Machine element life, S-N Wöhler curves, and fatigue guide
High-cycle fatigue analysis, S-N Wöhler curves, Marin modification factors, Goodman failure criterion, and Miner cumulative damage rule.
This guide outlines the analytical fatigue verification of dynamic components such as shafts, bolted joints, welded structures, and springs in accordance with DIN 50100, FKM guidelines, and Shigley mechanical design methodology. For mission-critical aerospace and pressure vessel systems, empirical component testing remains mandatory.
Problem / Objective
Over 80% of mechanical component failures occur not from monotonic static overloads, but from progressive fatigue fracture under cyclic stresses far below the material yield strength (S_y). Relying solely on static safety factors (S_static >= 1.5) leads to:
- Silent micro-crack initiation at stress concentrations (fillets, keyways, weld toes),
- Sudden catastrophic brittle fracture with zero warning or macroscopic plastic deformation,
- Costly downtime and cascading mechanical equipment destruction.
Assumptions
- High Cycle Fatigue (HCF,
N > 10^4cycles) governs, where macroscopic bulk behavior remains within the nominal elastic regime. - The base fatigue limit is derived from rotating beam tests (RR Moore) and corrected via Marin surface, size, and loading factors.
- Mean stress influence on endurance limit is evaluated using the modified Goodman or Soderberg failure criteria.
- Variable amplitude spectrum loads follow the Palmgren-Miner linear cumulative damage hypothesis.
Step by step method
- Calculate Alternating and Mean Stress Components:
From peak cyclical stress levels (
σ_maxandσ_min):
σ_a = fracσ_max - σ_min2, σ_m = fracσ_max + σ_min2
Stress ratio: R = fracσ_minσ_max (Fully reversed: R = -1; zero-to-tension: R = 0).
- Compute Corrected Endurance Limit (
S_e) via Marin Equation: Modify laboratory endurance limit (S_e' ≈ 0.5 * S_ut) to represent actual hardware conditions:
S_e = k_a * k_b * k_c * k_d * k_e * S_e'
k_a: Surface finish factor (ground, machined, hot-rolled, as-forged),k_b: Size modification factor (k_b ≈ 1.24 * d^-0.107ford > 8 mm),k_c: Load factor (1.0 for bending, 0.85 for axial, 0.577 for pure torsion),k_e: Fatigue notch sensitivity factor (k_e = 1 / K_f).
- Determine Fatigue Safety Factor (
n_f) Using Modified Goodman: Using ultimate tensile strengthS_utand corrected fatigue limitS_e:
(σ_a / S_e) + fracσ_mS_ut = (1 / n_f) implies n_f = frac1(σ_a / S_e) + fracσ_mS_ut
Achieving n_f >= 1.3 - 2.0 guarantees infinite life (N >= 10^6 - 10^7 cycles) under uninterrupted constant amplitude cycling.
- Evaluate Cumulative Damage via Miner's Rule:
For multi-stage spectrum loading with applied cycles
n_iat stress amplitudeihaving fatigue lifeN_i:
D = sum_i=1^k (n_i / N_i) <= 1.0
- Industrial Worked Example:
A gearbox input shaft manufactured from quenched & tempered AISI 4140 steel (
S_ut = 900 MPa,S_y = 650 MPa, Marin corrected endurance limitS_e = 360 MPa) operates under alternating bending stresses betweenσ_max = 220 MPaandσ_min = -80 MPa:- Stress amplitude:
σ_a = (220 - (-80) / 2) = mathbf150 MPa - Mean stress:
σ_m = (220 + (-80) / 2) = mathbf70 MPa - Modified Goodman factor of safety:
- Stress amplitude:
(1 / n_f) = (150 / 360) + (70 / 900) = 0.4167 + 0.0778 = 0.4945 implies n_f = mathbf2.02
- Verdict:
n_f = 2.02 > 1.50, confirming safe operation in the infinite life domain.
Common mistakes
- Equating Static Yield Safety with Fatigue Life: A shaft with a static safety factor of 3.0 may exhibit a fatigue safety factor of 0.8 and fracture after only 50,000 cycles.
- Specifying Sharp Fillet Radii at Steps: Increasing the fillet transition radius from
r/d = 0.05to0.15slashes stress concentration by nearly 50%. - Neglecting Surface Roughness in Severe Service: Rough-machined finishes reduce fatigue endurance limits by up to 35% compared to polished specimens.
- Overlooking Environmental Corrosion: In corrosive environments, ferrous alloys exhibit no true fatigue limit (
S_e); the S-N curve drops continuously without a horizontal plateau.
Related calculators
- Shaft Torsion and Stress Analysis
- Bolt Strength and Preload Calculator
- Fillet Weld Sizing
- Bearing Life Calculator (ISO 281)
Fatigue Failure Mechanisms
Fatigue failure proceeds through three distinct stages: crack initiation, crack propagation, and final fracture. Understanding these stages determines which engineering intervention prevents failure.
Stage 1 — Initiation: Begins at a stress concentration — machined notch, corrosion pit, inclusion, or surface scratch. Even below the yield strength, cyclic loading causes local plastic slip along slip planes, forming a persistent slip band. A micro-crack nucleates at the surface or just below it. Shot peening and surface hardening are effective here: they introduce compressive residual stress at the initiation site, suppressing early crack opening.
Stage 2 — Propagation: The crack grows incrementally with each load cycle. The Paris-Erdogan law describes the growth rate:
da/dN = C * (Delta K)^m
Where:
da/dN= crack growth per cycle (m/cycle)Delta K= stress intensity factor range =Y * Delta sigma * sqrt(pi * a)(MPa·√m)C, m= material constants (for steel: C ≈ 6×10^-12, m ≈ 3)
Stage 3 — Fracture: When the remaining cross-section can no longer support peak load, sudden overload fracture occurs. On the fracture surface, beach marks (visible concentric arcs) mark each crack front position during propagation — a definitive indicator of fatigue failure in post-fracture analysis.
S-N Curve Interpretation
The S-N (Wöhler) diagram plots alternating stress amplitude vs. cycles to failure. Three regions:
- Low cycle fatigue (LCF, N < 10^4): Plastic strains dominate; material yields per cycle. Use strain-life (epsilon-N) approach. Relevant for pressure vessels and engine components under infrequent overload.
- High cycle fatigue (HCF, 10^4 < N < 10^7): Elastic behavior with damage accumulation. The S-N curve is valid here; design stress must remain below the curve for the required cycle count.
- Endurance limit region (N > 10^7): For ferrous materials (steels), the S-N curve becomes horizontal — the endurance limit Se. Below Se, infinite life is theoretically achieved. Aluminum alloys do NOT have a true endurance limit — the S-N curve continues to decline, requiring a finite-life approach for all aluminum structural applications.
Endurance limit estimation (Shigley / MIL-HDBK-5):
Se' ≈ 0.5 * Su for Su < 1400 MPa (steel)
Se' ≈ 700 MPa for Su > 1400 MPa
These are material endurance limits. The corrected endurance limit Se = ka * kb * kc * kd * Se' applies Marin factors for surface finish (ka), size (kb), load type (kc), and temperature (kd).
Surface Treatments for Fatigue Life
Post-machining surface condition significantly alters fatigue performance:
| Treatment | Surface Roughness Ra | Fatigue Life Factor | Mechanism |
|---|---|---|---|
| Ground (Ra 0.4 µm) | 0.2–0.4 µm | 1.0 (baseline) | Smooth surface reference |
| Shot Peened | 0.8–2.0 µm (increased) | 1.2–1.5× | Compressive residual stress |
| Nitrided (gas/plasma) | 0.4–0.8 µm | 1.5–2.5× | Hard case + compressive stress |
| Hard Chrome Plated | 0.3–0.8 µm | 0.7–0.9× | Tensile residual, hydrogen embrittlement risk |
| Electroless Nickel | 0.4–0.6 µm | 0.85–1.0× | Slightly beneficial if baked |
Key insight: Grinding produces tensile residual stress in the surface if performed aggressively (burn). This reduces fatigue life below the calculated value. Always specify "no thermal damage" or verify with Barkhausen noise inspection for fatigue-critical components.
Frequently Asked Questions
What happens when a component exceeds the endurance limit for a single cycle? Nothing irreversible if the exceedance is brief and the stress stays below yield. However, the cumulative damage model (Palmgren-Miner, D = Sum of ni/Ni) accumulates damage from each cycle above the limit. If D ≥ 1.0, failure is predicted. For safety-critical parts, design to D ≤ 0.3–0.5 to account for statistical scatter.
Does aluminum have a fatigue endurance limit? No. Unlike steel, aluminum alloys show a continuously declining S-N curve with no horizontal asymptote. Design aluminum structures for a specific life (typically 10^7 or 10^8 cycles) with a defined stress level, never for "infinite life." This is critical for aircraft structures, bicycle frames, and rotating aluminum components.
How can I reduce Kf without changing geometry? Increasing the notch radius (r) reduces Kt — machine a generous fillet radius at all transitions, keyways, and shoulder steps. Additionally, shot peening the notch region reduces effective Kf by introducing compressive residual stress. Specifying a lower surface roughness at the notch (Ra < 0.8 µm) further reduces the surface factor ka, improving the corrected endurance limit Se.
Related standards
- DIN 50100: Testing of Materials - Continuous Vibration Fatigue Testing (Wöhler Test).
- FKM Guideline: Analytical Strength Assessment of Mechanical Components (VDI).
- ISO 1099: Metallic Materials - Fatigue Testing - Axial-Force-Controlled Method.
- Eurocode 3 (EN 1993-1-9): Fatigue Assessment of Steel Structures.
Related engineering links
Calculators, terms, and companion guides that share the same engineering concepts.
Related calculators
Bolt Size & Torque
Calculate stress area and preload from nominal diameter, pitch, and grade.
Tensile Stress
Compute tensile stress from force and cross-section area.
Bending & Deflection
Bending stress and max deflection for a simply supported beam.
Bolt Data Center
Quick access to metric and imperial bolt sizes.
Guides
Welded design safety and cost balance
Engineering evaluation of load path, inspection methods, distortion, and cost when choosing fillet or full-penetration welds. AWS D1.1 and EN 1993 referenced.
Helical Compression Spring Design and Force-Deflection Guide: EN 13906-1
Comprehensive engineering guide on helical compression spring rate k, deflection, Wahl stress correction, solid length, and fatigue limits under EN 13906-1.
Key design 101: parallel keys and mechanical wedges
DIN 6885 shaft-hub parallel keys, shear and bearing crushing stress verification, wedge mechanics, and self-locking conditions.