Sign in to save

Bookmark this page so you can find it later.

Sign in to save

Bookmark this page so you can find it later.

Failure analysis is the engineering process of figuring out why a part, structure, or system broke or stopped working. It matters because failures can cause injuries, expensive repairs, lost production, and damage to public trust. By studying broken components carefully, engineers can prevent the same problem from happening again.

This work connects physics, materials science, design, manufacturing, and real-world safety.

Engineers analyze failure by combining visual inspection, measurements, material testing, and calculations of stress and loading. They look for clues such as crack shape, corrosion, wear patterns, deformation, and signs of overload or fatigue. A good analysis separates the root cause from secondary damage that happened later during the final break.

The results are then used to improve design, material choice, maintenance schedules, and operating conditions.

Understanding Failure Analysis: Why Things Break

A crack is more dangerous than its visible length suggests. Near the sharp crack tip, the material carries load through a very small region. The local pulling effect can become far larger than the value calculated from the whole cross section.

Each loading event can extend the crack by a tiny amount. Once the remaining uncracked section becomes too small, the part can break suddenly. Brittle materials, including some hardened steels and glass, may show little bending before fracture.

Ductile metals often stretch and form a narrowed region before they separate. The fracture surface can preserve evidence of the final event, but it may not show where the original damage began.

Fatigue is especially important because everyday equipment experiences repeated loading. A bicycle crank turns thousands of times. A bridge moves slightly as vehicles pass.

A washing machine shaft vibrates during every cycle. Small surface scratches, machining marks, weld edges, and threads can become starting points for fatigue cracks. The number and size of load cycles matter, not only the largest load.

A part that survives one heavy test can still fail after months of smaller repeated loads. Smooth surfaces, rounded transitions, good weld quality, and compressive surface treatments can slow crack growth. Regular inspection is needed when a crack could grow in a hidden location.

Corrosion changes both the shape and the strength of a component. Water, oxygen, salts, acids, and industrial chemicals can react with metals. Rust on steel often spreads over a wide area, while pitting corrosion creates deep narrow holes.

A pit can act like a sharp notch and start a fatigue crack. Two different metals joined in a wet environment can form an electrical cell. One metal then corrodes faster than expected.

This can happen around fasteners, pipes, boats, vehicles, and outdoor electrical equipment. Paint and plating help, but design matters too. Water must be able to drain away, and sealed gaps must not trap moisture.

Overload failures happen when a part is asked to carry more than it can support at that moment. The extra load may come from impact, misuse, a blocked machine, incorrect assembly, or an unexpected pressure rise. Buckling is different because a slender part can suddenly bend sideways while the material itself has not reached its usual breaking strength.

A thin ruler pushed end to end shows this effect. Column length, end support, straightness, and sideways bracing strongly affect buckling risk. Wear adds another problem.

It removes material, changes clearances, and can make loading uneven. A worn bearing can cause vibration that then speeds up fatigue damage elsewhere.

When learning failure analysis, pay attention to sequence. The final break is often the most dramatic feature, yet it may be the last step in a long chain. Engineers compare the failed part with an undamaged example, check its service history, and examine whether the actual loads matched the design assumptions.

They must consider manufacturing defects, material mixups, maintenance actions, temperature, and human operation. Measurements need realistic uncertainty because a small error in thickness or crack size can change the conclusion. The strongest explanation is one that fits the physical evidence, the loading history, and the known behavior of the material.

Key Facts

  • Stress is force per area: sigma = F/A
  • Normal strain measures deformation: epsilon = Delta L/L0
  • Hooke's law in the elastic range: sigma = E epsilon
  • Fatigue failure can occur at stresses below yield strength after many load cycles.
  • Stress concentration near holes, notches, and sharp corners raises local stress above the average value.
  • A factor of safety is often written as N = failure strength/working stress

Vocabulary

Fracture surface
The exposed surface created when a material cracks or breaks, often containing clues about how the failure happened.
Fatigue
Progressive damage caused by repeated loading and unloading that can lead to crack growth over time.
Stress concentration
A local increase in stress caused by geometry changes such as holes, threads, grooves, or sharp corners.
Yield strength
The stress at which a material begins to deform permanently instead of returning to its original shape.
Root cause
The primary underlying reason a failure occurred, not just the visible final event.

Common Mistakes to Avoid

  • Assuming the final visible break is the root cause, which is wrong because the last fracture may only be the end result of earlier fatigue, corrosion, or design errors.
  • Using average stress only, which is wrong because local stress concentrations at notches or threads can be much higher and start cracks.
  • Ignoring service history, which is wrong because load cycles, temperature, vibration, and environment often explain why a part failed in actual use.
  • Confusing ductile and brittle fracture signs, which is wrong because each fracture mode points to different material behavior and different corrective actions.

Practice Questions

  1. 1 A steel rod carries a tensile force of 12000 N and has a cross-sectional area of 300 mm^2. Calculate the average stress in MPa.
  2. 2 A metal bar with original length 2.00 m stretches by 1.0 mm under load. Calculate the strain.
  3. 3 A machine shaft repeatedly fails near a shoulder where the diameter changes suddenly, even though the average stress is below the yield strength. Explain the most likely failure mechanism and one design change that could reduce the problem.