UK Steel Fatigue Assessment -- EN 1993-1-9 Fatigue Design of Steel Structures

Fatigue is the progressive, localised, and permanent structural damage that occurs when a material is subjected to repeated or fluctuating loads at stress levels below the material's static yield strength. In steel structures, fatigue manifests as the initiation and propagation of cracks at stress concentrations -- typically at weld toes, bolt holes, cope holes, and abrupt changes in section. BS EN 1993-1-9 provides the framework for assessing fatigue resistance in steel structures, applicable whenever a member or detail is subjected to repeated loading. This reference presents the fatigue design method, the detail category system, the S-N curve approach, and a worked example for a crane runway beam.

When Fatigue Assessment Is Required

Fatigue assessment is required per EN 1993-1-9 when:

• The number of stress cycles N during the design life exceeds:
  • 10,000 cycles for details with stress range above the constant amplitude fatigue limit (CAFL)
  • 2 x 10^6 cycles for details subject to variable amplitude loading evaluated by the damage equivalent method

UK steel structures requiring fatigue assessment include:

• Crane runway beams and their connections (every crane passage = 1 cycle)
• Bridge decks and girders (every vehicle passage = 1 cycle)
• Wind-sensitive structures (vortex shedding-induced vibration)
• Machinery support structures (rotating or reciprocating equipment)
• Highway gantries and sign supports (wind buffeting + vehicle-induced gusts)
• Chimneys and masts (wind-induced oscillation)
• Offshore structures (wave loading, typically >10^8 cycles in design life)

Building structures not subject to repeated loading from cranes, machinery, or wind-induced vibration do not require fatigue assessment -- static strength governs.

Detail Categories

The detail category (Delta_sigma_C) is the fatigue strength at 2 x 10^6 cycles for a constant amplitude stress range, expressed in MPa. It is the cornerstone of the EN 1993-1-9 fatigue assessment method. A higher number indicates better fatigue performance.

Common UK Structural Details and Their Categories

A detail category of 160 is effectively fatigue-proof for most building structures (stress range at 2 x 10^6 cycles of 160 MPa is above typical design stress ranges for building dead and imposed loads).

Selecting the Detail Category

The detail category is selected from EN 1993-1-9 Tables 8.1-8.10 based on:

• The geometric configuration of the detail
• The weld type (fillet, butt, partial penetration)
• The welding process (manual, automatic)
• Any post-weld treatment (toe grinding, peening, TIG dressing)
• The inspection standard (visual, MPI, ultrasonic)

Post-weld treatment upgrades the detail category by 1-2 classes (e.g., toe grinding can upgrade a transverse fillet weld from Category 80 to Category 100, a 25% increase in fatigue strength).

S-N Curves (Fatigue Strength Curves)

The fatigue strength of a detail is represented by an S-N curve (stress range vs. number of cycles to failure) on a log-log scale. EN 1993-1-9 defines S-N curves for each detail category:

For N <= 5 x 10^6: Delta_sigma_R^3 x N = Delta_sigma_C^3 x 2 x 10^6

For N > 5 x 10^6 (constant amplitude fatigue limit, CAFL): Delta_sigma_R^5 x N = Delta_sigma_D^5 x 5 x 10^6

Where Delta_sigma_D = (2/5)^(1/3) x Delta_sigma_C = 0.737 x Delta_sigma_C (the CAFL)

The change in slope from m=3 to m=5 at N=5x10^6 reflects the fatigue endurance limit -- below the CAFL, fatigue cracks do not propagate in structural steels under constant amplitude loading. However, for variable amplitude loading, the CAFL does not apply and the m=3 slope continues (cumulative damage rule).

Cut-Off Limit

For constant amplitude loading, the cut-off limit at N = 10^8 cycles is: Delta_sigma_L = (5/100)^(1/5) x Delta_sigma_D = 0.549 x Delta_sigma_D = 0.404 x Delta_sigma_C

Below the cut-off limit, fatigue damage is considered to be zero for constant amplitude loading.

Damage Accumulation -- Palmgren-Miner Rule

For variable amplitude loading (most real structures), fatigue damage from each stress cycle accumulates linearly per the Palmgren-Miner rule:

Dd = SUM(n_i / N_i) <= 1.0

Where n_i is the number of applied cycles at stress range Delta_sigma_i, and N_i is the number of cycles to failure at that stress range from the S-N curve.

In the EN 1993-1-9 implementation, the variable amplitude loading is converted to an equivalent constant amplitude stress range at 2 x 10^6 cycles:

Delta_sigma_E,2 = lambda x Delta_sigma_max

Where lambda is the damage equivalent factor that accounts for:

• The stress spectrum shape (the distribution of stress ranges over the design life)
• The S-N curve slope (m=3 for the damage calculation)
• The total number of cycles N during the design life

EN 1993-1-9 Annex A provides lambda values for cranes and bridges. For a building crane with a known usage class, lambda values range from 0.5 (light intermittent use, 50,000 cycles) to 1.0 (heavy continuous use, 2 x 10^6 cycles).

Fatigue Assessment Procedure

Safe-Life Method

1. Identify the governing detail category for each potential fatigue-critical location
2. Determine the design stress range Delta_sigma_E,2 from the frequent load combination
3. Calculate the damage equivalent factor lambda from EN 1993-1-9 Annex A (for cranes or bridges) or from first principles (for other structures)
4. Verify: Delta_sigma_E,2 / (Delta_sigma_C / gamma_Mf) <= 1.0

Where gamma_Mf is the partial factor for fatigue:

• gamma_Mf = 1.00 for safe-life method with low consequences of failure (most building structures)
• gamma_Mf = 1.15 for safe-life method with high consequences of failure
• gamma_Mf = 1.35 for damage-tolerant method

Damage-Tolerant Method

The damage-tolerant method recognises that fatigue cracks may initiate during the design life but will be detected by periodic inspections before reaching critical size. This requires:

• A defined inspection programme (intervals, methods, acceptance criteria)
• Redundancy in the structure so that failure of one element does not cause progressive collapse
• The residual life after crack detection must exceed the inspection interval

The damage-tolerant method allows a higher design stress range (lower gamma_Mf) at the cost of a mandatory inspection regime.

Worked Example -- Crane Runway Beam Fatigue Assessment

A Class S3 overhead crane runway beam in a UK steel stockholder's warehouse has:

• Crane class S3 (approximately 250,000 stress cycles in 50 years per EN 1991-3)
• Crane capacity: 100 kN, self-weight + crab: 105 kN
• Runway beam: 457 x 191 x 82 UB in S355J2, simply supported span 7.5 m
• Rail: 50 kg/m clipped to top flange with discontinuous fillet welds (Detail Category 160)
• Maximum static bending moment per wheel pass: 315 kN.m
• Section modulus Wy: 1,830 x 10^3 mm^3

Step 1 -- Frequent load combination for fatigue: phi_1,fat = 1.0 (vibration factor for fatigue = 1.0) phi_2,fat = 1.0 (hoisting factor for fatigue = 1.0, average load assumption) Frequent wheel load Qfat = 1.0 x 105/4 + 1.0 x 100/4 = 26.25 + 25.0 = 51.25 kN per wheel (compared to ULS wheel load of 58 kN)

Step 2 -- Stress range under frequent load: Delta_sigma = 2 x 51.25 / 58 x 315 x 10^6 / (1,830 x 10^3) = 2 x 0.884 x 172.1 = 304.2 MPa

Wait -- for a crane runway beam, the stress range is from the unloaded condition to the loaded condition for each wheel passage. If the beam self-weight is negligible compared to the crane load (reasonable for a 457 UB at 7.5 m span), the stress range per passage is approximately:

Delta_sigma = Mfat / Wy = (2 x 51.25 x 2.85) x 10^6 / 1,830,000 = 159.6 MPa

Step 3 -- Damage equivalent factor: For Class S3, EN 1993-6 Annex A gives lambda = 0.75 (approximate, depends on stress spectrum) Delta_sigma_E,2 = 0.75 x 159.6 = 119.7 MPa

Step 4 -- Fatigue verification: Detail Category 160 for clipped rail attachment. gamma_Mf = 1.00 (safe-life, low consequence) Delta_sigma_C / gamma_Mf = 160 / 1.00 = 160 MPa Utilisation = 119.7 / 160 = 0.75 -- OK.

The runway beam has adequate fatigue life at the rail attachment detail. However, note that the beam copes at the end connections (Detail Category 71) and any welded attachments (cleats, stoppers) in the tension zone must also be verified.

Design Resources

• UK Crane Runway Beam Design — Complete crane beam to EN 1993-6
• UK Weld Design Guide — Weld sizing and detail categories
• UK Portal Frame Design — Portal frames with crane loads
• UK Bolt Capacity Tables — Bolted connection fatigue (Category F bolts)
• UK Steel Design EN 1993 Guide — Complete Eurocode 3 overview

Frequently Asked Questions

When can I ignore fatigue in UK building design?

Fatigue assessment is NOT required for most building structures. EN 1993-1-9 allows fatigue to be deemed satisfied without calculation when: the member is not subject to repeated loading from cranes, vibrating machinery, or wind-induced oscillations; the number of stress cycles is below 10,000 over the design life; or the maximum stress range is below 26 MPa for Detail Category 160 (the CAFL cut-off limit). For a typical UK multi-storey office or residential building, none of the structural steel elements require fatigue verification.

What post-weld treatments improve fatigue life?

Toe grinding (grinding the weld toe profile to a smooth radius, removing slag intrusions and undercut) upgrades the detail category by 1-2 classes. TIG dressing (remelting the weld toe with a TIG torch to create a smooth profile) achieves a similar upgrade. Hammer or needle peening (cold-working the weld toe to introduce compressive residual stress) upgrades by 2-3 classes for details loaded in tension. Ultrasonic impact treatment (UIT) is the most effective post-weld treatment, potentially upgrading Category 71 to Category 125 or higher for certain details.

How does the UK NA modify EN 1993-1-9?

The UK NA adopts EN 1993-1-9 without significant modification. The partial factors gamma_Mf are adopted as recommended (1.00 safe-life, 1.15 safe-life high consequence, 1.35 damage-tolerant). The UK NA does not add any national-determined parameters beyond those in the main standard. UK practice follows the SCI publication "Fatigue Design of Steel Bridges" (P352) for bridge applications, which provides UK-specific guidance on the damage equivalent factors for highway loading.

What is the difference between the safe-life and damage-tolerant methods?

Safe-life: Design to prevent fatigue crack initiation within the design life. Requires a lower design stress range (more conservative) but no mandatory inspection regime. Used for building cranes where periodic inspection of runway beams is impractical once the building is operational. Damage-tolerant: Accept that cracks may initiate but will be detected before failure by scheduled inspections. Allows a higher design stress range but requires an inspection plan, redundancy in the structure, and a defined crack detection and repair protocol. Used for bridges and offshore structures where access for inspection is designed in.

Educational reference only. All design values are per BS EN 1993-1-9:2005 + UK National Annex. Fatigue assessment must be based on the actual stress spectrum for the specific structure, not generic values. Designs must be independently verified by a Chartered Structural Engineer registered with the Institution of Structural Engineers (IStructE) or the Institution of Civil Engineers (ICE). Results are PRELIMINARY -- NOT FOR CONSTRUCTION without independent professional verification.