5.2     The Induction Period

    The induction period, labeled Stage 2 in Figure 5-1, is a 1-2 hour period of inactivity that separates the initial short burst of reaction that occurs when cement and water first come into contact from the main hydration period that leads to set. This behavior is quite important because it prevents the cement from setting too quickly. The fact that a freshly mixed paste remains fluid and workable as the concrete is being placed and then sets shortly thereafter is one of the keys to the versatility and widespread use of portland cement concrete.
    The reason for the induction period, that is, the microstructural or chemical basis for the observed behavior, has been the subject of much speculation and debate by cement researchers. Before discussing the possibilities, lets review some relevant and well-established facts. Tricalcium silicate (C3S) has a very high intrinsic solubility, as evidenced by the rapid but short-lived release of heat during Stage 1. The main hydration period, Stage 3, occurs via a process of nucleation and growth, and is associated primarily with C3S hydration. Thus small nuclei of C-S-H and CH precipitate out of the pore solution and then grow in size. The more nuclei there are to grow, the greater the rate of hydration. Thus the increase in the rate of hydration during the first part of Stage 3 occurs because the number of nuclei is increasing. Finally, it should be noted that both OPC pastes and C3S pastes exhibit an induction period.
Taylor [ ] has divided the various arguments that have been proposed for the induction period into four main hypotheses. These are:
1) The induction period is simply an integral part of the nucleation and growth process, representing the period of time when the initial nuclei are forming. This is essentially arguing that there is, from a mechanistic point of view, no induction period after all.
2) The induction period occurs because the initial nuclei that form are "poisoned" by SiO2 and cannot grow until the solution becomes supersaturated.
3) A continuous membrane surrounding the particles forms during Stage 1, enclosing a small region of inner solution. The membrane prevents the ions dissolved in the inner solution from reaching the main pore solution (except perhaps very slowly). The advantage of this hypothesis is that there is a natural explanation for the end of the induction period, which is that osmotic pressure caused by the difference in concentrations inside and outside the membrane causes it to burst.
4) A thin, continuous layer of hydration product forms on the particles during Stage 1, preventing the C3S from dissolving, except perhaps very slowly. This is similar to argument (3) above except the layer forms directly on the surface of the particles. As with (3), nucleation and growth out in the pore solution cannot occur until this layer is removed or made permeable.
    Experimental evidence has allowed some of these arguments to be eliminated. While hypothesis (1) is partially correct, in that a nucleation and growth process can give the appearance of an initial period of no reaction while the first nuclei form, in the case of C3S and OPC hydration the observed period of slow reaction is not compatible with the very high intrinsic solubility of C3S, as the concentration of calcium and silicon in solution increase only very slowly during the induction period. Fitting of the hydration kinetics of C3S measured by thermal calorimetry with a nucleation and growth equation has shown that there is indeed a separate induction period that becomes shorter at higher temperatures [ ]. Furthermore, retarding chemicals such as simple sugars can greatly extend the induction period, for days or even weeks. Even after such a long dormant period, the subsequent Stage 3 hydration occurs at the normal rate. This is discussed further in Chapter 9.
    Hypothesis 2 is directly countered by observations that the addition of finely divided silica, such as silica fume, actually accelerates (rather than retards) the rate of early reaction by providing nucleation sites. As with the first hypothesis, this hypothesis is not consistent with the very high solubility of C3S. Hypothesis 3 is a good example of a beautiful and well thought out theory (proposed by Double [ ]) that simply does not pan out. Microscopic examination of C3S pastes show no evidence of formation of such a membrane.
    This leaves hypothesis 4, the formation of a thin layer directly on the cement or C3S particles, as the favored explanation. This layer, called a protective layer since it keeps C3S from dissolving, is believed to be a metastable calcium-silicate-hydrate phase with a different structure and morphology than that of C-S-H gel. With this hypothesis, the end of the induction period is initiated by the formation of nuclei of C-S-H gel on the outside of the protective layer. Since the C-S-H gel is more thermodynamically stable than the protective layer, these nuclei grow and convert the layer to C-S-H gel. This changes the morphology of the layer, making it more permeable, and the rate of hydration increases as more C-S-H gel nuclei form and grow.
    The above description explains much of the experimental evidence but is neither complete nor universally accepted even in such a brief form. The induction period remains a subject of both experimental and theoretical investigation, but new theories do not always account for all of the facts. For example, it has been proposed that the induction period is caused by a calcium sulfoaluminate layer [ ], but this cannot explain the presence of an induction period in pure C3S pastes that contain no sulfur or aluminum.

<-previous    next->


Home Pictures Site Map About the Authors Links