What is Ultrasonic Testing (UT)? Detecting Invisible Rail Defects

A technical overview of Ultrasonic Testing (UT) in railway maintenance. Learn how high-frequency sound waves detect invisible internal rail defects like kidney fractures and bolt hole cracks, ensuring the structural integrity of the track network.

November 27, 2025 6:14 am | Last Update: March 20, 2026 10:12 am

⚡ In Brief

Ultrasonic Testing (UT) is the primary method for detecting internal rail defects — cracks and flaws invisible to the eye — by transmitting high-frequency sound waves through steel and analysing the echoes that bounce back from discontinuities.
A rail break caused by an undetected internal defect can derail a train travelling at full speed; UT is the only reliable technology for finding these defects before they propagate to failure.
Modern automated ultrasonic test trains operate at 40–80 km/h and can inspect 300–500 km of track per day using arrays of 12–24 probes per rail, covering the full rail cross-section at multiple angles simultaneously.
The three most dangerous rail defects — tache ovale (kidney fracture), transverse defects, and detail fractures — are all internal and entirely invisible without UT inspection.
Phased Array Ultrasonic Testing (PAUT) — which steers and focuses ultrasonic beams electronically without moving the probe — is replacing conventional single-crystal UT as the standard for weld and complex defect inspection, improving detection probability while reducing inspection time.

On 17 October 2000, a high-speed GNER train derailed at Hatfield, North Yorkshire, killing 4 people and injuring 70. The cause was a catastrophic rail fracture caused by gauge corner cracking — a form of rolling contact fatigue that had progressed undetected to a critical depth. Post-accident investigation found that the failed rail section contained approximately 300 individual cracks, and that the infrastructure manager had known about the deterioration for months but had not removed the rail from service.

Hatfield was not a failure of ultrasonic testing technology — it was a failure of maintenance management. But it permanently changed how the railway industry approached rail defect detection. Network Rail’s annual ultrasonic testing mileage increased dramatically in the years that followed, and the development of more sensitive automated inspection systems accelerated. Understanding ultrasonic testing is understanding the primary technical defence against one of the most catastrophic failure modes in railway infrastructure.

What Is Ultrasonic Testing?

Ultrasonic Testing (UT) is a Non-Destructive Testing (NDT) method that uses high-frequency sound waves — typically in the range of 0.5 MHz to 10 MHz, well above the audible range — to detect internal discontinuities in solid materials. The fundamental principle is that sound waves travel at a known velocity through a uniform material, but are reflected (echoed) when they encounter a boundary between materials with different acoustic properties — such as the boundary between steel and an air-filled crack.

By measuring the time taken for a transmitted pulse to return as an echo, the distance from the probe to the reflecting discontinuity can be calculated precisely. By analysing the amplitude of the echo (how strongly it reflects), engineers can estimate the size and orientation of the defect. By using multiple probes at different angles, the full three-dimensional extent of a defect can be characterised.

The Pulse-Echo Principle: How UT Finds Defects

Step	What Happens	Engineering Detail
1. Transmission	Piezoelectric transducer converts electrical pulse to ultrasonic wave	Frequency 2–5 MHz for rail inspection; higher frequency = better resolution but less penetration
2. Coupling	Couplant (water or oil) fills gap between probe and rail surface	Air gaps absorb the wave — couplant ensures acoustic transmission into the rail
3. Propagation	Sound wave travels through rail steel at ~5,900 m/s	Velocity is constant in homogeneous steel — enables precise depth calculation
4. Reflection	Wave reflects from defect or rail bottom (backwall echo)	Amplitude of echo indicates defect size; total reflection from base confirms calibration
5. Reception	Same or separate transducer receives returning wave	Pulse-echo (one transducer) or pitch-catch (separate transmitter/receiver)
6. Analysis	Time of flight → depth; amplitude → severity; pattern → defect type	A-scan (single point), B-scan (cross-section), C-scan (plan view) displays used

Critical Rail Defects Detected by UT

Defect	Location in Rail	Origin	Risk Level	Detection Method
Tache ovale (kidney fracture)	Rail head interior	Hydrogen flakes from manufacturing; pre-existing from production	Very high — can cause sudden transverse break	70° angled probe; characteristic oval echo pattern
Transverse defect (TD)	Rail head, perpendicular to rail axis	Fatigue crack growing transversely from squat or surface defect	Very high — direct path to rail break	0° vertical probe + 70° angled probe
Detail fracture	Rail head, growing from surface defect	Rolling contact fatigue progressing from shelling or head check	High	Angled probes; may be missed if surface defect shields signal
Bolt hole crack	Rail web, around fishplate bolt holes	Fatigue from stress concentration at bolt holes in jointed track	Medium-high	Web probes angled toward bolt hole zone
Squat	Rail head surface, growing downward	Rolling contact fatigue; begins as surface depression	Medium (low until deep stage)	UT detects sub-surface growth; surface visible by camera
Weld defect	Thermite or flash-butt weld zone	Porosity, inclusions, or lack of fusion during welding	Variable — depends on defect size and location	Multiple angled probes; PAUT for complex weld geometry
Foot crack	Rail foot (base)	Corrosion-fatigue under fastening clip contact zone	Medium — can progress rapidly in aggressive environments	Foot probe array; difficult to detect due to fastening obstruction

Manual vs Automated UT Inspection: Comparison

Method	Equipment	Speed	Coverage per Day	Best Application
Handheld (walking stick)	Single probe pushed by operator on foot	Walking pace (~4 km/h)	~3–8 km	Verification of flagged defects; switch areas; short inaccessible sections
Hi-rail push trolley	Multi-probe trolley pushed on track by hi-rail vehicle	5–15 km/h	20–60 km	Lightly trafficked lines; focused inspection campaigns
Ultrasonic test train (USTT)	Dedicated vehicle; 12–24+ probes per rail in water-filled shoe	40–80 km/h	300–500 km	Network-wide mainline inspection; primary high-volume method
Combined inspection vehicle	UT + LiDAR + camera + geometry measurement on one vehicle	60–120 km/h	500–800 km	High-efficiency full infrastructure health profile in single pass

Probe Angles: Why Multiple Angles Are Required

No single probe angle detects all defect types. Different defects are oriented differently within the rail cross-section, and a sound wave only reflects strongly from a defect surface that is perpendicular (or close to perpendicular) to the incoming beam. A probe array on a test train therefore uses multiple simultaneous probes at different angles:

0° (vertical): Detects horizontal laminations and delaminations, foot cracks, and some head defects. The backwall echo (reflection from the rail bottom) confirms the probe is coupled and calibrated.
70° forward and backward: The primary probe angle for detecting transverse defects and tache ovale in the rail head. The angled beam intersects the typical orientation of these dangerous cracks.
45°: Detects bolt hole cracks and web defects at intermediate angles.
Head-to-web transition probes: Specific probe positions targeting the complex geometry at the rail head-to-web junction, where stress concentrations make defects common.

A full-coverage inspection requires a minimum of 6–8 independent probe channels per rail — the reason modern test trains carry water-filled probe shoes containing probe arrays that maintain consistent acoustic coupling at inspection speeds.

UT vs Eddy Current Testing: Complementary Technologies

Parameter	Ultrasonic Testing (UT)	Eddy Current Testing (ECT)
Detection depth	Full rail cross-section — deep internal defects	Surface and near-surface only (0–5 mm)
Best for	Transverse defects, tache ovale, bolt hole cracks, weld flaws	Head checks, gauge corner cracking, surface shelling
Physics	Sound wave reflection from acoustic impedance boundaries	Electromagnetic induction — surface cracks disrupt eddy current flow
Couplant required	Yes — water or gel between probe and rail	No — electromagnetic; lift-off tolerant
Effect of surface contamination	Reduced sensitivity if rail surface not clean	Less affected by surface contamination
Standard use	Primary internal defect detection on all networks	Supplementary surface crack characterisation; head check quantification

Phased Array Ultrasonic Testing (PAUT): The Next Generation

Conventional UT uses individual probes at fixed angles. Phased Array Ultrasonic Testing (PAUT) uses an array of small transducer elements that can be electronically timed to steer and focus the ultrasonic beam at multiple angles simultaneously — without mechanically moving the probe. The beam can be swept through a range of angles, producing a cross-sectional image of the rail (S-scan or sector scan) that shows the entire internal structure in a single pass.

PAUT advantages for railway inspection include:

Single probe replaces multiple: A PAUT probe can cover the angular range previously requiring 4–6 separate conventional probes, reducing probe array complexity.
Better defect characterisation: The sector scan image allows operators to see the shape and orientation of a defect in cross-section, improving sizing accuracy and reducing false calls.
Improved weld inspection: Weld geometry is complex and variable. PAUT’s ability to steer the beam to the optimal angle for each weld zone improves detection probability significantly.
Digital data storage: PAUT produces volumetric scan data that can be stored and compared with future inspections to track defect growth.

Inspection Frequency: How Often Must Rails Be Tested?

Inspection frequency depends on traffic density (measured in Million Gross Tonnes, MGT, per year), rail condition, and regulatory requirements. Higher traffic accelerates fatigue crack growth, requiring more frequent inspection:

Track Category	Typical Annual Traffic	Typical UT Frequency	Notes
High-speed line (200+ km/h)	10–50 MGT	Every 3–6 months	High speed increases consequence of defect; frequent inspection
Busy mainline (mixed traffic)	30–100 MGT	Every 3–6 months	High tonnage accelerates fatigue growth
Secondary mainline	5–30 MGT	Every 6–12 months	Annual cycle standard
Light branch line	<5 MGT	Annually or biennial	Low fatigue loading; longer interval acceptable

Editor’s Analysis

Ultrasonic testing is mature, proven, and irreplaceable — but the railway industry’s challenge is not whether to use it but how to use it more effectively. Two converging pressures are driving change. First, the growth of traffic on existing infrastructure increases the rate of fatigue crack growth and shortens the safe interval between inspections. On the busiest European mixed-traffic mainlines, the question is no longer “how often should we test” but “is annual testing frequent enough.” Continuous monitoring — using sensors permanently attached to the rail that monitor acoustic emissions as cracks propagate — is an active research area that could provide real-time warning of accelerating defect growth between inspection cycles. Second, the data revolution is transforming how inspection results are used. Traditionally, a test train pass produced a paper record of flagged locations that maintenance teams then investigated by hand. Modern digital systems produce a continuous record of the entire inspection, geotagged to millimetre accuracy, that feeds directly into asset management systems. Defect records from successive inspections are compared automatically — tracking crack growth, predicting time to critical size, and generating targeted intervention orders. The combination of better sensors, better data management, and predictive analytics is moving UT from a periodic check to a continuous risk management tool. The Hatfield lesson — that knowing about a defect is not the same as acting on it — is now built into the data systems that process inspection results. — Railway News Editorial

Frequently Asked Questions

Q: Why can’t visual inspection replace ultrasonic testing for rail defects?: The most dangerous rail defects — tache ovale, transverse defects, detail fractures — are entirely internal and have no visible surface indication until they are very close to critical size, at which point the rail may break with very little warning. A rail can look perfectly normal on the surface while containing a crack that occupies 50% of its cross-section internally. Visual inspection is important for detecting surface defects (squats, shelling, head checks, corrosion) but cannot detect sub-surface fatigue cracks. Ultrasonic testing is the only reliable method for detecting internal defects at early stage, when there is sufficient time to take the rail out of service before it breaks.
Q: What happens when a rail defect is found during UT inspection?: The response depends on defect severity, location, and the infrastructure manager’s response protocols. An indication that exceeds the defined amplitude threshold is flagged by the test train recording system with its track location. Depending on the defect classification, the response may be: immediate temporary speed restriction and hand verification inspection within 24 hours (for high-severity flags); scheduled hand verification within the next inspection cycle (for lower-severity flags); or marking for monitoring at the next scheduled test pass (for very low amplitude indications). If hand verification confirms a significant defect, the options are rail replacement (the defective section is cut out and replaced), speed restriction (maintained until replacement), or in some cases, controlled rail grinding to remove surface initiating features before sub-surface propagation.
Q: How deep can ultrasonic testing detect defects in a rail?: Ultrasonic testing can detect defects throughout the full depth of the rail — from just below the surface to the rail foot, typically 130–180 mm deep depending on rail section. Different probe angles target different depths and zones: vertical probes inspect the full depth for horizontal defects; angled probes at 70° target the rail head interior where tache ovale and transverse defects typically originate. The minimum detectable defect size depends on frequency, probe design, and the inspection standard being applied, but typical calibration targets are equivalent to a 6 mm diameter flat-bottomed hole — meaning defects of this size or larger at any depth should be reliably detected.
Q: What is the difference between a squat and a tache ovale?: A squat begins at the rail head surface as a rolling contact fatigue defect — a small depression visible to the eye — and grows downward into the rail head, sometimes developing a two-branch (W-shape) crack pattern. It is a progressive surface-to-internal defect. A tache ovale (French for “oval mark,” also called a kidney fracture) is a defect that originates internally in the rail head, typically from a hydrogen flake or inclusion present since manufacture. It has no surface expression until it is very large, and grows as a characteristic oval or kidney-shaped crack in the transverse plane. Both can eventually lead to transverse rail fracture, but tache ovale is considered more dangerous because it gives no warning signs until it is potentially critical.
Q: Can ultrasonic testing be used on rail welds?: Yes — weld inspection is one of the most important UT applications on continuously welded rail (CWR) networks. Both thermite welds (used for joining rails in the field) and flash-butt welds (made in a welding plant and delivered as welded panels) are inspected ultrasonically. Weld inspection is technically more challenging than rail body inspection because the weld zone has a different microstructure (coarser grain, residual stresses) that affects sound transmission, and the weld geometry may include features that produce confusing echoes. Phased Array UT (PAUT) is increasingly used for weld inspection because its ability to steer and focus the beam improves detection probability in the complex weld geometry. Newly installed welds are typically tested before the line is returned to service.