A bridge deck can look sound even when the concrete beneath its surface has delaminated. A dull response to chain dragging may point to a shallow separation, but it cannot show how deep the damage runs or whether corrosion is active. That is why non-destructive testing (NDT) starts with a question, not an instrument. Each method measures a physical response rather than the defect itself; uncertain results need to be checked against other evidence.
Start with the defect and the decision
An inspection may cover reinforced-concrete decks, prestressed girders, steel connections, bearings and foundations. No single test suits them all. Before collecting data, the team needs to identify the suspected damage mechanism, the component’s geometry and the decision the findings must support. Mapping likely delamination for targeted repair is a different task from estimating section loss in steel or investigating an inaccessible foundation.
The survey plan should record accessible faces, expected material layers, reinforcement density, surface condition and known repairs. As-built drawings are useful, but construction changes and later interventions can limit their accuracy. The team should also decide what size of defect matters: a method that maps a broad deteriorated zone may miss a narrow crack at a connection.
Concrete decks: faster coverage, better interpretation
Ground-penetrating radar and electromagnetic methods
Ground-penetrating radar (GPR) sends electromagnetic pulses into a deck and records reflections where electrical properties change. Vehicle-mounted systems can gather closely spaced measurements across traffic lanes with less disruption than a grid of stationary readings. Reflection patterns help interpreters estimate deck thickness, locate reinforcement and flag unusually high signal attenuation.
Attenuation does not directly measure corrosion. Moisture, chloride-bearing material, asphalt overlays, reinforcement spacing and surface condition can all affect the signal, leaving uncertain boundaries on an otherwise precise-looking map. When comparing surveys across years, the team needs to account for survey speed, antenna position and processing choices.
Electrical resistivity and half-cell potential testing address other parts of the corrosion question. Resistivity indicates how readily current passes through concrete; half-cell readings indicate the likelihood of active corrosion under suitable test conditions. Both depend on moisture and require appropriate electrical contact. Neither measures remaining bar diameter on its own. Alongside GPR, they can help distinguish a geometric anomaly from an area that warrants closer corrosion investigation.
Thermal imaging and acoustic response
Infrared thermography records surface temperature differences. Shallow delamination can change heat flow as a deck heats or cools, allowing rapid screening over a large area. Solar exposure, wind, surface moisture, overlays and time of day also affect the signal. Log those conditions before interpreting an anomaly; do not treat its outline as an automatic repair boundary.
Impact-echo, impulse-response and other acoustic tests use waves generated by a local impact to investigate suspected voids, thickness changes and delamination. Reinforcement and complex geometry can make the response harder to interpret. Automated tapping or instrumented impact systems improve the consistency of measurement locations, but the suspected flaw still needs verification.
Steel bridges and difficult connections
Ultrasonic testing remains important for locating internal discontinuities in steel and estimating thickness where corrosion has reduced a section. Phased-array ultrasonic testing steers and focuses sound beams electronically, offering a more informative view of some weld regions than a single conventional probe position. Results still depend on coupling, surface preparation, access and an accurate understanding of weld geometry. A complex signal is not necessarily a crack.
Magnetic particle testing can reveal surface and near-surface discontinuities in suitable ferromagnetic steel, particularly at fatigue-prone details. Eddy-current techniques can screen for near-surface cracking and, in some configurations, work through thin coatings. Both have narrower roles than a structural assessment: their indications still need sizing, classification and evaluation in relation to the member’s load path.
Guided-wave ultrasonics can screen lengths of members or components from fewer access points. Reflections point to places worth investigating, but attachments, section changes and supports also reflect waves. Broad screening coverage does not prove that every small defect has been excluded.
Where newer tools change the inspection workflow
Robotic platforms, drones and high-resolution imaging extend access to deck undersides, towers and other difficult surfaces. Often, their main value is safer, more repeatable data collection. Photogrammetry can place visible cracking or spalling in a spatial model, while computer vision can sort large image sets for review. Lighting, scale, occlusion and surface staining can cause false classifications. A detected line still needs engineering interpretation.
Combined datasets are most useful when each layer keeps its measurement meaning. A thermal anomaly, a GPR attenuation zone and a visible crack may appear on the same deck map, but their overlap does not establish a common cause. Survey dates, positioning accuracy and spatial resolution need to remain clear; neat alignment can otherwise suggest more certainty than the data support.
Turn indications into defensible findings
A staged investigation helps limit both missed damage and unnecessary invasive work:
- Screen: choose methods suited to broad coverage and record environmental and operating conditions.
- Target: examine anomalies and apparently sound comparison areas with a more specific method.
- Verify: where warranted, check the interpretation against a core, exposed detail or other direct observation.
- Report: distinguish measured values from interpreted defect boundaries and state what remains uncertain.
Repeat surveys need stable reference points, comparable settings and retained raw data. Without them, an apparent change may come from equipment setup or moisture rather than deterioration. That matters when findings feed maintenance forecasts: predictive analytics for transport infrastructure monitoring relies on measurements that can be compared over time, not just convincing maps.
Suppose a deck shows a radar attenuation zone but no clear thermal anomaly. Neither a clean bill of health nor an immediate removal boundary follows from that result. Record the survey conditions, compare reinforcement reflections with adjacent areas, and make a small number of targeted checks—including one outside the anomaly—before defining the repair extent.
