A suspension bridge is a long-span structure that carries its deck loads through tension in a pair of main cables draped between two towers and anchored at each end. The main cable is the critical component — it converts the deck's vertical weight into axial tension, which the anchorages resist as horizontal pull-out force. This design lets you span distances no truss or beam can reach economically. The Akashi Kaikyō in Japan uses this exact mechanism to clear 1,991 m in a single span.
Suspension Bridge Interactive Calculator
Vary span, deck load, and cable sag to see how horizontal cable tension and anchorage load change.
Equation Used
The calculator uses the suspension bridge cable tension relation H = wL²/(8f). Increasing sag f lowers the horizontal pull H on the anchorages, while longer spans and heavier deck loads increase it rapidly. The vertical tower reaction shown is V = wL/2, and the cable tension near the tower is estimated from T = sqrt(H² + V²).
- Uniform deck load carried by one main cable.
- Cable shape is approximated as parabolic under uniform vertical load.
- Dead-load static case only; wind, live-load redistribution, and cable self-weight are not included.
How the Suspension Bridges Works
The deck hangs from vertical suspender hangers, which clip onto the main cable above. That main cable drapes between two towers in a curve close to a catenary, with extra parabolic shaping under uniform deck load. The towers push the cable up at the saddles, redirecting the cable angle so the tension stays continuous through the saddle without bending the tower. From there the cable runs down to the anchorage block on each shore, where it's splayed out into individual strands and locked into bedrock or a gravity mass. Every bit of deck weight ends up as horizontal pull on those two anchorages — that's the whole load path in one sentence.
The sag ratio — the cable's mid-span dip divided by the main span length — is the single number that controls how the bridge behaves. Drop it below about 1:12 and the cable tension shoots up, the towers get punished, and the anchorages need to be massive. Push it above about 1:8 and the towers grow uncomfortably tall and the deck ends up too high above the water. Most long-span suspension bridges land between 1:9 and 1:11. The Golden Gate sits at roughly 1:11. Get this wrong on a 1,000 m span and you'll be designing for 30% more cable tension than necessary.
If the stiffening truss or stiffening girder is undersized, you get aerodynamic instability — Tacoma Narrows in 1940 is the textbook failure. The deck twisted itself apart at 64 km/h winds because the open plate-girder cross-section caught vortex shedding at its torsional natural frequency. Modern decks use either deep stiffening trusses (Forth Road, Mackinac) or aerodynamically shaped closed steel box girders (Humber, Great Belt East) to push flutter speed well above the design wind. Skip the wind-tunnel testing and you'll find out the hard way which mode your deck wants to dance in.
Key Components
- Main Cable: Two parallel cables that carry the deck weight as axial tension. Each cable is spun in place from thousands of 5 mm galvanised high-strength steel wires — Akashi Kaikyō used 36,830 wires per cable at 1,800 MPa tensile strength. The cable diameter typically lands between 0.6 m and 1.2 m on long spans.
- Tower (Pylon): Carries the vertical reaction from the main cable into the foundation. Modern towers are reinforced concrete or stiffened steel cells, sized so that the cable saddle sits 100 m to 300 m above deck level. The tower must also handle wind moments and slight cable-angle imbalances during construction.
- Saddle: A grooved casting at the tower top that supports the cable and lets it change direction without bending. The radius is matched to the cable diameter — typically 8× to 10× the cable diameter — to keep individual wire bending stresses below the fatigue limit. Saddles slide slightly on rollers during cable spinning to balance unequal side-span and main-span loads.
- Suspender Hangers: Vertical wire ropes or rigid rods that hang the deck from the main cable, spaced typically 10 m to 20 m apart. They carry pure tension under dead load but see fatigue cycling from traffic and wind. Replacement intervals run 30 to 50 years on heavily trafficked spans like the Severn Bridge.
- Anchorage: The gravity or rock-tunnel block that resists the horizontal pull of the main cable. On the George Washington Bridge each anchorage holds back roughly 32,000 tonnes of cable tension. Concrete gravity anchorages weigh 200,000 tonnes and up on long spans; rock anchorages tunnel directly into competent bedrock and are far cheaper when geology allows.
- Stiffening Truss or Box Girder: Distributes concentrated live loads along the deck so the main cable doesn't see point loads, and provides the torsional and bending stiffness needed to suppress aerodynamic flutter. Depth-to-span ratios run 1:60 to 1:180 depending on whether the deck is a truss or a streamlined box.
Industries That Rely on the Suspension Bridges
Suspension bridges show up wherever the span exceeds what a cable-stayed or arch bridge can economically reach, or where deep water rules out intermediate piers. The crossing length sets the choice — under about 600 m, cable-stayed wins on cost and stiffness; above 1,000 m, suspension is usually the only option that pencils out. The mechanism scales to spans that would require impossible foundation work for any other bridge type, which is why every span over 1,500 m on Earth is a suspension bridge.
- Highway Infrastructure: Akashi Kaikyō Bridge in Japan — 1,991 m main span, the longest suspension bridge span in the world until 2022, carrying a six-lane expressway between Kobe and Awaji Island.
- Highway Infrastructure: Golden Gate Bridge in San Francisco — 1,280 m main span, carrying US Route 101 with two main cables 0.92 m in diameter spun from 27,572 wires each.
- Rail and Road Combined: Tsing Ma Bridge in Hong Kong — 1,377 m span carrying both the Tung Chung Line MTR rails and a six-lane highway on a double-deck stiffening truss.
- Estuary Crossings: Humber Bridge in England — 1,410 m main span using a streamlined steel box girder deck, the first long-span suspension bridge to abandon the stiffening truss in favour of an aerodynamically shaped closed deck.
- Strait Crossings: Great Belt East Bridge in Denmark — 1,624 m main span connecting Funen and Zealand, with anchorage blocks each weighing 325,000 tonnes.
- Pedestrian and Light Vehicle: Capilano Suspension Bridge in North Vancouver — 137 m span at 70 m above the river, a tourist crossing that uses the same load path on a smaller scale with rope main cables.
- Mountain Crossings: Çanakkale 1915 Bridge in Turkey — 2,023 m main span over the Dardanelles, currently the world's longest, using twin towers 318 m tall.
The Formula Behind the Suspension Bridges
The horizontal component of main cable tension is the number that drives anchorage size, tower base reaction, and cable diameter selection. It depends on the distributed deck load, the main span, and the sag ratio. At a tight sag ratio of 1:12 the cable tension runs roughly 50% higher than at a loose 1:8 ratio for the same load. The sweet spot for most long spans sits at 1:10 to 1:11 — tight enough to keep towers reasonable, loose enough to keep cable steel and anchorages from blowing the budget. Push outside that band and one cost line dominates the project.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| H | Horizontal component of main cable tension (per cable) | N | lbf |
| w | Uniform distributed load along the deck per unit length, per cable | N/m | lbf/ft |
| L | Main span length between towers | m | ft |
| f | Cable sag at mid-span (vertical drop from saddle to low point) | m | ft |
Worked Example: Suspension Bridges in a 1,200 m highway suspension crossing
A bridge engineering team in Busan South Korea is sizing the main cable horizontal tension for a proposed 1,200 m main-span suspension bridge over the Nakdong estuary. The combined dead and live load on the deck works out to 320 kN/m total, which means 160 kN/m per cable. They want to test how cable tension responds across the realistic sag-ratio range of 1:8 to 1:12 to pick the optimum.
Given
- L = 1200 m
- w = 160 kN/m per cable
- fnominal = 120 (sag ratio 1:10) m
- flow = 100 (sag ratio 1:12) m
- fhigh = 150 (sag ratio 1:8) m
Solution
Step 1 — compute the nominal horizontal tension at the design sweet spot of 1:10 sag ratio, where f = 120 m:
That's 240 MN of horizontal pull per cable that the anchorage must resist — roughly 24,500 tonnes of force, equivalent to the weight of a fully loaded Panamax cargo ship pulling on each shore.
Step 2 — at the low end of the realistic range, a tight sag ratio of 1:12 (f = 100 m):
Drop the sag by 20 m and cable tension jumps 20% — you'll need a thicker main cable, beefier saddles, and significantly larger anchorage blocks. The towers also feel a bigger horizontal kick at the saddle from the side-span imbalance during construction.
Step 3 — at the high end of the realistic range, a loose sag ratio of 1:8 (f = 150 m):
Cable tension drops 20% from nominal, but the towers grow about 30 m taller to accommodate the deeper sag, and tower steel/concrete cost rises faster than cable cost falls on this span length. The deck also sits higher above the water, which can hit navigation clearance limits depending on the channel.
Result
Nominal horizontal cable tension is 240 MN at the 1:10 sag ratio sweet spot. That sets the main cable diameter at roughly 1.0 m of compacted galvanised wire bundle and dictates an anchorage block in the 250,000 to 300,000 tonne range, similar to the Great Belt East anchorages. Across the realistic operating range the tension swings from 192 MN at 1:8 down to 288 MN at 1:12 — a 50% spread driven entirely by sag choice, which is why this is the first decision on the design board. If your built measurement of cable tension during proof loading deviates more than 5% from the calculated value, the most common causes are: (1) saddle friction locking the cable from sliding to its true equilibrium position, leaving residual side-span/main-span imbalance, (2) anchorage strand splay angles set wrong during cable spinning so individual strand tensions are uneven, or (3) deck dead load coming in heavier than estimate because the asphalt overlay or stiffening-truss connections were under-counted at design stage.
When to Use a Suspension Bridges and When Not To
Suspension is one of three long-span bridge families that engineers compare on every major crossing project. Cable-stayed and arch bridges cover overlapping span ranges, but each has a clear cost and stiffness profile. Here's how they line up on the dimensions that matter for a project decision.
| Property | Suspension Bridge | Cable-Stayed Bridge | Steel Arch Bridge |
|---|---|---|---|
| Practical maximum main span | 2,023 m (Çanakkale 1915) | 1,104 m (Russky Bridge) | 552 m (Pingnan Third) |
| Economic span range | 1,000 m to 2,000+ m | 200 m to 1,100 m | 100 m to 550 m |
| Deck stiffness (live load deflection) | Low — needs stiffening truss or box girder | High — direct cable triangulation | Very high — rigid arch action |
| Anchorage requirement | Massive gravity or rock anchorages each end | Self-anchored at tower bases — no external anchorage | Thrust into abutments, no cable anchorage |
| Construction time on 1 km span | 5 to 7 years typical | 3 to 5 years typical | Not feasible at this span |
| Cost per square metre of deck (1 km span) | High — anchorages dominate | Medium — no anchorage cost | Not applicable |
| Aerodynamic flutter risk | High — wind tunnel testing mandatory | Moderate — stays add damping | Low — closed deck section |
| Foundation tolerance for deep water | Excellent — towers can be far apart | Good — limited by stay length | Poor — needs firm abutments |
Frequently Asked Questions About Suspension Bridges
A free-hanging cable under its own weight only forms a true catenary. On a real suspension bridge, the deck load carried through the suspender hangers is much larger than the cable's self-weight — typically 10 to 20 times larger on a long span. That deck load is uniformly distributed along the horizontal axis, not along the cable arc, and a uniform horizontal load produces a parabola.
The cable shape ends up being a parabola very slightly modified by the cable's own weight near the saddles. For sag-ratio calculations and tension estimates the parabolic equation H = wL²/8f is accurate to within 1% on most spans. Only when you're spinning the cable empty during construction does the catenary shape actually matter, and that's why aerial spinning targets a different reference shape than the final loaded shape.
It comes down to span length, wind environment, and depth budget. Stiffening trusses give you torsional stiffness through depth — typically 1:80 to 1:120 depth-to-span — and they're forgiving of aerodynamic uncertainty because the open lattice doesn't shed coherent vortices. The Forth Road Bridge and Mackinac Bridge both went this route.
Streamlined steel box girders, pioneered on the Humber Bridge, use shape rather than depth to suppress flutter. They're shallower (1:300 depth-to-span on Humber), lighter, and cheaper in steel tonnage, but they demand serious wind-tunnel testing because flutter speed is highly sensitive to edge-fairing geometry. If you're building in a high-wind environment with limited vertical clearance, the box girder usually wins. If wind data is uncertain or the deck must carry rail traffic with concentrated loads, the truss is the safer call.
The H = wL²/8f formula assumes the cable is perfectly flexible and the deck transfers load uniformly through the hangers. In a real bridge, the stiffening truss or box girder carries a fraction of the live load itself through bending, reducing the cable tension increment under traffic. On a long span this fraction is small (5-10%), but on shorter suspension spans below 400 m the deck stiffness can carry 20-30% of live load directly.
The other typical source of disagreement is suspender hanger elongation. The simple formula treats the hangers as rigid, but they stretch under load and let the deck drop a few centimetres independent of cable sag change. If your FE deflection runs higher than hand calc, check whether you've included hanger axial stiffness — and whether the model captures the correct construction stage of the cable (the unstressed cable length is set during spinning, not during service).
The crossover sits around 600 to 900 m of main span, and the deciding factor is usually anchorage geology rather than the deck itself. Cable-stayed bridges are self-anchored — the horizontal cable force resolves into deck compression, so you don't need a 250,000-tonne anchorage block at each shore. That single fact makes cable-stayed cheaper below about 800 m almost universally.
Above 1,000 m the deck compression in a cable-stayed scheme becomes excessive — you'd need a deck section so heavy that the back-stays can't economically support it. The Russky Bridge at 1,104 m is currently the longest cable-stayed span built and is widely considered to be at the practical limit. If your crossing exceeds 1,200 m, suspension is the only realistic option. Between 600 and 1,000 m, run both schemes through a cost estimate — soil conditions at the anchorage sites usually break the tie.
Tacoma Narrows failed in 1940 because the deck was an H-shaped plate girder only 2.4 m deep on a 853 m span — a depth-to-span ratio of 1:350, far shallower than anything built before. The shape caught vortex shedding at low wind speeds and the torsional natural frequency happened to align with the vortex shedding frequency at 64 km/h winds. The bridge entered self-excited torsional flutter and tore itself apart in under an hour.
The reason it still gets cited isn't that we don't understand the physics now — we do, and every long-span bridge since the 1950s has gone through wind tunnel testing. It gets cited because the failure mode (aerodynamic flutter) is fundamentally different from static load failure, and it's a reminder that a suspension bridge that passes every static check can still fail catastrophically if the deck cross-section is wrong. The lesson is procedural: wind tunnel testing is not optional, and analytical flutter prediction alone is not sufficient on a novel deck shape.
It matters more than most people realise because it controls the unbalanced horizontal force at the tower saddle. A side-span to main-span ratio between 0.25 and 0.5 is the practical range. At 0.25 the side spans are short and stiff, and the saddle sees significant horizontal imbalance under main-span live load — the saddle has to slide on its rollers or the tower top deflects.
At 0.5 the side spans match the main span more closely in stiffness behaviour, the saddle stays balanced, and the tower sees mostly axial load. The Golden Gate uses 0.23 (a tight side span), which is why its towers were designed with significant longitudinal bending capacity. The Akashi Kaikyō uses 0.48 for cleaner load balance. If you're early in a design and choosing site geometry, push the side spans longer when you can — it simplifies the tower design and reduces saddle slip requirements during construction loading.
References & Further Reading
- Wikipedia contributors. Suspension bridge. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
