- Email: [email protected]
Combined effect of expansive, shrinkage reducing and hydrophobic admixtures for durable self compacting concrete
Construction and Building Materials 36 (2012) 758–764
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Combined effect of expansive, shrinkage reducing and hydrophobic admixtures for durable self compacting concrete Valeria Corinaldesi ⇑ Università Politecnica delle Marche, Ancona, Italy
h i g h l i g h t s " Study of white self-compacting concretes for architectural structures. " Rhelogical study on both cement pastes and fresh SCCs. " Study of free and restrained drying shrinkage behaviour. " Achievement of shrinkage-free SCC with CaO-based expansive agent and SRA. " Effectiveness of the combined use of SRA, expansive agent and hydrophobic admixture.
a r t i c l e
i n f o
Article history: Received 10 November 2011 Received in revised form 7 March 2012 Accepted 25 April 2012 Available online 15 July 2012 Keywords: Architectural concrete Crack-free concrete Durability Expansive agent Hydrophobic admixture Rheology Self-compacting concrete Shrinkage reducing admixture Thermogravimetric analysis
a b s t r a c t This paper presents the results of an investigation carried out to develop white self-compacting concretes (SCCs) especially devoted to durable architectural structures. In fact, shrinkage-free SCC mixtures were studied, obtained by combining CaO-based expansive agent and shrinkage reducing admixture. Also a hydrophobic admixture was added to SCC in order to preserve the white surface from the growth of micro-organisms as well as to increase concrete durability. Rheological tests were carried out on cement pastes, as well as on fresh SCCs, containing the three admixtures in order to study their influence (alone or in combination) on cement paste and concrete rheology. Also thermogravimetric analysis was carried out on cement pastes in order to evaluate the influence of various admixtures on portlandite solubility. Hardened SCCs were characterized by means of compression tests, free dying shrinkage, and restrained expansion measurements. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction The development of coloured self-compacting concrete opens new fields of application of SCC, since it adds attractive alternatives for challenging architectural designs in terms of shapes, textures and colours, to the advantages of fluidity and filling capacity of SCC. For example, the use of white SCC could allow to obtain a marble-like effect of the skin of the concrete, even if placed in the absence of vibration (situation very common for structures with very congested reinforcement). As it is well known the design of SCC requires a careful combination of the various material components of the mixture. SCC must achieve high workability and flow into the formwork under ⇑ Tel.: +39 071 2204428; fax: +39 071 2204719. E-mail address: [email protected] 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2012.04.129
its own weight without compaction and with no segregation. In rheological terms, it is accepted that SCC has a low (but positive) yield stress, while the plastic viscosity can vary significantly [1]. However, an appropriate combination of the two parameters is required to obtain a concrete with adequate fluidity and stability. A recent example of application of architectural SCC is that of the very prestigious Museum of Modern Arts, in Rome, designed by Zaha Hadid Limited, London, UK. For this architectural structure a very special shrinkage compensating SCC was studied by Collepardi et al. [2], in order to avoid the risk of cracks in some special walls (8 m high and 55 m long), without constructions joints. In particular, in that case, CaO-based expansive agent in combination with shrinkage reducing admixture (SRA) was used. The technology of shrinkage-compensating concrete is based on the use of special products, such as calcium sulpho-aluminates or calcium oxide, which react with water and produce a restrained
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
expansion in reinforced concrete structures. In particular, the transformation of calcium oxide to calcium hydroxide by reacting with water causes its volume increase of about 90% [3]. This technology has been invented many years ago [4], but its use has been very limited in practice, due to the difficulty in adopting a continuous water curing, absolutely needed in the early ages after setting. Consequently, from a practical point of view, this technology can be adopted only in some special constructions such as concrete floors or slab foundations. However, Collepardi et al. [5] found that there is a synergistic effect in the combined use of SRA and a CaO-based expansive agent in terms of more effective expansion, even in the absence of wet curing. In fact, due to the additional effect of SRA, shrinkage-compensating concrete was used to produce crack-free concrete without wet curing for an outside industrial floor of 800 m2 in absence of contraction joints [6]. The synergistic effect of expansive and shrinkage reducing admixtures was also more recently studied by Meddah et al. [7]. According to scientists [8–10], the main mechanism through SRAs are able to reduce plastic, autogeneous and especially drying shrinkage of concrete seems to be the lowering of surface tension of the water in capillary pores. In fact, as water-filled pore begin to loose moisture, curved menisci are formed, and the surface tension of water pulls the walls of the pores. With the reduced surface tension of water, the force pulling the walls of the pores is probably decreased, and the resultant shrinkage strain is reduced in turn. Nevertheless, this hypothesis cannot justify the slight early-age expansion due to the addition of SRA to the concrete mixture. According to other scientists [11,12], the presence of SRA in the concrete mixture could reduce the water solubility of calcium hydroxide: the lower the solubility, the less diffused is the migration of calcium ions, the lower is the distribution of calcium hydroxide, which can result in higher stresses due to crystallization, and could lead to an early-age expansion. Also Chatterji [3] found that the lower is the solubility of calcium hydroxide, the higher was the concrete expansion. Another addition which can be useful to preserve the marblelike effect of the skin of the SCC from humidity, and then from the growth of organic micro-organisms such as fungi, lichens, etc. can be hydrophobic admixture [13]. In this work, it was used but not in the most common way of application, that is as surface treatment. In fact, it was directly introduced in the concrete mixture, with the aim of making the bulk concrete itself hydrophobic instead of the only surface, and so that improving its effectiveness and resistance to the aggressive agents in the atmosphere [14]. The use of hydrophobic admixture in reinforced concrete could be useful not only for maintaining its aesthetics but also for durability reasons. In fact, in sound concrete specimens, exposed to chloride solution, the use of hydrophobic admixture can be able to prevent the corrosion of reinforcing steel [14]. This effect is due to the very low water absorption, and then chloride ingress, through the pores of the hydrophobized cementitious matrix.
useful in order to preserve the white surface from humidity and then from the growth of unaesthetic micro-organisms, as well as to increase concrete durability.
3. Experimental 3.1. Materials White coloured portland-limestone blended cement type CEM II/B-LL 42.5 R according to the European Standards EN-197/1 [15] was used, its content of white clinker was in the range 65–79% by weight. The Blaine fineness of cement was 0.42 m2/g and its relative specific gravity was 3.04. Moreover, limestone powder (95% CaCO3 and other minor constituents) was used as filler, obtained as a by-product of quarry activity. In limestone quarries, considerable amounts of limestone powders are being produced as by-products of stone crushers. High amounts of powders are being collected and utilisation of this by-product is a big problem from the aspects of disposal, environmental pollution and health hazards. Previous studies showed the feasibility of the use of limestone powders for SCC [16–18]. The Blaine fineness of limestone powder was 0.58 m2/g and its relative specific gravity was 2.68. As aggregate, carbonatic gravel (2–16 mm) and carbonatic sand (0–4 mm) were used. The gradation of both aggregate fractions, evaluated according to UNI EN 9331 [19], are shown in Fig. 1, and their physical properties, evaluated according to UNI EN 1097-6 [20], are 2.66 and 2.63 relative specific gravity and 2.4% and 3.2% water absorption, respectively. A superplasticiser was used, which consisted of carboxylic acrylic ester polymer in the form of 30% aqueous solution with a relative specific gravity of 1.09 ± 0.02. As shrinkage-reducing admixture (SRA), polyethylene glycol was used, while as expansive agent, dead burnt lime (CaO) was employed. Finally, hydrophobic admixture in form of 45% aqueous emulsion of butyl-ethoxy-silane was added to some mixtures. The alkyl group size influences the decrease of surface tension and thus the treatment effectiveness. As a consequence, relatively high-molecular weight groups, such as butyl (in this case), significantly reduce the substrate water absorption [14]. The same low value of water to cement ratio (0.45) was adopted for durability reasons, in order to prepare all cement pastes and SCCs.
3.2. Rheological characterization of cement pastes The study of the rheological behaviour of cement pasts is certainly an essential step towards the optimization of SCC production [21]. Concerning the influence of the fine aggregate particles on the rheology of the ‘carrying phase’ (i.e. cement paste) within SCC, in this case it should not be considered, because no material is passing the sieve of 0.150 mm, as it can be seen from the grain size distribution curves of sand and gravel used (see Fig. 1). For this purpose, eight cement pastes were prepared. The proportions of these paste mixtures are shown in Table 1. Firstly, a reference paste was prepared (REF) with a water to cement ratio of 0.45, in which water, cement, limestone powder and superplasticizing admixture were added at dosages corresponding to those of the related SCC. Then, the effect of either calcium oxide or SRA or hydrophobic admixture on the rheology of cement paste was studied by suitably preparing further three cement pastes, in which every addition was made on the basis of its dosage on the final SCC mixture.
2. Research significance This paper presents the results of an investigation carried out to develop white self-compacting concretes (SCCs) by using a suitable combination of three admixtures in order to produce durable concrete for architectural structures. The combined effects of these three admixtures on cement paste and SCC rheology were studied, as well as their combined effects on mechanical performance, drying shrinkage behaviour and cracking tendency of concretes. In fact, shrinkage-free SCC mixtures were studied, obtained by combining CaO-based expansive agent and shrinkage reducing admixture. Moreover, also the influence on the concrete performance of a hydrophobic admixture was tested. Its addition to SCC could be
759
Fig. 1. Cumulative grain size distribution curves of the aggregates.
760
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
Table 1 Mixture proportions of the cement pastes (g).
REF REF + CaO REF + SRA REF + Hyd REF + CaO + SRA REF + CaO + Hyd REF + SRA + Hyd REF + CaO + SRA + Hyd
Water
Cement
Superpl. adm.
Limestone powder
CaO
SRA
Hydroph. adm.
50
110
1.83
44.5 35.3 44.5 44.5 35.3 35.3 44.5 35.3
0 9.2 0 0 9.2 9.2 0 9.2
0 0 1.05 0 1.05 0 1.05 1.05
0 0 0 0.26 0 0.26 0.26 0.26
Finally, the effect of the combined use of calcium oxide, SRA and hydrophobic admixture was studied by preparing the last four cement pastes: in three pastes they were added two by two at their respective dosage and, finally, the last mixture (REF + CaO + SRA + Hyd) was prepared by using all of them together. The rheological behaviour of these cement pastes was tested at 10 min after ingredients’ mixing and then after 25 min: data reported in this paper are those measured after 25 min (i.e. data collected during the second cycle of test). The apparatus was a rotating rheometer based on coaxial rotary cylinders with a slowly increasing shear rate (D), ranging from 1 to 100 s1. Bui et al. [22] found that the rate of 1–100 s1 was the most suitable for rheological model of SCCs, while higher rotation rates were considered too fast, and rates limited to 50 s1 did not produce consistent results. The walls of the concentric cylinders were not smooth but roughened in order to reduce (if not completely eliminate) the ‘‘slip’’ phenomenon, i.e. the development of a water-rich layer close to the inner surface of the rotating cylinder, which produces a lubricating effect, making flow easier, and not representative of the bulk material [23]. The rheological behaviour was described by means of the Bingham flow model:
In order to optimize the grain size distribution of the solid particles in the concrete, sand and gravel were suitably combined, taking into account also the suggestions reported in the literature concerning the mixture proportion of SCC, particularly in terms of maximum volume of coarse aggregate, that is 340 l/m3 [24,25]. The recommended volume of fine particles (including cement, fly ash, ground limestone, etc.) in SCC is in the range of 170–200 l/m3. In this work a very high volume of very fine particles was chosen (200 l/m3), due to the absence of Viscosity Modifying Agents in the SCC mixtures. In particular, in order to achieve a volume of very fine particles of about 200 l/m3, besides to 420 kg/m3 of cement it was necessary to include limestone powder in the mixture, at a dosage of 170 kg/m3. Finally, when the expansive agent (CaO) was added to the mixtures (35 kg/m3), the content of limestone powder was reduced to 135 kg/m3, in order to left unchanged the dosage of very fine materials into the mixture. Finally, shrinkage reducing admixture and hydro-phobic admixture were added at a dosage of 4 kg/m3 and 1 kg/m3, respectively.
3.5. Preparation, curing and testing of concrete specimens
s ¼ sy þ g D where s is the shear stress [Pa], sy is the yield stress [Pa], g is the plastic viscosity [Pa s] and D is the shear rate [s1]. The slope of the down-curve (decreasing shear rate) was used to calculate the plastic viscosity, while the intercept at zero shear rate was used to calculate the yield stress. 3.3. Thermogravimetric measurements on cement pastes Thermogravimetic analysis was carried out on four cement pastes in order to evaluate the influence of the various admixtures on portlandite solubility. This information is essential in order to interpret the results obtained by drying shrinkage tests in the presence of expansive agent. In fact, Chatterji [3] described the expansion in terms of delayed calcium hydroxide formation. This compound, having a molar volume higher than calcium oxide, gives origin to a crystal growth pressure. According to the scientists [11,12], the expansion is strictly related to the solubility of calcium hydroxide: the higher the solubility, the more diffused is the migration of calcium ions, the higher is the distribution of calcium hydroxide and the lower is the expansion. Tests were carried out after one day of wet curing on three samples for each of the cement pastes labelled ‘REF + CaO’, ‘REF + CaO + SRA’, ‘REF + CaO + Hyd’ and ‘REF + CaO + SRA + Hyd’ (their composition is the same reported in Table 1). 3.4. SCC mixtures proportions The SCC mixtures proportions are reported in Table 2. All concretes were prepared with the same w/c of 0.45, the same content of cement (420 kg/m3), water (190 kg/m3), sand (910 kg/m3) and gravel (670 kg/m3), as well as the same dosage of superplasticizer (1.2% by weight of cement plus limestone powder).
As a first step, properties of the fresh concrete other than slump were evaluated according to Italian Standards UNI 11041 [26], since in this case the slump value is not relevant due to very fluid concrete. Therefore, the attention was focused on the measurement of the slump flow, which is the medium diameter (Ufin) of the slumped concrete. Then, also the elapsed time to gain the mean diameter of 500 mm (t500) and the elapsed time to gain the final configuration (tfin) were detected. Moreover, time elapsed for the SCCs passing through V-funnel was detected, according to Italian Standards UNI 11042 [27]. Several concrete specimens were cast for tests on hardened concrete. In particular, for compression tests (carried out according to EN 12390-3 [28]), nine cubic specimens (100 mm in size) were manufactured, then wet-cured at 20 °C, and tested at right angles to the position of casting. Therefore the bearing faces were plane and smooth enough as to require no capping or grinding. The specimens were loaded at a constant strain rate until failure. In addition, for free drying shrinkage monitoring, three prismatic specimens (100 by 100 by 500 mm) were prepared for each concrete mixture according to Italian Standard UNI 6555 ‘Hydraulic Shrinkage Determination’ [29]. After one day of wet curing, the specimens were stored at constant temperature (20 ± 2 °C) and constant relative humidity (50 ± 2 %), while measuring drying shrinkage at different curing times up to 60 days of exposure. Finally, for restrained shrinkage evaluation, further three prismatic specimens (50 by 50 by 300 mm) were prepared for each concrete mixture according to Italian Standard UNI 8148 [30]. Fresh concrete was cast in metallic formworks by embedding a steel bar reinforcement (6 mm in diameter, 300 mm in length). After 6 h they were demoulded, then immersed under water up to 36 h and then exposed to air with constant temperature (20 ± 2 °C) and constant relative humidity (50 ± 2 %). The length-change of the steel bar was then recorded as a function of the curing time up to 60 days of exposure.
Table 2 Mixture proportions of the self-compacting concretes (kg/m3).
REF REF + CaO REF + SRA REF + Hyd REF + CaO + SRA REF + CaO + Hyd REF + SRA + Hyd REF + CaO + SRA + Hyd
Water
Cement
Superpl. adm.
Sand
Fine gravel
Limestone powder
CaO
SRA
Hydroph. adm.
190
420
7
910
670
170 135 170 170 135 135 170 135
0 35 0 0 35 35 0 35
0 0 4 0 4 0 4 4
0 0 0 1 0 1 1 1
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
Fig. 2. Measured yield stress values of the SCC paste fractions.
761
Fig. 4. Calculated thixotropic values of the SCC paste fractions.
Fig. 3. Measured plastic viscosity values of the SCC paste fractions.
4. Results and discussion
Fig. 5. Results of the thermogravimetric analysis (temperature ranging from 200 °C to 500 °C).
4.1. Rheological tests on cement pastes In Fig. 2 the mean values, calculated on three samples for each mixture, of the measured yield stress values are reported. The values obtained are very similar and only slight difference can be noticed among the various pastes. However, the target of sufficiently low values (but still positive) of the yield stress, in order to obtain fluid SCC without segregation [31], was achieved. In particular, data shown in Fig. 2 evidence that the addition of SRA reduces the yield stress value, while the addition of CaO (even if replacing limestone powder) causes it is increase, as well as the addition of the hydrophobic admixture. In Fig. 3 the mean values, calculated on three samples for each mixture, of the measured plastic viscosity values are reported. The values obtained are all included in a very narrow range (0.010– 0.055 Pa s), thus implying that no significant change happens by adding the various ingredients to the mixture. Only a slight increase can be detected when the hydrophobic admixture was used. However, the plastic viscosity values obtained are very low and the Bingham line obtained were practically horizontal. This fact is positive for SCC, when the yield stress value is low but not too much, as in this case, because it means that the same level of cohesion is maintained independently on the concrete flow velocity, by assuring sufficient mobility without internal segregation, as well as flow segregation [31]. Thixotropy is the property of certain gels, such as cement paste, which are rigid when left standing but increase their fluidity when
put into movement. In Fig. 4 the mean values, calculated on three samples for each mixture, of the measured thixotropy values are reported (where thixotropy was calculated as the area included between the up-curve and the down-curve). This measure can give an estimate of the energy necessary to move the SCC and even an estimate of the lateral formwork pressure that SCC will exert after casting, being the lower the pressure the higher the thixotropy value [26]. For the cement pastes tested in this work the value of thixotropy was always negative, thus showing that all these cement pastes have anti-thixotropic behaviour. In particular, the addition of either CaO or SRA seems to enhance the thixotropic behaviour, while the addition of the hydrophobic admixture seems to enhance the opposite.
4.2. Thermogravimetric measurements on cement pastes Thermal analysis carried out show that the examined cement pastes contain different amount of portlandite, Ca(OH)2. For the cement pastes labelled ‘REF + CaO’, ‘REF + CaO + Hyd’, ‘REF + CaO + SRA’ and ‘REF + CaO + SRA + Hyd’ the measured values (expressed in% by weight) were 12.72%, 12.48%, 9.39% and 9.78%, respectively. The first evidence is that the addition of SRA hindered portlandite formation (about 25%), and this result is also confirmed by the literature [11,12]. In fact, as it can be observed in Fig. 5, the weight losses (corresponding to the flex of the DTA curves, for simplicity
762
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
not reported in Fig. 5), which occur from about 400 °C to 450 °C (typical range for the decomposition reaction of calcium hydroxide), are less pronounced for the two mixtures containing SRA. Chatterji [3] stated that the lower is the solubility of calcium hydroxide, the higher was the concrete expansion. SRA could reduce the water solubility of calcium hydroxide, promoting the material expansion and compensating more efficaciously concrete shrinkage. On the other hand, only an insignificant effect on the content of portlandite was detected when the hydrophobizing admixture was used. On the basis of the results obtained, differently from SRA, there is not evidence of any hindering action of the hydrophobizing admixture on the calcium hydroxide formation. Consequently, according to the scientists [11,12], its addition to the concrete mixture should not significantly modify the drying shrinkage behaviour of the SCC containing calcium oxide, while the SRA addition should improve expansion of concrete, and consequently the effectiveness of the calcium oxide itself.
4.3. Slump flow and V funnel tests As shown in Table 3, all concretes had a sufficient deformability under their own weight (strictly related to the Ufin value, included in the range 620–640 mm). The presence of bleeding water was never observed at the periphery of the slumped material. Time elapsed for the SCCs passing through V-funnel was in the range 9–11 s in all cases, within the acceptance limits (5–12 s) [27]. The only significant effect on concrete rheology, which can be observed, is that of the hydrophobic admixture. In fact, its addition produced slightly higher tendency of the SCC to be viscous (longer time to gain the final configuration and to pass trough the V-funnel). This evidence was also detected by means of the rheological tests on cement pastes (see Figs. 2 and 3).
Fig. 6. SCC compressive strength vs. curing time.
The use of hydrophobizing admixture at low dosage (1 kg per cubic meter of concrete, corresponding to 0.25% by weight of cement) did not influence compressive strength at early age, and only a certain reduction can be detected after 7 and 28 days of curing (9%). As expected on the basis of the results obtained when used singularly, the worst result in terms of mechanical performance was obtained by contemporarily using SRA and hydrophobizing admixture. However, the use of the three additions all together (mixture ‘REF + CaO + SRA + Hyd’) allowed to obtain compressive strength values close to those of the reference mixture (around -5% independently on the age). In conclusion, the target concrete strength class C 37/45 was reached by the mixtures: ‘REF’, ‘REF + CaO’, ‘REF + CaO + SRA’ and ‘REF + CaO + SRA + Hyd’. On the other hand, a lower class strength C 32/ 40 was reached by the mixtures prepared by using the SRA and the hydrophobizing admixture alone, and an even lower strength class (C 30/37) characterized the mixture in which SRA and the hydrophobizing admixture were used together without expansive agent.
4.4. Compression tests Compressive strength of concrete was determined at 1, 7, and 28 days of curing and the results obtained are reported in Fig. 6. With respect to the reference mixture, the addition of the expansive agent (CaO) caused an evident strength increase after 1 day of curing (about +20%), then this effect disappeared with increasing curing time. On the other hand, the use of SRA, caused an evident strength loss, particularly at early age (about 20% after 1 day), but also after 28 days of curing (about 10%). The reason for this slower strength development seems to be that the addition of SRA to the mixing water depresses the dissolution of alkalis in the pore fluid [33]. This fact results in a pore fluid with lower alkalinity, which causes a reduction in the rate of cement hydration. However, if used in combination with CaO this negative effect on concrete compressive strength is practically negligible (see mixture ‘REF + CaO + SRA’).
4.5. Free drying shrinkage tests Results obtained are reported in Fig. 7. With respect to the reference mixture, the addition of the SRA showed to be very effective in reducing drying shrinkage of concrete, in fact the ultimate strain of the mixture ‘REF + SRA’ is about 40% lower than that of the reference mixture ‘REF’. In addition, after two days of exposure a slight expansion was detected in the presence of SRA. Sant et el. [12] demonstrated that the presence of SRA increases the portlandite oversaturation level in solution, which can result in higher crystallization stresses, which in turn could lead to an early-age expansion.
Table 3 Results of tests on fresh SCCs. Slump flow
REF REF + CaO REF + SRA REF + Hyd REF + CaO + SRA REF + CaO + Hyd REF + SRA + Hyd REF + CaO + SRA + Hyd
V-funnel
Ufin (mm)
t500 (s)
tfin (s)
tfin (s)
640 620 640 630 630 630 640 630
2 2 2 2 2 2 2 2
10 9 9 10 8 10 9 10
10 10 9 11 9 11 10 11
Fig. 7. SCC free drying shrinkage vs. time of exposure.
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
The addition of the hydrophobizing admixture seems not to affect drying shrinkage: in fact, quite the same behaviour can be detected for the mixtures ‘REF + Hyd’ and ‘REF’. On the other hand, when SRA and hydrophobizing admixture are used together (mixture ‘REF + SRA + Hyd’), the effectiveness of SRA is hindered: the initial expansion was not present anymore, and the reduction of strain is only about 13% instead of 40%. When calcium oxide was added to the mixture ‘REF + CaO’, the initial expansion of almost 0.4 mm/m allow to reach a very low value of drying shrinkage after 60 days of curing (about 75% with respect to the reference mixture). Moreover, the contemporaneous use of SRA and CaO proved to be very effective, and it caused strong initial expansion of SCC (+0.5 mm/m), which gradually decreased, but still remained positive after 60 days of exposure to 50% relative humidity. The addition of the hydrophobizing admixture to the mixture prepared with the expansive agent (mixture ‘REF + CaO + Hyd’) seems not to affect drying shrinkage, as expected on the basis of the results obtained by thermogravimetric analysis: in fact, quite the same behaviour can be detected for the mixtures REF + CaO + Hyd’ and ‘REF + Hyd’. On the other hand, also in the presence of CaO, when SRA and hydrophobizing admixture are used together (mixture ‘REF + CaO + SRA + Hyd’), the effectiveness of CaO + SRA is reduced: the initial expansion is reduced from 0.5 mm/m to 0.42 mm/m and the same values of strain were reached in advance by the mixture containing hydrophobizing admixture (for example + 0.03 mm/m was already measured after 35 days instead of 60 days). By comparing the results obtained for the mixtures ‘REF + SRA + Hyd’, ‘REF + CaO + Hyd’ and ‘REF + CaO + SRA + Hyd’, it can be noticed that the negative effect of the hydrophobizing admixture is evident only in the presence of SRA. Data obtained by means of thermogravimetric analysis showed only slight reduction of the amount of portlandite (about 2%) if the hydrophobizing admixture was added in the absence of SRA, while the amount of portlandite slightly increased (about 4%) when its was added in the presence of SRA, but this difference is probably too small for justifying such a macroscopic effect. Results obtained suggest a likely interference between SRA and hydrophobizing admixture, both acting on capillary pore forces, by reducing the surface tension of water (from 72 dyn/cm to approximately 40 dyn/cm), in the case of SRA, and by changing the sign of the contact angle (from hydrophilic to hydrophobic), in the case of the hydrophobizing admixture. In fact, although the mechanisms of drying shrinkage of concrete are not fully understood, literature suggests that, when considering concrete shrinkage in the 45–90% relative humidity range, capillary stress appears to be the predominant mechanism [10]. When pore water evaporates from capillary pores in hardened concrete during drying, tension in the liquid is transferred to the capillary walls, resulting in shrinkage. For a given pore size distribution, the internal stress generated upon evaporation is proportional to the surface tension of the pore water solution. However, the shrinkage measured for the mixture ‘REF + CaO + SRA + Hyd’ after 60 days of exposure can be considered negligible (0.060 mm/m) and the appearance of cracking is strongly unlikely also for this mixture. 4.6. Restrained drying shrinkage tests Results obtained are reported in Fig. 8. With respect to the reference mixture, the addition of the SRA showed to be very effective in reducing drying shrinkage of concrete, in fact the final strain of the mixture ‘REF + SRA’ is about 50% lower than that of the reference mixture ‘REF’. The addition of the hydrophobizing admixture seems not to affect drying shrinkage: in fact, quite the same behaviour can be detected for the mixtures ‘REF + Hyd’
763
Fig. 8. Restrained drying shrinkage vs. time of exposure.
and ‘REF’. On the other hand, when SRA and hydrophobizing admixture are used together (mixture ‘REF + SRA + Hyd’), the effectiveness of SRA is hindered and the reduction of strain is only about 30% instead of 50%. The reason for that was already discussed above. When calcium oxide was added to the mixture ‘REF + CaO’, the initial expansion of about 0.3 mm/m allow to reach a very low value of drying shrinkage after 60 days of curing (about 0.150 mm/m). Moreover, the contemporaneous use of SRA and CaO proved again to be very effective, and it caused strong initial expansion of SCC (+0.36 mm/m), which gradually decreased, but still remained positive after 60 days of exposure to 50% relative humidity. The addition of the hydrophobizing admixture to the mixture prepared with the expansive agent (mixture ‘REF + CaO + Hyd’) seems not to affect drying shrinkage: in fact, quite the same behaviour can be detected for the mixtures ‘REF + Cao + Hyd’ and ‘REF + Hyd’. On the other hand, also in the presence of CaO, when SRA and hydrophobizing admixture are used together (mixture ‘REF + CaO + SRA + Hyd’), the effectiveness of CaO + SRA is reduced: the initial expansion is slightly reduced from 0.36 mm/m to 0.32 mm/m, and the same values of strain were reached in advance by the mixture containing hydrophobizing admixture (for example + 0.06 mm/m was already measured after 20 days instead of 60 days). By comparing the results obtained by free drying shrinkage test and restrained shrinkage test, the same trend were found and similar discussions can be carried out, as well as the same conclusions can be drawn. 5. Conclusions Based on the results obtained by means of rheological tests, the following conclusions are drawn: 1. Concerning the yield stress the addition of SRA reduces its value, while the addition of CaO (even if replacing limestone powder) causes it is increase, as well as the addition of the hydrophobic admixture. 2. Concerning the plastic viscosity, all the values obtained are very low and the Bingham line obtained were practically horizontal. 3. On the basis of the thixotropy values detected, all the cement pastes have anti-thixotropic behaviour: in particular, the addition of either CaO or SRA seems to enhance the thixotropic behaviour, the opposite for the hydrophobic admixture. 4. Generally, the same level of shear stress is maintained independently on the paste flow velocity, which should assure sufficient mobility, without internal segregation and flow segregation, for the related SCCs.
764
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
5. In fact, concerning the tests carried out on the fresh SCCs, all concretes showed good flowability in absence of segregation. Moreover, based on the results obtained by means of tests on hardened concretes, the following conclusions are drawn:
[3] [4] [5]
[6]
1. The addition of the expansive agent (CaO) caused an increase of mechanical strength at early age; on the other hand, the use of SRA, caused an evident strength loss, particularly at early age, but also later; however, if SRA and CaO are used in combination the negative effect of SRA on concrete compressive strength is practically negligible; moreover, the use of hydrophobizing admixture at low dosage only caused a late reduction of strength. 2. Concerning the combined effects (of either SRA plus CaO, or SRA plus hydrophobizing admixture, or CaO plus hydrophobizing admixture) on mechanical performance, as predictable on the basis of the results obtained when used singularly, the worst result was obtained by using together SRA and hydrophobizing admixture; however, the contemporary use of all the three additions allowed to obtain the same strength class values of the reference mixture. 3. In terms of drying shrinkage (free or restrained), the addition of the SRA showed to be very effective in reducing it, while the addition of the hydrophobizing admixture seems not influent; on the other hand, when SRA and hydrophobizing admixture are used together the effectiveness of SRA is hindered. 4. In terms of drying shrinkage (free or restrained), the use of the expansive agent and, in particular, the contemporaneous use of SRA and CaO proved to be very effective, and it caused strong initial expansion of SCC, which gradually decreased, but still remained positive after 60 days. 5. The addition of the hydrophobizing admixture to the mixture prepared with the expansive agent seems not to affect drying shrinkage, but when CaO, SRA and hydrophobizing admixture are used all together, the effectiveness of CaO + SRA is reduced. 6. In conclusion, the negative effect of the hydrophobizing admixture on drying shrinkage is evident only in the presence of SRA: this result suggests a likely interference between SRA and hydrophobizing admixture, both acting on capillary pore forces. 7. In fact, data obtained from the thermogravimetric analysis only partially confirm the hypotesis that the increase of the portlandite oversaturation level in solution, which causes higher cristallization stresses, can be the main factor causing early-age expansion (in fact, such hypothesis is able to explain the concrete shrinkage behaviour in the presence of SRA, but not that in the presence of both SRA and hydrophobizing admixture together).
References [1] Níelsson I, Wallevik ÓH. Rheologycal evaluation of some empirical test methods-preliminary results. In: Wallevik ÓH, Níelsson I, editors. 3rd International RILEM symposium. RILEM Pub. PRO. vol 33; 2003. p. 59–68. [2] Collepardi M, Collepardi S, Troli R. Pratical applications of SCC in European works. In: Kraus Rudolph N, Naik Tarun R, Claisse Peter, Sadeghi-Pouya,
[7]
[8] [9] [10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
[18] [19] [20] [21] [22] [23] [24]
[25]
[26] [27] [28] [29] [30] [31]
[32]
editors. In: Proc int conf: sustainable construction materials and technologies, special papers proceedings. Milwaukee; Pub UW, CBU; 2007. p. 51–58. Chatterji S. Mechanism of expansion of concrete due to the presence of dead burnt CaO and MgO. Cem Concr Res 1995;25:51–6. Neville AM. Properties of concrete. 4th ed. Longman Group Limited; 1995. Collepardi M, Borsoi A, Collepardi S, Ogoumah Olagot JJ, Troli R. Effects of shrinkage-reducing admixture in shrinkage compensating concrete under non wet curing conditions. Cem Concr Compos 2005;27(6):704–8. Collepardi M, Troli R, Bressan M, Liberatore F, Sforza G. Crack-free concrete for outside industrial floors in the absence of wet curing and contraction joints. Cem Concr Compos 2008;30(10):887–91. Meddah MS, Suzuki M, Sato R. Influence of a combination of expansive and shrinkage-reducing admixture on autogenous deformation and self-stress of silica fume high-performance concrete. Constr Build Mater 2011;25:239–50. Rongbing B, Jian S. Synthesis and evaluation of shrinkage-reducing admixture for cementitious materials. Cem Concr Res 2005;35:445–8. Folliard KJ, Berke NS. Properties of high-performance concrete containing shrinkage-reducing admixture. Cem Concr Res 1997;27:1357–64. Bentz DP, Geiker MR, Hansen KK. Shrinkage-reducing admixtures and earlyage desiccation in cement pastes and mortars. Cem Concr Res 2001;31:1075–85. Maltese C, Pistolesi C, Lolli A, Bravo A, Cerulli T, Salvioni D. Combined effect of expansive and shrinkage reducing admixtures to obtain stable and durable mortars. Cem Concr Res 2005;35(12):2244–51. Sant G, Lothenbach B, Juilland P, Le Saout G, Weiss J, Scrivener K. The origin of early age expansions induced in cementitious materials containing shrinkage reducing admixtures. Cem Concr Res 2011;41:218–29. Collepardi M, Passuelo A. The best SCC: stable, durable and colourable. In: Helene P, Holland TC, Pazzini E, Bittecourt R, editors. Proc IV int ACI/CANMET conference on quality of concrete structures and recent advances in concrete materials and testing, furnas centrais Elqtricas S.A. Goidnia, Brasil: Civil Engineering Technological Center; 2005. Moriconi G, Tittarelli F, Corinaldesi V. Review of silicone-based hydrophobic treatment and admixtures for concrete. Indian Concr J 2002; 76(10):637–42. EN 197-1. Cement – Part 1: composition. Specifications and conformity criteria for common cements; 2000. Felekoglu B. Utilisation of high volumes of limestone quarry wastes in concrete industry (self-compacting concrete case). Resour Conserv Recycl 2007;51:770–91. El Barrak M, Mouret M, Bascoul A. Self-compacting concrete paste constituents: hierarchical classification of their influence on flow properties of the paste. Cem Concr Compos 2009;31:12–21. Corinaldesi V, Moriconi G, Naik TR. Characterization of marble powder for its use in mortar and concrete. Constr Build Mater 2010;24(1):113–7. EN 933-1. Tests for geometrical properties of aggregates – determination of particle size distribution – sieving method; 1997. EN 1097-6. Tests for mechanical and physical properties of aggregates – determination of particle density and water absorption; 2000. Saak AW, Jennings HM, Shah SP. New methodology for designing selfcompacting concrete. ACI Mater J 2001;98(6):429–39. Bui VK, Akkaya Y, Shah SP. Rheological model for self-consolidating concrete. ACI Mater J 2002;99(6):549–59. Saak AW, Jennings HM, Shah SP. The influence of wall slip on yield stress and viscoelastic measurements of cement paste. Cem Concr Res 2001;31:205–12. Bui VK, Montgomery D. Mixture proportioning method for self-compacting high performance concrete with minimum paste volume. In: Skarendahl A, Petersson O, editors. Self compacting concrete, PRO 007. Bagneux, France: RILEM Publication s.a.r.l; 1999. p. 373–84. Jacobs F, Hunkeler F. Design of self-compacting concrete for durable concrete structures, in ‘self-compacting concrete’. In: Skarendahl A, Petersson O, editors. Proc of the first int. RILEM symp, Stockholm, Sweden; 1999. p. 397– 407. Assaad J, Khayat KH, Mesbah H. Variation of formwork pressure with thixotropy of self-consolidating concrete. ACI Mater J 2003;100(1):29–37. UNI 11041. Prova sul calcestruzzo autocompattante fresco – determinazione dello spandimento e del tempo di spandimento; 2003. UNI 11042. Prova sul calcestruzzo autocompattante fresco – determinazione del tempo di efflusso dall’imbuto; 2003. UNI 6555. Hydraulic shrinkage determination; 1973. UNI 8148. Agenti espansivi non metallici per impasti cementizi – determinazione dell’espansione contrastata del calcestruzzo; 2008. Wallevik ÓH. Rheology – a scientific approach to develop self compacting concrete. In: Wallevik ÓH, Níelsson I, editors. Self compacting concrete, PRO 33. Bagneux, France: RILEM Publication s.a.r.l; 2003. p. 23–32. Rajabipour F, Sant G, Weiss J. Interactions between shrinkage reducing admixtures (SRA) and cement paste’s pore solution. Cem Concr Res 2008;38(5):606–15.
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Combined effect of expansive, shrinkage reducing and hydrophobic admixtures for durable self compacting concrete Valeria Corinaldesi ⇑ Università Politecnica delle Marche, Ancona, Italy
h i g h l i g h t s " Study of white self-compacting concretes for architectural structures. " Rhelogical study on both cement pastes and fresh SCCs. " Study of free and restrained drying shrinkage behaviour. " Achievement of shrinkage-free SCC with CaO-based expansive agent and SRA. " Effectiveness of the combined use of SRA, expansive agent and hydrophobic admixture.
a r t i c l e
i n f o
Article history: Received 10 November 2011 Received in revised form 7 March 2012 Accepted 25 April 2012 Available online 15 July 2012 Keywords: Architectural concrete Crack-free concrete Durability Expansive agent Hydrophobic admixture Rheology Self-compacting concrete Shrinkage reducing admixture Thermogravimetric analysis
a b s t r a c t This paper presents the results of an investigation carried out to develop white self-compacting concretes (SCCs) especially devoted to durable architectural structures. In fact, shrinkage-free SCC mixtures were studied, obtained by combining CaO-based expansive agent and shrinkage reducing admixture. Also a hydrophobic admixture was added to SCC in order to preserve the white surface from the growth of micro-organisms as well as to increase concrete durability. Rheological tests were carried out on cement pastes, as well as on fresh SCCs, containing the three admixtures in order to study their influence (alone or in combination) on cement paste and concrete rheology. Also thermogravimetric analysis was carried out on cement pastes in order to evaluate the influence of various admixtures on portlandite solubility. Hardened SCCs were characterized by means of compression tests, free dying shrinkage, and restrained expansion measurements. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction The development of coloured self-compacting concrete opens new fields of application of SCC, since it adds attractive alternatives for challenging architectural designs in terms of shapes, textures and colours, to the advantages of fluidity and filling capacity of SCC. For example, the use of white SCC could allow to obtain a marble-like effect of the skin of the concrete, even if placed in the absence of vibration (situation very common for structures with very congested reinforcement). As it is well known the design of SCC requires a careful combination of the various material components of the mixture. SCC must achieve high workability and flow into the formwork under ⇑ Tel.: +39 071 2204428; fax: +39 071 2204719. E-mail address: [email protected] 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2012.04.129
its own weight without compaction and with no segregation. In rheological terms, it is accepted that SCC has a low (but positive) yield stress, while the plastic viscosity can vary significantly [1]. However, an appropriate combination of the two parameters is required to obtain a concrete with adequate fluidity and stability. A recent example of application of architectural SCC is that of the very prestigious Museum of Modern Arts, in Rome, designed by Zaha Hadid Limited, London, UK. For this architectural structure a very special shrinkage compensating SCC was studied by Collepardi et al. [2], in order to avoid the risk of cracks in some special walls (8 m high and 55 m long), without constructions joints. In particular, in that case, CaO-based expansive agent in combination with shrinkage reducing admixture (SRA) was used. The technology of shrinkage-compensating concrete is based on the use of special products, such as calcium sulpho-aluminates or calcium oxide, which react with water and produce a restrained
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
expansion in reinforced concrete structures. In particular, the transformation of calcium oxide to calcium hydroxide by reacting with water causes its volume increase of about 90% [3]. This technology has been invented many years ago [4], but its use has been very limited in practice, due to the difficulty in adopting a continuous water curing, absolutely needed in the early ages after setting. Consequently, from a practical point of view, this technology can be adopted only in some special constructions such as concrete floors or slab foundations. However, Collepardi et al. [5] found that there is a synergistic effect in the combined use of SRA and a CaO-based expansive agent in terms of more effective expansion, even in the absence of wet curing. In fact, due to the additional effect of SRA, shrinkage-compensating concrete was used to produce crack-free concrete without wet curing for an outside industrial floor of 800 m2 in absence of contraction joints [6]. The synergistic effect of expansive and shrinkage reducing admixtures was also more recently studied by Meddah et al. [7]. According to scientists [8–10], the main mechanism through SRAs are able to reduce plastic, autogeneous and especially drying shrinkage of concrete seems to be the lowering of surface tension of the water in capillary pores. In fact, as water-filled pore begin to loose moisture, curved menisci are formed, and the surface tension of water pulls the walls of the pores. With the reduced surface tension of water, the force pulling the walls of the pores is probably decreased, and the resultant shrinkage strain is reduced in turn. Nevertheless, this hypothesis cannot justify the slight early-age expansion due to the addition of SRA to the concrete mixture. According to other scientists [11,12], the presence of SRA in the concrete mixture could reduce the water solubility of calcium hydroxide: the lower the solubility, the less diffused is the migration of calcium ions, the lower is the distribution of calcium hydroxide, which can result in higher stresses due to crystallization, and could lead to an early-age expansion. Also Chatterji [3] found that the lower is the solubility of calcium hydroxide, the higher was the concrete expansion. Another addition which can be useful to preserve the marblelike effect of the skin of the SCC from humidity, and then from the growth of organic micro-organisms such as fungi, lichens, etc. can be hydrophobic admixture [13]. In this work, it was used but not in the most common way of application, that is as surface treatment. In fact, it was directly introduced in the concrete mixture, with the aim of making the bulk concrete itself hydrophobic instead of the only surface, and so that improving its effectiveness and resistance to the aggressive agents in the atmosphere [14]. The use of hydrophobic admixture in reinforced concrete could be useful not only for maintaining its aesthetics but also for durability reasons. In fact, in sound concrete specimens, exposed to chloride solution, the use of hydrophobic admixture can be able to prevent the corrosion of reinforcing steel [14]. This effect is due to the very low water absorption, and then chloride ingress, through the pores of the hydrophobized cementitious matrix.
useful in order to preserve the white surface from humidity and then from the growth of unaesthetic micro-organisms, as well as to increase concrete durability.
3. Experimental 3.1. Materials White coloured portland-limestone blended cement type CEM II/B-LL 42.5 R according to the European Standards EN-197/1 [15] was used, its content of white clinker was in the range 65–79% by weight. The Blaine fineness of cement was 0.42 m2/g and its relative specific gravity was 3.04. Moreover, limestone powder (95% CaCO3 and other minor constituents) was used as filler, obtained as a by-product of quarry activity. In limestone quarries, considerable amounts of limestone powders are being produced as by-products of stone crushers. High amounts of powders are being collected and utilisation of this by-product is a big problem from the aspects of disposal, environmental pollution and health hazards. Previous studies showed the feasibility of the use of limestone powders for SCC [16–18]. The Blaine fineness of limestone powder was 0.58 m2/g and its relative specific gravity was 2.68. As aggregate, carbonatic gravel (2–16 mm) and carbonatic sand (0–4 mm) were used. The gradation of both aggregate fractions, evaluated according to UNI EN 9331 [19], are shown in Fig. 1, and their physical properties, evaluated according to UNI EN 1097-6 [20], are 2.66 and 2.63 relative specific gravity and 2.4% and 3.2% water absorption, respectively. A superplasticiser was used, which consisted of carboxylic acrylic ester polymer in the form of 30% aqueous solution with a relative specific gravity of 1.09 ± 0.02. As shrinkage-reducing admixture (SRA), polyethylene glycol was used, while as expansive agent, dead burnt lime (CaO) was employed. Finally, hydrophobic admixture in form of 45% aqueous emulsion of butyl-ethoxy-silane was added to some mixtures. The alkyl group size influences the decrease of surface tension and thus the treatment effectiveness. As a consequence, relatively high-molecular weight groups, such as butyl (in this case), significantly reduce the substrate water absorption [14]. The same low value of water to cement ratio (0.45) was adopted for durability reasons, in order to prepare all cement pastes and SCCs.
3.2. Rheological characterization of cement pastes The study of the rheological behaviour of cement pasts is certainly an essential step towards the optimization of SCC production [21]. Concerning the influence of the fine aggregate particles on the rheology of the ‘carrying phase’ (i.e. cement paste) within SCC, in this case it should not be considered, because no material is passing the sieve of 0.150 mm, as it can be seen from the grain size distribution curves of sand and gravel used (see Fig. 1). For this purpose, eight cement pastes were prepared. The proportions of these paste mixtures are shown in Table 1. Firstly, a reference paste was prepared (REF) with a water to cement ratio of 0.45, in which water, cement, limestone powder and superplasticizing admixture were added at dosages corresponding to those of the related SCC. Then, the effect of either calcium oxide or SRA or hydrophobic admixture on the rheology of cement paste was studied by suitably preparing further three cement pastes, in which every addition was made on the basis of its dosage on the final SCC mixture.
2. Research significance This paper presents the results of an investigation carried out to develop white self-compacting concretes (SCCs) by using a suitable combination of three admixtures in order to produce durable concrete for architectural structures. The combined effects of these three admixtures on cement paste and SCC rheology were studied, as well as their combined effects on mechanical performance, drying shrinkage behaviour and cracking tendency of concretes. In fact, shrinkage-free SCC mixtures were studied, obtained by combining CaO-based expansive agent and shrinkage reducing admixture. Moreover, also the influence on the concrete performance of a hydrophobic admixture was tested. Its addition to SCC could be
759
Fig. 1. Cumulative grain size distribution curves of the aggregates.
760
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
Table 1 Mixture proportions of the cement pastes (g).
REF REF + CaO REF + SRA REF + Hyd REF + CaO + SRA REF + CaO + Hyd REF + SRA + Hyd REF + CaO + SRA + Hyd
Water
Cement
Superpl. adm.
Limestone powder
CaO
SRA
Hydroph. adm.
50
110
1.83
44.5 35.3 44.5 44.5 35.3 35.3 44.5 35.3
0 9.2 0 0 9.2 9.2 0 9.2
0 0 1.05 0 1.05 0 1.05 1.05
0 0 0 0.26 0 0.26 0.26 0.26
Finally, the effect of the combined use of calcium oxide, SRA and hydrophobic admixture was studied by preparing the last four cement pastes: in three pastes they were added two by two at their respective dosage and, finally, the last mixture (REF + CaO + SRA + Hyd) was prepared by using all of them together. The rheological behaviour of these cement pastes was tested at 10 min after ingredients’ mixing and then after 25 min: data reported in this paper are those measured after 25 min (i.e. data collected during the second cycle of test). The apparatus was a rotating rheometer based on coaxial rotary cylinders with a slowly increasing shear rate (D), ranging from 1 to 100 s1. Bui et al. [22] found that the rate of 1–100 s1 was the most suitable for rheological model of SCCs, while higher rotation rates were considered too fast, and rates limited to 50 s1 did not produce consistent results. The walls of the concentric cylinders were not smooth but roughened in order to reduce (if not completely eliminate) the ‘‘slip’’ phenomenon, i.e. the development of a water-rich layer close to the inner surface of the rotating cylinder, which produces a lubricating effect, making flow easier, and not representative of the bulk material [23]. The rheological behaviour was described by means of the Bingham flow model:
In order to optimize the grain size distribution of the solid particles in the concrete, sand and gravel were suitably combined, taking into account also the suggestions reported in the literature concerning the mixture proportion of SCC, particularly in terms of maximum volume of coarse aggregate, that is 340 l/m3 [24,25]. The recommended volume of fine particles (including cement, fly ash, ground limestone, etc.) in SCC is in the range of 170–200 l/m3. In this work a very high volume of very fine particles was chosen (200 l/m3), due to the absence of Viscosity Modifying Agents in the SCC mixtures. In particular, in order to achieve a volume of very fine particles of about 200 l/m3, besides to 420 kg/m3 of cement it was necessary to include limestone powder in the mixture, at a dosage of 170 kg/m3. Finally, when the expansive agent (CaO) was added to the mixtures (35 kg/m3), the content of limestone powder was reduced to 135 kg/m3, in order to left unchanged the dosage of very fine materials into the mixture. Finally, shrinkage reducing admixture and hydro-phobic admixture were added at a dosage of 4 kg/m3 and 1 kg/m3, respectively.
3.5. Preparation, curing and testing of concrete specimens
s ¼ sy þ g D where s is the shear stress [Pa], sy is the yield stress [Pa], g is the plastic viscosity [Pa s] and D is the shear rate [s1]. The slope of the down-curve (decreasing shear rate) was used to calculate the plastic viscosity, while the intercept at zero shear rate was used to calculate the yield stress. 3.3. Thermogravimetric measurements on cement pastes Thermogravimetic analysis was carried out on four cement pastes in order to evaluate the influence of the various admixtures on portlandite solubility. This information is essential in order to interpret the results obtained by drying shrinkage tests in the presence of expansive agent. In fact, Chatterji [3] described the expansion in terms of delayed calcium hydroxide formation. This compound, having a molar volume higher than calcium oxide, gives origin to a crystal growth pressure. According to the scientists [11,12], the expansion is strictly related to the solubility of calcium hydroxide: the higher the solubility, the more diffused is the migration of calcium ions, the higher is the distribution of calcium hydroxide and the lower is the expansion. Tests were carried out after one day of wet curing on three samples for each of the cement pastes labelled ‘REF + CaO’, ‘REF + CaO + SRA’, ‘REF + CaO + Hyd’ and ‘REF + CaO + SRA + Hyd’ (their composition is the same reported in Table 1). 3.4. SCC mixtures proportions The SCC mixtures proportions are reported in Table 2. All concretes were prepared with the same w/c of 0.45, the same content of cement (420 kg/m3), water (190 kg/m3), sand (910 kg/m3) and gravel (670 kg/m3), as well as the same dosage of superplasticizer (1.2% by weight of cement plus limestone powder).
As a first step, properties of the fresh concrete other than slump were evaluated according to Italian Standards UNI 11041 [26], since in this case the slump value is not relevant due to very fluid concrete. Therefore, the attention was focused on the measurement of the slump flow, which is the medium diameter (Ufin) of the slumped concrete. Then, also the elapsed time to gain the mean diameter of 500 mm (t500) and the elapsed time to gain the final configuration (tfin) were detected. Moreover, time elapsed for the SCCs passing through V-funnel was detected, according to Italian Standards UNI 11042 [27]. Several concrete specimens were cast for tests on hardened concrete. In particular, for compression tests (carried out according to EN 12390-3 [28]), nine cubic specimens (100 mm in size) were manufactured, then wet-cured at 20 °C, and tested at right angles to the position of casting. Therefore the bearing faces were plane and smooth enough as to require no capping or grinding. The specimens were loaded at a constant strain rate until failure. In addition, for free drying shrinkage monitoring, three prismatic specimens (100 by 100 by 500 mm) were prepared for each concrete mixture according to Italian Standard UNI 6555 ‘Hydraulic Shrinkage Determination’ [29]. After one day of wet curing, the specimens were stored at constant temperature (20 ± 2 °C) and constant relative humidity (50 ± 2 %), while measuring drying shrinkage at different curing times up to 60 days of exposure. Finally, for restrained shrinkage evaluation, further three prismatic specimens (50 by 50 by 300 mm) were prepared for each concrete mixture according to Italian Standard UNI 8148 [30]. Fresh concrete was cast in metallic formworks by embedding a steel bar reinforcement (6 mm in diameter, 300 mm in length). After 6 h they were demoulded, then immersed under water up to 36 h and then exposed to air with constant temperature (20 ± 2 °C) and constant relative humidity (50 ± 2 %). The length-change of the steel bar was then recorded as a function of the curing time up to 60 days of exposure.
Table 2 Mixture proportions of the self-compacting concretes (kg/m3).
REF REF + CaO REF + SRA REF + Hyd REF + CaO + SRA REF + CaO + Hyd REF + SRA + Hyd REF + CaO + SRA + Hyd
Water
Cement
Superpl. adm.
Sand
Fine gravel
Limestone powder
CaO
SRA
Hydroph. adm.
190
420
7
910
670
170 135 170 170 135 135 170 135
0 35 0 0 35 35 0 35
0 0 4 0 4 0 4 4
0 0 0 1 0 1 1 1
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
Fig. 2. Measured yield stress values of the SCC paste fractions.
761
Fig. 4. Calculated thixotropic values of the SCC paste fractions.
Fig. 3. Measured plastic viscosity values of the SCC paste fractions.
4. Results and discussion
Fig. 5. Results of the thermogravimetric analysis (temperature ranging from 200 °C to 500 °C).
4.1. Rheological tests on cement pastes In Fig. 2 the mean values, calculated on three samples for each mixture, of the measured yield stress values are reported. The values obtained are very similar and only slight difference can be noticed among the various pastes. However, the target of sufficiently low values (but still positive) of the yield stress, in order to obtain fluid SCC without segregation [31], was achieved. In particular, data shown in Fig. 2 evidence that the addition of SRA reduces the yield stress value, while the addition of CaO (even if replacing limestone powder) causes it is increase, as well as the addition of the hydrophobic admixture. In Fig. 3 the mean values, calculated on three samples for each mixture, of the measured plastic viscosity values are reported. The values obtained are all included in a very narrow range (0.010– 0.055 Pa s), thus implying that no significant change happens by adding the various ingredients to the mixture. Only a slight increase can be detected when the hydrophobic admixture was used. However, the plastic viscosity values obtained are very low and the Bingham line obtained were practically horizontal. This fact is positive for SCC, when the yield stress value is low but not too much, as in this case, because it means that the same level of cohesion is maintained independently on the concrete flow velocity, by assuring sufficient mobility without internal segregation, as well as flow segregation [31]. Thixotropy is the property of certain gels, such as cement paste, which are rigid when left standing but increase their fluidity when
put into movement. In Fig. 4 the mean values, calculated on three samples for each mixture, of the measured thixotropy values are reported (where thixotropy was calculated as the area included between the up-curve and the down-curve). This measure can give an estimate of the energy necessary to move the SCC and even an estimate of the lateral formwork pressure that SCC will exert after casting, being the lower the pressure the higher the thixotropy value [26]. For the cement pastes tested in this work the value of thixotropy was always negative, thus showing that all these cement pastes have anti-thixotropic behaviour. In particular, the addition of either CaO or SRA seems to enhance the thixotropic behaviour, while the addition of the hydrophobic admixture seems to enhance the opposite.
4.2. Thermogravimetric measurements on cement pastes Thermal analysis carried out show that the examined cement pastes contain different amount of portlandite, Ca(OH)2. For the cement pastes labelled ‘REF + CaO’, ‘REF + CaO + Hyd’, ‘REF + CaO + SRA’ and ‘REF + CaO + SRA + Hyd’ the measured values (expressed in% by weight) were 12.72%, 12.48%, 9.39% and 9.78%, respectively. The first evidence is that the addition of SRA hindered portlandite formation (about 25%), and this result is also confirmed by the literature [11,12]. In fact, as it can be observed in Fig. 5, the weight losses (corresponding to the flex of the DTA curves, for simplicity
762
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
not reported in Fig. 5), which occur from about 400 °C to 450 °C (typical range for the decomposition reaction of calcium hydroxide), are less pronounced for the two mixtures containing SRA. Chatterji [3] stated that the lower is the solubility of calcium hydroxide, the higher was the concrete expansion. SRA could reduce the water solubility of calcium hydroxide, promoting the material expansion and compensating more efficaciously concrete shrinkage. On the other hand, only an insignificant effect on the content of portlandite was detected when the hydrophobizing admixture was used. On the basis of the results obtained, differently from SRA, there is not evidence of any hindering action of the hydrophobizing admixture on the calcium hydroxide formation. Consequently, according to the scientists [11,12], its addition to the concrete mixture should not significantly modify the drying shrinkage behaviour of the SCC containing calcium oxide, while the SRA addition should improve expansion of concrete, and consequently the effectiveness of the calcium oxide itself.
4.3. Slump flow and V funnel tests As shown in Table 3, all concretes had a sufficient deformability under their own weight (strictly related to the Ufin value, included in the range 620–640 mm). The presence of bleeding water was never observed at the periphery of the slumped material. Time elapsed for the SCCs passing through V-funnel was in the range 9–11 s in all cases, within the acceptance limits (5–12 s) [27]. The only significant effect on concrete rheology, which can be observed, is that of the hydrophobic admixture. In fact, its addition produced slightly higher tendency of the SCC to be viscous (longer time to gain the final configuration and to pass trough the V-funnel). This evidence was also detected by means of the rheological tests on cement pastes (see Figs. 2 and 3).
Fig. 6. SCC compressive strength vs. curing time.
The use of hydrophobizing admixture at low dosage (1 kg per cubic meter of concrete, corresponding to 0.25% by weight of cement) did not influence compressive strength at early age, and only a certain reduction can be detected after 7 and 28 days of curing (9%). As expected on the basis of the results obtained when used singularly, the worst result in terms of mechanical performance was obtained by contemporarily using SRA and hydrophobizing admixture. However, the use of the three additions all together (mixture ‘REF + CaO + SRA + Hyd’) allowed to obtain compressive strength values close to those of the reference mixture (around -5% independently on the age). In conclusion, the target concrete strength class C 37/45 was reached by the mixtures: ‘REF’, ‘REF + CaO’, ‘REF + CaO + SRA’ and ‘REF + CaO + SRA + Hyd’. On the other hand, a lower class strength C 32/ 40 was reached by the mixtures prepared by using the SRA and the hydrophobizing admixture alone, and an even lower strength class (C 30/37) characterized the mixture in which SRA and the hydrophobizing admixture were used together without expansive agent.
4.4. Compression tests Compressive strength of concrete was determined at 1, 7, and 28 days of curing and the results obtained are reported in Fig. 6. With respect to the reference mixture, the addition of the expansive agent (CaO) caused an evident strength increase after 1 day of curing (about +20%), then this effect disappeared with increasing curing time. On the other hand, the use of SRA, caused an evident strength loss, particularly at early age (about 20% after 1 day), but also after 28 days of curing (about 10%). The reason for this slower strength development seems to be that the addition of SRA to the mixing water depresses the dissolution of alkalis in the pore fluid [33]. This fact results in a pore fluid with lower alkalinity, which causes a reduction in the rate of cement hydration. However, if used in combination with CaO this negative effect on concrete compressive strength is practically negligible (see mixture ‘REF + CaO + SRA’).
4.5. Free drying shrinkage tests Results obtained are reported in Fig. 7. With respect to the reference mixture, the addition of the SRA showed to be very effective in reducing drying shrinkage of concrete, in fact the ultimate strain of the mixture ‘REF + SRA’ is about 40% lower than that of the reference mixture ‘REF’. In addition, after two days of exposure a slight expansion was detected in the presence of SRA. Sant et el. [12] demonstrated that the presence of SRA increases the portlandite oversaturation level in solution, which can result in higher crystallization stresses, which in turn could lead to an early-age expansion.
Table 3 Results of tests on fresh SCCs. Slump flow
REF REF + CaO REF + SRA REF + Hyd REF + CaO + SRA REF + CaO + Hyd REF + SRA + Hyd REF + CaO + SRA + Hyd
V-funnel
Ufin (mm)
t500 (s)
tfin (s)
tfin (s)
640 620 640 630 630 630 640 630
2 2 2 2 2 2 2 2
10 9 9 10 8 10 9 10
10 10 9 11 9 11 10 11
Fig. 7. SCC free drying shrinkage vs. time of exposure.
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
The addition of the hydrophobizing admixture seems not to affect drying shrinkage: in fact, quite the same behaviour can be detected for the mixtures ‘REF + Hyd’ and ‘REF’. On the other hand, when SRA and hydrophobizing admixture are used together (mixture ‘REF + SRA + Hyd’), the effectiveness of SRA is hindered: the initial expansion was not present anymore, and the reduction of strain is only about 13% instead of 40%. When calcium oxide was added to the mixture ‘REF + CaO’, the initial expansion of almost 0.4 mm/m allow to reach a very low value of drying shrinkage after 60 days of curing (about 75% with respect to the reference mixture). Moreover, the contemporaneous use of SRA and CaO proved to be very effective, and it caused strong initial expansion of SCC (+0.5 mm/m), which gradually decreased, but still remained positive after 60 days of exposure to 50% relative humidity. The addition of the hydrophobizing admixture to the mixture prepared with the expansive agent (mixture ‘REF + CaO + Hyd’) seems not to affect drying shrinkage, as expected on the basis of the results obtained by thermogravimetric analysis: in fact, quite the same behaviour can be detected for the mixtures REF + CaO + Hyd’ and ‘REF + Hyd’. On the other hand, also in the presence of CaO, when SRA and hydrophobizing admixture are used together (mixture ‘REF + CaO + SRA + Hyd’), the effectiveness of CaO + SRA is reduced: the initial expansion is reduced from 0.5 mm/m to 0.42 mm/m and the same values of strain were reached in advance by the mixture containing hydrophobizing admixture (for example + 0.03 mm/m was already measured after 35 days instead of 60 days). By comparing the results obtained for the mixtures ‘REF + SRA + Hyd’, ‘REF + CaO + Hyd’ and ‘REF + CaO + SRA + Hyd’, it can be noticed that the negative effect of the hydrophobizing admixture is evident only in the presence of SRA. Data obtained by means of thermogravimetric analysis showed only slight reduction of the amount of portlandite (about 2%) if the hydrophobizing admixture was added in the absence of SRA, while the amount of portlandite slightly increased (about 4%) when its was added in the presence of SRA, but this difference is probably too small for justifying such a macroscopic effect. Results obtained suggest a likely interference between SRA and hydrophobizing admixture, both acting on capillary pore forces, by reducing the surface tension of water (from 72 dyn/cm to approximately 40 dyn/cm), in the case of SRA, and by changing the sign of the contact angle (from hydrophilic to hydrophobic), in the case of the hydrophobizing admixture. In fact, although the mechanisms of drying shrinkage of concrete are not fully understood, literature suggests that, when considering concrete shrinkage in the 45–90% relative humidity range, capillary stress appears to be the predominant mechanism [10]. When pore water evaporates from capillary pores in hardened concrete during drying, tension in the liquid is transferred to the capillary walls, resulting in shrinkage. For a given pore size distribution, the internal stress generated upon evaporation is proportional to the surface tension of the pore water solution. However, the shrinkage measured for the mixture ‘REF + CaO + SRA + Hyd’ after 60 days of exposure can be considered negligible (0.060 mm/m) and the appearance of cracking is strongly unlikely also for this mixture. 4.6. Restrained drying shrinkage tests Results obtained are reported in Fig. 8. With respect to the reference mixture, the addition of the SRA showed to be very effective in reducing drying shrinkage of concrete, in fact the final strain of the mixture ‘REF + SRA’ is about 50% lower than that of the reference mixture ‘REF’. The addition of the hydrophobizing admixture seems not to affect drying shrinkage: in fact, quite the same behaviour can be detected for the mixtures ‘REF + Hyd’
763
Fig. 8. Restrained drying shrinkage vs. time of exposure.
and ‘REF’. On the other hand, when SRA and hydrophobizing admixture are used together (mixture ‘REF + SRA + Hyd’), the effectiveness of SRA is hindered and the reduction of strain is only about 30% instead of 50%. The reason for that was already discussed above. When calcium oxide was added to the mixture ‘REF + CaO’, the initial expansion of about 0.3 mm/m allow to reach a very low value of drying shrinkage after 60 days of curing (about 0.150 mm/m). Moreover, the contemporaneous use of SRA and CaO proved again to be very effective, and it caused strong initial expansion of SCC (+0.36 mm/m), which gradually decreased, but still remained positive after 60 days of exposure to 50% relative humidity. The addition of the hydrophobizing admixture to the mixture prepared with the expansive agent (mixture ‘REF + CaO + Hyd’) seems not to affect drying shrinkage: in fact, quite the same behaviour can be detected for the mixtures ‘REF + Cao + Hyd’ and ‘REF + Hyd’. On the other hand, also in the presence of CaO, when SRA and hydrophobizing admixture are used together (mixture ‘REF + CaO + SRA + Hyd’), the effectiveness of CaO + SRA is reduced: the initial expansion is slightly reduced from 0.36 mm/m to 0.32 mm/m, and the same values of strain were reached in advance by the mixture containing hydrophobizing admixture (for example + 0.06 mm/m was already measured after 20 days instead of 60 days). By comparing the results obtained by free drying shrinkage test and restrained shrinkage test, the same trend were found and similar discussions can be carried out, as well as the same conclusions can be drawn. 5. Conclusions Based on the results obtained by means of rheological tests, the following conclusions are drawn: 1. Concerning the yield stress the addition of SRA reduces its value, while the addition of CaO (even if replacing limestone powder) causes it is increase, as well as the addition of the hydrophobic admixture. 2. Concerning the plastic viscosity, all the values obtained are very low and the Bingham line obtained were practically horizontal. 3. On the basis of the thixotropy values detected, all the cement pastes have anti-thixotropic behaviour: in particular, the addition of either CaO or SRA seems to enhance the thixotropic behaviour, the opposite for the hydrophobic admixture. 4. Generally, the same level of shear stress is maintained independently on the paste flow velocity, which should assure sufficient mobility, without internal segregation and flow segregation, for the related SCCs.
764
V. Corinaldesi / Construction and Building Materials 36 (2012) 758–764
5. In fact, concerning the tests carried out on the fresh SCCs, all concretes showed good flowability in absence of segregation. Moreover, based on the results obtained by means of tests on hardened concretes, the following conclusions are drawn:
[3] [4] [5]
[6]
1. The addition of the expansive agent (CaO) caused an increase of mechanical strength at early age; on the other hand, the use of SRA, caused an evident strength loss, particularly at early age, but also later; however, if SRA and CaO are used in combination the negative effect of SRA on concrete compressive strength is practically negligible; moreover, the use of hydrophobizing admixture at low dosage only caused a late reduction of strength. 2. Concerning the combined effects (of either SRA plus CaO, or SRA plus hydrophobizing admixture, or CaO plus hydrophobizing admixture) on mechanical performance, as predictable on the basis of the results obtained when used singularly, the worst result was obtained by using together SRA and hydrophobizing admixture; however, the contemporary use of all the three additions allowed to obtain the same strength class values of the reference mixture. 3. In terms of drying shrinkage (free or restrained), the addition of the SRA showed to be very effective in reducing it, while the addition of the hydrophobizing admixture seems not influent; on the other hand, when SRA and hydrophobizing admixture are used together the effectiveness of SRA is hindered. 4. In terms of drying shrinkage (free or restrained), the use of the expansive agent and, in particular, the contemporaneous use of SRA and CaO proved to be very effective, and it caused strong initial expansion of SCC, which gradually decreased, but still remained positive after 60 days. 5. The addition of the hydrophobizing admixture to the mixture prepared with the expansive agent seems not to affect drying shrinkage, but when CaO, SRA and hydrophobizing admixture are used all together, the effectiveness of CaO + SRA is reduced. 6. In conclusion, the negative effect of the hydrophobizing admixture on drying shrinkage is evident only in the presence of SRA: this result suggests a likely interference between SRA and hydrophobizing admixture, both acting on capillary pore forces. 7. In fact, data obtained from the thermogravimetric analysis only partially confirm the hypotesis that the increase of the portlandite oversaturation level in solution, which causes higher cristallization stresses, can be the main factor causing early-age expansion (in fact, such hypothesis is able to explain the concrete shrinkage behaviour in the presence of SRA, but not that in the presence of both SRA and hydrophobizing admixture together).
References [1] Níelsson I, Wallevik ÓH. Rheologycal evaluation of some empirical test methods-preliminary results. In: Wallevik ÓH, Níelsson I, editors. 3rd International RILEM symposium. RILEM Pub. PRO. vol 33; 2003. p. 59–68. [2] Collepardi M, Collepardi S, Troli R. Pratical applications of SCC in European works. In: Kraus Rudolph N, Naik Tarun R, Claisse Peter, Sadeghi-Pouya,
[7]
[8] [9] [10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
[18] [19] [20] [21] [22] [23] [24]
[25]
[26] [27] [28] [29] [30] [31]
[32]
editors. In: Proc int conf: sustainable construction materials and technologies, special papers proceedings. Milwaukee; Pub UW, CBU; 2007. p. 51–58. Chatterji S. Mechanism of expansion of concrete due to the presence of dead burnt CaO and MgO. Cem Concr Res 1995;25:51–6. Neville AM. Properties of concrete. 4th ed. Longman Group Limited; 1995. Collepardi M, Borsoi A, Collepardi S, Ogoumah Olagot JJ, Troli R. Effects of shrinkage-reducing admixture in shrinkage compensating concrete under non wet curing conditions. Cem Concr Compos 2005;27(6):704–8. Collepardi M, Troli R, Bressan M, Liberatore F, Sforza G. Crack-free concrete for outside industrial floors in the absence of wet curing and contraction joints. Cem Concr Compos 2008;30(10):887–91. Meddah MS, Suzuki M, Sato R. Influence of a combination of expansive and shrinkage-reducing admixture on autogenous deformation and self-stress of silica fume high-performance concrete. Constr Build Mater 2011;25:239–50. Rongbing B, Jian S. Synthesis and evaluation of shrinkage-reducing admixture for cementitious materials. Cem Concr Res 2005;35:445–8. Folliard KJ, Berke NS. Properties of high-performance concrete containing shrinkage-reducing admixture. Cem Concr Res 1997;27:1357–64. Bentz DP, Geiker MR, Hansen KK. Shrinkage-reducing admixtures and earlyage desiccation in cement pastes and mortars. Cem Concr Res 2001;31:1075–85. Maltese C, Pistolesi C, Lolli A, Bravo A, Cerulli T, Salvioni D. Combined effect of expansive and shrinkage reducing admixtures to obtain stable and durable mortars. Cem Concr Res 2005;35(12):2244–51. Sant G, Lothenbach B, Juilland P, Le Saout G, Weiss J, Scrivener K. The origin of early age expansions induced in cementitious materials containing shrinkage reducing admixtures. Cem Concr Res 2011;41:218–29. Collepardi M, Passuelo A. The best SCC: stable, durable and colourable. In: Helene P, Holland TC, Pazzini E, Bittecourt R, editors. Proc IV int ACI/CANMET conference on quality of concrete structures and recent advances in concrete materials and testing, furnas centrais Elqtricas S.A. Goidnia, Brasil: Civil Engineering Technological Center; 2005. Moriconi G, Tittarelli F, Corinaldesi V. Review of silicone-based hydrophobic treatment and admixtures for concrete. Indian Concr J 2002; 76(10):637–42. EN 197-1. Cement – Part 1: composition. Specifications and conformity criteria for common cements; 2000. Felekoglu B. Utilisation of high volumes of limestone quarry wastes in concrete industry (self-compacting concrete case). Resour Conserv Recycl 2007;51:770–91. El Barrak M, Mouret M, Bascoul A. Self-compacting concrete paste constituents: hierarchical classification of their influence on flow properties of the paste. Cem Concr Compos 2009;31:12–21. Corinaldesi V, Moriconi G, Naik TR. Characterization of marble powder for its use in mortar and concrete. Constr Build Mater 2010;24(1):113–7. EN 933-1. Tests for geometrical properties of aggregates – determination of particle size distribution – sieving method; 1997. EN 1097-6. Tests for mechanical and physical properties of aggregates – determination of particle density and water absorption; 2000. Saak AW, Jennings HM, Shah SP. New methodology for designing selfcompacting concrete. ACI Mater J 2001;98(6):429–39. Bui VK, Akkaya Y, Shah SP. Rheological model for self-consolidating concrete. ACI Mater J 2002;99(6):549–59. Saak AW, Jennings HM, Shah SP. The influence of wall slip on yield stress and viscoelastic measurements of cement paste. Cem Concr Res 2001;31:205–12. Bui VK, Montgomery D. Mixture proportioning method for self-compacting high performance concrete with minimum paste volume. In: Skarendahl A, Petersson O, editors. Self compacting concrete, PRO 007. Bagneux, France: RILEM Publication s.a.r.l; 1999. p. 373–84. Jacobs F, Hunkeler F. Design of self-compacting concrete for durable concrete structures, in ‘self-compacting concrete’. In: Skarendahl A, Petersson O, editors. Proc of the first int. RILEM symp, Stockholm, Sweden; 1999. p. 397– 407. Assaad J, Khayat KH, Mesbah H. Variation of formwork pressure with thixotropy of self-consolidating concrete. ACI Mater J 2003;100(1):29–37. UNI 11041. Prova sul calcestruzzo autocompattante fresco – determinazione dello spandimento e del tempo di spandimento; 2003. UNI 11042. Prova sul calcestruzzo autocompattante fresco – determinazione del tempo di efflusso dall’imbuto; 2003. UNI 6555. Hydraulic shrinkage determination; 1973. UNI 8148. Agenti espansivi non metallici per impasti cementizi – determinazione dell’espansione contrastata del calcestruzzo; 2008. Wallevik ÓH. Rheology – a scientific approach to develop self compacting concrete. In: Wallevik ÓH, Níelsson I, editors. Self compacting concrete, PRO 33. Bagneux, France: RILEM Publication s.a.r.l; 2003. p. 23–32. Rajabipour F, Sant G, Weiss J. Interactions between shrinkage reducing admixtures (SRA) and cement paste’s pore solution. Cem Concr Res 2008;38(5):606–15.