Practical Guide to Non Destructive Weld Testing

A sound weld can still contain a defect that is not visible from the surface. Lack of fusion at a root, internal slag inclusion or cracking beneath a weld cap may not show up until the joint is loaded, pressurised or put into service. This guide to non-destructive weld testing explains how fabrication teams can inspect weld quality without cutting up the component, and how to select a method that matches the material, joint design and level of risk.

Non-destructive testing, usually shortened to NDT, is not a substitute for qualified welding procedure control or competent workmanship. It is a verification tool. Used properly, it confirms whether a completed weld meets the acceptance criteria set by the drawing, contract, weld procedure specification or relevant standard.

What non-destructive weld testing can detect

Different methods reveal different types of discontinuity. Surface-breaking cracks, undercut, overlap and porosity may be found during visual inspection or penetrant testing. Near-surface cracking in carbon steel is well suited to magnetic particle inspection. Internal planar defects such as lack of fusion, incomplete penetration and cracking are often assessed with ultrasonic testing, while radiography is particularly useful for volumetric defects including porosity and slag inclusions.

A discontinuity is not automatically a rejectable defect. Weld standards define limits for size, location, quantity and orientation. A small isolated pore may be acceptable in one structural weld yet unacceptable in a pressure-retaining joint. Inspection therefore starts with the required quality level, not with the test equipment available in the workshop.

Start with visual testing

Visual testing is the first stage of every sensible inspection plan. It is fast, economical and capable of finding many issues before further testing is required. A clean weld, good lighting and access to both sides of the joint are essential.

The inspector checks weld profile, leg length, throat size where applicable, reinforcement, root condition, stop-start areas and overall alignment. They also look for visible cracks, spatter where it matters, overlap, arc strikes, undercut, incomplete fill and excessive convexity or concavity. Weld gauges, a torch, mirror, magnification and accurate measuring tools are often all that is required.

Visual inspection should begin before welding. Fit-up, root gap, joint preparation, tack weld quality and cleanliness have a direct effect on the finished weld. Catching a poor bevel or contaminated root face before welding is cheaper than identifying incomplete fusion after the assembly has been painted or installed.

Dye penetrant testing for surface-breaking defects

Dye penetrant testing, also called liquid penetrant testing, identifies fine cracks and other defects open to the surface. The process uses a low-viscosity penetrant that is drawn into a discontinuity by capillary action. After excess penetrant is removed, a developer draws the trapped material back to the surface to form a visible indication.

It works on most non-porous materials, including stainless steel, aluminium and many non-magnetic alloys. That makes it useful where magnetic particle testing cannot be used. It is particularly effective for finding crater cracks, grinding cracks and fatigue cracking around weld toes.

Surface condition is the main limitation. Paint, oil, moisture, rust, scale and weld spatter can mask defects or create false indications. The weld must be thoroughly cleaned before testing, and the specified dwell and development times must be followed. Penetrant testing only finds defects that break the surface. It cannot assess internal fusion or root penetration.

Magnetic particle testing for ferromagnetic steels

Magnetic particle testing is a practical option for carbon steel and other ferromagnetic materials. The test area is magnetised, then fine magnetic particles are applied. Where a surface or near-surface discontinuity disrupts the magnetic field, particles gather and form an indication.

This method is generally more sensitive than visual inspection for tight surface cracks and can reveal defects just below the surface. It is commonly used around weld toes, attachment welds, repaired areas and highly stressed details. Dry powder, wet visible ink and fluorescent systems are available, with the method chosen around site conditions, required sensitivity and available lighting.

It cannot be used on aluminium, austenitic stainless steel or other non-magnetic materials. It also requires correct field direction. A crack running parallel to the magnetic field may not produce a clear indication, so multi-directional magnetisation is often needed. Components may require demagnetisation afterwards, particularly where residual magnetism could affect machining, instrumentation or service performance.

Ultrasonic testing for internal weld quality

Ultrasonic testing sends high-frequency sound waves into the weld and base material. Reflections from interfaces or discontinuities are displayed and interpreted by a trained operator. For butt welds in plate, pipe and structural sections, it is one of the most effective methods for locating and sizing internal planar defects.

Ultrasonic testing is portable, provides immediate results and avoids the radiation controls associated with radiography. It is well suited to thicker sections and can inspect from one side where access is limited. Modern phased array systems can produce detailed scan data, but conventional ultrasonic testing remains widely used when the joint and acceptance standard allow it.

The trade-off is operator skill and procedure control. Weld geometry, material thickness, grain structure, surface condition and probe angle all influence results. Thin material, complex weld profiles, coarse-grained material and restricted access can make interpretation more difficult. The test procedure must define calibration blocks, scanning coverage, recording thresholds and acceptance criteria rather than relying on a general scan.

Radiographic testing and its limits

Radiographic testing uses X-rays or gamma radiation to create an image of internal weld condition. It is valued for producing a permanent record and for showing volumetric defects clearly. Porosity, wormholes, slag inclusions and some incomplete penetration indications can be readily identified where image quality is sufficient.

Radiography is often specified for pipework, pressure work and critical butt welds, but it is not automatically the best method. It can be less sensitive than ultrasonic testing for planar defects aligned unfavourably to the radiation beam, including certain lack-of-fusion defects and cracks. Access is normally required to both sides of the weld, and exposure zones can interrupt nearby work.

Radiation safety arrangements, competent personnel and controlled exclusion areas are non-negotiable. In a busy fabrication shop or live maintenance environment, these controls can affect programme planning. Digital radiography can improve image handling and reduce processing time, but it does not remove the need for disciplined radiation management.

Choosing the right test method

The correct method depends on what must be found and what the component will do in service. Material type matters first. Magnetic particle inspection is restricted to ferromagnetic materials, while penetrant testing is suitable for non-porous non-magnetic alloys. Joint type and thickness then determine whether ultrasonic or radiographic testing is practical.

Service conditions should drive the inspection level. A handrail bracket and a cyclically loaded lifting attachment should not be treated the same way. Fatigue-sensitive details require close attention to weld toe condition and surface cracking. Pressure systems and safety-critical structures may need a combination of visual examination, surface testing and volumetric inspection.

Access, production timing and repairability also matter. A weld that can be tested before final assembly is usually easier and less costly to inspect. If the job requires coating, galvanising or site installation, arrange inspection hold points before those stages. Testing after paint removal or after a component has been placed in a congested structure adds avoidable delay.

Control the inspection process, not just the result

Reliable NDT begins with clear documentation. The drawing or inspection and test plan should identify which welds are to be examined, the test method, inspection extent, acceptance standard and timing. Weld identification and traceability are essential where multiple welders, shifts or batches are involved.

Before testing, confirm that the weld has cooled as required, the surface is accessible and any temporary attachments or slag have been removed. For ultrasonic testing, ensure sufficient smooth contact area for probe movement. For penetrant and magnetic particle testing, do not grind aggressively simply to make the weld look neat. Grinding can remove evidence of a crack or alter the weld profile being assessed.

Where an indication is found, evaluate it against the specified acceptance level before deciding on repair. Repairs should be controlled in the same way as original welding: remove the defect fully, prepare the area, reweld to an approved procedure and re-inspect. Repeated local repairs can introduce distortion, hard zones or new cracking if they are not managed carefully.

Competence and acceptance criteria matter

The most expensive error is treating an NDT report as a simple pass or fail certificate. The result is only meaningful when the technician is competent for the method used, the procedure is suitable for the joint and the acceptance criteria are correctly applied. A good report records the component reference, weld location, method, equipment settings where relevant, coverage, indications and final disposition.

For fabrication managers, the practical objective is straightforward: specify the inspection needed for the risk, build it into the production sequence and provide clean, accessible welds for examination. That approach reduces rework, prevents late-stage disputes and gives the finished fabrication a defensible quality record. The strongest inspection result is one that confirms sound welding without slowing the job unnecessarily.