Intergranular corrosion

An illustrative coupon losing 0.040 g over 20.0 cm² at assumed 8.00 g/cm³ density gives 0.00250 mm equivalent average thickness loss. Over 240 h, linear annualization on a 365-day year gives 0.09125 mm/year. This averages the lost volume over the entire exposed area. It does not measure maximum grain-boundary penetration or predict service life.

Intergranular attack follows susceptible grain boundaries. In austenitic stainless steel, carbide precipitation can deplete chromium nearby. Corrosive exposure is still needed for attack. BSSA, intergranular corrosion.

Other alloy systems can have different susceptibility mechanisms and tests.

Susceptibility, exposure and observed damage

Outokumpu describes stainless-steel sensitization through chromium-rich grain-boundary carbides and adjacent depleted regions, identifying welding and inappropriate hot-forming conditions as possible causes. Its general guidance gives a 550–850 °C range, but that is not a universal safe/unsafe boundary or a time-independent process limit. Actual response depends on the alloy and thermal history. Outokumpu, corrosion resistance, intergranular-corrosion section.

A component may look satisfactory externally while having a susceptible microstructure. Conversely, an etched grain-boundary appearance or a failed laboratory screening result is not, by itself, a measurement of the component’s future service life. Observation, susceptibility assessment and service performance are distinct conclusions.

Low carbon and stabilization reduce particular risks

SSINA explains that reduced carbon delays sensitization and that 304L, 316L and 317L are designed to resist sensitization during typical welding. It also explicitly warns that these grades do not resist sensitization from all long-term exposures in the critical range. “L” is therefore not a universal immunity designation. SSINA, intergranular-corrosion guidance.

Titanium or niobium stabilization binds carbon in more stable carbides and can reduce the chromium-carbide problem. Outokumpu also discusses limiting time in the susceptible range and solution annealing followed by suitable rapid cooling. These mechanisms explain candidate controls, while suitable treatment depends on the alloy’s qualified heat-treatment requirements. Outokumpu, sensitization controls.

A generic post-weld treatment does not establish preservation of the specified mechanical condition, dimensions or surface state. Cleaning or passivation also does not, by itself, establish removal of a chromium-depleted subsurface region. These are different process objectives from reversing sensitization.

Test scope depends on alloy and purpose

ASTM A262-15(2021) addresses susceptibility to intergranular attack in austenitic stainless steels. Its public scope distinguishes oxalic-acid etch screening from hot-acid methods and limits screening to the grades covered by the applicable practices. A “suspect” screening classification is not a universal service-failure prediction. ASTM A262, public scope.

The practices do not all produce the same response quantity. ASTM describes mass-loss-based methods and a copper–copper sulfate–sulfuric-acid method using visual examination of bend specimens. The scope also says that results do not necessarily measure performance in other environments and do not predict resistance to general corrosion, pitting or stress-corrosion cracking. Consequently, the phrase “IGC test” alone leaves the response quantity and specimen condition undefined. ASTM A262-15(2021).

An important cross-alloy boundary is ASTM G67-24a, the nitric-acid mass-loss test (NAMLT) for 5XXX-series aluminium alloys. Its named scope is not an alternative A262 practice for stainless steel. Its numerical criteria belong to the defined alloy system and test basis. ASTM G67-24a, primary product record.

The test specimen may represent the supplied condition, a specified sensitizing treatment or a welded condition. These conditions answer different questions. Preparation, orientation and sampling location affect the relationship between a specimen and the material it represents; a parent-stock certificate does not automatically describe a later welded assembly.

Named supplier example: delivery limits are condition-specific

Alleima’s 2RE10 seamless tube and pipe datasheet identifies UNS S31002 and EN 1.4335. It describes Huey testing in boiling 65% nitric acid for five 48-hour periods. In delivery testing, the stated guaranteed maximum corrosion rate is 0.12 mm/year in the solution-annealed condition and 0.20 mm/year after sensitization at 675 °C. These are named supplier/test/condition limits, not recommended corrosion allowances or general service rates for every stainless steel. Alleima 2RE10 datasheet, updated June 12, 2024, printed pp. 1, 3–4.

The guarantee is tied to the stated delivery condition and test basis. A different sensitizing cycle, product form or test environment lies outside that comparison. These supplier figures are distinct from evidence that an individual heat or component meets its governing specification.

Original example: mass loss is not maximum attack depth

Assume a hypothetical laboratory coupon loses 0.040 g, has a defined total exposed area of 20.0 cm², and an assumed density of 8.00 g/cm³. Exposure lasts 240 hours. These invented inputs illustrate unit conversion; they do not reproduce an ASTM procedure or claim to be an Alleima test result.

equivalent lost volume = 0.040 / 8.00 = 0.00500 cm³.

equivalent average thickness = 0.00500 / 20.0 = 0.000250 cm = 0.00250 mm.

Using an explicitly assumed 365-day year, linear annualization gives:

0.00250 × (8,760/240) = 0.09125 mm/year.

That is an equivalent average mass-loss rate under the assumed conversion. It is not the deepest grain-boundary penetration, the minimum remaining wall thickness or a validated annual service prediction. Localized attack and grain detachment can produce a mass loss that does not correspond to uniform thinning.

Doubling the entered area to 40 cm² would halve the calculated rate to 0.045625 mm/year without changing the specimen’s damage. This illustrates why exposed area, excluded/masked surfaces, density, cleaning and weighing conventions belong with a result. Comparison with a supplier guarantee would require equivalent method, condition and reporting basis. A pass/fail decision requires the applicable procedure and acceptance requirement, not the arithmetic alone.

Interpretation of damage and process changes

Compare grain-boundary morphology, thermal history and corrosive exposure. Other corrosion or cracking mechanisms can coexist.

A prevention or rework result is limited to the material and condition actually evaluated. Heat treatment or material removal can alter dimensions and mechanical properties as well as corrosion behavior. A favorable corrosion result alone therefore does not establish that a modified component still satisfies all of its original requirements. The represented material, test edition and specimen condition define the reach of the evidence.

Source versions

Reviewed 2026-11-02 PST. The ASTM primary records identify A262-15(2021) and G67-24a; G67’s record displays an update date of November 8, 2024, not the earlier G67-24 edition. Alleima’s PDF repeatedly prints 6/12/2024 6:23:56 AM, consistent with its June 12, 2024 dated filename. No separate publication date is assigned to the Outokumpu, BSSA or SSINA web guidance. The original example is a bounded dimensional calculation, not a corrosion test protocol.