Guide

Beginner's Guide to Material Fatigue and Fracture

Beginner fatigue and fracture guide for cyclic stress, Goodman screening, Miner damage, critical crack size, inspection interval, environment, NDE, and validation.

Branch: Materials Engineering
Content: Guide
Updated: Jun 23, 2026
Revision: v1.0.0 · reviewed

Fatigue and fracture explain why a component can pass a static strength check and still fail in service. Fatigue is driven by repeated loading, local stress raisers, surface condition, environment, and time. Fracture is driven by crack size, stress, toughness, geometry, temperature, and inspection capability. Together they form the structural-integrity workflow used for shafts, aircraft structures, bridges, ships, pressure equipment, welds, fasteners, implants, springs, rotating machinery, and composite structures.

This guide organizes the fatigue and fracture cluster for engineering students and early-career engineers. It does not replace the detailed topic, formula sheet, exercise set, reliability guide, materials selection guide, characterization guide, processing guide, corrosion guide, or case studies. It shows how to learn them as one workflow: define the local stress history, identify the critical detail, choose fatigue or fracture evidence, calculate first-pass margins, define inspection capability, and document validation limits.

The key habit is to stop asking whether a material is “strong enough” in a static sense. Fatigue and fracture ask a more useful question: can this material, geometry, surface, environment, defect population, and inspection plan tolerate the service mission without crack initiation, unstable growth, or loss of residual strength?

1. Separate Static Strength, Fatigue, and Fracture

Static strength checks compare a load case with yield strength, ultimate tensile strength, bearing strength, compressive strength, or allowable stress. Fatigue checks repeated stress cycles and damage accumulation. Fracture checks whether a crack-like flaw becomes unstable under a stress field.

Integrity question	Main inputs	Typical evidence
Static strength	peak stress, yield or ultimate strength, load factor	stress analysis, proof load, tensile test
Fatigue initiation	stress range, mean stress, notch, surface, cycle count	S-N curve, detail category, strain history
Fatigue crack growth	initial crack, growth law, stress intensity range	crack-growth data, inspection plan, NDE capability
Fracture	crack size, peak stress, toughness, geometry factor	fracture toughness, flaw size, residual strength screen
Environmental damage	corrosion, hydrogen, temperature, moisture, fretting	exposure evidence, surface condition, inspection interval

A fatigue and fracture decision should state which of these questions governs. A single high strength value does not answer all of them.

2. Calculate Cyclic Stress Quantities First

Before using an S-N curve or Goodman diagram, define the stress cycle.

Worked example: stress range, amplitude, mean stress, and stress ratio

A machined component has:

S_{max} = 180\ \text{MPa}

and:

S_{min} = 20\ \text{MPa}

Stress range:

\Delta S = S_{max} - S_{min} = 180 - 20 = 160\ \text{MPa}

Stress amplitude:

\displaystyle S_a = \frac{S_{max} - S_{min}}{2} = \frac{160}{2} = 80\ \text{MPa}

Mean stress:

\displaystyle S_m = \frac{S_{max} + S_{min}}{2} = \frac{180 + 20}{2} = 100\ \text{MPa}

Stress ratio:

\displaystyle R = \frac{S_{min}}{S_{max}} = \frac{20}{180} = 0.11

Engineering comment. The stress is not fully reversed because $R$ is positive and the mean stress is tensile. This matters because tensile mean stress usually reduces fatigue margin. A fatigue report should state whether stresses are nominal, local, hot-spot, principal, equivalent, notch, or measured strain-derived values.

3. Use Mean-Stress Screens With Clear Limits

Mean stress corrections are screening tools. They are useful when the fatigue data and service condition justify the model, but they do not remove the need for local stress, surface, notch, environment, residual stress, and validation review.

Worked example: Goodman screening

A detail has:

S_a = 70\ \text{MPa}

and:

S_m = 80\ \text{MPa}

Use a corrected endurance-limit screen:

S_e = 160\ \text{MPa}

and ultimate tensile strength:

S_u = 500\ \text{MPa}

The Goodman utilization is:

\displaystyle U_G = \frac{S_a}{S_e} + \frac{S_m}{S_u}

\displaystyle U_G = \frac{70}{160} + \frac{80}{500} = 0.4375 + 0.160 = 0.598

The reciprocal screening factor is:

\displaystyle n_G = \frac{1}{U_G} = \frac{1}{0.598} = 1.67

Engineering comment. The Goodman screen passes with a first-pass factor of about $1.67$ . That does not prove infinite life. The result still depends on surface finish, notch severity, size, reliability level, corrosion, residual stress, load spectrum, material condition, and whether $S_e$ is valid for the component detail.

4. Treat Cumulative Damage as a Model, Not a Fact

Miner damage is often used for variable-amplitude fatigue. It is simple and useful for screening, but it ignores load sequence effects, overload retardation, corrosion interactions, scatter, and many crack-growth details.

Worked example: Miner damage

A service spectrum has three load blocks:

Block	Applied cycles $n_i$	Allowable cycles $N_i$
A	$20{,}000$	$200{,}000$
B	$15{,}000$	$60{,}000$
C	$5{,}000$	$20{,}000$

Miner damage is:

\displaystyle D = \sum \frac{n_i}{N_i}

\displaystyle D = \frac{20{,}000}{200{,}000} + \frac{15{,}000}{60{,}000} + \frac{5{,}000}{20{,}000}

D = 0.10 + 0.25 + 0.25 = 0.60

If the screening criterion is:

D < 1.0

the spectrum passes the simplified damage screen.

Engineering comment. A damage value of $0.60$ is not a guarantee. If corrosion pits, weld defects, fretting scars, overloads, poor surface finish, or residual tensile stress exist, the real margin may be lower. Miner damage is best used as a decision gate for deeper review, not as standalone release evidence.

5. Use Fracture Toughness to Connect Stress and Crack Size

Fracture mechanics asks whether a crack can become unstable. A simple screen uses:

K = Y\sigma\sqrt{\pi a}

where $K$ is stress intensity, $Y$ is a geometry factor, $\sigma$ is stress, and $a$ is crack depth.

Worked example: critical crack size

Assume:

K_c = 45\ \text{MPa}\sqrt{\text{m}}

Y = 1.1

\sigma = 150\ \text{MPa}

Set $K = K_c$ and solve:

\displaystyle a_c = \frac{1}{\pi}\left(\frac{K_c}{Y\sigma}\right)^2

\displaystyle a_c = \frac{1}{\pi}\left(\frac{45}{1.1 \times 150}\right)^2 = 0.0237\ \text{m} = 23.7\ \text{mm}

If the qualified inspection can reliably detect a crack depth of:

a_d = 5.0\ \text{mm}

the simple crack-size margin is:

\displaystyle \frac{a_c}{a_d} = \frac{23.7}{5.0} = 4.74

Engineering comment. This screen suggests detectable cracks are well below the critical crack size. It still does not define inspection interval. The engineer also needs crack-growth rate, inspection access, probability of detection, stress spectrum, environment, geometry, and residual strength at the detectable flaw size.

6. Set Inspection Intervals From Crack-Growth Evidence

Damage tolerance depends on finding cracks before they become critical. Inspection interval should come from crack-growth analysis, detectable flaw size, uncertainty, consequence of failure, and a conservative inspection factor.

Worked example: inspection interval from growth life

A crack-growth calculation, using project-specific stress spectrum and crack-growth data, estimates:

N_{growth} = 180{,}000\ \text{cycles}

from the reliably detectable crack size to critical size. The program applies an inspection factor:

F = 3

The maximum inspection interval is:

\displaystyle N_{inspect} \leq \frac{N_{growth}}{F}

\displaystyle N_{inspect} \leq \frac{180{,}000}{3} = 60{,}000\ \text{cycles}

Engineering comment. The interval is only as good as the crack-growth model and inspection capability behind it. If environment accelerates growth, access is poor, probability of detection is weak, or load spectrum changes, the interval must be shortened or revalidated.

7. Include Environment, Surface, and Process History

Fatigue and fracture are sensitive to details that are easy to miss:

corrosion pits and galvanic attack can create crack starters;
hydrogen can reduce apparent fracture resistance in high-strength steels;
weld toes, undercut, lack of fusion, and residual stress can dominate fatigue;
quenching, heat treatment, grinding, machining, additive manufacturing, and cold work can change residual stress and defect population;
polymers and composites may be controlled by creep, delamination, impact damage, moisture, and matrix cracking;
ceramics are flaw-sensitive and may require proof screening or statistical reliability evidence.

A calculation that ignores the real surface and process state is often too optimistic.

8. Connect NDE to the Failure Mode

Non-destructive testing is not a generic safety step. The inspection method must match the flaw type, orientation, size, material, access, and acceptance rule.

Flaw or damage	Possible evidence
surface crack	visual, dye penetrant, magnetic particle, eddy current where appropriate
internal crack or lack of fusion	ultrasonic testing or radiography depending on geometry
porosity or additive defects	x-ray computed tomography or qualified radiography
delamination	ultrasonic inspection, tap test for limited cases, CT where justified
corrosion loss	thickness survey, corrosion-rate evidence, visual and NDE mapping
thread-root flaw	microscopy, magnetic particle, sectioning for investigation

The release basis should state detectable flaw size, calibration standard, inspector qualification, access limits, and action when an indication is found.

9. Learn the Cluster in a Practical Order

A good learning sequence is:

Read the fatigue and fracture topic to understand cyclic loading, crack initiation, crack growth, toughness, environment, and inspection.
Use the formula sheet for stress amplitude, stress ratio, Goodman screening, Miner damage, stress intensity, critical crack size, crack growth, and inspection intervals.
Work through the exercise set before trusting fatigue or fracture margins.
Study materials selection, processing, characterization, and corrosion because material condition, defects, surface, and environment control many failures.
Study case studies on aircraft fatigue, hydrogen embrittlement, weld cracking, quench cracking, ductile-brittle transition, composite delamination, polymer creep, and corrosion.
Connect to mechanical stress analysis, machine design, aircraft structures, marine structures, civil infrastructure, biomaterials, quality engineering, reliability, uncertainty, and systems requirements.

10. Common Beginner Mistakes

Common mistakes include:

using yield strength as if it were fatigue strength;
applying an S-N curve without checking material condition, surface, notch, stress ratio, environment, or survival probability;
treating Miner damage as a precise life prediction;
computing critical crack size but forgetting crack-growth time and inspection capability;
assuming NDE finds every dangerous flaw;
ignoring weld toes, undercut, threads, keyways, corrosion pits, fretting scars, and machining marks;
ignoring residual stress, heat treatment, hydrogen, corrosion, temperature, and moisture;
using nominal stress when local notch or hot-spot stress controls;
releasing a design without a defined inspection interval or revalidation trigger.

11. The Engineering Standard

Good fatigue and fracture engineering makes the integrity argument traceable. It states the stress history, local detail, material condition, surface condition, environment, flaw assumption, inspection capability, calculation model, validation evidence, and action limits.

The best beginner habit is to replace “the stress is below yield” with a stronger question: “can this detail survive the expected cycles and detected flaw population, in this environment, with this inspection plan, for the required service life?”

REF

Disciplines