alt Hacker News> It turns out that another chemical reaction, known as carbonation, might also contribute to Roman concrete’s longevity.
Roman concrete was made lime cement (calcium dioxide); which cures via carbonation (hardens with carbon oxide). And adding pozzolan to lime makes it hydrolic (hardens with water). Is it surprising that it can still carbonate some? Modern concrete has steel which rust and crack concrete. You can use fiberglass rebar for longevity, or build without rebar even, but that is more costly and and less efficient.
As I understand it, concrete has excellent resistance to compression but fails easily on traction, while steel bars are exactly the opposite. That is why you put rebar in concrete: the steel handles the traction loads and the concrete handles the compression. This works well because both materials have similar coefficients of thermal expansion, so as the temperature changes they both expand and contract at the same rate. I suppose you can engineer fiberglass to have the same thermal expansion coefficient and use it to replace steel (assuming it is just as strong on traction). But how would you "build without rebar even"? Wouldn't your beams start cracking at the bottom, where they are subject to traction?