1. Why Rails Need Internal Inspection
A rail carries a train on its surface, but its safety depends on what is inside. During manufacture, rolling and cooling can leave pores, inclusions, shrinkage and micro-cracks in the steel; during service, fatigue growth can turn a tiny internal flaw into a transverse fracture, and transverse fractures are the rail defect class that most often leads to rail breaks. Surface inspection cannot see these internal features, which is why ultrasonic testing is the backbone of rail quality control in the mill, at the construction site and on the track.
Ultrasonic testing is a non-destructive method based on mechanical waves above 20 kHz. The waves travel through the rail steel with good directivity; when they meet an interface with different acoustic impedance, such as a crack, pore or inclusion, part of the energy is reflected. The probe emits pulses, receives the echoes, and the instrument displays them as an A-scan signal or as a B-scan/C-scan image. The position of the echo gives the depth of the defect, the amplitude gives its size class, and the echo shape helps the operator judge the nature of the reflector. The general principles of the method follow ISO 16810, and testing of rails in track follows EN 16729-1 where the railway application specification applies.
2. Defect Classes and Test Configuration
| Defect class | Typical location | Detection set-up |
|---|---|---|
| Pores, blowholes, inclusions | Anywhere in the head, web or foot (rolling defects) | Normal (0 degree) probes, full section scan at the mill |
| Head checks / head surface cracks | Gauge corner of the head | Angled probes (typically 37-70 degrees), in-track testing |
| Bolt-hole cracks | Around fishplate bolt holes | Angled probes focused on the hole area, joint testing |
| Web fractures | Web centre, often near welds or joints | 0 degree and angled probes, in-track testing |
| Weld defects | Flash butt or aluminothermic weld zone | Full weld scan with angled and 0 degree probes after welding |
Probe frequency is chosen by resolution versus penetration: 2-4 MHz probes are common for full section scans where penetration is needed, while higher frequencies give better resolution of small near-surface defects. The actual set-up, including probe angles, frequencies and acceptance levels, is defined by the project specification, EN 16729-1 for in-track testing, and the rail manufacturer's mill testing procedure for new rails.
3. Role in Production Quality Control
At the mill, rails are scanned over their full length and full cross-section after final rolling and cooling. The scan detects rolling defects such as pores, shrinkage and inclusions that no amount of surface inspection would reveal. Defects above the acceptance level are marked, cut out and the rail section downgraded or scrapped; smaller indications are recorded in the rail's traceability record. The ultrasonic scan is part of the release test together with dimensional, hardness and tensile checks, and the result follows the rail to the site through the heat number and rail number traceability system. This is the first line of defence: defects stopped at the mill never reach a track.
4. Role in Construction and Welded Joint Acceptance
After laying, rails are damaged by transport, handling, lifting and welding. The welded joint is the most critical zone: flash butt and aluminothermic welds can contain incomplete fusion, pores, slag and cracks that are invisible from the outside. Ultrasonic testing of every welded joint after welding, usually the same day or within the acceptance window, is the standard acceptance step. The scan covers the weld zone and the heat affected zone with angled and normal probes; indications above the acceptance level require the weld to be cut out and re-welded. Full-length scanning of newly laid rail catches handling damage and confirms that the rail delivered is the rail certified at the mill.
5. Role in Service Monitoring and Rail Break Prevention
In service, the rail is struck by every wheel and subjected to bending, temperature stress and, on heavy-haul lines, very high axle loads. Internal defects grow: a small head check or web flaw can become a transverse fracture within a few million gross tonnes. Regular ultrasonic testing compares the echo pattern of each rail length over time; a growing indication is re-tested at a shorter interval, and rails with indications above the limit are replaced before they break. On heavy-haul corridors, testing is scheduled by tonnage rather than by calendar: for example a patrol every few million gross tonnes with walking-speed trolleys, complemented by self-propelled ultrasonic vehicles that scan both rails at higher speed. The inspection record is part of the track asset database and feeds the rail replacement planning cycle.
6. Common Misconceptions
"If the rail surface looks good, the rail is sound." False. Fatigue cracks and manufacturing defects grow internally; the surface appearance is not a measure of internal integrity, which is why ultrasonic testing is mandatory for rail acceptance and monitoring.
"Ultrasonic testing finds every defect." No method finds every defect. The sensitivity depends on defect orientation, position, probe selection and operator skill; indications below the acceptance level are allowed by design, and the limits are set by the specification.
"A welded joint that passed ultrasonic testing never fails." False. The test confirms the joint state at the time of testing; fatigue, corrosion and poor maintenance can create new defects later, so in-service re-testing of welds remains necessary.
"Testing frequency can be set by the calendar alone." False. On heavy-haul lines, accumulated tonnage is the correct driver of testing intervals; a calendar-based schedule either over-tests quiet lines or under-tests busy ones.
"Higher probe frequency is always better." False. Higher frequency improves resolution but reduces penetration and is more affected by surface condition; the frequency is chosen for the defect class and depth range of the inspection task.
FAQ
Q1: What does ultrasonic testing detect inside a rail?
Internal reflectors such as pores, inclusions, shrinkage, head checks, web fractures, bolt-hole cracks and weld defects. The echo position gives the depth, the amplitude the size class, and the shape helps identify the nature of the reflector.
Q2: Which standards govern ultrasonic rail testing?
General ultrasonic testing principles follow ISO 16810; in-track rail inspection requirements follow EN 16729-1 where applicable. Mill testing and weld acceptance follow the relevant product and construction specifications of the railway authority.
Q3: Why are welded rail joints tested immediately after welding?
Because weld defects such as incomplete fusion, pores and cracks form during the weld process and are invisible from the surface. Testing within the acceptance window catches defective welds while they can still be cut out and re-welded economically.
Q4: How often are rails tested on a heavy-haul line?
Testing is scheduled by accumulated tonnage, typically every few million gross tonnes, with walking-speed trolleys for detailed sections and higher-speed ultrasonic vehicles for long stretches; the interval is shortened where indications are found or track conditions worsen.
Q5: Can ultrasonic testing prevent rail breaks?
It cannot stop crack growth, but it detects growing indications in time to schedule replacement before the crack reaches the critical size for a transverse fracture. Regular testing with documented trend data is the core of rail break prevention on main lines.
Q6: Who performs ultrasonic testing and what qualification is needed?
Testing is performed by certified NDT personnel (typically qualified to ISO 9712 level 2 or equivalent for ultrasonic testing) using calibrated equipment; the operator records the indications and the track authority decides the response, from re-testing to immediate replacement.