Fatigue failure is the process by which a material breaks after being subjected to repeated cycles of stress, even when that stress is well below the level needed to break it in a single pull. It’s the most common mode of failure in mechanical components, and it’s dangerous precisely because it gives little visible warning before the final break.
Unlike a one-time overload that bends or snaps a part immediately, fatigue works slowly. A tiny crack forms, grows a little with each loading cycle, and eventually reaches a critical size where the remaining material can no longer hold. The part then fractures suddenly, often catastrophically.
The Three Stages of Fatigue Failure
Fatigue failure unfolds in three distinct stages: crack initiation, crack propagation, and final rupture.
In Stage I, a microscopic crack forms at a point of high stress, typically at the surface of the material. This is where the majority of a component’s fatigue life is spent. The cracks at this stage are tiny, usually less than 1 mm, and essentially invisible without specialized inspection. They tend to start at surface imperfections, grain boundaries, or areas where the geometry of the part creates stress concentrations.
In Stage II, the crack begins growing in a more stable, predictable pattern with each loading cycle. Engineers can model this growth rate mathematically, and it follows well-established relationships between the stress intensity at the crack tip and the rate of advance. If you’ve ever seen the smooth, shell-like markings on a broken metal surface (called “beach marks”), those are the visual signature of Stage II growth. Each mark represents a period of crack advancement.
Stage III is the final, sudden fracture. Once the crack has grown large enough, the remaining cross-section of material can’t support the load, and the part breaks in a single overload event. This final zone looks rough and granular compared to the smoother fatigue region, which is one way investigators identify fatigue as the cause of a failure after the fact.
Low-Cycle vs. High-Cycle Fatigue
Not all fatigue is the same. Engineers divide it into two broad categories based on the stress levels involved and how many cycles it takes to fail.
High-cycle fatigue (HCF) involves relatively low stress amplitudes that keep the material in its elastic range, meaning it flexes but returns to its original shape each cycle. Because the stress per cycle is low, it takes a very large number of cycles to cause failure, often millions or tens of millions. A vibrating turbine blade or a rotating shaft under a consistent load are classic examples. Engineers typically aim to identify the highest stress a material can withstand for beyond 10 million cycles without cracking. For some alloys, this threshold is remarkably specific: one aluminum alloy, for instance, can handle alternating stresses up to 70 ksi for 10 million cycles without breaking.
Low-cycle fatigue (LCF) is the opposite scenario. Here, the stresses are high enough to cause plastic deformation, meaning the material is permanently distorted with each cycle. Because the damage per cycle is much greater, failure comes after relatively few cycles, typically in the hundreds to tens of thousands. Thermal cycling in jet engine components is a common source of LCF: the part heats and cools with each flight, expanding and contracting enough to cause plastic strain.
Why Geometry Matters So Much
The shape of a part has an outsized influence on where fatigue cracks start. Any abrupt change in geometry, such as a sharp corner, a hole, a groove, or a thread, creates a stress concentration where local stress can be several times higher than the average stress in the part. These are the spots where cracks initiate first.
Research on stainless steel specimens illustrates this clearly. In smooth, uniform samples, fatigue cracks tend to start at multiple points around the perimeter and propagate inward from several directions. But in notched samples, the crack initiates at just one or two points, right at the stress concentration. That single point of origin means the crack grows faster and more directionally, shortening the total fatigue life of the part. This is why engineers spend significant effort on fillet radii, smooth transitions, and avoiding sharp internal corners in parts that will see cyclic loading.
How Corrosive Environments Accelerate Failure
Fatigue in a clean, dry lab is one thing. Fatigue in a real-world environment with moisture, salt, or chemical exposure is another entirely. Corrosion fatigue is the combined attack of cyclic stress and a corrosive environment, and the two mechanisms reinforce each other in ways that are worse than either one alone.
In a corrosive environment, chemical reactions at the crack tip can weaken the material between loading cycles. Chloride ions, for example, can enable stress-corrosion cracking to occur at stress levels well below the threshold that would cause cracking in dry conditions. Research on high-strength stainless steels has shown that even very small cyclic loads (called “ripple loads”) can rupture protective films at the crack tip, exposing fresh metal to chemical attack and accelerating crack growth through hydrogen embrittlement. The result is that parts exposed to seawater, de-icing salts, or industrial chemicals can fail at stress levels and cycle counts that would be perfectly safe in a benign environment.
The Role of Pre-Existing Defects
One of the more striking findings in fatigue research is just how much of what gets labeled “fatigue failure” actually traces back to manufacturing defects. Pre-existing flaws in the material, particularly thin oxide films trapped inside cast metals during manufacturing (called bifilms), can account for up to 90% of the actual failure mechanism. In one analysis of a failed wind turbine bearing, researchers found that the manufacturing defect contributed over 90% of the failure, while true cyclic fatigue contributed roughly 1%. A study of helicopter magnesium gearbox casings found that more than 70% of fatigue failures could be attributed to these trapped oxide films and associated porosity.
This doesn’t mean cyclic loading is irrelevant. It means that in many real-world failures, the crack didn’t have to initiate from scratch. It started at a flaw that was already there, skipping much of Stage I and jumping straight into crack propagation. This is why material quality and casting processes matter enormously for fatigue-critical parts.
How Engineers Prevent Fatigue Failure
Prevention starts at the design stage. Engineers use tools like the Goodman diagram, which plots the relationship between the alternating stress (the cyclic part) and the mean stress (the constant baseline load) to predict whether a part will survive a given number of cycles. The basic principle is straightforward: as the constant background stress on a part increases, the amount of additional cyclic stress it can tolerate before fatigue failure decreases. By plotting a component’s expected loading conditions against the material’s known limits, engineers can determine whether a design includes an adequate safety margin.
Beyond design, surface treatments can significantly extend fatigue life. Shot peening, a process where small metal or ceramic beads are blasted at the surface, introduces compressive residual stresses that make it harder for cracks to initiate. The improvement varies widely depending on conditions. Mild shot peening of low-carbon steel can increase fatigue life by a modest 1.4%, but combining severe shot peening with nitriding (a surface-hardening treatment) has been shown to improve fatigue life by over 51% in low-alloy steels. The key is that both treatments push the surface into compression, forcing any would-be crack to first overcome that compressive stress before it can open and grow.
Other prevention strategies include choosing materials with known endurance limits (the stress level below which they can theoretically cycle forever), avoiding sharp geometric transitions, specifying tight manufacturing tolerances to minimize initial defects, and implementing regular inspection schedules for critical components. Industries like aerospace rely on standardized test methods, such as ASTM E647, which provides procedures for measuring fatigue crack growth rates across the full range from near-threshold to rapid instability. These standardized tests allow engineers to compare materials and predict service life with confidence.
What a Fatigue Fracture Looks Like
If you’re examining a broken part and wondering whether fatigue caused the failure, the fracture surface tells the story. A fatigue fracture typically has two visually distinct zones. The fatigue zone, where the crack was slowly growing, appears relatively smooth and may show concentric beach marks radiating from the initiation point. The final fracture zone, where the part gave way all at once, looks rough, jagged, and crystalline. The ratio between these two zones tells you something about the loading: a large smooth zone and a small rough zone means the cyclic stress was low and the crack grew for a long time before final failure. A small smooth zone and a large rough zone means the stress was high and failure came quickly.
The initiation point itself is often traceable to a surface scratch, a machining mark, a corrosion pit, or an internal void. Finding that origin is usually the first step in a failure investigation, because it reveals whether the root cause was a design problem, a manufacturing defect, or an unexpected service condition.