DNV-RP-C203 Fatigue Analysis: S-N Curves, SCF Derivation, DFF Selection, and the Mistakes That Fail Reviews

1. Why Fatigue Governs Many Offshore Structures

Offshore structures experience cyclic loading continuously throughout their design life. Wave action applies tens of millions of stress cycles over a 20–30 year life. Even when peak stresses are well within yield, the cumulative effect of repeated loading at welded connections can initiate and propagate cracks — ultimately leading to failure at stress levels that a static analysis would flag as acceptable.

For tubular jacket structures, semi-submersibles, and offshore cranes, fatigue frequently governs member and joint sizing rather than ultimate strength. A joint that passes every ULS check comfortably can still fail the fatigue limit state if its detail class is wrong, the SCF is underestimated, or the DFF is set too low for its inspection regime.

DNV-RP-C203 is the detailed guidance document for fatigue design calculations in the offshore industry. It provides the S-N curves, SCF parametric equations for tubular joints, hot-spot stress methodology, and DFF requirements that form the basis of virtually every fatigue check on a DNV-governed offshore structure.

2. S-N Curves: Families, Classes, and Environments

An S-N curve defines the relationship between cyclic stress range S and the number of cycles to failure N for a given weld or structural detail. DNV-RP-C203 Section 2 provides a library of S-N curves for different environments and detail classes.

S-N curve equation

Each S-N curve has a bilinear form — slope m₁ = 3 below the knee point (typically at N ≈ 10⁶–10⁷ cycles) and a shallower slope m₂ = 5 above it (low stress cycles contribute less damage per cycle). Using the single-slope approximation for all cycles underestimates cumulative damage at low stress amplitudes.

Detail class naming and resistance ranking

Environment correction

Three environmental S-N curve sets exist in DNV-RP-C203:

• In-air: highest fatigue resistance — applies above waterline or in dry spaces
• Seawater with cathodic protection (CP): approximately factor 2–3 reduction in life vs in-air at the same stress range; applies to submerged details with functioning CP
• Seawater free corrosion (no CP): further reduction — approximately factor 3–4 reduction in life vs in-air; applies where CP is absent or has failed

3. Nominal Stress vs Hot-Spot Stress Method

DNV-RP-C203 supports two primary stress calculation approaches Section 4:

Nominal stress method

The nominal stress is the section stress in the structural member remote from the discontinuity — calculated from section forces using standard beam theory. SCF is then applied explicitly to convert nominal stress to the stress at the weld toe. This method is used for classified joints where the geometry matches a defined detail class closely.

Hot-spot stress method

The hot-spot stress is the structural stress at the weld toe, derived by extrapolating the stress gradient from two reference points on the surface of the detail. It captures the effect of structural geometry (thickness transitions, cutouts, attachments) but not the weld toe notch itself — that is embedded in the T-class S-N curve.

The hot-spot stress method is preferred for complex geometries modelled with FEA. It eliminates the need to classify every weld detail individually when geometry makes classification ambiguous.

4. Stress Concentration Factors (SCF)

For tubular joints, the SCF is the ratio of the hot-spot stress at the weld toe to the nominal stress in the brace:

Parametric equations (Efthymiou)

DNV-RP-C203 Appendix B provides parametric SCF equations for T, Y, X, K, and KT tubular joints, based on the Efthymiou equations. The key non-dimensional parameters are:

For a T-joint under axial brace load, SCF on the chord side typically ranges from 1.5 to 8 depending on joint geometry. Slender chords with high γ values give the highest SCFs — a chord with γ = 30 can have SCF ≈ 5–7 compared to γ = 15 giving SCF ≈ 2–3 for the same β.

5. Design Fatigue Factor (DFF): Inspection Access Classes

The Design Fatigue Factor is a multiplier on the required fatigue life. It accounts for the consequence of failure and the possibility of inspection and repair during service.

DFF selection by inspection access and consequence

The NORSOK N-001 safety class framework also influences DFF. High safety class details (consequence of failure = risk to life or total loss) require higher DFF than normal or low safety class details. When both safety class and inspection access criteria apply, the higher of the two DFF values governs.

6. Miner's Rule and Damage Accumulation

DNV-RP-C203 uses the Palmgren-Miner linear damage rule to accumulate fatigue damage from all stress cycles across the design life §2.3:

The fatigue utilisation check requires:

Sensitivity to stress range distribution

Fatigue damage accumulation is highly sensitive to the shape of the stress range distribution. Because the S-N curve slope is m = 3 (damage ∝ Δσ³), doubling the stress range increases damage by a factor of 8. This makes SCF accuracy critical — an SCF underestimate of 20% translates to a 73% underestimate of fatigue damage at that stress level.

7. Spectral vs Deterministic Fatigue Analysis

Deterministic approach

A single design wave (or a small number of regular waves) is selected, the structural response computed, and the stress range derived as a function of wave height. The stress range distribution is then constructed from the wave scatter diagram by scaling:

Deterministic analysis is acceptable for simple jacket structures where the structural response is near-linear and a few dominant wave directions cover the loading. It is computationally inexpensive but can miss resonance effects and non-linear wave-structure interactions.

Spectral approach

Spectral fatigue analysis treats the sea state as a stochastic process. The wave energy spectrum (JONSWAP, Pierson-Moskowitz — per DNV-RP-C205) is combined with the structural stress transfer function H(ω) to derive the stress response spectrum:

From the stress spectrum, the Rayleigh-distributed stress range probability density is derived, and the damage integral is computed directly over the sea state scatter diagram. Spectral analysis is required when structural resonance occurs near dominant wave frequencies (as in deep-water risers, flexible moorings, and slender topsides).

8. Cross-reference Map

Fatigue analysis under DNV-RP-C203 draws on several other standards for structural safety classes, load inputs, and governing requirements.

9. Common Pitfalls That Fail Fatigue Reviews

S-N curve and detail class errors

• Using class D for all weld details — fillet welds and load-carrying attachments are typically class F or F3; overestimating detail class can underestimate damage by 5–10×
• Applying in-air S-N curves to submerged details — seawater with CP reduces fatigue life by factor 2–3; wrong environment class gives unconservative results
• Ignoring the bilinear S-N curve — using a single slope for the full stress range spectrum underestimates damage contribution from the high-cycle, low-amplitude end of the distribution

SCF and stress derivation errors

• Missing SCF for thickness transitions (chord can/stub welds) — a 25–50% thickness step at a can transition adds an SCF contribution typically 1.1–1.3 that is routinely missed
• Using SCF for axial load only — ignoring IPB and OPB contributions to hot-spot stress; for K-joints with moment loading this can underestimate hot-spot stress by 30–50%
• SCF from parametric equations outside their validity range — the Efthymiou equations have explicit limits on β, γ, τ; extrapolating outside these ranges gives unreliable SCF values

DFF and utilisation errors

• Setting DFF = 1 for all jacket nodes — submerged nodes require DFF ≥ 2 (accessible with in-water inspection) or DFF = 3 if not inspected; using DFF = 1 understates required life by 2–3×
• Not checking DFF against both inspection access AND safety class — forgetting that the more conservative of the two governs; high safety class forces DFF up regardless of accessibility
• CP assumption extending beyond anode design life — using seawater-with-CP curves when CP anode life is shorter than structure life misclassifies the environment for the tail end of service