The Verification Venue · pointed at a thing everyone gets wrong
The Stone That Drinks Its Water
A slab does not dry to get hard. It reacts: cement and water grow into interlocking crystals that consume the water. Take the water away too early and you don't speed it up; you stop the reaction and leave it permanently weak. Concrete poured underwater cures just fine.
The mental model everyone carries ("wet concrete dries out and becomes stone") is backwards about the main event. Hardening is hydration: tricalcium silicate + water → C-S-H, the calcium-silicate-hydrate gel that laces the paste together. That reaction needs water and it locks water in. About 0.23 g of water is chemically bound per gram of cement forever. Set the mix and the curing below, and watch the strength come, not from water leaving, but from water being eaten.
Kept-wet at 28 days
– MPa
degree of hydration –
Air-dried at 28 days
– MPa
hydration stalled at –
Curing penalty (air vs wet)
– %
strength lost by letting it dry
Below 0.42 there isn't enough water to fully hydrate the cement: the slab can't reach full strength no matter how you cure it. Above it, the excess is what genuinely evaporates. More water is not better: higher w/c = weaker.
Highlights one curve. Both are always drawn so you can see the kept-wet slab pull above the air-dried one: the opposite of "drying makes it hard."
The strength law is Powers' gel-space ratio: the more of the space between grains that fills with C-S-H gel, the stronger the paste, as the cube of that fraction. As hydration α climbs, the gel-space ratio r = 0.68·α / (0.32·α + w/c) climbs, and strength follows fc = 234·r³ MPa. Keep the slab wet and α keeps rising for months. Let it dry and hydration stalls where the moisture ran out; the curve flattens. The wet slab doesn't just win eventually; it wins because it kept the very water the "drying" story says it should lose.
So where does "drying" come from? Two real places.
The drying model isn't purely wrong: it's pointed at the wrong quantity
1. Excess water really does evaporate. Full hydration only needs about w/c ≈ 0.42 (≈0.23 bound + ≈0.19 gel). Almost every real mix is wetter than that for workability, so the surplus above 0.42 has nowhere to react; it stays as free water in the capillary pores and, yes, slowly evaporates, leaving the pore space behind (which is exactly why a wetter mix is a weaker one). That part of the folk model is real. It's just not what makes the slab hard.
2. Flooring installers really do wait for a slab to "dry." Before you glue down vinyl or pour epoxy, the slab has to pass a moisture test, using relative-humidity probes (ASTM F2170) or emission testing, because trapped moisture ruins the adhesive. A cured, full-strength slab is still emitting moisture for weeks to months. So a contractor genuinely "waits for the concrete to dry," and is completely right.
Both are true, and they don't contradict the reversal, because they're about a different quantity than strength. Name which one you mean and the verdict flips:
Structural strength = reaction
Set by how much cement has hydrated. Wants water kept in. Cures underwater. Keeps rising for years. Premature drying permanently weakens it.
Floor-covering readiness = emission
Set by how much free moisture has left the surface. Wants water to go. This is the real "drying," gated by an RH test, a distinct process that happens after the concrete is already strong.
Same slab, two clocks. Ask "is it strong yet?" and the answer is reaction (needs water). Ask "can I lay the floor yet?" and the answer is emission (needs drying). The folk model collapses both into one: that's the whole confusion.
The 28-day number is a test date, not a finish line
Labs test a standard cylinder at 28 days because that's a practical age where most of the strength is in and the schedule can move, an acceptance milestone, not the end of hydration. Keep it moist and it keeps gaining strength for months and years; a well-kept old slab is often stronger than the day it passed its test. Drag the slider and watch the kept-wet curve keep climbing right past the 28-day line.
| age | kept-wet α | kept-wet fc | air-dried α | air-dried fc | % of 28-day (wet) |
|---|
The check: every number recomputed in front of you
Nothing here is a stored figure. For your current w/c, the page recomputes the water bookkeeping and the strength curve from the Powers model, live:
The offline gate recomputes all of this two independent ways: node research/does-concrete-dry/verify-does-concrete-dry.mjs. Free choices & uncertainty: the bound/gel split (0.23 / 0.19 g/g) is the Powers model for ordinary Portland cement; real values shift a few percent with cement type and temperature. The maturity time-constants (τ, β) and the "dries at 3 days" cutoff are a representative model, not a lab guarantee; the ordering (wet > air, and strength keeps climbing) is robust; the exact MPa are illustrative. Powers' intrinsic strength is quoted as 234 MPa (34,000 psi); some texts use 240.
What's exactly true here, and what's a model
Exactly true (chemistry & bookkeeping). Concrete hardens by hydration, not drying; the reaction consumes water into C-S-H; roughly 0.23 g of water per gram of cement is chemically bound (non-evaporable) at full hydration, with about another 0.19 g held as gel water, so full hydration of the reaction system needs w/c ≈ 0.42. Below that, the cement cannot fully hydrate regardless of curing (self-desiccation), so more water is not "more strength"; a higher w/c leaves more evaporable pore water and gives a weaker stone. Concrete cures submerged. 28 days is an acceptance-test age, not the end of strength gain.
The Powers strength law. r = 0.68α/(0.32α + w/c) and fc = 234·r³ MPa are the classic Powers (1958) gel-space relations for hardened cement paste. Real concrete strengths are lower (aggregate, the interfacial transition zone, entrained air, testing), so read the curve for its shape and ordering, not as a delivery ticket. Strength is monotone in α, so more hydration is always more strength: the kept-wet slab, which hydrates further, is stronger at every age and pulls away once the air-dried one stalls.
A model, not a measurement (the time axis). The degree of hydration follows a maturity curve α(t) = α_max·exp(−(τ/t_eff)^β) with representative τ = 0.6 d, β = 0.42; the air-dried case freezes its effective maturity after a ~3-day drying point (real hydration doesn't stop dead; it crawls once internal humidity falls below ~80%, so we let it creep, not halt). Actual rates vary strongly with cement chemistry, temperature and member size. What the model is for is the qualitative truth the folk story gets backwards: keeping water in beats letting it out.
The honest concession. Surface-moisture "drying" (the RH/emission clock that gates floor coverings) is a real, distinct process. See the panel above. It is not the thing that makes concrete strong, and treating the two as one is the misconception.