Fatigue Design Guide

Cyclic loads.
Endurance limits.
Design for infinity.

Q: Does aluminum have endurance limit?

Aluminum alloys do not have true endurance limit — S-N curve continues declining even at very high cycles. "Fatigue strength" for aluminum typically specified at 10^7 or 10^8 cycles as design value. For aerospace aluminum (7075-T6), typical design stress 100-150 MPa at 10^8 cycles. For infinite life, use factor of safety on this value.

Q: Shot peening benefit?

Shot peening creates compressive residual stress in surface layer (typically 0.1-0.5 mm deep). Benefits: (1) 2-3× improved fatigue life in high-cycle fatigue. (2) Crack initiation delayed — stress must overcome compressive residual before crack can start. (3) Improved stress corrosion cracking resistance. Cost: $10-50 per part depending on geometry. For fatigue-critical parts (springs, aerospace, automotive), routine shot peening is standard.

Q: Fatigue inspection?

For critical parts in service: periodic NDT inspection for cracks. Eddy current (near-surface), ultrasonic (subsurface), dye penetrant (surface). Inspection interval set by fracture mechanics — time to grow from detectable crack size to critical size. For aerospace: mandatory fatigue inspections at specified intervals. For commercial: inspect at maintenance shutdowns.

Q: What's better: rolled or cut threads for fatigue?

Rolled threads significantly better for fatigue. Thread rolling creates compressive residual stress in thread roots, improved fatigue life 2-4× vs cut threads. Also better surface finish in thread roots. For high-cycle fatigue applications (aerospace fasteners, engine parts), rolled threads specified. For general industrial, cut threads adequate. Thread-rolling equipment cost higher but part life dramatically better.

Fatigue causes more structural failures than any other mechanism. A part can survive a million static loads but fail after 100,000 cycles. This guide covers the essentials of designing for fatigue life.

Free DFM Review Material selection

01 · Fatigue basics

How parts fail cyclically.

Under repeated loading, cracks nucleate at stress concentrations, propagate slowly, then accelerate to failure. Total life is dominated by crack initiation — surface quality matters enormously.

Initiation

60-90% of life

Crack nucleation at surface defects, scratches, inclusions, or stress concentration features. Most of fatigue life is spent here. Surface finish dramatically affects this phase.

Propagation

Stable growth

Crack grows one increment per cycle — predictable by fracture mechanics. Paris' law: da/dN = C×(ΔK)^m. Design inspections during this phase catch cracks before failure.

Final fracture

Rapid

When crack reaches critical size, remaining ligament fails statically. Last cycle failure usually sudden and catastrophic.

Endurance limit

Infinite life

For ferrous metals: stress below which infinite cycles without failure. Typically 40-50% of UTS for steel. Non-ferrous metals often have no true endurance limit — design to specific life (10^7 cycles typical).

Stress ratio R

Min/max

R = σmin/σmax. R = -1: fully reversed loading. R = 0: zero to peak. R = +1: static load. Fatigue life depends on both amplitude and mean stress.

Mean stress effect

Goodman diagram

Tensile mean stress reduces fatigue life. Compressive mean stress increases life. Goodman/Soderberg/Gerber diagrams quantify this for design.

02 · Typical values

Endurance limits.

Approximate fatigue strengths for common materials at 10^7 cycles, R = -1 (fully reversed), polished specimens. De-rate for real conditions.

Material	UTS (MPa)	Endurance limit (MPa)	Ratio
Mild steel 1018	440	205	0.47
Medium carbon 1045	630	290	0.46
4140 Q&T	1000	450	0.45
4340 Q&T	1100	495	0.45
17-4 PH (H1025)	1070	480	0.45
304 stainless	505	240	0.47
316L stainless	485	220	0.45
Aluminum 2024-T3	485	140 (10^8)	0.29
Aluminum 6061-T6	310	95 (10^8)	0.31
Aluminum 7075-T6	572	160 (10^8)	0.28
Ti-6Al-4V	900	500	0.55
Inconel 718	1241	620	0.50

03 · Stress concentration

The #1 cause of fatigue failure.

Sharp corners, holes, notches concentrate stress. Local stress can be 2-5× nominal. Fatigue cracks always start at stress concentrations.

Kt = 2-3

Small holes

Transverse hole in tension-loaded member. Kt typically 2.5 for small hole. Minimize hole size and round corners.

Kt = 1.5-2

Fillet radius

Shoulder transition. Kt depends on ratio of fillet radius to section change. Generous fillet (r/D > 0.2) gives low Kt.

Kt = 3+

Sharp corners

Internal square corners. Kt can exceed 3. Always add fillet — improves fatigue life 2-3×.

Kt = 2-4

Threads

Thread roots concentrate stress. Rolled threads (cold formed) have residual compression in roots — better fatigue than cut threads.

Kt = 3-5

Keyways

Keyway corners are severe stress concentrations. Use splines instead of keyways for high-cycle fatigue applications.

Kt = 2

Welds

Weld toes and roots. Post-weld heat treatment and grinding improve fatigue life significantly.

04 · Design practice

Fatigue design rules.

Surface finish: Rough surfaces reduce fatigue life significantly. Polished surfaces can have 30% better fatigue life than machined. Corrosion pits can reduce fatigue life 50%+. For critical fatigue parts, specify Ra 0.4 µm or better, inspect for surface defects, consider shot peening.

Residual stresses: Compressive residual stress in surface layer improves fatigue dramatically. Shot peening produces 300-600 MPa compressive residual at surface — can double fatigue life. Cold forming (rolling threads, cold drawing, burnishing) also creates beneficial residual stress. Hot-worked or machined surfaces often have tensile residual stress — reduces fatigue life.

Material selection: For fatigue, don't automatically pick highest UTS. Higher strength often means less ductility and less tolerance to stress concentrations. 4140 Q&T at 40 HRC often gives better fatigue life than 4340 Q&T at 50 HRC. Test for your specific loading. For aerospace: Ti Gr.5 has excellent fatigue life and damage tolerance.

Goodman diagram: For mean stress different from R=-1: Modified Goodman: σa/Se + σm/Sut ≤ 1/FS. σa = stress amplitude, σm = mean stress, Se = endurance limit, Sut = ultimate tensile. For FS = 2, keep operating point well inside the diagram. Goodman is the most common mean stress correction, slightly conservative.

Testing vs calculation: Fatigue analysis has inherent scatter — factor of 2-10 variation in life for same nominal stress is normal. For critical parts, calculate conservatively AND test. Accelerated testing: higher stress amplitudes to compress testing time, then extrapolate to service conditions. Finite element analysis with fatigue post-processor (nCode, FEMFAT) common for complex geometry.

FAQ

What safety factor for fatigue?

Typical factors of safety on fatigue: 1.5 for well-characterized loading and proven material (aerospace production). 2.0 for typical structural (industrial machinery). 3.0 for uncertain loads or critical safety. 4.0+ for life-critical or inspection-limited parts. Apply SF to stress, not to life — fatigue life is highly sensitive to stress (S^5 to S^7 relationship). Small stress reduction gives large life improvement.

High-cycle vs low-cycle fatigue?

High-cycle fatigue: >10^5 cycles, stress below yield, elastic behavior. Design with S-N curves and endurance limits. Typical applications: rotating shafts, vibrating components, wind turbine parts. Low-cycle fatigue: <10^4 cycles, stress above yield, plastic strain. Design with strain-life approach (ε-N curves). Typical applications: pressure vessels cycled through pressure, thermal cycling parts, ductile failure analysis.

Does aluminum have endurance limit?

Aluminum alloys do not have true endurance limit — S-N curve continues declining even at very high cycles. "Fatigue strength" for aluminum typically specified at 10^7 or 10^8 cycles as design value. For aerospace aluminum (7075-T6), typical design stress 100-150 MPa at 10^8 cycles. For infinite life, use factor of safety on this value.

Shot peening benefit?

Shot peening creates compressive residual stress in surface layer (typically 0.1-0.5 mm deep). Benefits: (1) 2-3× improved fatigue life in high-cycle fatigue. (2) Crack initiation delayed — stress must overcome compressive residual before crack can start. (3) Improved stress corrosion cracking resistance. Cost: $10-50 per part depending on geometry. For fatigue-critical parts (springs, aerospace, automotive), routine shot peening is standard.

Fatigue inspection?

For critical parts in service: periodic NDT inspection for cracks. Eddy current (near-surface), ultrasonic (subsurface), dye penetrant (surface). Inspection interval set by fracture mechanics — time to grow from detectable crack size to critical size. For aerospace: mandatory fatigue inspections at specified intervals. For commercial: inspect at maintenance shutdowns.

What's better: rolled or cut threads for fatigue?

Rolled threads significantly better for fatigue. Thread rolling creates compressive residual stress in thread roots, improved fatigue life 2-4× vs cut threads. Also better surface finish in thread roots. For high-cycle fatigue applications (aerospace fasteners, engine parts), rolled threads specified. For general industrial, cut threads adequate. Thread-rolling equipment cost higher but part life dramatically better.