Specific surface areas cements

Clinkers with increased C2S and reduced C3S contents are employed in the production of cements with reduced hydration heat evolution (see section 19), such as moderate heat of hardening cement (corresponding to ASTM Type II cement) and low-heat cement (corresponding to ASTM Type IV cement). To slow down the rate of heat evolution, cements employed for this purpose are usually ground to a relatively low specific surface area. Cements of this type are used in applications in which a reduced release of hydration heat is required, as in the constraction of dams and other bulk concrete structures (Kelham and Moir, 1992 Sone et al, 1992). High-C2S cUnkers, especially those with a reduced C3A content, are also constituents of some oil well cements (see section 27). L0W-C3S clinkers are less suitable for the production of blended cements, owing to the reduced amount of free calcium hydroxide produced in the hydration of such clinkers. 

Magnesium oxide is always blended with the zinc oxide prior to ignition. Magnesium oxide promotes densification of the zinc oxide, preserves its whiteness and renders the sintered powder easier to pulverize (Crowell, 1929). The sintered mixed oxide has been shown to contain zinc oxide and a solid solution of zinc oxide in magnesium oxide (Zhuravlev, Volfson Sheveleva, 1950). Specific surface area is reduced compared with that of pure zinc oxide and cements prepared from the mixed oxides are stronger (Crowell, 1929 Zhuravlev, Volfson Sheveleva, 1950). 

Recent applications of relaxation dispersion measurements to concrete or cement-based materials are promising for characterizing reactive nanopor-ous materials, the structure of which may evolve over time (75-78). The MRD profiles have provided, for the first time, a direct means for characterizing the specific surface area, Sp, of a hydrated cement-based material (79), without exposing the sample to extremes of temperature or pressure (80-83). The interest in such a surface area is to provide information on the microsctruc-ture and its impact on macroscopic or structural properties. The method is based on a clear separation of surface and bulk contributions of the overall... 

In the absence of knowledge of the surface area of cement hydrates available for adsorption at the time of addition, it is difficult to estimate how many layers of water-reducing admixture molecules are adsorbed, but attempts have been made [40] indicating that over 100 layers may be formed with calcium lignosulfonate and salicylic acid at normal levels of addition. However, these calculations were based on specific surface areas of 0.3-1.0 m g-l, whereas other studies [27, 38, 39] have indicated... 

Water-cement ratio (by wt) Nature of mix Air content (%) Paste content (%) Voids in concrete (%) Specific surface area (mm) Void spacing factor (mm)... 

Table 3.4 The effect of various air-entraining agents at different concentrations on the specific surface area and computed spacing factor of air bubbies in cement paste...

An increase in the water-cement ratio of cement pastes leads to greater air entrainment and a decrease in the specific surface area of bubbles. However, the spacing factor is relatively unchanged, as shown in Table 3.5 [14]. 

Table 3.5 The effect of water-cement ratio of cement pastes on the air content, specific surface area and computed spacing factor...

Cement Specific surface area Quantity of admixture required (ml 0.0368 m batch) for 4% air... 

The adsorption of superplasticizer to spherical cement particle surface decreases by 40% because of a decrease of the specific surface area and localization of the intestitial phase with gypsum. 

The tobermorite obtained in the hydration of tricalcium silicate (Ca3Si02), / -dicalcium silicate (/ -Ca2Si04), portland cement, and concrete is a colloid, with a specific surface area of the order of 300 sq. meters per gram. To give an idea of how the elementary particles of tobermorite look, Figure 7 is an electron micrograph of a few particles (obtained by L. E. Copeland and Edith G. Schulz at the Portland Cement Association Research and Development Laboratories). These particles look like fibers, but if you watch them closely, you see that they are very thin sheets, rolled up as one would roll up a sheet of paper. At the lower end the sheets are partly unrolled. When one prepares tobermorite by the reaction of lime and silica, one usually obtains crumpled sheets, which are not rolled up. The electron microscopists tell us that the sheets are very thin, of the order of a single unit cell in thickness. 

Fig. 4.1 Rosin-Rammler plot of the particle size distribution of a typical Portland cement. 5 = specific surface area attributable to particles of size smaller than. v. Open circuit grinding based on the data of Sumner et al. (S25).

The specific surface area of eement is eommonly determined directly by air. permeability methods. In the Lea and Nurse method (LI 5). a bed of cement / of porosity 0.475 is eontained in a cell through which a stream of air is f passed, and steady flow established. The specific surface area is caleulated ( from the density of the eement, the porosity and dimensions of the bed of j powder, the pressure differenee aeross the bed, and the rate of flow and ] kinematie viscosity of the air. In the Blaine method (B36), a fixed volume of I air passes through the bed at a steadily deereasing rate, whieh is controlled / and measured by the movement of oil in a manometer, the time required i being measured. The apparatus is ealibrated empirically, most obviously / using a cement that has also been examined by the Lea and Nurse method. The two methods gave elosely similar results. The Blaine method, though not absolute, is simpler to operate and automated variants of it have been devised. 

Other methods that have been used to determine specific surface areas of cements include the Wagner turbidimeter (W16) and BET (Brunauer-Emmett-Teller) gas adsorption. The former, as eonventionally used, gives very low results beeause of a false assumption that the mean diameter of the particles smaller than 7.5 pm is 3.8 pm, which is much too high. The BET method gives results two to three times higher than the air permeability... 

The specific surface area, like the PSD, is thus a quality whose value depends on how it is defined, and is liable to be affected by any pretreatment or conditions affecting the degree of flocculation. In practice, air permeability methods are widely used. Typical values are 300-350 m kg for modern ordinary Portland cements and 400-450 m kg " for rapidhardening Portland cements. 

Because the constituent phases of a cement are not equally easy to grind, different particle size fractions differ in composition. Gypsum, and its dehydration products, are concentrated in the finer fractions. Osbaeck and Jons (08) concluded that each 1% of gypsum contributed about 10m kg to the specific surface area in a typical case, some 15% of the total specific surface area is thus due to gypsum. The content of alite decreases, and that of belite increases, with increasing particle size (R12,G30), the contents of aluminate and ferrite phases being little affected. 

By definition, the kinetic curve of a cement is the weighted sum of the curves for its constituent phases as they occur in that cement. The reactivities of individual clinker phases were considered in Section 4.5 and some effects of particle size distribution, which is a particularly important variable, in Section 4.1.4. Although many data relating particle size distribution directly to strength exist, much less is known about its relation to degrees of reaction. Parrott and Killoh (P30) presented data indicating that the rate of hydration, as represented by that of heat evolution, was proportional to the specific surface area during the period of hydration in which the rate was controlled by diffusion. 

Brunauer and co-workers (B55,BI08) considered that the gel particles of the Powers-Brownyard model consisted of either two or three layers of C S-H, which could roll into fibres. D-drying caused irreversible loss of interlayer water, and the specific surface area could be calculated from water vapour sorption isotherms, which gave values in the region of 200m g for cement paste. Sorption isotherms using N2 give lower values of the specific surface area this was attributed to failure of this sorbate to enter all the pore spaces. 

In principle, isotherms at low partial pressures of the sorbate may be used to determine specific surface areas by the Brunauer-Emmett-Teller (BET) method (G64). In this method, it is assumed that molecules of the sorbate are adsorbed on surfaces that can include the walls of pores, provided that the distance between molecules on opposing walls is large compared with molecular dimensions. From a plot derived from the isotherm, and given the effective cross-sectional area of the sorbate molecule, the specific surface area of the sorbent and the net heat of adsorption are obtained. Using water as sorbate, specific surface areas of about 200 m per g of D-dried paste have typically been obtained for mature cement pastes of normal w/c ratios... 

Of the pfa characteristics that influence reactivity, the glass content appears to be much the most important, but specific surface area, glass composition and the effect of stress in the glass caused by the crystalline inclusions may also be relevant (U17). Of external factors, the RH, temperature (C43) and alkali content of the cement are probably the most important. Sulphate ion may also enhance reactivity by promoting the removal of AF from the glass (U17). The rates of the pozzolanic reaction and of strength development are more sensitive to temperature than are those of hydration and strength development for pure Portland cements (e.g. Ref. H52). 

Costa and Massazza (C44) concluded from a study of natural pozzolanas of varied types that reactivity in mixtures with CH at w/s = 2 and 40 C depends during the first 28 days on the specific surface area and at later ages on the contents of Si02 and AI2O3 in the active constituents. A comparative study of five natural pozzolanas and three low-CaO pfas in pastes with cement showed that the CH contents of the pozzolanic cements were considerably lower than those of the pfa cements at 3-60 days, but virtually the same at 90 days, the pozzolanas thus appearing to react more rapidly than the pfas at early ages but more slowly later. Determinations of the unreacted mineral admixture in pastes with CH showed that at 90 days 23-30% of the natural pozzolana had reacted, compared with 11-15% for the pfas. The similarity in CH contents suggests, however, that these values may not apply to mixtures with cement. 

Regulated-set cement and jet cement are modified Portland cements in which the normal aluminate phase is replaced by CuA CaF, through the use of a raw mix containing CaF,. Uchikawa and Tsukiyama (U2I) gave chemical (Table 10.4) and phase compositions of two jet cements. Botli contained approximately 60% of alite, 20% of C, i A CaF,. 1% of belite and 5% of ferrite. Admixtures are required to control the rate of reaction of the C, 1 A CaFj and the nature of the products. One of the cements included a proprietary retarder based on citric acid, togetlier with 2"/o of CaC O, . The other contained 2.5% of hemihydrale. In each case, Na,S04 (T o) and anhydrite were also present. The specific surface areas were around 550 m kg . 

Cement Ordinary portland cement (specific gravity 3.16, specific surface area 327,000 mm2/g)... 

---