Portland Cement Hydration, Part 2

Hydration of Portland cement

The hydration of Portland cement is rather more complex than that of the individual constituent minerals described above. A simplified illustration of the development of hydrate structure in cement paste is given in Figure 1.9.

When cement is first mixed with water some of the added calcium sulfate (particularly if dehydrated forms are present, and most of the alkali sulfates present, dissolve rapidly.

If calcium langbeinite is present then it will provide both calcium and sulfate ions in solution, which are available for ettringite formation.

The supply of soluble calcium sulfate controls the C3A hydration, thus preventing a flash set. Ground clinker mixed with water without added calcium sulfate sets rapidly with heat evolution as a result of the uncontrolled hydration of C3A.

The cement then enters a dormant period when the rate of loss of workability is relatively slow. It will be

more rapid, however, at high ambient temperatures (above 25°C).

Setting time is a function of clinker mineralogy (particularly free lime level), clinker chemistry and fineness. The finer the cement and the higher the free lime level, the shorter the setting time in general.

Cement paste setting time is arbitrarily defined as the time when a pat of cement paste offers a certain resistance to penetration by a probe of standard cross-section and weight.

Setting is largely due to the hydration of C3S and it represents the development of hydrate structure, which eventually results in compressive strength.

The C–S–H gel which forms around the larger C3S and C2S grains is formed in situ and has a rather dense and featureless appearance when viewed using an electron microscope.

This material is formed initially as reaction rims on the unhydrated material but as hydration progresses the anhydrous material is progressively replaced and only the largest particles (larger than ~30 microns) will retain an unreacted core after several years, hydration. This dense hydrate is referred to as the ‘inner product’.

The ‘outer hydration product’ is formed in what was originally water-filled space and also space occupied by the smaller cement grains and by interstitial material (C3A and C4AF).

When viewed using an electron microscope this material can be seen to contain crystals of Ca(OH)2 , AFm/AFt and also C–S–H with a foil-or sheet-like morphology. The structure of the outer product is strongly influenced by the initial water-to-cement ratio, which in turn determines paste porosity and consequently strength development.

Figure 1.9 Simplified illustration of hydration of cement paste.

The hydration of Portland cement involves exothermic reactions, i.e. they release heat. The progress of the reactions can be monitored using the technique of isothermal conduction calorimetry (Killoh, 1988).

Figure 1.10 Heat of hydration of a cement paste determined by conduction calorimetry at 20°C.

The shoulder on the main hydration peak which is often seen at ~16 hours is associated with renewed ettringite formation which is believed to occur as a result of instability of the ettringite protective layer. In some cements with a low ratio of SO3 to C3A it may be associated with the formation of monosulfate.

 

The heat release is advantageous in cold weather and in precast operations where the temperature rise accelerates strength development and speeds up the production process.

However, in large concrete pours the temperature rise, and in particular the temperature difference between the concrete core and the surface can generate stresses which result in ‘thermal cracking. Figure 1.11 illustrates the influence of concrete pour size on concrete temperature for a typical UK Portland cement. The data were obtained using the equipment described by Coole (1988).

The temperature rise experienced depends on a number of factors, which include:

  • concrete placing temperature
  • cement content
  • minimum pour dimensions
  • type of formwork
  • cement type (fineness, C3S and C3A contents)

Cement heat of hydration (during the first ~48 hours) is highest for finely ground cements with a high C3S content (>60%) and a high C3A content (>10%).

By 28 days a typical Portland cement cured at 20°C can be expected to be ~90% hydrated.

The extent of hydration is strongly influenced by cement fineness and in particular the proportions of coarse particles in the cement. Cement grains which are coarser than ~30 microns will probably never fully hydrate.

Thus, cement particle size distribution has a strong influence on long-term compressive strength. Cement produced in an opencircuit mill with a 45 micron sieve residue of 20% may give a 28-day strength ~10% lower than that of a cement produced from the same clinker but ground in a closed-circuit mill with a 45 micron sieve residue of 3% (Moir, 1994).

Elevated temperature curing, arising from either the semi-adiabatic conditions existing in large pours or from externally applied heat, is associated with reduced ultimate strength.

This is believed to be due to a combination of microcracks induced by thermal stresses but also a less dense and ‘well-formed’ microstructure.