INFLUENCE OF TEMPERATURE, CEMENT AND PLASTICIZER TYPE ON THE RHEOLOGY OF PASTE

Transcription

1 INFLUENCE OF TEMPERATURE, CEMENT AND PLASTICIZER TYPE ON THE RHEOLOGY OF PASTE H. Vicar SINTEF Building and Infrastructure, Norway Abstract The effects of temperature on the rheological behaviour of cement pastes with three different types of plasticizers namely lignosulphonate, naphthalene sulphonate formaldehyde condensate and polyacrylate have been investigated. One type of cement was used in order to study the influence of hydration, while limestone was used as a non-reactive model system. The adsorbed amounts of plasticizers were measured for pastes of cement and limestone within a temperature range of 11 to 40 o C. The type of plasticizer was found to determine its affinity towards the limestone or cement surface at a given temperature. The plasticizer absorption seemed, moreover, to be governed by both kinetic energies and entropic effects of the organic molecules. Keywords: rheology, plasticizers, temperature, cement, limestone 1. INTRODUCTION The effect of temperature on the rheology of Portland cement is important not only because of climatic variations prevailing throughout the world, but also because of the exothermic nature of early hydration reactions involved. There have been few studies made on the effect of temperature on the rheology of cement pastes and its effect on the adsorption of superplasticizers. This paper investigates therefore the effect of temperature on the rheological behaviour of cement pastes with lignosulphonate (LS), naphthalene sulphonate formaldehyde condensate (SNF) and polyacrylate (PA) as plasticizers. Limestone was used as a non-reactive model system for cement. Adsorbed amounts of on cement and limestone as a function of temperature have been studied. 2. EXPERIMENTAL 371

2 SCC 2009-China, June , Beijing China A Portland cement CEM I 52.5 R-LA was used for the study. It consisted of 50.4 % C 3 S, 22.0 % C 2 S, 0.5% C 3 A, 15.6 % C 4 AF and 4.9 % C S according to Bogue calculations and had a Blaine surface fineness equal to 364 m 2 /kg. High purity limestone powder (98-99 % CaCO 3 ) with a Blaine surface fineness 486 m 2 /kg was used as a non-reactive model material. Three different categories of plasticizers have been used namely Sodium lignosulphonate (LS), sugar reduced and molecular size enriched by ultra filtration. Molecular weight of the polymer was 51,000. Sodium naphthalene sulphonate formaldehyde condensate (SNF) in water solution with 42% solids. The number of structural units (n) in the SNF molecule was presumably between 10 and 20. The molecular weight is 2,350 for n = 10 (number of structural units) and 4,700 for n = 20. Polyether grafted polyacrylate (PA) in water solution containing 18 % solids. The molecular weight of the polymer was 220,000. Pastes of limestone powder were prepared with a particle volume fraction of 0.60 (corresponding to a w/ls ratio of 0.25) and artificial pore water which was prepared of NaOH and KOH in distilled water with a K/Na molar ratio equal to 2 and ph = The water was heated or cooled to obtain paste temperatures in the range of o C. The plasticizers were added to the limestone together with the artificial pore water in the concentration 0.32% by weight for SNF and LS and 0.10% by weight of PA. Cement pastes were prepared with a particle volume fraction equal to (w/c=0.40) with distilled water and SNF, LS and PA as plasticizers. The water was heated or cooled in order to produce pastes with varying temperatures. The plasticizers were added to the cements together with the water. Table 1 lists the various experimental combinations of materials and temperatures. Table 1: Combinations of materials, admixtures and temperatures used for the experiments Adsorbents Plasticizer Plasticizer dosage (% of cement weight) Paste temperature ( o C) SNF , 25, 30, 33, 40 Limestone LS , 25, 33, 43 PA , 24, 34 SNF , 18, 27, 39 Cement LS , 26, 32, 40 PA , 18, 31 The cement and limestone pastes were mixed in a high shear blender from Braun (MR5550CA) for ½ minute and rested for 5 minutes before they were blended once more for 1 minute. The paste temperatures were measured after the last mixing sequence. A sample of the paste was transferred to a parallel plate rheometer MCR 300 produced by Physica. The temperature of the rheometer was set equal to the temperature of the paste and the rheological properties were recorded 11 minutes after water addition. The following measures were made in order to minimize paste dehydration during the measurements: The parallel plate measuring system of the rheometer was covered with a plastic ring and a metallic lid during the 372

3 measurements. The upper plate of the rheometer had furthermore a water trap filled with water to ensure saturated water pressure. The rheological measurement sequence is described below: 1. 1 minute with constant shear rate ( γ& ) of 100 s -1 to stir up the paste 2. Stress (τ) shear rate ( γ& ) curve with linear sweep of γ& from 200 down to 2 s -1 in 30 points lasting 6 s each (down curve). 3. Stress (τ) shear rate ( γ& ) curve with linear sweep of γ& from 2 up to 200 s -1 in 30 points lasting 6 s each (up curve). Flow resistance was calculated by integrating the area under the down curve within a shear rate range of s -1 : The pore water of the remaining paste was filtrated from the pastes through 0.45μm filters 11 minutes after water addition. The concentration of plasticizer left in solution was measured by UV adsorption spectrum for pastes with added SNF and LS (292 and 283 nm were used respectively). Total Organic Carbon (TOC) measurements were performed on pastes with added PA. The densities of the extracted pore water were measured with pycnometers and the kinematic viscosities (m 2 /s) were measured with an Ostwald viscometer produced by Schott-Geräte. The apparent viscosity (Pa s) was derived by multiplying the measured density and the kinematic viscosity. 3. RESULTS 3.1 Limestone paste rheology The flow curves of limestone powder pastes with SNF are given in Figure 1, while the calculated flow resistances are given in Figure 2. The figures illustrate that the shear stresses do not increase steadily with increasing temperature. It can, moreover, be seen that pastes with SNF and LS follow the same trends although pastes with SNF generally had higher flow resistances pastes with LS. The shear stresses of pastes with PA seem to increase linearly with increasing temperature. It is, however, important to notice that fewer measurements are made for pastes with PA than for than for pastes with SNF and LS. 373

4 250 Shear Stress (Pa) oc 33 oc 30 oc 25 oc 16 oc Shear Rate (s -1 ) Figure 1: Flow curve for limestone powder pastes as a function of temperature Flow Resistance (Pa/s) SNF LS PA Temperature ( o C) Figure 2: Flow Resistance for limestone pastes with SNF, LS and PA as a function of temperature Increased flow resistance with increasing temperature cannot be explained by increased solubility of limestone in water since the limestone solubility is known to decrease with increasing temperature [1]. Reduced repulsive potential of the limestone particles, V R, and thus increased degree of flocculation with increasing temperature can also be ruled out since a temperature change from o C would increase the repulsive potential by a factor of only (311/288) 2 = 1.17 when equation (1) is used. The equation is a general expression for the repulsive potential between two spherical particles (radii a 1 and a 2 ) with distance H were T is 374

5 the temperature, ε is the permittivity of the liquid, k is Boltzmann s constant, γ is a parameter of the surface potential, e is the ion charge, z is the ionic valence and κ expresses the effect of the ion charge and concentration [2]. V R π ε a a k T 1 2 γ 1 γ 2 = ( H ) ( a a ) e z exp κ (1) Nawa et al. [3] similarly found that the degree of particle flocculation hardly varies in the temperature range of 10 o C to 30 o C. Altered flowability with increasing temperature should thus be caused by altered properties of the pore water and/or the plasticizer molecules. 3.2 Properties of pore water extracted from limestone paste The apparent viscosity of the pore water decreased as much as 42% when the temperature increased from 15 to 43 o C as shown by Table 2. These results are found to be consistent with measured viscosities of pure water. Thus, the viscosity of the pore water does not seem to depend upon how the behavior of polymers might alter with increasing temperature. Decreased apparent viscosity of the continuous phase should according to the Krieger- Dougherty equation [4] result in the same reduction in paste viscosity when all other factors are equal. The flow curves illustrate, however, that the effect of the reduced viscosity of the pore water is suppressed by other factors resulting in increased flow resistance with increasing temperature. A possible explanation of these results are given by Figure 3 which shows the relative adsorbed amounts of SNF, LS and PA plasticizers on limestone as a function of temperature. The figure illustrates that the adsorbed amounts of SNF reach a maximum at approximately 25 o C before it start to decrease. The adsorbed amounts of LS are, on the other hand, generally found to decrease in the whole temperature range, while the adsorption of PA is found to be independent of temperature within the range of o C. The shape of the adsorption curves for limestone pastes may partly be explained by the entropy of the polymers. The entropy of a polymer molecule decreases when it adsorbs onto a solid surface because the configuration of the polymer becomes restricted. Nawa et al. [3] have found that that the adsorption of polymer onto a solid surface is an exothermic reaction which causes the adsorption to decrease with increasing temperature. Loss of entropy by adsorption onto a solid surface is, moreover, found to be less for large macromolecules than for smaller ones [2]. This might explain why the adsorption of the relatively small SNF molecules is more temperature dependent than the larger LS and PA molecules. As the temperature increases, smaller molecules will, additionally, obtain a higher kinetic energy than bigger ones and causing them to dislocate in a higher degree from the limestone surface. 375

6 Table 2: Viscosity of pore water extracted from limestone pastes as a function of temperature. Viscosities of pure water as a function of temperature are included for comparison. Limestone -Lignosulphonate Pure water Temperature ( o C) Measured viscosity Reduction of viscosity Measured viscosity Reduction of viscosity (%) (mpa s) (%) (mpa s) Consumed (% of added) Temperature ( o C) SNF LS PA Figure 3: Adsorbed amounts SNF, LS and PA on limestone as a function of temperature 3.3 Cement paste rheology Figure 4 illustrates that the flow resistances of cement pastes grows exponentially with increasing temperature. The volume concentration of solids will increase in a twofold manner as the temperature increases; through the formation of hydration products and consumption of water. The Krieger-Dougherty equation [4] correspondingly postulates that the viscosity of a particle suspension increases exponentially with increasing particle concentration. Justnes et al. [5] calculated that even a minor degree of hydration will markedly affect the paste workability since 3% hydration of cement paste with w/c = 0.40 increased the apparent viscosity by 22%. 376

7 Flow Resistance (Pa/s) LS SNF PA Temperature ( o C) Figure 4: Flow resistance measured for cement pastes with SNF, LS and PA as a function of temperature The literature reports somewhat contradicting effects of temperature on the rheology of paste: Heikal et al. [6] measured the rheology of cement pastes (w/c = 0.30) with the Rheotest cell with and without 1% by cement weight polycarboxylic acid (PC) in the temperature range of o C. They found that increased temperature resulted in a sharp decrease of the shear stresses of cement pastes both with and without superplasticizer. Nawa et al. [3] measured the fluidity (slump spreading area) of Portland cement pastes containing grafted co-polymers, SNF and SMF as a function of temperature ranging from 10 to 30 o C. They found that the flow values for cement pastes with co-polymers clearly depended on the temperature and that it generally exhibited a minimum at 20 o C. A similar trend was found for SMF. For SNF, however, the mode of temperature dependence was found to vary with the added dosage. Golaszwski and Szwabowski [7] investigated similarly the influence of temperature (10, 20 and 30 o C) on the rheological properties of mortars with different w/c ratios and with SNF and polycarboxylate ester as superplasticizers. They found that an increase of temperature generally resulted in increased g-value (yield stress) while the h-value (plastic viscosity) decreased. Jolicoeur et al. [8] found, on the other hand, that the rheological response to altered temperature depended on the composition of the cementitious paste. 3.4 Properties of the pore water extracted from cement paste The viscosity of the pore water decreased with increasing temperature in a similar manner as for limestone pastes as shown in Table 4. This illustrates that increased flow resistance with increased temperature is governed by the particle phase of the cement paste. Amounts of plasticizer consumed (adsorbed and intercalated) by the cement as a function of temperature are depicted in Figure 5. It can be seen that the SNF and LS consumption curves reach a plateau. This might be caused by two opposing effects, namely increased number of adsorption sites and reduced adsorption with increased temperature (as seen and 377

8 discussed for limestone pastes). Reduced plasticizer adsorption with increasing temperature might be caused by altered thermodynamic properties of the polymer but also different plasticizer adsorption capacity of unhydrated cement and hydration products. It is unclear if the hydration products adsorb more or less plasticizer than the clinker minerals: Uchikawa et al. [9] claimed that the hydration products including CSH and ettringite adsorb less SNF and LS than cement. Prince et al. [10] suggested, on the other hand, that SNF might adsorb somewhat stronger on the initial ettringite germs than on the anhydrous constituents of the cement paste. The morphology of the hydration products will, moreover, alter with increased temperature. Verbeck and Helmuth [11] believed that dense zones of hydration products would form around the hydrating grains at higher temperatures. Kjellsen et al. [12,13] confirmed the existence of the "shells" by backscattered electron images of mature pastes. The shells were distinguishable in cement pastes hydrated at 20 o C and were denser at higher temperatures. It is not clear how the hydration products formed at elevated temperatures effect the plasticizer consumption. Table 4: Viscosity of pore water extracted from pastes of cement with SNF, LS and PA as a function of temperature Cement Plasticizer Viscosity type (Pa s) 12 SNF SNF SNF SNF LS LS LS LS PA PA PA Temperature ( o C) Reduction of viscosity (%) Figure 5 shows that the amounts of PA consumed by cement increased linearly with increasing temperature. No plateau was reached as seen for SNF and LS. This might be explained by the corresponding limestone curves (Figure 3) which show that the adsorption of PA is independent of the given temperature range. The adsorptive behavior of PA molecules are therefore mainly governed by increased number of adsorption sites created by cement hydration and not so much by the thermodynamic properties of the polymer. 378

9 95 Consumed (% of added) Temperature ( o C) SNF LS PA Figure 5: Amounts SNF consumed by cement as a function of temperature Limestone was used as a non-reactive model material for cement. Flow resistance values measured on limestone pastes were somewhat fluctuating, but were generally found to increase with increasing temperature. The adsorbed amounts of SNF and lignosulphonate on limestone were found to decrease after reaching a maximum which occurred at approximately 25 o C. SNF seemed to have a stronger dependency of the temperature than lignosulphonate. Decreased amounts of adsorbed plasticizer with increasing temperature might be explained by the molecules obtaining increased kinetic energy and/or by an entropic effect. Adsorbed amounts of polyacrylate on limestone seemed to be independent of paste temperature in the range of o C which might be caused by low reduction of entropy at adsorption due to its short backbone and long, grafted chains. The flow resistance of cement paste increased exponentially with increasing temperature. Amounts of consumed (adsorbed and intercalated) plasticizer by cement reached, however, a plateau in the case of SNF and lignosulphonate. This finding might be caused by two opposing effects namely: Increased number of adsorption sites due to increased hydration rate and reduced adsorption due to increased kinetic energy and/or reduced entropy of the plasticizer. Amounts of polyacrylate consumed by cement increased linearly with increasing temperature. Increased consumption of PA by the cements is, thus, probably governed by the increased number of adsorption cites due to increased hydration rate and not so much by the thermodynamic properties of the polymer. REFERENCES [1] D.E. Macphee, and E.E. Lachowski, in Lea s Chemistry of Cement and Concrete, Fourth Edition, Edited by P.C. Hewlett, Arnold 1998, London [2] Geffroy C., Persello J., Foissy A., Lixon P., Tournilhac F., Cabane B., Molar mass selectivity in the adsorption of polyacrylates on calcite, Colloids and Surfaces A 162 (2000) [3] Nawa, T., Ichiboji, H., Kinoshita, M., Influence of Temperature on Fluidity of Cement Paste Containing Superplasticizer with Polyethylene Oxide Graft Chains, Proceedings the sixth 379

10 CANMET/ACI International Conference on Superplasticizers and other Chemical Admixtures in Concrete, Ed. V.M. Malhotra, Nice, 2000, pp [4] Krieger, I.M. and Dougherty, T.J. (1959): A Mechanism for Non-Newtonian Flow in Suspensions of Rigid Spheres, Trans. Soc. Rheol. 3, (1959) [5] Justnes, H., Van Dooren, M. and Van Gemert, D.: Reasons for Workability Loss in Cementitious Binders, 7 th CANMET/ACI Intl. Conf. on Super-plasticizers and Other Chemical Admixtures in Concrete, Berlin, October 22-24, 2003, Supplementary paper. [6] Heikal, M., Morsy, M.S., Aiad, I., Effect of treatment temperature on the early hydration characteristics of superplasticized silica fume blended cement pastes, Cem. Concr. Res. 4 (4) (2005) [7] Golaszewski, J. and Szwabowski, J. The Influence of superplasticizers on the rheological behaviour of fresh cement mortars, Cem Concr Res 34 (2004) [8] Jolicoeur, C., Sharman, J., Otis, N., Lebel, A., Simard, M.-A., Pagé, M., The Influence of Temperature on the Rheological propreties of Superplasticized Cement Pastes, Proceedings the Fifth CANMET/ACI International Conference on Superplasticizers and other Chemical Admixtures in Concrete, Ed. V.M. Malhotra, Rome (1997) [9] Uchikawa, H., Hanehara, S., Shirasaka, T., Sawaki, D., "Effect of admixture on hydration of cement, adsorptive behavior of admixture and fluidity and setting of fresh cement paste", Cem. Concr. Res. 22 (6) (1992), [10] Prince, W., Edwards-Lajnef, M., Aïtcin, P.-C., Interaction between ettringite and a polynaphtalene sulfonate superplasticizer in a cementitious paste, Cem Concr Res 32 (2002) [11] Verbeck, G.J., and Helmuth, R.H., "Structures and Physical Properties of Cement Pastes", Proceedings of the 5 th International Congress on the Chemistry of Cement, Toky, 1968 [12] Kjellsen, K.O., Detwiler, R.J., Gjörv, O.E., "Development of Microstructure in Plain Cement Pastes Hydrated at Different Temperatures", Cem. Conr. Res., 21 (1) (1991) [13] Kjellsen, K.O., Detwiler, R.J., Gjörv, O.E., "Backscattered Electron Imaging of Cement Pastes Hydrated at Different temperatures", Cem. Conr. Res. 20 (2) (1990)