Ideally, it should only take about 3 gallons of water to hydrate a sack of cement. To obtain adequate workability, however, a mix design must typically employ between 5 to 6 gallons of water per sack—even when a water reducer is used. Strictly in chemical terms, therefore, every slab mix is always going to contain somewhere between 67% and 100% too much water. The necessary inclusion of this additional water of convenience has enormous consequences. Indeed, it is the root cause of all of the “Big Three” slab performance problems: cracking, curling, and delamination.

This complex story begins with the character of the paste—the water and cement portion of the mix. Because portland cement does not dissolve in water, the paste is not a homogeneous solution, but rather a special kind of heterogeneous mixture—termed an emulsoid—wherein the tiny insoluble solid particles are both suspended in and strongly attracted to the dispersing liquid. Indeed, the ability of the cement particles to generate heat either by reacting with or adsorbing the water molecules in which they are immersed is the engine that drives the entire concrete hardening process. Critically, however, until it becomes involved with the cement in such hydration chemistry, any “free” water molecule can be mechanically removed from the system at any time, for example, by evaporation.

A dimensional perspective on the two bits of matter that make up the paste may be helpful. A single water molecule measures about 1/100 of 1/1000 of 1/1000 of an inch across. The suspended cement particles, in contrast, are gigantic in comparison—most spanning between 20,000 and 200,000 individual H2O molecules. Each irregularly shaped cement particle is itself just a colossal conglomeration of different molecules—the four principal ones being oxides of calcium, silicon, aluminum, and iron. These molecules are not present in isolation, but rather are variously combined to form the four distinct compounds that comprise portland cement: tricalcium silicate, dicalcium silicate, tricalcium aluminate, and tetracalcium aluminoferrite. The basic picture of fresh paste, then, is one of massive solid cement chunks suspended in a sea of miniscule water molecules with which the four different molecules at the surfaces of the chunks very much want to combine.

Now nobody knows for sure exactly what happens next. Although we know what we start with, and end up with, precisely how we get from one to the other remains one of those fortuitous mysteries—like the way a car battery works—that is best viewed as simply miraculous. This much is known: for the necessary chemical processes to occur at the surfaces of the cement particles, the reacting water and surface molecules must be in contact. Because the newly created cement + water molecules at the particle’s surface occupy less volume than do the constituent cement and water molecules individually, voids tend to be created around the cement particles as the free water molecules get used up. Unless additional water molecules from the surrounding pools of surplus liquid continue to move into contact with the particle surfaces, the relative humidity at the surfaces will drop below 100%, and the cement’s hydration will be starved into inactivity.

The process by which the cement’s hydration persists in depleting the supply of adjacent free water molecules is called “self desiccation,” and in the absence of wet curing, this internal drying will inevitably lead to less than optimal conversion of the raw cement. Once the water-cement ratio reaches 0.50, however, enough surplus water is available for the hydration to continue at the same rate obtained with wet curing. But even under these most favorable conditions, because each bit of hydrate that forms at the cement particle’s surface creates a new barrier between the surrounding free water molecules and the raw cement inside the particle, the rate of hydration always will continue to decrease irregularly. It is entirely normal, therefore, for a large amount of raw cement to remain unhydrated for years.