Non-destructive testing (NDT): which method for which defect?

Non-destructive testing (NDT, also written NDE for non-destructive examination) covers the techniques that look for a defect in a part without damaging it or taking it out of service. The real question is never which method is best, but which method for which defect. Each one sees certain families of defect and stays blind to the rest. Choosing well comes down to matching three things: the type of defect you are looking for, the material being tested, and where the defect sits.

Surface or volume: the split that governs everything else

Before comparing methods, you have to understand the split that structures the whole field. A defect belongs to one of two families.

It breaks the surface: an open crack, surface porosity, corrosion, a weld defect reaching the surface. Or it lives within the thickness of the wall: an inclusion, a gas pocket, lack of fusion, a delamination, an internal crack, metal loss on the inaccessible side. The methods line up along this divide, and it is this split that rules out half the options straight away.

A second criterion sits on top of the first: the material. A ferromagnetic material such as a carbon steel allows magnetic particle testing. A conductive material allows eddy current testing. Ultrasonic testing and radiography, for their part, apply to most materials. The defect sought and the material, together, almost always narrow the choice down to one or two relevant methods.

The main non-destructive testing methods

Visual testing (VT)

The oldest and most widely used of the methods. It detects defects visible at the surface: open cracks, corrosion, distortion, surface-breaking weld defects. Assisted by borescopes or cameras (remote visual inspection), it reaches areas the naked eye cannot. Its strength is being fast, cheap and consumable-free, and being the precondition for every other method. Its limit is seeing only what is visible at the surface, with a heavy dependence on access, lighting and the operator's experience.

Penetrant testing (PT)

A coloured or fluorescent penetrant is applied and seeps by capillary action into surface-breaking defects, then a developer draws the indication back out and makes it visible. Penetrant testing detects cracks and porosity that break the surface, on all non-porous materials: ferrous and non-ferrous metals, some plastics and ceramics. It is a simple, low-cost method, sensitive to fine cracks and applicable to complex shapes. In return, it sees only what breaks the surface, requires a clean, non-porous surface, and demands careful cleaning before and after.

Magnetic particle testing (MT)

The part is magnetised: a surface-breaking or slightly sub-surface defect creates a magnetic flux leakage that holds a revealing powder. Magnetic particle testing detects surface and near-surface cracks, but only on ferromagnetic materials (carbon steels, cast irons). It is highly sensitive to surface cracks, fast, and tolerant of a thin film. Its limits: it does not apply to non-magnetic materials such as austenitic stainless steels, the orientation of the defect relative to the field matters, and demagnetisation may be needed after testing.

Eddy current testing (ET)

A coil carrying an alternating current induces eddy currents in the part; a defect alters the way they flow, which the instrument reads as a change in impedance. The method detects surface and near-surface defects, measures coating thickness, sorts material grades and inspects heat-exchanger tubes. It applies to conductive materials, ferrous as well as non-ferrous. With no contact and no couplant, fast and easy to automate, it excels on tubes and surfaces. Its weakness is the limited depth of investigation and a sensitivity to many parameters that calls for expert interpretation.

Ultrasonic testing (UT)

An ultrasonic wave is sent into the part; its reflections off a defect or off the far wall report the position and size of the anomaly. Ultrasonic testing detects internal defects (inclusions, cracks, delaminations), measures thickness and inspects welds, on most metals and many composites. It penetrates the volume, quantifies depth and often needs only one accessible face, with no radiation. Its constraints: a couplant is needed, the surface has to be prepared, the operator has to be qualified, and complex geometries remain awkward. The advanced variants, TOFD and phased array, offer far better imaging and coverage, to the point of replacing radiography on certain welds. Ultrasonic thickness measurement is a special case of this family, dedicated to tracking metal loss.

Radiographic testing (RT)

X-ray or gamma radiation passes through the part and exposes a film or a digital detector; defects show up as contrast on the image. Radiography reveals internal volumetric defects (gas pockets, inclusions, lack of fusion, porosity) and reads particularly well on welds. Applicable to most materials, it produces a permanent, self-explanatory image that can be archived. Its limits weigh heavily in operation: radiation safety imposes an exclusion zone and certified personnel, access to both faces is often required, sensitivity to fine, badly oriented cracks is poor, and the logistics remain costly. Digital radiography now shortens turnaround and does away with silver-based consumables.

Acoustic emission and complementary methods

Acoustic emission (AE) listens to the waves given off by a defect that is evolving under load, for example during a pressure test, to locate the active zones of a large item of equipment such as a tank or a sphere. Infrared thermography reveals disbonds, delaminations in composites and hot spots. Leak testing looks for leaks on pressurised or evacuated circuits. These methods do not replace the ones above but complement them for specific needs.

Which method for which defect: the matching table

The table below sums up what each method is for. Read it first down the defect column, never by cost.

Table scrolls horizontally on small screens.

How to choose in practice

The choice always follows the same reasoning, from the defect to the method and never the other way round.

• Choosing the method out of habit,because radiography has always been done, without going back to the defect actually being sought.
• Believing a surface method sees inside,or the reverse: it is the most common and most costly confusion.
• Overlooking the material,and specifying magnetic particle testing on a non-magnetic austenitic stainless steel where it reveals nothing.
• Forgetting the operator's qualification.A sensitive method poorly applied is less reliable than a simple method well executed.
• Underestimating surface preparation,which directly governs penetrant, magnetic particle and ultrasonic testing.
• Treating the test report as an end-of-campaign formalityinstead of a piece of data to keep and compare from one year to the next.

After the test: put the reports to work, not just produce them

Every NDT campaign produces a deliverable: a test certificate, a radiographic film or image, a thickness map, a penetrant report. Across a fleet of equipment, these documents pile up by the thousand, in mixed formats, held by different contractors, year after year.

Yet the value of a test does not stop at the pass-or-fail verdict. It lies in being able to compare an indication with the one seen three years earlier, to find every time a defect was flagged on a given item, to reconstruct a complete history in front of an auditor. This is where simply filing PDFs shows its limit: the information exists, but it stays impossible to find at the moment you need it.

A documentary intelligence tool for inspection reports such as Integrity Loop takes in these existing NDT reports, extracts the technical data, links them to the equipment and makes them searchable in seconds. It does not carry out the test and does not replace the inspector: it unlocks the data trapped in the reports, so that years of NDT become a genuinely usable technical base. That is the subject of centralising inspection reports and turning PDF reports into usable data.