Fatigue, Damage Tolerance & Aeroelasticity

Structures that last and don't shake themselves apart. Learn fatigue in aircraft (S-N curves, Miner's rule, spectra), the safe-life, fail-safe and damage-tolerance philosophies, fracture mechanics (stress intensity, fracture toughness, Paris law crack growth), inspection intervals and widespread fatigue damage, and aeroelasticity — divergence, control reversal, flutter and gust response — with Python integrating crack growth.

The de Havilland Comet crashes (1954) taught the industry that repeated pressurisation cycles can grow cracks from tiny corners until the fuselage fails. Modern aircraft are designed so that cracks are found long before they become dangerous — damage tolerance.

Aeroelasticity is the interaction of aerodynamic forces with a flexible structure. Wings bend and twist under load, which changes the load — sometimes stabilising, sometimes catastrophically divergent. Flutter, a self-excited vibration, can destroy an aircraft in seconds and must be excluded throughout the flight envelope.

An analogy

Fatigue is like bending a paper clip back and forth: each bend is harmless, but eventually it breaks. Flutter is like a flag flapping in the wind — except the flag is a wing and the flapping grows until something breaks.

Fatigue in aircraft

Every flight applies load cycles: ground–air–ground pressurisation, gusts, manoeuvres, landing impacts, engine vibration. Cracks start at stress concentrations (holes, corners, fastener rows) and grow. Aluminium has no fatigue limit, so life is finite; designers track flight cycles and flight hours.

Design philosophies

Three approaches:

Fracture mechanics

The stress-intensity factor K captures crack size, stress and geometry. When K reaches the material's fracture toughness K_IC the crack runs unstably. The Paris law describes stable growth in between. Materials are chosen for toughness (2024-T3 fuselage skins) as well as strength; residual strength must exceed limit load with the largest crack that could be missed.

Aeroelasticity

Static: divergence (twisting increases lift, which increases twist — sweep-forward wings are prone) and control reversal (aileron twists the flexible wing so much that the roll reverses — at high dynamic pressure). Dynamic: flutter is a coupling of bending and torsion modes that extracts energy from the airflow; it is prevented by stiffness, mass balancing of control surfaces, and certified by ground vibration tests, analysis and flight flutter tests. Gust and buffet response and limit-cycle oscillations are also checked.

Formulas

Worked example: Critical crack length

Given: K_IC = 30 MPa√m, σ_max = 120 MPa, Y = 1

Find: a_c.

Answer: ≈ 19.9 mm

Worked example: Miner's rule

Given: 10⁴ cycles at N = 10⁵, 5×10⁵ cycles at N = 10⁷

Find: Damage.

Answer: D = 0.15 (15% of life used)

Worked example: Flutter margin

Given: V_D = 180 m/s EAS

Find: Required flutter-free speed.

Answer: 207 m/s EAS

Crack growth by the Paris law

import math
# Fatigue crack growth (Paris law) in an aluminium skin: da/dN = C (dK)^m
C, m_ = 1e-11, 3.0             # m/cycle, MPa*sqrt(m) units
dS = 100.0                     # stress range MPa per pressurisation cycle
a, a_c, N = 0.002, 0.050, 0    # initial 2 mm, critical 50 mm
Y = 1.0
while a < a_c:
    dK = Y * dS * math.sqrt(math.pi * a)
    a += C * dK ** m_ * 100        # integrate in blocks of 100 cycles
    N += 100
print(f"cycles from 2 mm to 50 mm: {N:,}")
print(f"inspection interval with a scatter factor of 3: {N//3:,} cycles")
# cycles from 2 mm to 50 mm: 642,800
# inspection interval with a scatter factor of 3: 214,266 cycles

Most of the life is spent while the crack is small (growth ∝ a^1.5). Damage-tolerance analysis finds the time from a detectable crack to the critical size and sets inspection intervals as a fraction of it, so a crack gets several chances to be found. (Constants are illustrative; real analysis uses material data and load spectra, e.g. NASGRO.)

Common mistakes

Applications

Key takeaways

Going further