The Critical Connection: Fishplates (Rail Joints) Explained

On 28 December 1879, during a fierce winter storm, the Tay Bridge in Scotland collapsed as a passenger train crossed it. Seventy-five people died. The subsequent inquiry found multiple causes, but among them was a detail of the track structure on the bridge: the rail joints. Under the storm’s wind loading, combined with the weight and dynamic forces of the train, the track structure on the bridge had insufficient lateral stiffness — a condition that the fishplated joints, each a mechanical discontinuity in the rail string, had contributed to by allowing relative movement between adjacent rail ends that a welded connection would have prevented.

The Tay Bridge disaster did not directly cause the invention of CWR — that would wait another sixty years for welding technology to mature sufficiently for railway use. But it established an engineering truth that has only deepened with each subsequent decade of track engineering: the rail joint is the weakest point in the track structure. Every fishplate joint is a location where continuity of the rail section is broken, where bending stiffness is reduced, where stress concentrations develop around bolt holes, and where wheel impacts are amplified by the small gap between rail ends. The history of railway track engineering since the 1850s has been, in significant part, the history of managing the consequences of that weakness — and eventually, on high-speed and mainline track, eliminating it.

What Is a Fishplate?

A fishplate (the name comes from the fish-belly profile of the plate’s cross-section, curved to match the rail web) is a steel plate typically 400–600 mm long, designed to be bolted through the web of two abutting rail ends. Fishplates are installed in pairs — one on each side of the rail web — and secured with 4–6 high-strength bolts passing through matching holes in the fishplate and the rail web.

The fishplate maintains mechanical continuity across the joint by acting as a beam splice: the bending moment at the joint is carried partially by the fishplates, partially by the rail section’s residual stiffness. It is not a rigid connection — the fishplated joint has significantly less bending stiffness than an equivalent length of continuous rail, and the small gap between rail ends (typically 4–8 mm on conventional jointed track) allows both differential movement and the characteristic wheel impact that produces the “clickety-clack” of traditional railway travel.

The Anatomy of a Fishplate Joint

Why Fishplate Joints Fail: The Mechanics of Bolt-Hole Cracking

The bolt holes drilled through the rail web to accept fishbolts are stress concentrations — points where the local stress in the steel is significantly higher than the nominal stress in the surrounding material. Under the bending loads imposed by passing wheels, the stress at the edge of each bolt hole cycles between tension (as the wheel approaches) and compression (as the wheel recedes). This cyclic stress drives fatigue crack initiation and propagation.

The typical fatigue crack pattern at a fishplate joint:

1. Initiation: A microscopic fatigue crack initiates at the bolt hole edge — typically at the surface of the hole, where stress concentration and surface roughness from drilling are highest.
2. Propagation: Under continued cyclic loading, the crack propagates across the rail web in a roughly diagonal direction, sometimes producing the characteristic “star crack” pattern where multiple cracks radiate from a single bolt hole.
3. Rail break: If undetected and unrepaired, the crack extends through the full web cross-section, connecting with cracks from adjacent bolt holes to produce a complete transverse fracture of the rail — a rail break at the joint.

The rate of crack propagation depends on traffic intensity (load cycles per unit time), axle load (higher loads → higher stress cycle amplitude → faster crack growth), rail section (heavier rail → lower nominal stress → slower crack growth), and joint geometry (well-supported, correctly spaced joint → lower bending moment at hole → slower crack growth). On a busy mainline with 25-tonne axle loads, bolt-hole cracks may develop within 2–3 years of joint installation if not managed by proactive inspection and rail end renewal.

The Dip Joint: Geometry Failure at the Fishplate

Beyond fatigue cracking, the second major failure mode of fishplate joints is the “dip joint” — progressive settlement of the ballast and subgrade beneath the joint location, causing the rail ends to sag below the level of the adjacent track. The dip joint produces amplified wheel impacts as trains cross it (the wheel “falls” into the dip and impacts the rising rail on the other side), which accelerates both ballast deterioration and rail end batter.

The dip joint is a self-reinforcing failure: settlement causes impact amplification, which causes further settlement, which causes worse impact, and so on. Management requires either frequent ballast tamping at joint locations (a maintenance cost significantly higher than equivalent CWR track sections) or rail end reconditioning to restore vertical alignment. On busy mainlines, joint locations require tamping at 3–5 times the frequency of equivalent CWR track — a maintenance cost differential that is one of the primary economic drivers for CWR conversion.

Insulated Rail Joints (IRJs): The Fishplate’s Modern Safety Role

While fishplate joints have been largely eliminated from mainline track through CWR conversion, one form of fishplate joint is not only retained on CWR track but is actively manufactured and installed on CWR lines daily: the Insulated Rail Joint. An IRJ is a fishplate assembly in which every metal-to-metal contact path between the two rail ends is interrupted by an electrically insulating material — glass-fibre reinforced epoxy composite, nylon, or polyurethane — creating a defined electrical boundary between the two adjacent rail sections.

IRJs are required at every track circuit section boundary. They define the electrical limits of each detection block — ensuring that the traction current and signalling current flowing in one section cannot pass electrically into the adjacent section. Without IRJs, the track circuit system cannot distinguish which section of track is occupied by a train.

IRJ Construction and Failure Modes

The Two IRJ Failure Directions

IRJ failures have two distinct electrical directions, each with a different safety consequence:

Short failure (conducting when should insulate): The insulating material between the two rail ends has broken down, allowing electrical current to flow across the joint. The track circuit on one side of the joint detects the return current from the other side as if a train’s axle were bridging the joint — producing a false “occupied” indication. Signals ahead of the affected section display red; trains are stopped. Service disruption without any train in the section. This is safe but disruptive.

Open failure (not conducting when should): In some track circuit configurations, the IRJ’s insulation has become intermittent — functioning correctly most of the time but failing to maintain insulation under specific conditions (wet weather, freeze-thaw, heavy traction current). This can produce intermittent false “clear” or “occupied” indications that are difficult to reproduce and diagnose. This failure mode is operationally more dangerous than a consistent short because it may not be immediately apparent and can produce unpredictable signalling behaviour.

Glued Insulated Rail Joints (GIRJs): The Modern Solution

Traditional bolted IRJs have significant mechanical limitations: the bolt holes create stress concentrations in the rail web at the most heavily loaded location on the track (a discontinuity in the rail section), and the mechanical joint has less bending stiffness than the surrounding CWR. Glued Insulated Rail Joints (GIRJs) address this by bonding the fishplate assembly to the rail using structural adhesive, eliminating the bolt holes and creating a near-rigid joint that approaches the bending stiffness of the continuous rail.

GIRJ construction process:

1. Rail ends are precisely cut and ground to ensure flush abutment.
2. A fibreglass end post is inserted in the gap to maintain insulation between the rail ends.
3. Pre-formed fibreglass-reinforced composite fishplates are bonded to the rail web using a two-part structural epoxy adhesive.
4. The adhesive is cured under controlled conditions (temperature-maintained for defined cure time).
5. The completed GIRJ is installed in the track; no bolt holes are drilled through the rail.

GIRJs typically achieve bending stiffness within 10–15% of continuous rail — compared to 50–60% for a bolted IRJ — significantly reducing the dynamic impact at the joint location and extending both the GIRJ’s own service life and the life of the surrounding track structure.

Fishplate Joints vs CWR Welds: Complete Comparison

Where Fishplates Remain Essential Today

Despite the dominance of CWR on mainlines, fishplate joints remain in use or essential in several specific contexts:

• Insulated Rail Joints on CWR lines: IRJs and GIRJs are required at track circuit boundaries on every CWR mainline in the world. CWR eliminates mechanical joints but cannot eliminate the signalling need for electrical joints.
• Switches and crossings: The complex geometry of turnouts and crossings includes numerous short rail sections and closures that cannot easily be CWR-welded into continuous strings. Fishplate joints within switch and crossing assemblies remain common, though bolted joints in S&C are increasingly being replaced by specially welded assemblies.
• Temporary repairs: A broken CWR rail in service is repaired as an emergency with a welded insert, but as a very short-term temporary repair, fishplates are sometimes used to restore running connectivity while the weld crew prepares. Temporary fishplated repairs carry a speed restriction until properly welded.
• Depot and siding track: Low-speed, low-traffic maintenance depot and siding track still uses jointed track in many networks — the lower cost of fishplated jointed construction is justified where speeds and loads are too low to warrant CWR.
• Thermal adjustment joints (breather joints): Some networks use specially designed expansion joints at defined intervals in long CWR strings to allow controlled thermal movement and provide maintenance access points.

Editor’s Analysis

Frequently Asked Questions

Q: Why are fishplate joints called “fishplates”?

The name derives from the cross-sectional profile of the joint bar, which curves inward at the middle in a shape resembling the belly of a fish. The fishplate’s bearing faces are profiled to match the underside of the rail head and the upper surface of the rail foot, with the middle section curved away from the rail web to create clearance. When viewed in cross-section, this curved profile was thought to resemble the shape of a fish, giving rise to the “fishplate” name that has persisted for nearly 200 years. In North American usage, the component is more commonly called a “joint bar” or “splice bar” — functional names that are more descriptive if less evocative.

Q: What is a “star crack” at a rail joint and how dangerous is it?

A star crack is a fatigue crack pattern that develops from a fishbolt hole in the rail web, characterised by multiple cracks radiating outward from the bolt hole in multiple directions — producing the star-like appearance that gives the defect its name. Star cracks develop because the bolt hole creates a stress concentration at which fatigue crack initiation occurs, and the cyclic bending stresses at the joint location are high enough to drive crack propagation in multiple directions simultaneously. The danger level of a star crack depends on its size and rate of growth: small, stable cracks detected early are manageable through rail end renewal (cutting out the joint section and rewelding or replacing the rail); advanced star cracks that have propagated across a significant portion of the rail web cross-section can lead to sudden rail fracture if not acted upon. Rail breaks at joint locations are responsible for a significant proportion of track-caused derailments on jointed-track networks. Ultrasonic testing of joint sections, and rail end visual inspection, are standard maintenance activities specifically targeting bolt-hole and star crack detection.

Q: Can an IRJ failure be detected before it causes a signalling problem?

Yes — and regular IRJ testing is a standard maintenance activity on most networks. The primary test method is measurement of insulation resistance between the two rail sections at the joint: a properly functioning IRJ should have a resistance of many kilohms or more between the two rail sections; a degraded or failing IRJ will show reduced resistance as the insulating material deteriorates. Some networks use automated rail-mounted test equipment that measures IRJ resistance at line speed without requiring a track possession. Visual inspection can identify obvious mechanical damage — cracked fishplates, missing end posts, visible cracks in insulating liners — but cannot detect internal degradation of the adhesive bond in GIRJs or subsurface insulator cracking. IRJ replacement cycles (replacing IRJs after a defined service period, regardless of apparent condition) are used by some networks on critical joints where failure consequences are most severe — this is a cost-certain maintenance strategy that trades planned replacement cost against the risk of unplanned failure and service disruption.

Q: What is a “bond wire” and why does it matter?

A bond wire (or rail bond) is a copper braid cable attached to the rail web on each side of a fishplate joint, providing a low-resistance electrical path for traction return current to flow across the joint gap without depending on the rail-to-fishplate metal contact. On electrified railways, the traction return current flows through the running rails back to the traction substation. At each fishplate joint, the metal-to-metal contact between fishplate and rail provides some electrical conductivity — but this contact has higher resistance than the rail itself, and the gap between rail ends provides none. If bond wires are missing or broken at a joint, the traction return current is forced through the higher-resistance fishplate contact path, which can cause localised heating, accelerated corrosion at the joint, and — on DC electrified railways where stray current control is critical — increased stray current leakage from the track into the surrounding ground. Bond wire inspection and replacement is a standard maintenance activity on electrified networks; a missing bond wire is a defect requiring early repair rather than routine scheduling.

Q: How does the track circuit “know” when an IRJ has failed?

A track circuit failure caused by IRJ degradation typically manifests as an unexpected change in the track circuit’s normal electrical condition — either a false “occupied” signal (if the IRJ shorts, allowing the track circuit transmitter current from one section to reach the receiver of the adjacent section as if a train axle were present) or loss of the expected “clear” current level (if the IRJ opens intermittently). The signalling system’s track circuit relay or electronic receiver detects that its input current is outside the normal operating range and declares the section in an unsafe state, displaying a red signal ahead as a fail-safe response. The maintenance response is to measure the insulation resistance across the IRJ to confirm the failure, and then either replace the IRJ or carry out a temporary repair until planned replacement. In signalling maintenance statistics, IRJ-related track circuit failures are consistently one of the top causes of signal failures by incident count on most European mainline networks — a consequence of the large number of IRJs in service and their exposure to the mechanical stress of the joint location.