Non-destructive testing (NDT) checks welds, castings, structures and equipment for flaws without damaging them. The common methods are visual, liquid penetrant, magnetic particle, ultrasonic and radiographic testing. Each finds different kinds of flaws in different places, so the right choice depends on the material, the joint, the likely defect and what the governing code requires.
Surface methods
These methods find flaws that are on the surface or just below it. The main thing that separates NDT methods is how deep into the part they can see, which the diagram below compares.
Visual testing (VT)
The most basic and most used method. A trained inspector checks the surface with good lighting, gauges and magnifiers. It finds surface cracks, porosity, undercut and profile problems, and confirms weld size. It cannot see inside the part. Our guide to visual weld inspection covers it in detail.
Liquid penetrant testing (PT)
A coloured or fluorescent liquid is applied to a clean surface and drawn into any surface-breaking crack or pore. After the excess is removed, a developer draws the penetrant back out, showing the flaw as a visible mark. PT works on most non-porous materials, including stainless steel and aluminum. It only finds flaws open to the surface, and the surface must be thoroughly cleaned before and after.
Magnetic particle testing (MT)
The part is magnetized and fine iron particles are applied. A crack at or just below the surface disturbs the magnetic field and collects the particles, making it visible. MT is fast and sensitive for surface cracks in ferromagnetic materials such as carbon steel. It does not work on austenitic stainless steel or aluminum, and it finds flaws best when they lie across the direction of the magnetic field.
Volumetric methods
These methods look inside the material.
Ultrasonic testing (UT)
High-frequency sound is sent into the material through a probe. Flaws reflect the sound back, and the instrument shows their position and size. UT is good at finding planar flaws such as cracks and lack of fusion, and it can measure thickness to check for corrosion. Phased array ultrasonic testing (PAUT) uses many elements to sweep the beam and build an image, and is increasingly used for weld inspection. UT needs a skilled technician and reference standards, and results depend heavily on technique.
Radiographic testing (RT)
X-rays or gamma rays pass through the part onto film or a digital detector. Denser or thicker areas let less radiation through, so voids show up as darker areas. RT is good at finding volumetric flaws such as porosity and slag and gives a permanent record. It is less reliable for cracks that do not line up with the beam. Because it uses ionizing radiation, the area must be controlled and the work done by certified operators following federal and provincial radiation safety rules.
Radiography means a controlled area. When RT is scheduled, the site needs a plan for barricades, timing and who must stay clear. Nobody should enter the controlled area until the radiography crew says it is safe.
Comparing the methods
| Method | Finds best | Main limits |
|---|---|---|
| Visual (VT) | Surface defects, weld size and profile | Surface only |
| Penetrant (PT) | Fine surface-breaking cracks, any non-porous metal | Must be open to surface; surface prep critical |
| Magnetic particle (MT) | Surface and near-surface cracks in ferromagnetic steel | Not for aluminum or austenitic stainless |
| Ultrasonic (UT/PAUT) | Internal planar flaws, thickness | Operator skill, geometry, coarse-grained materials |
| Radiographic (RT) | Internal porosity and inclusions, permanent record | Radiation safety, crack orientation, access to both sides |
Other methods you may hear about
- Eddy current testing uses electromagnetic induction to find surface flaws and measure coating thickness on conductive materials, and is common for tubing.
- Leak testing checks that vessels and piping hold pressure or vacuum.
- Hardness testing can check that welding or heat treatment did not leave the material too hard.
- Ground penetrating radar and concrete X-ray are used on concrete rather than metal, mainly to locate reinforcement and services. See concrete X-ray vs GPR scanning.
Choosing a method
The governing code or project specification usually decides which methods apply and how much of the work is tested. Where there is a choice, consider:
- What defects are most likely for this material, process and joint.
- Whether you need to find surface flaws, internal flaws or both.
- Material type and thickness.
- Access: some methods need both sides of the part.
- Site constraints, such as radiation safety for RT.
- Whether a permanent record is needed.
A typical case: testing a repaired weld
Consider a steel beam-to-column connection with a full-penetration groove weld. Visual inspection found a crack at one end of the weld. The fabricator gouges out the crack, and before rewelding the excavation is checked with magnetic particle testing to confirm the crack is completely gone, since a surface method is quick and sensitive at that stage. The joint is then rewelded to an approved procedure. After the specified waiting time, the finished repair is inspected visually, then tested with ultrasonic testing through its full thickness to confirm there is no lack of fusion or a new crack inside. Each step uses the method that suits the question being asked at that moment, rather than one method for everything.
Planning NDT on a project
- Read the specification. List which welds or components need which methods, and what percentage of each is tested.
- Agree on timing. Some tests must wait a set time after welding, and some must be done before coatings or insulation go on.
- Plan access and surface preparation. UT needs a surface the probe can couple to, MT and PT need clean surfaces, and RT may need access to both sides and a cleared area.
- Confirm technician certification. Check that the level held matches what the specification requires for performing and interpreting each method.
- Track results and repairs. Rejected welds are repaired and retested. Where testing is done on a sample, many specifications require extra testing when rejects are found.
What an NDT report should include
- The component or weld identification and its location.
- The method, technique and equipment used, including calibration references.
- The code or specification and acceptance criteria applied.
- Each indication found, its size and position, and whether it was accepted or rejected.
- Any areas that could not be tested and why.
- The technician's name, certification level and the date.
NDT technicians in Canada are commonly certified to CAN/CGSB-48.9712 (ISO 9712) through Natural Resources Canada, at levels that reflect their training and experience. Specifications usually state the level required to perform and interpret each test.
How ehZee can help
ehZee carries out independent welding inspections and can help you work out which testing a job needs, coordinate qualified NDT technicians, and review results against the specification. See our welding inspections service for details.
Common questions
Is UT or RT better for welds?
Neither is better in every case. UT is generally stronger at finding cracks and lack of fusion, while RT is good at showing porosity and inclusions. Codes often allow either, and some jobs use both.
Does every weld need NDT?
No. The code and specification set how much testing is required, which may be a percentage of welds or full testing of critical joints. Visual inspection is normally applied to all welds.
Why does a report list areas that were not tested?
Geometry, coatings or access can stop a method from covering part of a weld. Those gaps should be stated so the owner knows what was and was not examined and can decide whether another method is needed.