Unlike concrete, steel has no internal chemistry that degrades over time: it doesn't carbonate, doesn't undergo expansive aggregate reactions, doesn't lose protective alkalinity. Its pathologies are physically simpler to describe, but no less dangerous for that — if anything, corrosion and fatigue are responsible for some of the most serious structural failures in recent engineering history.

Corrosion: not just uniform rust

Uniform atmospheric corrosion — the rust everyone knows — is actually the least insidious form, because it's visible and predictable: it proceeds at a rate that can be estimated from the environmental corrosivity class (EN ISO 12944, categories C1-C5) and is prevented with well-established protections — galvanizing, painting, duplex systems. Far more dangerous is localized corrosion: galvanic corrosion, triggered when two metals with different electrochemical potential are in contact in the presence of moisture (a carbon-steel bolt on a stainless-steel structure, for example, preferentially corrodes the less noble metal); crevice corrosion, which concentrates in narrow gaps where oxygen is scarce, typical of poorly sealed bolted joints; and stress corrosion cracking, which affects high-strength tensioned elements — structural cables, high-strength bolts — where the combination of mechanical stress and a corrosive environment generates cracks that propagate far faster than corrosion or stress alone would.

Fatigue: the failure you don't see coming

Fatigue is the mechanism by which a steel element, subjected to repeated load cycles even well below its static strength, develops a microscopic crack that progressively propagates until sudden fracture — with no visible plastic deformation to warn of the impending failure. It's the typical mechanism in bridges, cranes, and structures subject to vibration or loads that vary over time (wind, traffic, machinery). Eurocode 3 (EN 1993-1-9) classifies construction details into fatigue categories (from 160 down to 36, in N/mm²) precisely because it isn't the base material that's the weak point, but the geometry of the detail: a hole, a weld, a change in section concentrates stresses in a way that drastically lowers fatigue life compared with the homogeneous material. The 2007 collapse of the I-35W bridge in Minneapolis, which killed 13 people, was attributed by the National Transportation Safety Board to undersized gusset plates dating back to the original 1967 design — a textbook case of how a marginal structural detail, not the material itself, can cause a catastrophic failure decades later.

Steel pathologies — recognition and prevention
Uniform corrosionClasses C1-C5 (EN ISO 12944), galvanizing/painting
Galvanic corrosionContact between dissimilar metals — electrically isolate the joints
Stress corrosion cracking (SCC)High-strength cables/bolts — environmental control + inspection
FatigueEN 1993-1-9 categories — the construction detail, not the material
NDT for weldsUltrasound, magnetic particles, dye penetrant, radiography
Strength reduction at 500°C~50% (all structural grades)

Fire: perception versus full-scale tests

Steel's behavior in a fire is often described in simplified terms — it loses strength linearly with temperature — but full-scale tests have revealed a more complex picture. The famous Cardington Fire Tests, conducted by the UK's Building Research Establishment in the 1990s on an experimental eight-story building with a steel structure and composite concrete floors, showed that a real frame, with its structural redundancy and the membrane behavior of the composite concrete floors, can survive real fires significantly longer than predicted by calculating the isolated element in a furnace — because the structural system as a whole redistributes loads in ways a single-beam test doesn't capture. This doesn't mean passive fire protection (intumescent paints, encasement in gypsum board or sprayed concrete) is superfluous — it remains legally mandatory and necessary in a great many configurations — but it has pushed research toward "fire engineering" approaches based on the performance of the whole structural system, not just the strength of a single element read off prescriptive tables.

Non-destructive testing: reading steel without breaking it

Diagnostic tools for steel in service are grouped by the type of flaw they're looking for: ultrasound detects internal defects and delamination in welds; magnetic particle inspection finds extremely fine surface cracks in ferromagnetic materials; dye penetrant testing works on any metal for surface-breaking defects; industrial radiography, more expensive, remains the standard for verifying critical welds in high-responsibility works (bridges, plants). None of these techniques requires damaging the inspected element — a huge advantage over reinforced-concrete diagnostics, which often needs core sampling or localized demolition to check the condition of the internal rebar.

«Steel doesn't hide its pathologies inside a slow, invisible chemical reaction the way concrete does — it shows them on the surface, in a crack, in a rust stain. The problem is never how hard they are to see: it's the temptation to ignore them because "it's only surface-deep."» — Ing. Arch. Sara Conti