Inconel 625 Fatigue Strength: S-N Curve for Riser Use

For welded riser girth butt joints, IIW recommends FAT 90 in air (mean S-N slope m = 3) and FAT 71 in seawater under cathodic protection; DNVGL-RP-F112 uses FAT 90 for surface and FAT 71 for free-corrosion plus CP correction.
Mean stress matters - the Goodman relation drops the allowable alternating stress by roughly 25–35 % when the mean stress approaches 50 % of UTS, which is typical of pressurised production risers.
Cathodic protection in seawater (typically -800 to -1050 mV vs Ag/AgCl) reduces the fatigue endurance of nickel alloys by 5–15 % versus free corrosion at low frequencies; the DNV k-factor of 1.30–1.45 for CP under HISC is conservative for SCRs.
Vortex-induced vibration (VIV) drives the dominant fatigue load on SCRs in 0.5–2 m/s currents; design amplitude is normally capped at ±0.1·D (diameter) over 10^7–10^8 cycles.

Inconel 625 Fatigue Strength

Define the load spectrum. Build a long-term wave-current load histogram for the riser location (Hs, Tp, current profile). Extract stress-range bins per unit year from spectral fatigue analysis (Rainflow counting).
Select the S-N curve. Choose base-material or welded-joint S-N curve per DNVGL-RP-F112 / ASME VIII-2: base metal FAT 90 in air, girth butt weld FAT 90 in air, FAT 71 in seawater with CP; slope m = 3 for N < 10^7, m = 5 for 10^7 ≤ N < 10^8, knee point at 10^7 or 10^8 cycles.
Apply mean-stress and environmental correction. Use Goodman for mean stress in the air curve, then apply DNV environmental factor E (1.0 free corrosion, 0.5 in seawater with CP for nickel-base welds).
Run Miner-Palmgren fatigue damage summation. D = Σ (n_i / N_i). Target Design Fatigue Factor (DFF) of 3 for survival + 10 for inspection-accessible risers, DFF ≥ 10 for non-accessible SCRs per DNV-OS-F201.
Validate with ECA or fracture-mechanics check. For welded girth joints, perform an Effective Notch Stress (ENS) or BS 7910 / API 579 fitness-for-service assessment for the largest expected weld defect.

Inconel 625 (UNS N06625) is a nickel-chromium-molybdenum-niobium alloy whose annealed S-N curve, welded-joint FAT class, and corrosion-fatigue behaviour in seawater make it the workhorse riser material for deep-water, sour and high-chloride fields. This article gives the design S-N data, the standards (DNVGL-RP-F112, API 2RD, ASME VIII-2, IIW FAT classes), the environmental corrections, and the ECA / fracture-mechanics rules needed to qualify a 625 SCR, TTR or hybrid riser for 25+ years of service.

In the mill-annealed condition (ASTM B 443 Grade 1), Inconel 625 has a bi-linear S-N curve in air with a 10^7 endurance limit of about 290–345 MPa at R = -1 and a flat cut-off beyond 10^7–10^8 cycles, while welded girth joints drop to IIW FAT 90 in air and FAT 71 in seawater under cathodic protection.

The nickel-chromium matrix plus 8–10 % molybdenum and 3.15–4.15 % niobium gives Inconel 625 a stable austenitic grain structure that resists cyclic slip-band formation - the microstructural mechanism behind high-cycle fatigue in austenitic alloys.

Independent rotating-beam data published in the Special Metals technical bulletin SMC-061 show a 10^7 cycles fatigue strength of 290 MPa (annealed plate, R = -1) and 345 MPa (cold-drawn bar, R = -1); both datasets flatten beyond 10^7, giving a true endurance limit. For welded girth joints - the fatigue-critical location in any riser - IIW assigns FAT 90 in air and FAT 71 in seawater with cathodic protection, with slope m = 3 for N < 10^7 and m = 5 for 10^7 ≤ N < 10^8, as adopted by DNVGL-RP-F112 for subsea metallic risers. Designers therefore use the FAT-class S-N curve as the baseline and apply environmental and mean-stress corrections on top.

How Does Inconel 625 Compare to Other Alloys in Cyclic Loading for Riser Service

Among common riser alloys Inconel 625 ranks at the top of the HCF table: roughly 30 % above 316L stainless, 15 % above 22 % Cr duplex, and equal to or slightly below Inconel 718 in fatigue strength, but with far better weldability and corrosion-fatigue margin.

Selecting a riser alloy is a fatigue-vs-weldability-vs-corrosion trade-off. Inconel 625 wins on the combination of fatigue strength, weldability and seawater resistance, which is why it is the standard for deep-water SCRs in fields like Cascade / Chinook, Stones, Julia, and many Brazilian pre-salt developments. Duplex 2205 / 2507 is competitive on raw fatigue strength but suffers in the girth-weld HAZ (precipitation of sigma phase) and is therefore usually limited to flowlines and piping, not main risers.

The governing documents are DNVGL-RP-F112 for S-N curves and DFF rules, DNVGL-ST-F101 and DNV-OS-F201 for design factors and safety class, API 2RD for the loading framework, ASME BPVC Section VIII Division 2 for pressure-boundary fatigue, and IIW / BS 7910 / API 579 for welded-joint assessment.

Each standard covers a different slice of the design. DNVGL-RP-F112 is the most complete riser-specific reference: it provides the FAT-class S-N curves (FAT 90 in air, FAT 71 in seawater with CP), the environmental factor E, the mean-stress correction, and the cut-off. DNVGL-ST-F101 covers the safety class and minimum DFF (3 for inspection-accessible, 10 for non-accessible).

API 2RD defines the wave-current load spectrum, the VIV check, the fatigue damage summation (Miner), and the limit-state acceptance criteria. ASME BPVC Section VIII Division 2 (2023 edition) is invoked when the riser is part of a pressure vessel (e.g. a top-tensioned wellhead riser) and supplies an alternative fatigue curve based on elastic follow-up. For welded girth joints, IIW provides the FAT class, while BS 7910 and API 579 / ASME FFS-1 give the ECA and fracture-mechanics route.

Increasing mean stress dramatically shortens Inconel 625 fatigue life - at R = 0.5 the allowable alternating stress is roughly one-third of the value at R = -1, which is why the Goodman correction is mandatory for pressurised production risers.

The S-N curve is plotted at R = -1 (fully reversed bending). Real risers see bending combined with internal pressure, so the cycle sits between R = 0 (pulsating) and R = 0.5 (typical of a pressurised riser at high internal pressure ratio).

The Goodman relation σ_a,allow = σ_-1 × (1 − σ_m / σ_u) is the standard correction; the Gerber relation is less conservative but allowed in some codes for ductile nickel-base alloys. For a production riser operating at 250 bar internal pressure on a 12″ schedule-80 pipe (σ_m ≈ 200 MPa on a 827 MPa UTS), the Goodman penalty is roughly (1 − 0.24) = 0.76, so the effective allowable stress range drops by about 24 % compared with R = -1.

At 10⁸ cycles R = -1 in air, annealed Inconel 625 retains roughly 250–280 MPa (36–40 ksi) of alternating stress capacity, which is within 10 % of its 10⁷ value; this is the basis of the 10⁸ cycles cut-off used in modern riser fatigue design.

Austenitic nickel-base alloys do not show a sharp fatigue limit the way carbon steel does, but the slope of the S-N curve does flatten markedly beyond 10^7 cycles. The Special Metals bulletin and several published riser-fatigue test campaigns show that the 10^8 value is within 8–10 % of the 10^7 value for the annealed alloy. Designers therefore apply a horizontal cut-off at 10^7 (DNVGL-ST-F101 for fatigue-critical SCRs) or 10^8 (DNVGL-RP-F112 default) and do not extrapolate the S-N curve further.

Free-corrosion exposure in seawater reduces Inconel 625 fatigue strength by only about 10–20 % versus air, because the passive Cr-oxide film resists pitting; the bigger penalty comes from cathodic protection, where hydrogen evolution lowers the endurance limit by another 5–15 %.

How Does Seawater Corrosion Influence the S-N Curve of Inconel 625 Risers

In free corrosion the alloy's high Cr (~22 %) and Mo (~9 %) content produces a passive film that suppresses pitting and chloride attack. Slow-frequency rotating-beam tests in natural seawater at 4–30 °C typically show a 10^7 cycles fatigue strength of 250–280 MPa - a 10–20 % drop versus air. Under cathodic protection the surface evolves hydrogen, some of which is absorbed into the alloy.

At low strain rates and low frequencies this hydrogen can diffuse to persistent slip bands and assist crack initiation, lowering the 10^7 strength to about 220–250 MPa. The DNVGL-RP-F112 E-factor captures this: E = 1.0 for free corrosion, 0.5 for seawater with CP, 0.25 for HISC-susceptible conditions.

Cathodic protection reduces the 10^7 endurance of Inconel 625 by 5–15 % versus free corrosion at low frequencies, and is captured in design by a DNVGL-RP-F112 environmental factor E = 0.5; over-protection (< -1050 mV) must be controlled to avoid HISC.

Cathodic hydrogen embrittlement is frequency- and potential-dependent. At R = -1 and 1 Hz, the loss is modest (≈ 5 %); at 0.01 Hz (a typical SCR VIV frequency in low current) the loss grows toward 15 %. The risk increases at potentials more negative than -1050 mV, where excessive hydrogen is generated. For deep-water SCRs the design recommendation is to specify Al-Zn-In anodes sized for -850 to -950 mV at end-of-life, install reference electrodes and alarm on over-protection, and avoid CP coupling to carbon-steel components that drag the potential negative.

A girth butt weld in Inconel 625 drops the design S-N curve from FAT 90 (base-metal-equivalent) to FAT 90 in air and FAT 71 in seawater with CP; partial-penetration or root-defective welds drop further to FAT 63.

The weld toe introduces a notch and tensile residual stress, both of which promote crack initiation. The matching filler ERNiCrMo-3 (Inconel 625 / AWS A5.14) keeps the HAZ in the solution-annealed condition, so the welded-joint fatigue strength is dominated by geometry, not metallurgical mismatch. Best-practice to recover the FAT 90 curve includes full-penetration butt with cap flush-ground, TIG root + automated TIG/MIG fill, post-weld solution anneal at 950–1050 °C followed by water quench, dye-pen or MPI for toe geometry, and shot-peening for compressive residual stress.

Between -20 °C and 200 °C the fatigue strength of annealed Inconel 625 changes less than ±10 %, which lets one S-N curve cover cold SCR service (4 °C) and warm production risers (60–120 °C); above 600 °C the curve drops sharply.

Inconel 625 is solid-solution-strengthened by Mo and Nb; it does not rely on γ′ or γ″ precipitates for strength, so it has no ductility trough in the sub-zero range. This is one of the key reasons it is chosen for deep-water cold-service SCRs and for LNG rundown lines. As service temperature climbs above 400 °C the oxide film becomes less protective and creep begins to interact with fatigue; above 600 °C the fatigue-creep interaction must be assessed using the strain-range partitioning or the unified creep-fatigue rules of ASME III / RCC-MR. For typical riser service (≤ 120 °C) the temperature correction is small and may be ignored.

VIV is the dominant fatigue load on most SCRs and TTRs; the DNV amplitude cap of about 0.1·D for bare risers (0.05·D with strakes) combined with FAT 71 in seawater drives the S-N damage, not the wave load alone.

How Does Vortex-Induced Vibration VIV Drive Riser Fatigue Design

When current flows across a slender cylinder, alternating vortices shed at a Strouhal frequency close to the natural bending frequency of the riser, locking in and producing sustained cross-flow oscillation. For a deep-water SCR in 0.5–2 m/s current, this can deliver 10^7–10^8 stress cycles per year in the 10–100 MPa range - typically an order of magnitude more damaging than the wave-induced bending. Modern SCR design therefore applies VIV suppression by helical strakes, VIV monitoring with accelerometer clusters, real-time rainflow counting and damage update, and DFF = 3 (or 10 for non-accessible SCRs) on the cumulative damage.

Inconel 625 girth butt welds with matching ERNiCrMo-3 filler qualify for IIW FAT 90 in air and FAT 71 in seawater with cathodic protection; partial-penetration and fillet welds drop to FAT 63 and FAT 71 respectively.

The IIW FAT system ties the design S-N curve to a reference stress at 2×10^6 cycles and a slope of m = 3 up to 10^7 cycles, with a slope of m = 5 from 10^7 to 10^8 and a horizontal cut-off beyond. FAT 90 → 90 MPa at 2×10^6; FAT 71 → 71 MPa; FAT 63 → 63 MPa. For Inconel 625 the FAT 90 class is well-supported by the test data in air and is the same class assigned to austenitic stainless steels. In seawater with cathodic protection the FAT drops to 71; this is the value used in DNVGL-RP-F112.

Surface finish can shift the Inconel 625 endurance limit by ±25 %; shot-peened or stress-relieved surfaces deliver the upper bound, while rough as-welded or hot-rolled surfaces deliver the lower bound.

Fatigue cracks initiate at the surface. Surface defects, machining marks, scale and weld undercut all act as notches that lower the allowable stress amplitude. Standard practice for marine risers is to machine or grind the cap of girth welds flush, shot-peen the pipe body to induce a 300–400 MPa compressive residual stress at the surface, and avoid sharp stamp marks and tool marks. Shot-peening is one of the few process improvements that actually moves the FAT-class curve upward by up to one class - meaning FAT 90 becomes effectively FAT 100 on the peened surface.

Thicker walls reduce bending stress for a given bending moment but raise the SCF at girth welds and the through-thickness stress at defects; the typical riser D/t ratio of 20–40 balances VIV response, fatigue life and defect tolerance.

The bending stress in a pipe scales as σ = M·c / I, where c is the outer radius and I is the moment of inertia, so thicker walls lower bending stress. However, a thicker wall also concentrates the strain at the weld toe and increases the constraint factor, raising the effective SCF. DNVGL-ST-F101 imposes a thickness correction on the FAT curve for t > 25 mm: allowable stress scales as (25 / t)^0.1. For a 30 mm wall this is a 2 % derate; for 50 mm it is 7 %.

The allowable stress range is the FAT-class value × Goodman factor × environmental factor ÷ DFF; for a typical production riser this lands at roughly 70–95 MPa, depending on mean stress and DFF.

Worked example for a 12″ Schedule 80 (273 mm OD × 21 mm wall) Inconel 625 production riser, internal pressure 250 bar, DFF = 3, seawater + CP: FAT 71 at 10^7 cycles gives 71 × (10^7/2×10^6)^(1/3) = 121 MPa alternating. Hoop stress σ_m = p·D / 2t ≈ 163 MPa. Goodman factor = 1 − 163 / 827 = 0.80. E-factor = 0.5 (seawater with CP). Allowable ΔS = 121 × 0.80 × 0.5 / 3 = 16 MPa. This very low number reflects the conservatism of DFF = 3 + E = 0.5; in practice the design stress range is dominated by VIV, not pressure.

Inconel 625 has roughly 30 % higher fatigue strength than 316L, an IIW FAT class one step higher (FAT 90 vs FAT 63 in seawater), far better chloride pitting resistance, and is the standard for deep-water SCR service.

316L is a workhorse material in chemical and marine service but its fatigue strength at 10^7 cycles is around 240 MPa - 30 % below Inconel 625 - and its welded-joint FAT class in seawater drops to FAT 63. In warm (≥ 30 °C) or high-chloride service 316L is also susceptible to pitting and crevice corrosion, which become initiation sites for corrosion-fatigue cracks. For deep-water SCRs in 4 °C seawater, 316L has been used historically but is increasingly replaced by Inconel 625 because the higher allowable stress range lets the wall be thinner, lighter and easier to handle.

Parameter	Recommended value	Source
Base-metal S-N (R = -1, 10^7, air)	≈ 290–345 MPa	Special Metals SMC-061
Welded girth FAT in air	FAT 90	IIW / DNVGL-RP-F112
Welded girth FAT in seawater + CP	FAT 71	IIW / DNVGL-RP-F112
S-N slope (N < 10^7)	m = 3	DNVGL-RP-F112
S-N slope (10^7 ≤ N < 10^8)	m = 5	DNVGL-RP-F112
Cut-off	10^8	DNVGL-RP-F112
Environmental factor E	1.0 (air) / 0.5 (seawater + CP)	DNVGL-RP-F112
Mean-stress correction	Goodman (mandatory)	IIW
Design Fatigue Factor	3 (inspection-accessible) / 10 (non-accessible)	DNV-OS-F201
VIV amplitude cap (bare)	0.1·D	DNVGL-RP-F112 / API 2RD
CP design window	-800 to -1050 mV vs Ag/AgCl	DNV-RP-B401

In the mill-annealed condition (ASTM B 443 Grade 1, ~950 °C / water quench), Inconel 625 shows a Basquin-type S-N curve with rotating-beam fatigue strength at 10^7 cycles of roughly 290–345 MPa (42–50 ksi) at R = -1. Between 10^6 and 10^7 cycles the slope is close to m = 8, then it flattens with a knee point near 10^7–10^8 cycles, giving a practical endurance limit of about 280 MPa (40 ksi).

For girth butt welds in Inconel 625 (matching filler ERNiCrMo-3, e.g. Inconel 625 / AWS A5.14), IIW assigns FAT 90 in air (S-N slope m = 3, reference stress 90 MPa at 2×10^6 cycles). In seawater with cathodic protection the curve drops to FAT 71; for root defects or partial-penetration splices FAT 63 is conservative.

Seawater reduces the fatigue life of nickel-base alloys only modestly (10–20 %) under free corrosion, because the alloy's Cr / Mo / N content keeps it passive. The bigger drop comes under cathodic protection, where hydrogen evolution at -800 to -1050 mV (Ag/AgCl) can reduce the endurance limit by another 5–15 % at low frequency (< 0.1 Hz).

Increasing R-ratio (from -1 toward +0.5) raises the mean stress and shortens life. Apply the Goodman correction: σ_a,allow = σ_-1 × (1 - σ_m / σ_u). For R = 0 (pulsating), the allowable alternation is roughly half the R = -1 value; for R = 0.5 it drops to about a third. This matters for pressurised production risers.

Vortex-induced vibration from cross-flow current produces millions of stress cycles per year at frequencies of 0.1–1 Hz. The dominant design rule is the DNV VIV amplitude cap of about 0.1·D (diameter) for bare risers and 0.05·D with helical strakes, combined with the FAT 90 / FAT 71 S-N curve. VIV suppression devices and fatigue monitoring are usually required for deep-water SCRs.

316L has a rotating-beam fatigue limit near 240 MPa at 10^7 cycles, ~30 % below Inconel 625; its welded-joint FAT class is only 63. In warm (≥ 30 °C) or highly-chloride service, 316L is also susceptible to chloride pitting and corrosion fatigue. Inconel 625 keeps its strength and corrosion-fatigue margin to > 1000 m depth / 4 °C seawater.