The material ultimately created by the cement’s hydration—the gray stuff we know as hardened cement paste—is called a colloidal gel—a curious sort of “permanently damp solid” wherein very tiny bits of liquid (the molecules of water and dissolved ions of calcium and hydroxide) are evenly dispersed throughout a porous random nanostructure of insoluble hygroscopic solid material (the precipitated crystalline hydrates).

In the same way that the precise chemistry of the cement’s hydration remains a mystery, nobody knows for sure exactly how hardened cement achieves its strength. It is known that each leaf-like crystal of calcium silicate hydrate (C-S-H) remains exceedingly small; most measure less than 1/125 of 1/1000 of an inch. And though the bits of C-S-H do interlock, their mere physical entanglement does not appear to be the principal source of the paste’s strength.

The particles in a pound of dry portland cement powder have a total surface area of about 5000 square feet. In contrast, the much smaller C-S-H crystals which form on the surfaces of all these particles once they are immersed in water have a total area of about 1,000,000 square feet. This 200-fold increase in the internal solid surface area brought about by the cement’s hydration appears to be most significant factor leading to the strength development. It is this property, in concert with the conjoined water, that makes the paste behave like super-stiff Jello.

A pound of portland cement powder and a third of a pound of water each occupy about 9 cubic inches. The 11/3 pounds of 0.33 water-cement ratio (w/c) paste created by mixing the two together will have a volume of about 18 cubic inches—half of which will be solid material and half of which will be liquid water. If this paste were somehow able to avoid self-desiccating (see “That Pesky Moisture Gradient, Part 1” in the December 2010 issue) and hydrate completely, the volume of the resulting C-S-H gel would be enough to fill the entire mass; there would be no space left for anything else. A w/c of about 0.33 thus represents something of an ideal minimum, because it is around this value that all of the mix water is needed to convert all of the raw cement into gel.

If the w/c is above 0.33, increasing amounts of unused, and thus evaporable, water will be leftover after the cement has fully hydrated. Some of this unused water will occupy the remaining capillary pores in the paste that have not been filled by the newly created gel. The capillary pores are irregularly shaped, randomly distributed, and widely varying in size. Although most measure less than 400 water molecules across, a few do get as much as 10 times larger. The lingering interconnections that exist between the capillary pores in mature paste create the concrete’s permeability and consequent susceptibility to freeze/thaw damage.

Some of the leftover evaporable water—though in a much less mobile state— also will occupy the much smaller gel pores within the C-S-H itself. C-S-H gel is, in fact, highly porous—being riddled with myriad minute voids that comprise almost a third of its volume. Having nominal diameters in the range of only 8 to 12 water molecules, the gel pores are 1 to 2 orders of magnitude smaller than the capillary pores. The porosity of C-S-H gel remains fairly constant regardless of w/c or the degree of hydration: the more the hydration, the more the gel pore formation. The situation regarding the capillary pores is just the inverse: the more the hydration, the less the capillary pore volume.

Regardless of the actual w/c employed, the theoretical maximum volume to be occupied by the C-S-H gel upon complete hydration always will be about twice the volume of the raw cement powder. This makes the percentage of the fully hydrated paste occupied by the evaporable water solely a function of w/c: Percent of Paste Volume Containing Evaporable Water = 100 [ 1 - 2/(3 W/C + 1)] %

With regard to curling, cracking, and delamination, it is this remaining removable portion of the original mix water that causes all the problems.