Steel Cable Structures — Design Guide for Cable Roofs and Bridges
Steel cable structures use high-strength cables as primary load-carrying elements. This guide covers cable roofs, cable-stayed bridges, suspension bridges, and cable trusses.
Quick links: Cable sag → | Steel truss bridge → | Steel bridge girder →
Core calculations run via WebAssembly in your browser with step-by-step derivations across AISC 360, AS 4100, EN 1993, and CSA S16 design codes. Results are preliminary and must be verified by a licensed engineer.
Frequently Asked Questions
What types of steel cables are used in structural applications? Common structural cables: (1) Spiral strand — helically wound wires, good for suspension bridge main cables, (2) Locked coil strand — interlocking wire profiles, higher density, better corrosion resistance, (3) Parallel wire strand (PWS) — parallel wires in a PE sheath, used for cable-stayed bridges, (4) Full-locked coil rope — outer layers mechanically interlocked. Per EN 1993-1-11 Table 2.1, cable modulus ranges from 160±10 GPa (spiral strand) to 195±10 GPa (parallel wire). Ultimate tensile strength ranges from 1370-1770 MPa.
How are cable structures analyzed for nonlinear behavior? Cable structures exhibit geometric nonlinearity due to sag effects. The key parameters are: (1) Sag-to-span ratio — typically 1/8 to 1/12 for suspension bridges, 1/4 to 1/8 for cable roofs. (2) Catenary action — cable shape under self-weight follows cosh(x/c). (3) Stress stiffening — cable stiffness increases with tension. Analysis methods: nonlinear FEA (Newton-Raphson with large displacement), equivalent modulus approach (Ernst formula: Eeq = E/(1+(γ²L²E)/(12σ³))), and trial-and-error with equilibrium equations.
What are the design considerations for cable anchors? Cable anchors must be designed for: (1) Static strength — anchor capacity ≥ 1.5× cable MBL (minimum breaking load) per PTI recommendations, (2) Fatigue — socket connections designed for 2 million cycles at 100 MPa stress range (EN 1993-1-11), (3) Corrosion protection — Class A (greased and fully encapsulated) for permanent anchors, (4) Creep — strand relaxation loss of 2-8% of initial prestress over 50 years depending on steel grade and stress level. Bracket/cable anchorage zone must be designed for local stresses from the concentrated cable force.
How are cable vibrations controlled in cable-stayed structures? Cable vibration control is essential for fatigue performance and serviceability. Per EN 1993-1-11 Section 6.3 and PTI Recommendations: (1) Rain-wind induced vibration — the most critical type for stay cables, occurring when water rivulets form on the cable surface at wind speeds of 8-15 m/s. (2) Vibration mitigation devices — helical fillets (wire wound around the PE sheath disrupt water rivulets), dampers (internal or external viscous dampers, typically oil-filled with 0.5-2% critical damping), and cross-tie cables (secondary cables connecting primary cables). (3) Scruton number criterion — Sc = 2meδs/(ρD²) ≥ 10 for aerodynamic stability, where me = cable mass per unit length, δs = structural damping (log decrement), ρ = air density, D = cable diameter. Example: a 100 mm diameter cable with m = 50 kg/m, δs = 0.02, ρ = 1.2 kg/m³ gives Sc = 2×50×0.02/(1.2×0.1²) = 167 ≥ 10 — stable. (4) Fatigue analysis per EN 1993-1-11 — cables designed for 2×10⁶ cycles at stress ranges of 150-200 MPa at anchorages. (5) Parametric excitation — cable natural frequency should not coincide with deck or tower frequencies; avoid ratios of 1:1, 2:1, or 1:2 between cable and structural frequencies.
Cable Structure Analysis Methodology
The analysis of cable structures differs fundamentally from conventional steel frame analysis due to geometric nonlinearity. A suspension bridge main cable is perhaps the best example to illustrate the method.
Parabolic versus catenary theory. For a horizontal cable with uniform self-weight w (kN/m) spanning L between supports: (1) The cable shape is approximately parabolic if the sag d is less than 1/8 of the span: y(x) = 4d(x/L)(1 - x/L). (2) Horizontal tension component H = wL²/(8d). For a suspension bridge with L = 500 m, w = 40 kN/m (main cable + suspenders + superimposed dead load), and sag d = 50 m (1/10 span): H = 40 × 500²/(8 × 50) = 25,000 kN. (3) Maximum cable tension at the tower: Tmax = H/cosθ, where θ = angle at tower. If the cable enters the tower saddle at an angle of 30° from horizontal: Tmax = 25,000/cos(30°) = 28,868 kN. (4) Cable area required: for a cable with fpu = 1,770 MPa and φ = 0.65 (LRFD): A ≥ 28,868/(0.65 × 1,770/1,000) = 25,100 mm² — requiring approximately 60 parallel wire strands of 127 wires each at 5 mm diameter.
Equivalent modulus (Ernst formula). For preliminary analysis, the cable sag effect can be represented through an equivalent straight truss element with reduced modulus: Eeq = E/(1 + (γ²L²E)/(12σ³)). Example: for locked coil strand (E = 160 GPa), cable self-weight γ = 78.5 kN/m³, horizontal projected length L = 100 m, and tensile stress σ = 400 MPa: Eeq = 160,000/(1 + (78.5² × 100² × 160,000)/(12 × 400³ × 10⁶)) = 160,000/(1 + 0.080) = 148 GPa. The 7.5% reduction in stiffness represents sag effects and must be accounted for in global structural analysis.
Cable net and cable truss systems. For cable roof structures, two common configurations: (1) Cable net — two families of cables crossing at right angles, one acting as load-bearing cables (catenary downward) and the other as stabilizing cables (curved upward). A typical cable net for a 100 m span stadium roof uses 40-60 mm diameter locked coil strands at 2-3 m spacing in each direction with pretension of 200-400 kN per cable. (2) Cable truss — paired cables (upper and lower) separated by compression struts, forming a self-equilibrating system. The pretension in the lower cable creates upward camber that counteracts dead load deflection. Prestress levels typically range from 15-25% of the cable's ultimate capacity.
Construction sequence effects. Cable structures are erected in stages and the stress state changes at each stage. For a cable-stayed bridge: stage 1 — erect tower and anchor the back stays; stage 2 — install deck segments one at a time, tensioning each stay cable to 60-70% of final force; stage 3 — apply superimposed dead load (pavement, barriers); stage 4 — final cable tension adjustment. The construction analysis must track stress increments in each cable at each stage — a single stay cable in a 500 m span bridge may be re-tensioned 3-5 times during erection to achieve the target profile.
Structural damping of cable systems. The inherent damping of steel cables is very low, typically 0.1-0.2% of critical damping. This makes cables susceptible to wind-induced and rain-wind vibrations. Per EN 1993-1-11 Section 6.3 and PTI Guide Specification: (1) Internal dampers — viscous dampers installed inside the cable anchorage, typically oil-filled, providing 0.5-2.0% critical damping. For a 100 mm diameter cable with 50 m length, a damper with C = 10 kN·s/m provides approximately 1.5% critical damping. (2) External dampers — mounted on the cable near the deck anchorage (typically at 2-5% of cable length from the anchorage). Maximum damping efficiency when installed at 4-5% of cable length. (3) Cross-tie cables — secondary cables connecting adjacent stay cables, creating cable networks that increase system damping. A cross-tie at mid-span between two 60 m cables increases damping from 0.2% to 1.0-1.5%. (4) Active control systems — magnetorheological (MR) dampers for long cables > 100 m, where passive dampers are insufficient. MR dampers can achieve 3-5% critical damping with active control.
Cable saddle and tower top design. At the tower of a cable-stayed or suspension bridge, the cable passes over a saddle that transfers the cable force to the tower. Per EN 1993-1-11 Clause 6.5 and PTI Recommendations: (1) Saddle radius — minimum radius = 30× cable diameter for locked coil strand and 40× for spiral strand, to limit bending stress in the wires. For a 100 mm diameter cable: minimum saddle radius = 3,000-4,000 mm. (2) Friction between cable and saddle — the friction coefficient μ = 0.05-0.15 for greased saddles and 0.30-0.50 for ungreased saddles. (3) Out-of-balance force — the difference in cable tension on either side of the saddle must be resisted by friction or by mechanical anchors. Maximum out-of-balance force ΔT = T1 - T2 = T1(1 - e^(-μθ)) where θ is the wrap angle in radians. (4) Fatigue at the saddle — wires at the saddle exit point experience bending stress concentration. Per test data, the fatigue life at the saddle can be 30-50% lower than the free cable.
Cable replacement design. Per EN 1993-1-11 Clause 6.4, structures must be designed to permit cable replacement. Design considerations: (1) Adjacent cables must carry the load from the removed cable with a load factor of 1.35. (2) Temporary jacking points required at anchorages. (3) For a 100-cable bridge, the redundancy check assumes any single cable can be removed without collapse. (4) Replacement tensioning sequence must limit deck deflection to a maximum of L/500 during cable change-out.
Use the cable sag calculator to compute cable forces and deflections for various cable configurations, and the beam capacity calculator for the deck girder elements between cable supports.
Related pages
Disclaimer (educational use only)
This page is provided for general technical information and educational use only. It does not constitute professional engineering advice. All results must be independently verified by a licensed Professional Engineer.