Key takeaways
- Penetrant testing reveals surface-breaking defects on all metals, whereas magnetic particle testing addresses ferromagnetic materials.
- Ultrasonic and radiographic testing detect volumetric internal defects such as blowholes or lack of penetration.
- Standard SN EN 1090 sets the scope of non-destructive testing according to the project execution class.
- Each NDT method requires a qualified certified inspector to formally validate the inspection report.
Identify the target defect to order the appropriate NDT method
The choice of a non-destructive testing (NDT) method on a structural steel or composite facade weld depends directly on the location of the sought imperfection. Two main families are distinguished: surface examination, which detects surface-breaking microcracks and crater pipes, and volumetric examination, which probes the core of the steel bead in search of slag inclusions, blowholes or lack of penetration.
- Surface testing (PT / MT)
- Non-destructive examination technique restricted to the detection of imperfections breaking the surface of the weld or located in the immediately adjacent sub-surface zone.
For surface testing, two processes dominate the steelwork workshop. Penetrant testing (PT) consists in applying a red or fluorescent liquid penetrant that seeps into open discontinuities, then developing it with a white powder. It works on all metallic materials (carbon steel, stainless steel, aluminium), but requires flawless degreasing cleaning beforehand. Magnetic particle testing (MT) applies exclusively to ferromagnetic materials. By creating a magnetic flux within the component, field leakage due to cracks attracts a black or fluorescent magnetic powder. Faster than penetrant testing, magnetic particle testing also detects sub-surface defects very close to the surface.
To explore the mass of the weld bead, volumetric examination becomes essential, particularly on full-penetration welds subject to fatigue or pure tension.
| NDT method | Target defect type | Compatible materials | Application limits |
|---|---|---|---|
| Penetrant testing (PT) | Surface-breaking cracks, shrinkage cavities and porosities | All metals (steel, stainless steel, aluminium) | Unable to detect internal non-surface-breaking defects |
| Magnetic particle testing (MT) | Surface and sub-surface cracks | Ferromagnetic steels only | Requires an angle grinder to remove thick mill scale |
| Ultrasonic testing (UT) | Lack of fusion, lack of penetration, internal cracks | Heavy-thickness steels (> 8 mm) | Complex interpretation on thin thicknesses and fillet geometries |
| Radiographic testing (RT) | Blowholes, slag inclusions, wormholes | All metals and butt joints | Rigorous management of the safety perimeter and X/gamma radiation |
Ultrasonic testing (UT) transmits a high-frequency acoustic wave that reflects off internal discontinuities. It is the preferred method for testing butt joints in heavy thicknesses and stressed fillet joints. Radiographic testing (RT) provides a photographic or digital image of the metal mass through attenuation of gamma or X-ray radiation. It excels at detecting three-dimensional defects (blowholes, porosities), but requires rigorous safety logistics in the workshop as well as on the construction site to protect erectors against radiation.
Determine the scope of testing according to the execution class of standard SN EN 1090
Within the framework of standard SN EN 1090, the testing requirement is not left to the free discretion of the workshop or the construction site: it is dictated by the execution class (EXC1 to EXC4) specified by the structural engineer in the tender documents. The higher the execution class, the greater the percentage of weld beads that must undergo non-destructive testing supplementary to visual testing (VT).
- Execution class EXC1Applies to secondary, lightly stressed structures. 100% visual testing is required. Supplementary tests (PT, MT, UT, RT) are only required upon specific request from the construction management.
- Execution class EXC2Standard class for building frames and load-bearing metal facades. 100% visual testing, supplemented by spot checks using magnetic particle or penetrant testing (generally 5% to 20% depending on the utilisation rate) and ultrasonic testing on full-penetration joints.
- Execution class EXC3Structures subject to fatigue or large public buildings. The scope of testing by ultrasonic and magnetic particle methods increases sharply (up to 20% to 50% of strategic joint beads).
- Execution class EXC4Exceptional structures with high human or environmental impact. NDT testing rates reach 100% across all structural welded joints.
For site supervision, the foreman or site manager must ensure that the operator performing the tests holds the required certification (ISO 9712 Level 2 minimum for results interpretation). Each test results in an inspection report appended to the as-built documentation and transmitted to the construction management to release the corrosion protection or facade encapsulation phase.
Surface methods: penetrant testing (PT) and magnetic particle testing (MT)
Surface testing aims to detect the presence of surface-breaking discontinuities such as solidification cracks, crater pipes, edge lack of fusion and toe microcracks. In steel construction according to standard SN EN 1090, these methods are systematically applied prior to any volumetric testing in order to verify the geometric and structural quality of the weld envelope.
Penetrant testing (PT) relies on the infiltration of a red or fluorescent liquid penetrant into surface-open imperfections, followed by its extraction using a developer sprayed after rinsing the excess. It applies to all metallic materials (carbon steels, stainless steels, aluminium). In contrast, magnetic particle testing (MT) is based on the deflection of a magnetic field generated by a portable magnet or electromagnet, causing the accumulation of magnetic particles at flux leakage locations. Magnetic particle testing is strictly reserved for ferromagnetic materials, but proves significantly faster and less sensitive to the residual cleanliness of pores than penetrant testing.
| Selection criterion | Penetrant testing (PT) | Magnetic particle testing (MT) |
|---|---|---|
| Permissible materials | All metals (steel, stainless steel, aluminium) | Ferromagnetic materials only |
| Type of defect revealed | Strictly surface-breaking defects | Surface-breaking and near-surface defects (< 2 mm) |
| Required surface condition | Surface perfectly degreased and dried | Tolerates a thin layer of paint or mill scale (< 50 µm) |
| Speed of execution | Mandatory dwell time (10-30 min) | Immediate reading under magnetic field |
Volumetric methods and scope according to execution classes
Internal, non-surface-breaking defects, such as deep blowholes, slag inclusions, root lack of penetration or internal cracks, require probing within the mass. The two primary volumetric methods are ultrasonic testing (UT) and radiographic testing (RT).
Ultrasonic testing (UT) uses a high-frequency acoustic wave emitted by a contact probe. The presence of a discontinuity returns an echo to the transducer. It is the method of choice for testing butt joints exceeding 8 mm in thickness on the construction site and for characterising the depth of lack of fusion. Radiographic testing (RT) (X-rays or gamma rays) projects the image of the internal structure onto film or a digital sensor. It produces indisputable visual archival evidence, particularly suited to thin to medium thickness butt joints, but proves complex to deploy on the construction site due to mandatory radiological safety perimeters.
The scope of non-destructive testing (NDT) to be ordered depends directly on the execution class (EXC) defined by the civil engineer or design office. While in class EXC1 100% visual testing combined with occasional spot checks is generally sufficient, class EXC2 imposes a fixed percentage of supplementary NDT on stressed fillet and butt welds. In classes EXC3 and EXC4 (civil engineering works, structures subject to fatigue), the rate of volumetric (UT/RT) and surface (PT/MT) testing reaches very extensive, or even complete, coverage on high stress concentration areas.
When planning fabrication and erection, the site manager ensures the inclusion of physical access routes and time windows necessary for NDT inspectors before applying any anti-corrosion paint or hot-dip galvanising.
Management of NDT non-conformities and site traceability
When non-destructive testing (NDT) reveals an anomaly in a welded joint of a steel frame or facade load-bearing structure, the immediate halt of the affected operations is required. Defect management requires absolute rigour to ensure structural safety and compliance with the requirements of standard SN EN 1090-2. Any indication exceeding the acceptance limits defined by the applicable standards must result in a written non-conformity report, supported by the precise survey of the certified inspector.
For surface-breaking defects detected by penetrant testing (PT) or magnetic particle testing (MT), such as cracks, lack of penetration or open porosities, repair requires careful gouging or grinding down to sound metal. Complete removal of the discontinuity must be confirmed by renewed surface testing prior to any weld re-execution. In the event of internal anomalies identified volumetrically by ultrasonic testing (UT) or radiographic testing (RT) — such as aligned blowholes, slag inclusions or full-penetration lack of fusion —, the excavation profile must be validated to avoid embrittling the HAZ (heat-affected zone). Weld build-up is carried out strictly according to the qualified welding procedure specification (WPS).
Criminal liability on the construction site remains engaged in the event of a serious breach of recognised trade practices. Art. 229 CP provides for a custodial sentence or a monetary penalty for anyone who breaches the rules of construction and thereby endangers the life of others. Full traceability (NDT reports, weld location plans, welder qualification records and repair reports) constitutes the as-built documentation file indispensable for acceptance.
Frequently asked questions
When should magnetic particle testing be preferred over penetrant testing on the construction site?
Magnetic particle testing (MT) is essential on ferromagnetic steels because it is faster, does not require a long dwell time, and also detects sub-surface defects very close to the surface. Penetrant testing (PT) is reserved for non-magnetic stainless steels, aluminium, or construction sites without direct access to a magnetisation source.
Why use ultrasonic testing rather than radiography to inspect a fillet weld?
Ultrasonic testing (UT) allows acoustic waves to be angled to inspect complex geometries such as fillet or T-joints. Radiographic testing (RT) creates variable thickness superimpositions on these joints, making the interpretation of films or digital images extremely difficult and unreliable, not to mention the constraints of safety cordoning.
How does the scope of testing vary between EXC2 and EXC3?
In execution class EXC2 (standard for buildings), volumetric and surface non-destructive testing is performed via statistical sampling, often between 5% and 20% depending on joint type. In EXC3 (fatigue-stressed structures or complex facades), the scope of testing increases significantly, regularly reaching 100% on butt welds subject to tension.
Who is authorised to carry out and validate NDT examinations on structural steelwork?
Non-destructive testing must be performed by personnel qualified and certified in accordance with standard SN EN ISO 9712 (Level 1 or 2). Only an inspector certified to Level 2 or 3 is authorised to interpret results, sign official inspection reports, and validate the conformity of welded joints.
Texts and standards cited
- SN EN 1090-2Execution of steel structures and aluminium structures - Technical requirements for steel structures
- SN EN ISO 9712Non-destructive testing - Qualification and certification of NDT personnel
- CP art. 229Breach of the rules of the art of building
