Most failure mechanisms are matters of degree: enough load, enough cycles, enough time. Stress-corrosion cracking is closer to a switch. It requires an alloy that is susceptible in one particular environment, a sustained tensile stress at or above a threshold, and that environment present at the metal surface — all three at once. Absent any one, cracking does not initiate, and cracks already running arrest. That conditional structure is inconvenient when it is missed during design and unusually useful afterwards, because it turns a metallurgical question into a set of discrete, documentable ones.
Three conditions, all at once
The three legs are usually stated as alloy, stress and environment, and the word carrying the weight is "and". A susceptible alloy in a benign environment does not crack. The same alloy in an aggressive environment with no tensile stress does not crack. Each leg is individually ordinary; the coincidence is what is rare, which is why SCC so often appears years into service rather than at commissioning.
Susceptibility is specific, not general
There is no such thing as an alloy susceptible to SCC in the abstract. Austenitic stainless steels resist a great many environments and crack readily in hot chlorides. Copper alloys are largely indifferent to chlorides and crack in ammonia. Carbon steel is unremarkable in most waters and cracks in concentrated caustic. The pairing is the mechanism, and a specification that selected a corrosion-resistant material without naming the service environment has not addressed SCC at all.
Condition matters as much as grade. An austenitic stainless steel that has been sensitised — chromium carbides precipitated at grain boundaries during welding or slow cooling — becomes vulnerable along those boundaries to environments that leave the same alloy untouched when solution annealed. ASTM A262 sets out the practices used to detect that susceptibility.
The stress leg is a threshold, not a load case
The stress that matters is tensile, sustained, and present at the surface where the environment is. It need not be the design load, and frequently is not: residual stress from welding, forming or machining routinely reaches a substantial fraction of yield and is enough on its own. Cyclic loading moves the problem into corrosion fatigue, a different mechanism with different fracture features. SCC is a static-stress phenomenon, which is why a vessel sitting idle at ambient pressure is not thereby safe.
The environment leg is local, not bulk
The chemistry that governs is the chemistry at the crack tip, not in the process stream. Evaporation at a hot surface, wet-dry cycling, capillary action in a crevice, and concentration under a deposit or beneath disbonded coating can each produce a local environment orders of magnitude more aggressive than the bulk. This is the leg most often argued about, because bulk logs are recorded and local conditions are not.
Removing one leg is a real engineering option
Because the three conditions are necessary rather than merely contributory, SCC is more tractable than mechanisms driven by accumulated damage. A different alloy removes susceptibility. Post-weld heat treatment or solution annealing removes the residual stress. Drying insulation, controlling chloride ingress, adjusting chemistry or eliminating an ammonia source removes the environment. Any one prevents the mechanism, and the choice between them is ordinarily a matter of cost and practicality rather than technical possibility.
That is also what makes SCC a specification question. The decision determining whether a component would crack was usually made in a materials selection document, a weld procedure or an insulation detail, years before anything went wrong.
What the three-leg test does in a dispute
Framing the investigation as three separate factual questions — which alloy and in what condition, what stress and from where, what species and at what concentration — separates issues that otherwise blur together. Each has its own evidence and often its own custodian: mill certificates and material verification for the first, fabrication and heat-treatment records for the second, process and inspection history for the third.
Testing that establishes susceptibility
The ASTM G-series provides standardised ways to hold exemplar material at a known stress in a known environment: G30 for U-bend specimens, G38 for C-rings, G39 for bent-beam and G49 for direct-tension loading, with G44 describing alternate-immersion exposure and G36 the boiling magnesium chloride test used to rank austenitic stainless steels. ASTM G123 covers testing of stainless alloys in boiling acidified sodium chloride. These establish that the alloy could crack in that environment. They do not establish that it did.
Standards that name the environment in advance
Some environments are aggressive enough that industry has written the constraint down. NACE/AMPP MR0175 with ISO 15156 governs materials selection for hydrogen sulfide-bearing oil and gas service, setting limits on alloy, hardness and condition; the hydrogen-dominated end of that spectrum is a distinct mechanism, addressed separately. API RP 571 catalogues damage mechanisms in refining equipment, including the SCC families, with the conditions under which each is expected. Where such a document applied and the selected material sat outside it, the specification question largely answers itself.
What the framing does not settle
It establishes that the mechanism was possible and identifies which condition could have been removed. It does not, by itself, establish when cracking initiated, how fast it grew, or whether the local environment that developed was foreseeable. Those remain matters of inspection history, operating records and the standard of care at the time, and they are where competent analyses most often disagree.
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.