orientation sensitive |
| Consumables | Particles, yoke power | Cleaner, penetrant, developer |
| Standard | AS 1171 | AS 2062 |
In plain terms:
- Carbon steel and you have a choice — magnetic particle, nearly always. It is quicker, it is not fooled by a rough weld cap, and it finds things just under the surface.
- Stainless, aluminium or any non-ferrous alloy — penetrant, because there is no choice.
- A stainless weld on a carbon steel component — the parent may be magnetic and the weld metal not, so the weld itself needs penetrant.
- Painted and you cannot remove the coating — magnetic particle may still give a usable result; penetrant will not.
- Working at height via rope access — either method works, but plan for daylight. Fluorescent techniques need darkness that a façade or a stack does not provide.
- Nothing found by either, and you still suspect a defect — the defect is internal, and the next step is ultrasonic testing.
Post-repair verification
The most common use of both methods on our sites is confirming a repair. A crack is excavated, the excavation is checked to confirm the defect has actually been removed — usually magnetic particle on steel — the weld is completed, and the finished weld is checked again. Two inspections, not one. Grinding out a crack until it disappears from view is not evidence that it is gone; the tail of a fatigue crack is far finer than the part you can see.
Frequently asked
Which is more sensitive?
For surface-breaking cracks in steel under good conditions they are comparable, and both are far more sensitive than visual inspection. Magnetic particle wins overall on steel because it also reaches slightly below the surface and is not defeated by a rough or smeared finish.
Can you test a crane hook or a lifting lug in place?
Yes. Yoke-technique magnetic particle is designed for exactly that — the component stays where it is, no electrical connection is made to it, and the inspection is done in position, including at height on rope access.
Does either method leave the component contaminated?
Penetrant and developer are removed after the inspection. Where the component is going back into food, oxygen or high-purity service, low-halogen and low-sulphur consumables are used and post-cleaning is part of the procedure — worth specifying at the enquiry, not after.
Do we need to be shut down?
No. Both are external surface methods carried out on cold, accessible steel. The practical constraints are surface temperature, access and having the area clear of other trades — not an outage.