Concrete Is Strong. Until You Pull on It.

Why does ordinary concrete resist compressive forces so well — routinely tested in the thousands of pounds per square inch — while its tensile strength lags an order of magnitude behind, and what does the materials science say about how engineers get around that weakness without simply making a stronger slab?

Where the physics gets genuinely elegant is in the pairing of concrete with steel reinforcement — rebar. Two coincidences make this partnership work rather than just patch over a weakness. First, steel and concrete have very similar coefficients of thermal expansion, meaning that as a structure heats up in a Chicago summer and contracts in a Chicago winter, the embedded steel and the surrounding concrete expand and shrink at nearly matching rates. If they didn't, decades of thermal cycling would tear the bond between them apart regardless of how well the concrete was poured.

Second, concrete's chemistry actively protects the steel inside it. Fresh and cured concrete is highly alkaline, typically with a pH in the range of 12 to 13, largely due to calcium hydroxide produced as a byproduct of cement hydration. That alkalinity causes a thin, tenacious oxide film to form on the surface of embedded steel, a phenomenon called passivation, which is the same basic principle that keeps stainless steel from rusting. As long as that alkaline environment holds and chlorides or carbonation don't disrupt it, steel rebar can sit inside concrete for a very long time without corroding.

Reinforced concrete puts steel bars where tensile stress will concentrate — the underside of a beam spanning a gap, for instance — so that the steel, which is excellent in tension, carries the pulling forces while the concrete around it carries the compression. Post-tensioned and prestressed concrete take this further: steel tendons are tensioned, either before the concrete cures (pretensioning, common in precast beams) or after (post-tensioning, common in bridge decks and parking structures), squeezing the concrete into a permanent state of compression. Because concrete handles compression so well, this preloading effectively cancels out tensile stresses before service loads ever get the chance to open a crack. The technique, pioneered in usable form by French engineer Eugène Freyssinet in the early 20th century, is why long-span bridges and thin, elegant concrete slabs are structurally possible at all.

None of this protection is permanent, and durability failure is where concrete science gets genuinely contentious among engineers arguing over inspection intervals and repair budgets. Chloride-induced corrosion is the big one in cold climates: road salt, sea spray, or de-icing chemicals introduce chloride ions that, given enough time and enough exposure, penetrate through concrete's pore network and reach the embedded steel. Once chloride concentration at the steel surface crosses a threshold, it destroys the passive oxide layer, and the steel begins to corrode. Corroding steel forms iron oxide, which occupies significantly more volume than the original metal, and that expansion cracks and eventually spalls the surrounding concrete from the inside out — the reddish-brown staining and crumbling edges familiar to anyone who has looked closely at an aging Chicago parking garage or an older highway overpass.

Freeze-thaw damage compounds the problem in exactly the climate Chicago has. Water that has worked its way into concrete's capillary pores expands roughly nine percent in volume when it freezes, and repeated freeze-thaw cycling generates internal pressure that gradually fractures the paste, particularly near the surface. Air-entrained concrete, which incorporates deliberately engineered microscopic air bubbles as pressure-relief valves for that expanding ice, is now standard practice for pavement and exposed structures in northern climates specifically because of this failure mode, a practice grounded in research going back to postwar-era studies by organizations including the Portland Cement Association and reinforced by ongoing NIST and university lab work on pore structure and freeze-thaw resistance.

The messiness comes from the fact that these failure modes interact, and predicting service life for any specific structure involves a lot of uncertainty about exposure history, mix quality, and construction quality that isn't fully knowable from a single inspection. Two overpasses built the same year with nominally similar specifications can age at dramatically different rates depending on how well they were cured, how much salt they've actually been exposed to, and whether cracks let water and chloride in early. Structural engineers manage this with probabilistic service-life models rather than fixed guarantees, and disagreement over how conservative those models should be is a live, ongoing debate in the field rather than settled science.