A practical specification guide for stay cables and their anchorages — written for bridge engineers who have already been handed a steel cable design and now need to make it hold.
Every long-span bridge conversation eventually arrives at the same place: the deck wants to be heavy enough to stay calm, and the cable system wants to be light enough to keep the towers economical. Steel strand solved that trade-off for most of the last century, and for spans under roughly a kilometre it still does. But on long crossings, in marine air, and at joints where a stainless fork meets a galvanised socket, the cable system starts to show a different set of problems — ones that have less to do with breaking load and more to do with durability, fatigue at the anchorage, and the slow creep of thermal movement through a structure that was designed around steel.
This article is about how carbon fiber composite products answer those problems. It is organised around the actual engineering decisions on a cable-stayed bridge: where the material earns its place, which product form fits which detail, what the properties mean numerically, and what goes wrong during installation if nobody thinks about it beforehand.
High-strength steel strand is protected by three layers: galvanising, a wax or grease fill, and an HDPE or PE sheath. Each layer is a barrier, and each barrier has an attack surface. The sheath cracks at the saddle, or at the point where the cable exits the anchorage, or where a maintenance vehicle has clipped it. Water enters, the fill degrades, and the strand begins to corrode from the inside where no inspection camera reaches.
Most bridge owners now accept that stay cables are consumable items on a 20–35 year cycle, not permanent structure. Replacement is planned, budgeted and disruptive. Reducing that cycle — or removing the corrosion mechanism entirely — is the first place where a composite cable earns attention.
Traffic loading produces small, high-cycle stress ranges in the cable. A steel strand is designed to carry these well in the free length, but the anchorage is a different story: the socket transition concentrates stress, and fretting between individual wires and the potting compound acts as a crack initiator. Allowable stress ranges at the anchorage are typically a fraction of the nominal cable strength, and that fraction governs the whole design.
A parallel element composite has no interstitial fretting between elements, and its termination can be designed so that load transfer is distributed rather than concentrated.
Steel has a coefficient of thermal expansion around 12 × 10-6/°C. Over a 400 m cable stay, a 30 °C daily swing produces roughly 144 mm of free-length movement before restraint is considered. That movement is absorbed by the tower, the deck, the expansion joints and the dampers. Carbon fiber composite along the fibre direction sits near 0 to 1 × 10-6/°C — practically an order of magnitude less. The structure becomes easier to reason about thermally, and the articulation hardware gets a smaller job.
Cable self-weight loads the tower, the anchorage zone and the foundation. CFRP at around 1.6 g/cm³ against roughly 7.85 g/cm³ for steel removes about 80% of that load for equal section capacity in tension. On long spans this ripples out into tower geometry, foundation size and the crane capacity needed for erection.
It would be dishonest to present a composite cable as a complete answer. The anchorage housing, the deck-level diaphragms, the inspection gantries and the wind fairings around the earlier steel solution all remain exposed to the same marine atmosphere. That is where the rest of the product family — fabric, plates and pultruded profiles — comes in.
Rather than describe the material in the abstract, it helps to walk the structure from the top of the tower down to the deck soffit and name the product that belongs at each point.
Carbon fiber long strand is produced as continuous tow with controlled diameter and high filament count, available in standard and high-modulus grades. In cable manufacture the tow is assembled into parallel-element rods, impregnated and sheathed. The roving form is what a cable maker buys; the design values that matter are tensile modulus in the fibre direction, strain to failure, and the consistency of the tow.
Twisting a strand introduces a helical path, which converts some axial load into a radial component and slightly reduces the effective modulus. A parallel-element cable keeps every fiber on the load line, and it lets the anchorage grip each element rather than squeeze a bundle. For a stay cable the benefit is a stiffer, more predictable cable and a termination that behaves better under fatigue.
Bridges that carry rail or trolleybus power, and crossings that sit under high-voltage transmission lines, deal with induced currents and stray current through the cable system. A composite cable with a properly designed insulating sheath interrupts that path. This is a secondary benefit, but on light-rail crossings it has decided more than one specification.
The zone where the cable meets the structure is geometrically the hardest part of the system. Carbon fiber tubes with thick walls and controlled winding angles are used as transition sleeves, guide tubes and the outer body of the anchorage assembly. Wound tube gives hoop and axial stiffness in a single part, resists the local bearing stresses from the wedge or potting system, and does not create a dissimilar-metal couple against stainless or galvanised steel.
Inspection walkways inside the anchorage chamber, cable-support saddles and temporary erection guides are also commonly built from pultruded tube — not for strength, but because they can be carried up by hand and will not corrode in a chamber that is permanently damp.
At deck level the cable force enters through a diaphragm. That region is congested with reinforcement, post-tensioning ducts and bearing plates, and adding steel stiffening is often impossible without cutting something else. Laminated carbon fiber plates, bonded and bolted to the concrete or steel diaphragm, add stiffness exactly where the local strut-and-tie model needs it, with a plate thickness measured in millimetres rather than centimetres.
The same plate is used for the cable anchorage inspection platforms and for local repair of spalled cover concrete around the anchor recess, where a thin bonded laminate restores confinement without adding significant dead load.
Carbon fiber fabric in woven 3K and 12K weights is the practical material for wrap-based protection: sealing the top of the anchor housing, repairing impact damage to the pylon fairing, or providing a durable skin over a composite access structure. Where the part needs to be moulded rather than wrapped — the aerodynamic fairing around a cable root, a cable guide shroud, a service hatch cover — carbon fiber prepreg gives a repeatable, low-void laminate with a paint-ready surface.
Material claims are only useful if they can be checked against design values. The following are typical figures for standard modulus carbon fiber composite against a conventional stay cable strand, and are given as ranges because they depend on fibre grade, resin system and fibre volume fraction.
Standard modulus carbon fiber composite reaches tensile modulus in the region of 140–160 GPa and tensile strength around 1,800–2,500 MPa at a density near 1.6 g/cm³. Steel strand for cable use sits near 195 GPa and 1,770 MPa at 7.85 g/cm³. The composite therefore carries comparable stress at about one fifth of the mass, and its specific modulus — modulus divided by density — is roughly four times that of steel.
What that means in design terms: for the same cable force, the composite cable's cross-section will be larger in absolute area but far lighter in total; and the sag, which scales with weight per unit length against tension, drops noticeably.
Unidirectional composite in the fibre direction does not accumulate fatigue damage in the way steel does, provided the load stays tensile and the matrix is not overheating. Stress ranges that would shorten a steel strand's anchorage life are absorbed without crack initiation at the element level. The design caution is at the termination and at any point where the cable is bent — compressive and transverse loads are where composites are least forgiving, and where the anchorage detail must be engineered.
Carbon fiber itself creeps negligibly under sustained tensile load. The resin matrix and the termination do creep, at a low but non-zero rate, and this is why a composite stay cable is normally proof-loaded and then re-tensioned after a settling period during erection. Design guidance for composite cables generally allows a small allowance for long-term relaxation, which is comparable to, or lower than, the relaxation allowance for steel strand.
Longitudinal coefficient of thermal expansion for the composite is between 0 and 1 × 10-6/°C, against about 12 × 10-6/°C for steel. The practical consequence is not only smaller movement at the deck joint, but a cable whose tension changes much less between a cold night and a hot afternoon — which reduces the demand on dampers and on the anchorage fatigue detail.
Carbon fiber does not corrode. The resin system and the sheath do age under UV and moisture, and that is why composite cables use a protective sheath with the same seriousness that a steel cable uses its HDPE. The difference is that a breach in a composite cable's sheath exposes a material that does not rust, whereas a breach in a steel cable's sheath starts a clock.
The table below is intended to be copied into a design review, not read as a catalogue. Priorities are ranked for a marine cable-stayed crossing in a seismic region; a dry inland bridge with a shorter span would weight them differently.
| Bridge Detail | Product Form | Primary Priority | Secondary Priority |
|---|---|---|---|
| Main stay cables | Carbon fiber long strand, parallel element | Fatigue performance at anchorage | Specific stiffness; corrosion immunity |
| Anchor transition sleeves | Wound carbon fiber tube, thick wall | Hoop bearing strength | Galvanic isolation from steel |
| Deck diaphragms and anchor recesses | Carbon fiber plate, UD or woven | Local stiffness gain with minimal added depth | Bond quality on prepared concrete |
| Cable root fairings and shrouds | Carbon fiber prepreg, moulded | Surface finish and dimensional repeatability | UV-stable coating compatibility |
| Access platforms and internal walkways | Pultruded carbon fiber tube and plate | Corrosion resistance in damp chambers | Hand-carried erection |
| Housing wrap and impact repair | Carbon fiber fabric, 3K woven | Conformability to complex geometry | Wet-out speed on site |
Composite cable systems do not usually fail in the free length. They fail at the four interfaces below, and each one has a known fix.
The termination must grip the parallel elements without crushing them and without letting them slip. Resin-potting into a cone, and wedge systems with a compliant liner, are both used; whichever is chosen, the design needs a proof load significantly above the working load, and the anchorage zone needs to be detailed so the load can be introduced gradually. Ask the cable supplier for the fatigue test data at the termination, not just at mid-length.
Carbon fiber is electrically conductive and nobly positioned relative to steel. Where a carbon composite part is bolted to a steel or aluminium structure in a wet environment, the steel becomes the anode. On a bridge this is solved with insulating washers and bushings at the fasteners, a bond gap filled with an isolating adhesive, and no bare carbon edge in contact with bare metal. It is a small detail that is routinely missed on site, and it is the single most common cause of premature corrosion on composite retrofits.
Unidirectional composite loses tensile capacity quickly when bent. Every change of cable direction — at the saddle, at a guide pipe, at the exit of an anchorage — must respect a minimum bend radius that the cable supplier states explicitly. Field modifications that add a slightly sharper bend are cheaper now and expensive later.
Because the termination and the resin matrix settle, composite cables are typically installed, proof-loaded, and re-stressed after a defined interval. The erection sequence should place this re-tensioning before the deck wearing course and the final survey, not after. Schedule it, and the finished cable has the tension the design assumed.
The resin system has a glass transition temperature, and the cable's design strength depends on staying below it. On most bridges this is not a concern; near a fire risk zone at the deck level, or where a cable passes close to an exhaust path, the sheath and the resin choice need to be reviewed. This is also where aramid reinforcement enters the conversation — aramid fiber is used in the protective wrap and in heat-resistant cable accessories where a non-conducting, flame-resistant high-performance fiber is needed alongside the carbon load path.
A cable-stayed bridge is a mixed-material structure by nature, and the composite solution connects to all of it. Three additional product families matter here.
Aramid fabric is used as a protective layer over the cable sheath in zones exposed to vandalism and vehicle impact, as a fire barrier at the anchor housing, and in the anchorage chamber where the fire load from a maintainer's equipment is considered. It is non-conductive and does not create a galvanic couple with anything.
Pre-oxidized fiber felt and fabric are used where the anchorage zone needs thermal insulation — protecting the resin in the termination from radiant heat, and lining service galleries that run close to hot services. The material is non-combustible and keeps its integrity at temperatures where polymer-based insulation would degrade.
Handrails, cable guide brackets, ladder rungs and junction boxes on the bridge are increasingly moulded from short-fiber reinforced compounds. Compounded carbon fiber short fiber gives those small parts a stiffness improvement over glass-filled equivalents while letting the moulder keep the existing tool design.
We manufacture the full chain that a cable-stayed bridge project needs: carbon fiber long strand and tow, carbon fiber fabric in woven and unidirectional forms, carbon fiber prepreg, pultruded and wound carbon fiber tube, laminated plate including machined finished parts, and the aramid and pre-oxidized fiber materials used around the anchorage and the protected zones. That matters because a bridge solution is not one product — it is a set of products that have to arrive with consistent properties and compatible sizing.
What we do not do is design your bridge. We supply material to the design, and we expect the design to have been checked by the engineer of record. What we can give you is the specification data, the test certificates, the sample panels and the pre-production parts that let you confirm the design before the first cable is made.
A sensible first step is to send us the cable force, the anchorage geometry and the environment. From those three inputs we can tell you which fibre grade and which cable construction to evaluate, and which of the surrounding products — tube, plate, fabric, prepreg, aramid or pre-oxidized fiber — you will need alongside it.
The bridge industry is conservative for good reasons, and it should be. A composite stay cable is not a novelty item being sold on enthusiasm; it is a material change that has to survive forty years of being rained on. If your project is at the stage where the cable is being specified, the useful conversation is about test data, not adjectives.
Contact our engineering team with your bridge span, cable force and environment, and we will send the material datasheets, sample laminates and a proposed qualification sequence for review.
Home
Call us