Pressure equipment fails in ways that look alike from the outside and are not alike at all. A split pipe, a torn vessel shell and a ruptured tank course can each be called a rupture, but the mechanism behind each is a different engineering question with different evidence and a different record trail. The distinctions that matter forensically are made early: whether the wall thinned, whether a crack grew, whether the material behaved in a ductile or brittle manner, and whether the pressure at failure was the pressure the system was built to carry.
Thinning versus cracking is the first split
Metal loss and crack growth both end in a wall that can no longer carry hoop stress, and they leave different signatures. Corrosion thins the wall over an area, and failure follows when the remaining ligament yields, typically as a bulged, thinned lip with a dull fibrous tearing surface. A crack instead produces a flatter fracture over part of its length, with a visible transition where stable growth ended and overload began.
The distinction is not academic. Wall loss points toward corrosion control, coating condition and inspection for metal loss. Crack growth points toward a different family of inspection tools, and toward records that were sometimes never collected.
What wall-loss failures leave behind
Remaining-thickness measurement around the opening is the core evidence, taken as a map rather than a single reading. The relationship between the thinnest surviving wall, the operating pressure and the material's strength establishes whether loss alone was sufficient. Morphology narrows it further: broad wastage, discrete pitting, preferential attack along a weld, or accumulation at the low point of a liquid line.
Damage mechanisms for fixed equipment in process service are catalogued in API RP 571, which gives an investigation a defensible vocabulary for what was attacking the wall.
Crack colonies versus a single dominant crack
Stress corrosion cracking and fatigue are routinely confused because both are subcritical growth ending in rupture. Under examination they rarely resemble each other. Stress corrosion cracking typically appears as colonies of shallow, roughly parallel cracks that interlink, often beneath disbonded coating on buried line pipe, with intergranular or branched transgranular paths visible in cross-section.
Fatigue is usually one dominant crack initiating at a stress concentrator — a weld toe, a pit, a gouge — with a beach-marked or striated surface recording the pressure cycles that drove it. Cross-sectioning through the origin and scanning electron microscopy settle the question. Surface appearance alone does not.
Ductile and brittle behaviour are separate findings
Fracture appearance describes how the component responded once the flaw became critical, a different question from how the flaw got there. Ductile failure absorbs energy, deforms and usually arrests within a limited length. Brittle fracture propagates at low absorbed energy, can run far beyond its origin, and travels through material containing no flaw of its own.
That difference bears mainly on the extent of damage. Brittle behaviour raises questions about toughness at the actual minimum operating temperature, about heat-affected zone hardness, and about whether the governing specification's impact requirements were met — answered by testing the failed material, not by appearance.
Overpressure as a competing explanation
A genuine excursion beyond the design basis produces a different picture: gross plastic deformation, thinning around the tear, and no meaningful pre-existing flaw at the origin. It moves the inquiry to overpressure protection — relief device sizing and set pressure under ASME Section VIII and the API 520 and 521 practices, relief path obstruction, control system behaviour.
Most ruptures are not overpressure events, but it is worth excluding rather than assuming, and the pressure history establishes that independently of the metallurgy.
Weld and fabrication flaws as ready-made origins
Lack of fusion, slag, porosity and hydrogen cracking in a seam or girth weld supply a flaw at manufacture that needs little growth to become critical. Where the origin sits in weld metal or the heat-affected zone, attention shifts to the weld procedure, the qualification records, and the examination the construction code required — ASME B31.3, B31.4 or B31.8 for piping, ASME Section VIII for a vessel.
Vintage construction complicates this. Methods that were compliant when installed may carry known susceptibilities, making the origin question partly a documentation exercise.
Damage that fails later rather than immediately
Excavation contact, dents and gouges create a stress concentration alongside cold-worked, locally embrittled metal. Some fail on contact. Others survive for years, then become the origin of a fatigue crack growing under ordinary pressure cycling, so a rupture with mechanical damage at its origin does not date itself to the damage event.
Coating disturbance, paint transfer, tool marks and dent geometry help establish whether the contact preceded failure by minutes or by seasons.
Mechanisms combine more often than they compete
Real failures stack. A corrosion pit becomes a fatigue initiation site, a dent accelerates local corrosion, a hard heat-affected zone hosts environmental cracking. Framing the analysis as a single-mechanism election tends to produce an answer that survives examination and not cross-examination.
The defensible form identifies the initiating flaw, the growth mechanism, and the condition under which the flaw became critical, and says which of those is measured and which inferred.
What makes a mechanism call hold up
These opinions are challenged on consistent grounds: that the origin was located by assumption rather than from the fracture surface, that the mechanism was named from morphology without metallography, that toughness and strength were taken from a specification rather than measured on the material that failed.
Work that shows the origin, the section taken through it, the test results and the calculation tying flaw size to failure stress answers those objections before they are made.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.