GUIDEResources / GuidesUpdated 29 Oct 2026 · 6 min read

NDT methods: non-destructive testing explained (VT, PT, MT, UT, RT)

Deskely Completions team Singapore

Non-destructive testing (NDT) finds defects in welds, castings and components without damaging them. On a capital project it is how every critical weld is proved sound before the line is pressure tested and put into service.

This guide covers the main NDT methods, what each one finds and where it is used, the standards behind them, how NDT technicians are qualified, and how NDT results are recorded against the weld.

Key takeaways
  • Surface methods (VT, PT, MT, ET) find defects at or near the surface; volumetric methods (UT, RT) find them inside.
  • MT works only on ferromagnetic materials; PT works on any non-porous material.
  • The code and specification set which welds get which method, how many, and the acceptance criteria.
  • Every NDT result belongs to a weld on the weld map, and open NDT holds the test pack back.

NDT methods at a glance

MethodFindsMaterialsMain limits
VT Visual testingSurface defects, profile, size, undercutAllSurface only, depends on access and the inspector
PT Penetrant testingSurface-breaking cracks and porosityNon-porous metals and plasticsDefect must be open to the surface; surface must be clean
MT Magnetic particle testingSurface and near-surface cracksFerromagnetic steels onlyNot for austenitic stainless, aluminium or copper
UT Ultrasonic testingInternal flaws, lack of fusion, wall thicknessMost metalsNeeds skilled operators; coarse-grained materials are hard
PAUT Phased array UTAs UT, with imaging and a recorded scanMost metalsEquipment cost, set-up and calibration
TOFD Time of flight diffractionSizing of internal flaws through the wall heightMost metalsDead zones near the surfaces
RT Radiographic testingPorosity, slag, lack of fusion, cracks in line with the beamAllRadiation safety, exclusion zones, planar flaws can be missed
ET Eddy current testingSurface cracks, tube wall loss, coating thicknessConductive materialsShallow penetration
LT Leak testingLeaks through the pressure boundaryAllFinds leaks, not flaws that do not leak

Visual testing (VT)

Visual testing is the first and most used method: every weld is visually inspected before any other NDT. The inspector checks the weld size and profile, undercut, porosity, cracks, spatter and misalignment, with gauges, a light and sometimes a borescope or camera. Weld visual inspection follows ISO 17637 or the code in use, such as AWS D1.1 or ASME B31.3.

Penetrant testing (PT)

A coloured or fluorescent liquid is applied to the clean surface, drawn into any surface-breaking defect by capillary action, the excess is removed, and a developer draws the penetrant back out to show the defect as a visible line or dot. It works on any non-porous material, including stainless steel and aluminium. Standards: ISO 3452 and ASTM E165, also called DPI (dye penetrant inspection) or LPI.

Magnetic particle testing (MT)

The part is magnetised with a yoke or current, and iron particles, dry or in a liquid (often fluorescent), are applied. A crack at or just below the surface disturbs the magnetic field and the particles gather along it. It is faster and more sensitive than PT on carbon and low-alloy steels, but does not work on austenitic stainless steel or non-ferrous metals. Standards: ISO 17638 and ASTM E709, also called MPI.

Ultrasonic testing (UT, PAUT, TOFD)

High-frequency sound is sent into the material from a probe; reflections from flaws, or the back wall, are received and shown on the screen. UT finds internal planar flaws such as lack of fusion and cracks well, and measures wall thickness for corrosion monitoring.

Phased array UT (PAUT) uses a probe with many elements fired in sequence to sweep the beam, producing an image of the weld cross-section and a recorded scan that can be reviewed later. TOFD uses a pair of probes and the diffracted signals from flaw tips to measure flaw height accurately. Both are widely used in place of radiography on thick welds, where they avoid radiation exclusion zones. Standards include ISO 17640 (UT of welds), ISO 13588 (PAUT) and ISO 10863 (TOFD).

Radiographic testing (RT)

X-rays or gamma rays (commonly from iridium-192 or selenium-75 sources) pass through the weld onto film or a digital detector. Denser or thinner areas show as differences in the image: porosity, slag, incomplete penetration and some cracks. The result is a permanent record of the weld.

RT needs a controlled area cleared of people while the source is exposed, which on a busy site is often done at night. Digital radiography (computed radiography and digital detector arrays) has largely replaced film. Standards: ISO 17636 and ASME Section V.

Other methods

  • Eddy current (ET): a coil induces currents in conductive material; flaws change the current. Used for heat exchanger tubes, surface cracks and coating thickness.
  • Leak testing (LT): bubble, pressure decay, helium and other methods prove the pressure boundary is tight. The hydrostatic test is the main strength and leak test for piping.
  • Hardness testing and PMI: not flaw detection, but usually managed alongside NDT. Positive material identification (PMI) confirms that alloy materials are what the certificate says.
  • Acoustic emission (AE): listens for growing defects while a vessel or tank is loaded.

NDT levels and qualification

NDT is only as good as the person doing it, so technicians are qualified by method and level:

SchemeHow it worksLevels
ISO 9712Central certification by an independent certification body, recognised internationally (for example PCN in the UK)Level 1, 2 and 3
ASNT SNT-TC-1AEmployer certification to a written practice, common in the USLevel I, II and III
ANSI/ASNT CP-189A standard, more prescriptive version of employer certificationLevel I, II and III

Level 1 technicians carry out tests to written instructions, Level 2 set up equipment, interpret and evaluate results and write instructions, and Level 3 are responsible for the techniques and procedures. Weld inspectors are qualified separately, for example AWS Certified Welding Inspector (CWI) or CSWIP.

How much NDT, and what is acceptable

The design code and the project specification decide which welds get which method, how many, and when. In ASME B31.3 normal fluid service, for example, at least 5 % of circumferential butt welds are examined by random radiography or ultrasonics, with more for severe cyclic, Category M and high pressure service. Pressure vessel codes, pipeline codes and structural codes have their own rules, and specifications often go further.

Acceptance criteria also come from the code: ASME B31.3 Table 341.3.2 for process piping, ISO 5817 quality levels B, C and D for welds to ISO standards, and AWS D1.1 for structural steel. A weld that fails is repaired and re-examined, and on random sampling a reject usually triggers extra examinations of the same welder's work.

NDT records on a project

Every weld is identified on a weld map (the isometric with weld numbers; see the piping isometric symbols) and recorded in a weld log: weld number, welder, WPS, date, visual result, NDT required, NDT request and report numbers, results and repairs. The welding symbols on the drawing say what weld is required; the weld log proves it was made and examined.

Before a line can be pressure tested, every weld in the test pack must have its NDT done and accepted. Missing or rejected NDT is one of the most common reasons a test pack is not ready.

In Deskely, inspection records, NDT results and punch items are held against the tag and the test pack, so what still stands between a pack and its test is visible on the pack.

The Deskely products for this

Frequently asked questions

What are the main NDT methods?

Visual (VT), penetrant (PT), magnetic particle (MT), ultrasonic (UT, including PAUT and TOFD), radiographic (RT), eddy current (ET) and leak testing (LT).

What is the difference between MT and PT?

Both find surface defects. MT uses magnetism and iron particles and only works on ferromagnetic steels, but also finds defects just below the surface. PT uses a liquid penetrant and works on any non-porous material.

Is UT better than RT?

They find different things. UT is better at planar flaws such as lack of fusion and cracks, and needs no radiation exclusion zone. RT is better at volumetric flaws such as porosity and slag and gives an image of the weld. Many specifications accept either for certain welds.

What are NDT levels?

Levels of qualification for NDT personnel: Level 1 does tests to instructions, Level 2 sets up, interprets and reports, and Level 3 is responsible for techniques and procedures. ISO 9712 and ASNT SNT-TC-1A are the main schemes.

What is PAUT?

Phased array ultrasonic testing: UT with a multi-element probe that steers and focuses the beam electronically, giving an image of the weld and a recorded scan.

How much NDT is required on piping welds?

It depends on the code and service. In ASME B31.3 normal fluid service, at least 5 % of circumferential butt welds are randomly examined; severe cyclic, Category M and high pressure service need more. Project specifications often require more again.

Test packs held back by missing NDT?Deskely holds inspection records, NDT results and punch items against the tag and the test pack, checked as they are signed, so what blocks each pack is visible before the test date.
  • In production since 2014
  • Thousands of daily users
  • On MODEC FPSO projects since 2016

Autonomous engineering runs on Deskely.

Talk to sales