Magnetic Particle Testing MPI
Magnetic particle inspection finds cracks that break the surface, or sit just beneath it, in steel and other ferromagnetic components. It is quick, it is sensitive to exactly the flaw type that matters most — tight, planar cracking — and it works on a component in place, in daylight, without cutting anything.
We carry out MPI on site across the UK and on the bench at our Lillyhall facility in Cumbria, using dry powder, wet visible and wet fluorescent techniques, to ISO 9934, ASME V Article 7 and customer specifications.
How magnetic particle inspection detects a crack
Magnetise a steel component and the magnetic field runs through it, largely unseen. Where the field meets a discontinuity — a crack, a lap, a lack of fusion at the toe of a weld — it cannot pass through the gap. It leaks out of the surface instead and back in again on the far side, forming two small opposing poles either side of the flaw. That escape is called flux leakage.
Apply fine ferrous particles and they are drawn to that leakage field and held there, bridging the flaw. The particles gather into a line several times wider than the crack itself, which is why MPI can show a crack far too tight to see with the naked eye. The indication you read is not the crack — it is the particles the crack collected.
Two consequences follow, and both matter more than most people expect.
Orientation decides everything. A flaw only leaks flux if it lies across the field. A crack running parallel to the field disturbs almost nothing and can be missed entirely. That is why a proper examination magnetises in two directions at roughly 90° to each other and inspects after each. A single-direction inspection is half an inspection, and it is the most common shortcut we see on work that arrives for a second opinion.
Depth is limited. Flux leakage from a flaw a few millimetres down is weak and spreads out, giving a faint, diffuse indication. MPI is a surface and near-surface method. For anything genuinely buried, it is the wrong tool — that is ultrasonic testing or radiography.
The limitation that decides whether you can use it at all
The component has to be ferromagnetic. Carbon and low-alloy steels, cast iron, and the martensitic and ferritic stainless grades are fine. Austenitic stainless — the 300 series, so 304, 316 and their variants — is not. It will not hold a useful magnetic field, and no amount of current will change that.
This gets written into specifications more often than it should be. If MPI has been called up on an austenitic weld, the technique is wrong and the answer is dye penetrant, which detects surface-breaking flaws by capillary action and does not care what the material is. We would rather tell you that before we mobilise than after.
Dry powder, wet visible or fluorescent
Three techniques, and the choice is about sensitivity, surface condition and where the work is happening.
Dry powder applies coloured magnetic powder from a bulb or puffer while the field is on. It suits rough, hot or heavily contoured surfaces — as-welded seams, castings, structural steel — and needs no cleaning up afterwards beyond brushing off. It is the least sensitive of the three to very fine cracking, and works best in good light.
Wet visible suspends the particles in a carrier fluid, usually applied by aerosol or from a tank, over a white contrast paint. The particles are far finer than dry powder, so sensitivity is better, and the black-on-white contrast is easy to photograph for a report. It is the everyday choice for weld inspection on site.
Wet fluorescent uses particles coated in a fluorescent dye, viewed under UV-A at 365 nm in a darkened area. This is the most sensitive technique by a clear margin: the eye picks out a bright green line against near-black at a contrast ratio nothing in visible light can match. It is what we use for aerospace and safety-critical work, and on the bench where the light can be controlled.
Fluorescent inspection has conditions attached that are easy to skip and invalidate the result. The viewing area has to be genuinely dark, the inspector needs a few minutes for their eyes to adapt, and the UV lamp's intensity has to be verified rather than assumed — lamp output falls off with age long before the bulb fails.
Magnetising equipment, and what each is for
Electromagnetic yoke. A hand-held U-shaped magnet placed on the surface, magnetising the region between its legs. Portable, no risk of arc strikes, and by far the most common method on site. Rotate it 90° and inspect again for the second direction.
Prods. Two contact tips passing current directly through the component, generating a circular field around the path. Higher sensitivity over a larger area, but the contacts can arc and leave a burn on the surface, so prods are excluded on many specifications for finished or high-integrity work.
Bench unit. A fixed machine that clamps the component between headstocks for a current-through shot, and encircles it with a coil for a longitudinal shot — both directions from one setup, with controlled current and repeatable results. This is how batch work should be done, and it is what our Lillyhall facility is set up for.
Central conductor. A bar passed through a bore, magnetising the wall circumferentially from the inside. The right method for ring forgings, flanges, hollow shafts and pipe ends.
You will still hear MPI called Magnaflux, after the equipment manufacturer whose name became shorthand for the method — in the same way people say Hoover for a vacuum cleaner. It refers to the technique, not to a particular brand of kit.
Why MPI, and where it stops
It earns its place because it is fast, it is cheap per component, the consumables are inexpensive, it needs far less surface preparation than penetrant, and results appear while you watch rather than after a dwell time. On a fillet weld toe or a shaft fillet radius, where fatigue cracking initiates, nothing else gives you an answer as quickly.
Where it stops: ferromagnetic materials only, surface and near-surface only, orientation-dependent so it needs two directions, and it can leave residual magnetism that interferes with machining or with instruments later — so demagnetisation may be required and should be specified up front, not discovered at assembly. Coatings matter too: paint much beyond about 50 µm starts to mask fine indications, and thick coatings have to come off.
On site or on the bench
On site we work from portable yokes with dry or wet visible technique, on structural steel, pressure equipment, lifting points, fabrication welds and plant during outages. Access and light are the constraints, so we plan the second magnetising direction around them rather than hoping.
In the laboratory at Lillyhall, a bench unit takes components up to the machine's capacity, gives both field directions from one setup at a controlled current, and allows fluorescent inspection in a properly darkened booth. For one-off components or batch work — forgings, castings, machined parts, welded assemblies that will fit — the bench gives better sensitivity and a more repeatable result than anything achievable in a fabrication shop. Send components to us and they come back with a report; we can also take a lab-standard setup to you where volume justifies it.
Certification and reporting
Our technicians hold PCN Level 2 in magnetic particle testing for welds, forgings and castings, and we provide PCN Level 3 services for procedure writing, technique approval and third-party review.
Every examination is reported against the acceptance criteria named in your specification, with indications located, sized and photographed where they are recordable. Reports are issued from our digital reporting system, so a component inspected repeatedly builds a comparable history rather than a stack of unrelated PDFs.
If you are unsure whether MPI is the right method for your material and flaw type — or you have had it specified on austenitic stainless and want a second opinion before mobilising — tell us what the component is and we will tell you straight.
