Fatigue Fracture: Beach Marks and Striations

Fatigue crack development is conventionally divided into three stages that differ in their geometry, their microscopic mechanism, and the markings they leave on the fracture surface.

1. Stage I: crack initiation and early growth
   Crack initiation occurs at a point of stress concentration: a surface scratch, a corrosion pit, a sharp notch, a rivet hole, or a weld toe. At this scale, slip bands form along maximum shear stress planes, typically at 45 degrees to the applied stress axis. The micro-crack grows along these slip planes for a few grain diameters. Stage I growth is slow, the fracture surface is smooth and featureless, and the crack path is oblique. In many components Stage I consumes the majority of the total fatigue life.
2. Stage II: stable crack propagation
   Once the crack is large enough to be dominated by the stress intensity factor rather than by crystallographic slip, it deflects to grow perpendicular to the maximum tensile stress. This is Stage II, the fatigue crack propagation phase. The crack advances by a blunting-resharpening mechanism: at peak load the crack tip blunts and stretches forward; on unloading it resharpens, creating a new surface increment each cycle. Each cycle leaves one striation on the fracture surface. The Paris law governs this stage: da/dN = C(ΔK)^m, where da/dN is the crack extension per cycle and C and m are material constants.
3. Stage III: rapid final fracture
   When the crack has grown until K_max reaches K_IC, the remaining ligament fractures by ductile overload or brittle cleavage depending on temperature and material. Stage III is rapid, often completing in a single or a few loading cycles. The fracture surface in Stage III shows overload features (dimples or cleavage) rather than striations, marking a sharp boundary visible macroscopically as the transition from the flat fatigue zone to the rougher or brighter final fracture zone.

Beach marks, also called clamshell marks or arrest marks, are macroscopic curved bands that arc around the fatigue origin on a well-preserved fracture surface. They are typically 0.1 to 1 mm wide and visible without magnification.

Beach marks form when the crack growth rate changes noticeably for a period long enough that the fresh crack surface oxidises or changes appearance relative to the preceding growth. Common causes include: overnight shutdown of a machine (the crack stops, oxidises, then restarts the next day), an unusually large overload cycle (rapid growth followed by retardation), a change in the corrosive environment, or a period of non-use. They are not formed during every cycle and are not the same as striations.

When beach marks are present, they are one of the most useful forensic features on a fracture surface because they carry timeline information. If the component had a known operating schedule, the number and spacing of beach marks can sometimes be matched to the schedule. A shaft that shows 52 distinct bands in a component that ran with weekly shutdowns grew its crack over roughly one year. This reconstruction can establish when a crack must have been present, which determines whether a prior inspection should have detected it.

Fatigue striations are the microscopic record of Stage II crack growth. Each striation is the boundary formed as the crack blunted at maximum load and then resharpened on unloading, advancing by one increment. In ductile metals the spacing between successive striations equals the crack extension per cycle at that point in the propagation history.

Under SEM at 2,000 to 10,000x magnification, striations appear as parallel ripple-like lines running perpendicular to the local crack propagation direction, which means they are parallel to the crack front. They are typically spaced from 0.1 micrometres to 10 micrometres apart. Closer to the origin, where ΔK was small, spacing is finer. Further from the origin, where ΔK had grown with the crack, spacing is wider.

Using the Paris law relationship (da/dN = C(ΔK)^m) in reverse, measured striation spacings can be used to estimate the stress amplitude at the time the crack was at that position. This is a quantitative technique used in fatigue life assessments for aerospace and power generation components, and it has been used in litigation to argue either that an overload event, rather than normal service, drove the crack, or that the crack was present before a component was sold.

Ratchet marks are radial steps or ridges running from the component surface into the fracture interior. They separate adjacent regions of the fracture surface that initiated at separate, closely spaced sites and grew independently before their crack fronts met. At the junction, one crack is slightly ahead of the other, so one crack front climbs up over the other, forming a step: the ratchet mark.

A single ratchet mark between two initiation sites is possible in relatively benign service where only two nearby stress concentrations both exceeded the fatigue threshold. Many ratchet marks distributed around the entire surface indicate many simultaneous initiation sites, which in practice means either a very high applied stress amplitude or a surface with numerous small defects from corrosion, fretting, or poor machining.

On 28 April 1988 Aloha Airlines Boeing 737-200 N73711 was operating its daily inter-island shuttle schedule in Hawaii, a route that averaged flights of around 20 minutes. The aircraft had completed 89,680 flight cycles since manufacture, making it one of the most-cycled 737s in the world. Each pressurisation cycle loaded the fuselage skin in tension and deflected the lap joints at the fastener rows. At 24,000 feet, the upper fuselage section above the wing failed and separated.

The National Transportation Safety Board investigation found that fatigue cracks had grown from the countersunk rivet holes at the S-10L lap splice joint along the entire circumference of the fuselage. In a lap joint, two sheets of skin overlap and are joined by rows of rivets. Each rivet hole is a stress concentration; the load transfer through the fastener causes local bending that amplifies the stress at the hole edge. After tens of thousands of cycles, small fatigue cracks initiated at multiple holes simultaneously, the phenomenon known as multi-site damage (MSD).

MSD is more dangerous than a single fatigue crack because cracks from neighbouring holes can link up far faster than a single crack can grow across an equivalent distance. Once linking begins, the residual strength of the joint drops catastrophically. On Flight 243, when cracks in adjacent holes finally bridged across the thin ligament between them, the failure propagated around almost the entire upper fuselage circumference in an instant. The fractographic examination of the failed sections showed beach marks and fatigue striations at the rivet holes, confirming that the cracks had been growing over many flights before the accident.

Striation analysis serves three broad purposes in failure investigations. First, it confirms fatigue as the failure mechanism, distinguishing it from stress-corrosion cracking or overload fracture, which can look similar macroscopically. Second, it provides a quantitative crack growth rate at any measurable point on the fracture surface, which feeds back-calculation of the stress amplitude. Third, it can give an estimate of the number of cycles spent in Stage II propagation, which combines with knowledge of the operating schedule to give a time-in-service estimate.

• Step 1: locate the fracture origin using macroscopic ratchet marks and beach mark convergence; this establishes the starting point for the growth history.
• Step 2: select multiple SEM fields at known distances from the origin along the crack propagation direction; measure striation spacing at each field as an average over 10-20 striations.
• Step 3: use the geometry of the component and the Paris law to calculate the K range (ΔK) at each sampling location; compare measured spacing to the Paris prediction for the material to check consistency.
• Step 4: integrate the crack growth rate over the total crack length to estimate the number of cycles in Stage II. If the component had a known cycles-per-day usage, this translates directly to a time period.
• Step 5: if beach marks are also countable, compare the beach mark count to the striation-based cycle estimate as a cross-check. Systematic disagreement between the two lines of evidence points to an unusual event, such as an accidental overload or an undisclosed repair.

Striation analysis is not a precise chronometer. Striation spacing measurement has uncertainties of 20-30% from field-to-field variability in local crack geometry. Paris law constants vary between heats of nominally the same alloy. And the Stage I portion of the life, in which no striations are visible, is not accounted for. Competent analysts present a range of estimated cycles and are transparent about all assumptions. Treating the output as a single precise number in court is a misuse of the technique.