Q.6 — Fatigue Cracks: Why They Occur, Where to Look, and Which Areas Need More Stringent Inspection

Why Fatigue Cracks Occur

Basic mechanism Fatigue cracking is fundamentally different from corrosion or overload failure — it results from repeated cyclic loading at stress levels well below the material’s ultimate tensile strength, and often below yield strength, applied over a very large number of cycles. On a ship’s hull structure, this cyclic loading arises continuously throughout the vessel’s operating life from:

1. Wave-induced hull girder bending — every wave encounter produces alternating hogging and sagging stresses in the hull girder, and over a 25-year design life a vessel may experience on the order of 10⁸ load cycles, an enormous number even at relatively low individual stress ranges.
2. Local dynamic pressure loading — slamming (bottom and bow flare slamming in ballast condition), sloshing loads in partially-filled tanks, and green water on deck all impose additional high-frequency local cyclic stresses superimposed on the global hull girder bending.
3. Thermal cycling — repeated heating/cooling of cargo tank structure (hot cargo loading, cold ballast passages) produces cyclic thermal stress in restrained structural members.
4. Vibration — machinery-induced and propeller-induced vibration adds high-cycle, low-amplitude stress reversals, particularly in way of the aft structure.

Why cracks initiate specifically where they do

• Fatigue cracks do not initiate randomly on a flat, unstressed plate; they initiate at points of local stress concentration, where the nominal (average) stress in a member is locally amplified by geometry, discontinuity, or a pre-existing flaw:

• Weld toes and weld defects — the weld toe geometry itself creates a natural stress-riser (the abrupt change in section and the residual tensile stress field from welding), and any weld defect (porosity, lack of fusion, undercut, slag inclusion) that acts as a micro-notch further concentrates stress at a sub-millimetre scale, providing the ideal crack-initiation site.

• Structural discontinuities — bracket toes, knuckles, hard points where stiffening arrangements change, scallops, and cut-outs all produce a local stress concentration factor (Kt) significantly above 1.0, meaning the local stress range experienced at these points is several times the nominal stress range elsewhere in the member.

• Corrosion pits and wastage — as discussed in Q.2 and Q.4, a corrosion pit acts exactly like a machined notch, providing a highly effective, often overlooked, fatigue crack initiation site — this corrosion-fatigue interaction is now recognized as one of the dominant failure mechanisms in aging tankers.

• Misalignment and fabrication tolerance — out-of-plane misalignment at butt welds or fillet welds introduces additional local bending stress (secondary bending) not accounted for in the nominal design stress, effectively increasing the actual stress range experienced at the joint beyond calculated values.

• Once a micro-crack initiates at such a site, crack propagation follows a well-established fatigue crack growth law (Paris’ Law regime) — the crack grows a small increment with each stress cycle, with the growth rate accelerating as the crack lengthens (because the effective stress intensity factor at the crack tip increases with crack length), until the remaining ligament of sound material can no longer sustain the load, at which point rapid/brittle final fracture occurs — often suddenly, with little visible warning beyond the crack itself if not caught during inspection.

Where to Look During Routine Inspection

Because fatigue cracks are inherently localized to specific geometric/structural features rather than distributed generally, inspection must be targeted rather than generic. Key locations include:

1. Weld connections at bracket toes — particularly transverse web frame brackets to longitudinals, at both cargo tank and ballast tank boundaries.
2. Longitudinal stiffener connections to transverse webs/floors, especially where the stiffener passes through a slot/cut-out in the transverse member — the slot edge is a classic high-stress-concentration detail.
3. Hopper tank knuckle (the junction between the inner bottom, hopper sloping plate, and side shell) — a primary hot-spot in double-hull tankers due to the abrupt change in structural stiffness and geometry at this line.
4. Deck plating at hatch corners (on vessels with deck openings) and at intersections of deck longitudinals with transverse deck structure.
5. Side shell frame connections to top-side and bottom brackets, particularly in the way of the ballast waterline where corrosion (wet/dry cycling) and stress concentration compound each other.
6. Manhole and lightening hole corners/edges in floors, girders, and web frames — any circular or elliptical cut-out edge in way of a heavily loaded plate field.
7. Toe of doubler plates, insert plates, and any welded attachment to primary structure (ladders, piping clips, staging brackets) — these are often overlooked “non-structural” attachments that nonetheless introduce local stress concentration into primary members.
8. Butt weld and block-erection joints in highly stressed strakes (sheer strake, bilge strake, deck stringer) where fabrication misalignment is most likely to have occurred.
9. Termination points of longitudinal bulkheads and girders, where a continuous member ends and load must transfer into adjoining structure through a bracket or soft-toe detail.

Inspection at these locations should specifically look for: hairline cracks in coating (which often show cracking before the underlying steel crack becomes visible), rust bleeding/weeping from a crack line (indicating water ingress through a through-thickness crack), local coating disbondment following a linear pattern, and, where coating is absent, visual crack lines often best revealed by close-up visual examination supplemented by dye penetrant or magnetic particle testing at suspect locations.

Areas Requiring More Stringent Inspection

Classification society Enhanced Survey Programme (ESP) requirements and industry experience (including from the casualties discussed in Q.4) direct particular attention to:

1. Ballast tank structure adjacent to cargo tanks — given the combined corrosion-fatigue interaction discussed above, these locations are specifically flagged for close-up survey (meaning the surveyor must be within arm’s reach of the structure, not just general visual survey from a distance) under IACS ESP requirements, especially as the vessel ages beyond 10 years.
2. Transverse bulkhead structure, particularly at the connection to the double bottom and to the deck, since transverse bulkheads carry significant racking/shear loads.
3. Fore end structure (particularly bottom shell forward, subject to slamming loads) and aft end/engine room structure (subject to propeller-induced vibration and higher local stress from hull form changes).
4. Any area with a known history of repair or previous crack findings — a location that has cracked once is statistically far more likely to crack again, both because the underlying stress concentration that caused the original crack remains, and because repair welding itself introduces new residual stress and potential defects.
5. Suspect areas identified through the vessel’s own trading history and loading pattern — vessels engaged in frequent alternate loading (partial cargo/ballast combinations that increase local sloshing and dynamic loading) or vessels with a documented history of heavy weather encounters warrant more targeted attention at the locations known to be sensitive to these load types.
6. Locations identified by finite element/fatigue strength assessment during design (CSR/HCSR fatigue analysis) as having a low calculated fatigue life or low safety factor — modern classification rules require designers to identify and report these “fatigue-critical” locations specifically, and ESP survey requirements are increasingly being tied to these designer-flagged hot-spots rather than relying solely on generic survey checklists.

The overarching survey principle is that because fatigue cracks are a statistically low-probability-of-detection defect at any single random inspection (a crack must be caught while it’s still small enough not to have caused failure, but often after it is large enough to be visible), inspection intervals and intensity must be weighted heavily toward the geometrically and historically known hot-spots rather than treated as a uniform, generic sweep of the entire structure.

Let me know when you’d like Q.7.