This research reports on the effects of including natural pozzolans in two Portland cements with different mineralogical compositions, with and without excess gypsum at amounts equivalent to 7.0% SO3. The main analytical techniques used to study these effects were: the amount of water needed to make a paste of normal consistency, the 2-day Frattini pozzolanicity test and conduction calorimetry. The results obtained showed that these natural pozzolans caused contradictory (accelerating and retarding) effects on the rheology of the resulting cements, depending on the mineralogical composition of the respective Portland clinkers as well as the reactive chemical composition of the pozzolans, in particular their reactive alumina content (Al2O3r−). The addition of gypsum also caused acceleration and delays in the calorimetric evolution of the resulting pastes, which proved to be heavily dependent upon the more or less aluminic chemical character of the natural pozzolans studied. This, in turn, was conditioned by the higher or lower Al2O3r− content (for the SiO2r− content was of a very similar order of magnitude in all three pozzolans analyzed). The Al2O3r− content was likewise responsible for paste behaviour in the afore-mentioned trials and analyses, and the pozzolanic activity exhibited by the compound was found to be more specific than generic, indirectly stimulating C3A hydration more intensely and rapidly than C3S hydration in PC1, one of the two Portland cements used. Indeed, when these natural pozzolans exhibited such prior pozzolanic activity in the second cement studied, PC2, the hydration of its 79.5% of C3S was not indirectly stimulated to the same degree; rather, the contrary effect was observed, i.e., this cement was physically diluted by the three pozzolans. Pozzolan O stimulated hydration directly and non-directly more than indirectly, while pozzolan C acted conversely, and A exhibited varying combinations of the two patterns. The physical state of the reactive alumina, Al2O3r−, in these three natural pozzolans, must be more amorphous than vitreous, i.e., resembling metakaolin more than fly ash in this regard. That notwithstanding, the reactive alumina content in each pozzolan must have conditioned the water/cementitious material ratio obtained for the respective blends with both types of Portland cement (a finding that could be used in future for speedy, simple, reliable and economical characterization), as well as their specific pozzolanicity developed and the rate and total heat of hydration generated by such blended cements.
Fly ashes from the combustion of coal thermal power stations are commonly incorporated into portland cements and/or concretes
and mortars. The chemical and morphological composition of fly ashes, together with their particle size, make them suitable
as pozzolanic(non-calcic) or pozzolanic/hydraulic(highly calcic) additions to manufacture such building materials. This work
focuses on the incorporation of two different fly ashes (non-calcic but of very different Fe2O3(%) contents, fineness and morphology) to two ordinary portland cements (of very different mineralogical composition as well),
to determine the effects those have and the interactions they produce in the hydration reactions of portland cement. The main
techniques employed for this study have been: conduction calorimetry and Frattini test; secondary techniques applied have
also been: determination of setting times and analysis by X-ray diffraction and SEM. Analysis of the results obtained permitted
to find different effects of fly ash addition on the hydration reactions of portland cements. Thus, dilution and stimulation
effects augment with the increased fly ash percentage. Delay and acceleration of the reactions depend mainly on the type of
portland cement and are accentuated with increased fly ash contents. Their behaviour as concerns heat dissipation mainly,
depends on the type of fly ash used and is more pronounced with increased cement replacement. On the other hand, the pozzolanic
activity of these fly ashes has been revealed at 7 and 28 days, but not at 2 days. Finally, pozzolanic cements can be manufactured
using different portland cements and/or types of fly ashes, in the appropriate proportions and compatible qualities, depending
on the effect(s) one wish to enhance at a specific age, which is according to previous general conclusions drew out of sulphate
attack and chloride attack researches.
This paper analyzes the effect of fly ash chemical character on early Portland cement hydration and the possible adverse effects
generated by the addition of gypsum. Behaviour was analyzed for pure Portland cements with varying mineralogical compositions
and two types of fly ash, likewise differing in chemical composition, which were previously characterized under sulphate attack
as: silicic-ferric-aluminic or aluminic-silicic ash in chemical character, irrespective if they are in nature, siliceous or
siliceous and aluminous materials according to the ASTM C 618-94a.
The experimental results showed that water demand for paste with a normal consistency increased with the replacement ratio
in fly ash with a more aluminic than silicic chemical character, whereas it declined when silicic-ferric-aluminic ash was
used. On the other hand, the differences between the total heat of hydration released at the first valley and the second peak
also clearly differentiated the two types of ash. While the relative differences increased in the more aluminic than silicic
ash, they declined in the more silicic than aluminic. In another vein, the findings indicate that within a comparable Blaine
fineness range, the reactive alumina (Al2O3r−) content in pozzolanic additions has a greater effect on mortar strength than the reactive silica (SiO2r−) content, at least in early ages up to 28 days. Finally, the adverse effect generated in the presence of excess gypsum is
due primarily to the chemical interaction between the gypsum and the C3A in the Portland cement and the reactive alumina (Al2O3r−) in the fly ash.
This new study must be regarded to be a direct outcome of two previous studies published by these same authors, which were
conducted to respond to interesting questions brought out about the effect of silica fume, SF and metakaolins, M and MQ, on
the heat of hydration of portland cements, PC, with very different C3A and C3S contents. The answer to these so interesting questions has been the primary objective of the present research. For this
purpose, the same PC, PC1 (14% C3A) and PC2 (≈0% C3A), metakaolins, silica fume and blended cements were once again used more 60/40 for sulphate attack, and the same analytical
techniques (CC, pozzolanicity and XRD analysis) and parameters determined as well. In this new research, the sulphate attack
was determined by two accelerated methods: Le Chatelier-Ansttet and ASTM C 452-68.
The experimental results of sulphate attack mainly, have demonstrated definitively that the high, rapid and early pozzolanic
activity exhibited by SF also is, as in the case of the two metakaolins, more specific than generic, for it indirectly stimulated
greater C3A than C3S hydration, but only in the first 16 h monitored in this study. Thereafter it is the contrary, i.e., anti- or contra-specific
for the same purpose. And the longer the hydration time, the more anti- or contra-specific it became, since, when exposed
to sulphate attack, SF blended cements resisted or even prevented the aggressive attack against PC1 which, with a higher C3A content than PC2, was the more vulnerable of the two. By contrast, metakaolin MQ not only failed to hinder or prevent the
attack, but heightened its effects, rendering it more intense, aggressive and rapid, leading to what could be called a rapid
In this work two aluminic pozzolans (metakaolins) and a non-pozzolan were added to two Portland cements with very different
mineral composition, to determine the effect on the rate of heat release and the mechanisms involved. The main analytical
techniques deployed were: conduction calorimetry, pozzolanicity and XRD.
The results showed that the two metakaolins induced stimulation of the hydration reactions due to the generation of pozzolanic
activity at very early stage, because of their reactive alumina, Al2O3r− contents, mainly. Such stimulation was found to be more specific than generic for more intense C3A hydration than C3S, at least at very early on into the reaction, and more so when 7.0% SO3 was added, and for this reason, such stimulation is described as ‘indirect’ to differentiate it from the ‘direct’ variety.
As a result of both stimulations, the heat of hydration released is easy to assimilate to a Synergistic Calorific Effect.
Authors:V. Rahhal, O. Cabrera, R. Talero, and A. Delgado
The use of active mineral additions is an important alternative in concrete design. Such use is not always appropriate, however, because the heat released during hydration reactions may on occasion affect the quality of the resulting concrete and, ultimately, structural durability. The effect of adding up to 20% silica fume on two ordinary Portland cements with very different mineralogical compositions is analyzed in the present paper. Excess gypsum was added in amounts such that its percentage by mass of SO3 came to 7.0%.
The chief techniques used in this study were heat conduction calorimetry and the Frattini test, supplemented with the determination of setting times and X-ray diffraction. The results obtained showed that replacing up to 20% of Portland cement with silica fume affected the rheology of the cement paste, measured in terms of water demand for normal consistency and setting times; the magnitude and direction of these effects depended on the mineralogical composition of the clinker. Hydration reactions were also observed be stimulated by silica fume, both directly and indirectly – the latter as a result of the early and very substantial pozzolanic activity of the addition and the former because of its morphology (tiny spheres) and large BET specific surface. This translated into such a significant rise in the amounts of total heat of hydration released per gram of Portland cement at early ages, that silica fume may be regarded in some cases to cause a synergistic calorific effect with the concomitant risk of hairline cracking. The addition of excess gypsum, in turn, while prompting and attenuation of the calorimetric pattern of the resulting pastes in all cases, caused the Portland cement to generate greater heat of hydration per gram, particularly in the case of Portland cement with a high C3A content.
Authors:V. Rahhal, O. Cabrera, A. Delgado, C. Pedrajas, and R. Talero
The need for cements or other cementitious materials that afford high early age mechanical strength has led to the use of
extremely reactive pozzolanic additions such as silica fume, nanosilica, metakaolin and similar. The inclusion of the right
proportion of such pozzolanic additions stimulates portland cement hydration, i.e., directly, as they are initially moistened by the mixing water, non-directly when they act as “seed crystals”, and indirectly, because of the pozzolanic reaction between the addition particles and the portlandite forming from the portland cement components
hydration; since this reaction is characterized by its intensity and speed, when its occurs it prevails over the other two.
Indirect stimulation also causes the fraction of portland cement in the blend to release more heat of hydration than pure portland
cement, and its does so on a scale consistent with the existence of a calorific synergic effect. Such greater heat is released in the early stages of hydration primarily by C3A and C3S that react with the mixing water to respectively generate ettringite and hydrated calcium silicates. When portland cements
have a low to nil C3A content, less heat of hydration is released due to the absence of an AFt phase that could be transformed into AFm. However, when extremely active pozzolanic additions, such as silica fume, are used, ettringite forms from C4AF, further contributing to origin amounts of hydration heat released comparable to the above calorific synergic effect.
Authors:V. F. Rahhal, E. F. Irassar, M. A. Trezza, and V. L. Bonavetti
The calorimetric technique provides continuous, direct, and general measurements of the course of coexisting reactions and their interactions during hydration of blended cement at early age. In this article, this technique is used to analyze the influence of compositional and process variables on the early age hydration of Portland limestone cements (PLC) made by intergrinding in a full size-cement plant. Eight cements, the vertices of 23 factorial design, were made with a limestone filler content (LF) of 0 and 24%, a gypsum content (GC) of 2.5, and 5.0%; and a fineness, measured as that fraction retained on a 45 μm sieve (R45), of 5 and 18%, to study their effects on the heat released. In addition, a PLC with a composition nearly to the center point of 23 designs was analyzed. Measurements were performed on cement pastes (w/cm = 0.4) using a semiadiabatic differential calorimeter operating at 20 °C during 48 h. At different time, the heat released was determined and it was modeled using a linear mathematical model including the three variables (LF, R45, CG) and their interactions. The significance of the model, the variables and the interactions was judged using the analysis of variance. Results of model show that heat released is reduced by LF due to physically dilution phenomenon, which is directly proportional to LF content. The R45 exerts its major influence during the development of second peak (12–21 h) but later its effect declines to null contribution. GC retards and attenuates the hydration reactions moderately until 30 h, and then its increase contributes to Qt due to the formation of ettringite and its transformation. The only significant interaction was LF with R45 during the second peak development. Results present good correlation with the isolate measurement of compressive strength at 12, 24, and 48 h.