Sensitization and intergranular corrosion

The mechanism

Between roughly 425 and 815 °C, carbon and chromium combine to precipitate M₂₃C₆ carbides on the grain boundaries. Because chromium diffuses slowly at these temperatures, the carbide draws its chromium from the immediately adjacent lattice faster than the bulk can replenish it, creating a narrow chromium-depleted band along each boundary:

Where the depleted zone falls below the ~12% chromium needed for passivity, it no longer resists corrosion and is attacked preferentially.

Why welding is the usual culprit

A weld drags the surrounding metal through the full temperature range. The fusion zone cools quickly, but a band a few millimetres away dwells in the sensitizing window long enough to precipitate carbides — the familiar "weld decay" line that corrodes in service while the weld itself looks fine.

Measuring it: degree of sensitization

Sensitization is quantified by the degree of sensitization (DOS), most commonly via the electrochemical potentiokinetic reactivation (EPR) test or the etch/immersion practices of ASTM A262. These reveal whether the chromium-depleted network is continuous enough to support intergranular attack.

Designing it out

The three defences are compositional and thermal: pick a low-carbon (L) grade to starve the carbides of carbon, pick a stabilized grade (321, 347) to lock carbon into stable titanium/niobium carbides, or solution-anneal and quench to dissolve the carbides and cool fast enough to avoid re-precipitation. For thick welded sections that cannot be post-weld solution-annealed, the L or stabilized grades are the practical answer.