International RILEM Conference on Material Science MATSCI, Aachen 21 Vol. III, AdIPoC 129 ENHANCING EARLY-AGE RESISTANCE TO CRACKING IN HIGH- STRENGTH CEMENT-BASED MATERIALS BY MEANS OF INTERNAL CURING USING SUPER ABSORBENT POLYMERS L. Dudziak, V. Mechtcherine, Institute of Construction Materials, TU Dresden, Germany ABSTRACT: This paper reports on the effectiveness of internal curing to reduce two major causes of early-age cracking in high-strength cementitious materials, i.e. autogenous and plastic shrinkage. As a measure to develop such curing, the materials, including fine-grained UHPC and cement paste, were enriched using varied amounts of Super Absorbent Polymers (SAP) and extra water whilst mixing. Investigations of shrinkage under free and restrained conditions demonstrated that internal curing can diminish the likelihood of early-age cracking significantly. The amount of SAP and extra water added was found to be decisive both in determining the magnitude of such changes and the history of evolution of autogenous shrinkage. Some deterioration of the early-age mechanical properties in internally cured materials resulted from the retarding of their setting behaviour and from particular changes in their porosity. 1 INTRODUCTION A number of high-strength cementitious materials such as Ultra High Performance Concrete (UHPC) have been increasingly promoted because of their advantages in comparison to ordinary concrete. In particular the combination of enhanced compressive strength with improved durability due to the dense microstructure has evidently made this group of materials attractive for use in harsh environments and under extreme operating conditions. However, the markedly lower resistance to tensile stresses than to compressive stresses left a permanent chink in the armour against early-age shrinkage cracking as a result of volume changes. As part of a reasonable solution to this, the sources of these volume changes must be eliminated to take advantage of the great potential promised by these advanced materials. At early ages, which could be considered as 24 hours after the cement begins to hydrate, shrinkage occurs as the result of a few complex physical-chemical phenomena. Particularly when dealing with low water-to-cement systems, some types of shrinkage are considerably more pronounced when compared to those of ordinary concrete; this is due to the large disproportion between binder content and water availability. With this in mind, in the development of both autogenous and plastic shrinkage a range of factors contributing to the entire shrinkage-cracking phenomenon must be considered. Conceptually autogenous shrinkage may be regarded as the common effect of two phenomena: Le Chatelier s contraction (so-called chemical shrinkage) and progressive desaturation of the porous microstructure (so-called self-desiccation) in the material. The role of each phenomenon depends on the form of the cement paste microstructure, where time zero, i.e. the point at which the hydrating system is able to support its own weight and can be regarded as solid, provides the decisive threshold mark. Time zero signals the concrete age, beginning with which increasing capillary stresses forcing the matrix to contract become relevant to its propensity to form cracks.
13 DUDZIAK, MECHTCHERINE: Enhancing Early-Age Cracking Resistance by Internal Curing Using SAP In contrast to autogenous shrinkage, which occurs without any mass exchange with its surroundings, plastic shrinkage results by definition from water loss. It occurs when the magnitude of the water losses due to evaporation is greater than that replenishing the surface moisture, e.g. bleeding or moist curing. Evaporation causes the buildup of negative capillary pressure, which leads to a volume decrease in the material at a stage when it is still easily plastically deformable and has not yet reached significant strength. Both types of shrinkage can lead to cracking in the presence of restraint (external: adjacent structure, formwork; internal: aggregates, reinforcement). Internal curing (IC) is an actively developing approach in concrete technology and is oriented on the uniform distribution of curing water over the entire volume of a particular concrete structure. First research efforts on this topic were summarized in the framework of RILEM TC 196-ICC [Kov7]. Common to IC methods is that prior to mixing the original compositions of high-performance materials are enriched by components with high water storing capacity (IC agents) and some extra water, which may be absorbed by the IC agent before its addition to the mix, or indeed during mixing. IC agents are added to store the absorbed water until capillary suction forces develop in the fine pore system of the surrounding matrix and gradually extract the water from these reservoirs. Among all the natural and artificial products which can be potentially utilized for the purpose of IC, Super Absorbent Polymers (SAP) appear presently to be the most promising water-regulating additive. These polymers consist of cross-linked chains having dissociated ionic functional groups which facilitate the absorption of large amounts of water. The advantage of SAP is that in principle they can be engineered to the particular needs of the IC problem at hand by designing the necessary size and shape of the particles, their water absorption capacity, and other properties. Moreover, SAP can be added in dry form to the mix as a powder, which makes the handling of IC measures easier. This is the one of the main attributes differentiating SAP from other proposed IC agents, e.g. light-weight aggregates (LWA). While a number of studies examined the effectiveness of IC using LWA or SAP on autogenous shrinkage, very limited studies have been performed on the effects of IC on plastic shrinkage in which only LWA was used [Hen1]. Further to that, there is still insufficient information on the effects of IC measures on properties of cementitious systems in their fresh and hardened states. This paper presents a brief review of the findings of a research project examining the use of SAP for internal curing in two high-strength cement-based materials: a cement paste and a fine-grained UHPC. It addresses the effectiveness of IC in mitigating free plastic shrinkage as well as autogenous shrinkage under both free and restrained conditions. Furthermore, possible contributions to changes observed in the properties of UHPC in its fresh and hardened states are discussed based on the results for the reference material (no SAP, no extra water), SAPenriched material (SAP, no or some extra w/c) and secondary reference material (no SAP but extra w/c). 2 MIX PROPORTIONS AND PROPERTIES IN THE FRESH STATE 2.1 Cement paste proportions and concrete compositions Two cement pastes were studied in collaboration between the TU Dresden and the Leipzig University of Applied Sciences: a plain cement paste with no internal curing (P-R) and a cement paste with IC (P-S.6.87). The water-to-cement ratios of these pastes were.265 and
International RILEM Conference on Material Science MATSCI, Aachen 21 Vol. III, AdIPoC 131.352, respectively. Cement CEM I 32.5 R was used. The designation of paste with IC was particularly chosen to yield information on the amount of SAP used (.6% by mass of cement) as well as the amount of additional water extra implemented for IC in reference to mass of cement (here.87). The suspension-polymerised SAP selected had spherical particles of average size 15 µm. For the production of pastes, a small dosage of superplasticizer was applied (.37% m.c.); a standard mortar mixer was used. The same workability of both pastes in the fresh state was achieved, i.e. slump flow tests performed with a small cone provided values between 19 and 195 mm. Table 2.1. Mixing compositions [kg/m³] 1 Mixture Cement Silica fume Water SAP 1 Extra water for IC Extra w/c in reference to F-R w/(c+s) (free water) 2 w/(c+s) (total, incl. IC water) 2 F-R 853.4 138.5 17.3 - - -.19.19 212.3 1.1 3.2 F-S.4 853.4 138.5 17.3.4 - -.19.19 212.3 1.1 3.2 F-R.4 824.5 133.8 197.5 - -.4.23.23 25.1 966.3 29.1 F-S.3.4 824.5 133.8 164.5.3 33..4.19.23 25.1 966.3 29.1 F-R.7-1 817.6 132.7 22.4 - -.7.24.24 23.4 958.1 11. F-R.7-2 84.2 13.5 216.8 - -.7.25.25 2.1 942.4 28.4 F-S.4.7 84.2 13.5 16.5-56.3.7.19.25 2.1 942.4 28.4 F-S.6.8 797.6 129.4 159.1.6 63.8.8.19.26 198.4 934.7 28.2 F-S1..16 748.7 121.5 149.4 1. 119.8.16.19.33 186.3 877.3 26.5 1 as an exception SAP content is given in % of m.c. 2 includes 65% water from superplasticizer The mixtures of finely grained UHPC examined were composed on the basis of the recipe developed and optimized in the framework of the Priority Program 1182 of the German Research Foundation (DFG) [Feh4]. In contrast to the original mixture, steel fibres were not used in the part of the investigation presented, cf. Table 2.1. In total nine mix compositions were experimentally evaluated in this study: one reference concrete mix with no internal curing (F-R), five concrete mixes with IC (F-S.4, F-S.3.4, F-S.4.7, F-S.6.8 and F-S1..16) and three concrete mixes with an equivalent amount of extra added water (F-R.4, F-R.7-1 and F-R7-2). In contrast to the cement paste study, CEM I 52.5 R-HS/NA was used for the production of UHPC specimens. Similarly to cement paste, the designations used for UHPC with IC indicate the amount of SAP and/or extra water implemented. With regard to IC, the variable parameters were the amount of IC agent (.3,.4,.6 and 1. % SAP, respectively, related to the mass of cement) and the amount of additional water (.4 to.16 in reference to the mass of cement). An exception is the SAP-enriched mix identified as F-S.4, to which no extra water for IC was introduced. The selfsame SAP material as that used in the investigations of cement pastes was used. In the group of mixtures with extra w/c but no IC agent, a specific case is concrete F-R.7-1 which, unlike all the other mixtures examined, contained a decreased amount of superplasticizer. To this particular change was attributed the expected loss of homogeneity of the mixture with increased w/c (excessive, not IC water) if the amount of superplasticizer was not reduced. For the production of UHPC, two high intensity mixers (HIM), including one with a vacuum unit (HIMvac) were used depending on the investigation. With regard to workability, the same values of slump flow testing according Quartz flour Quartz sand Superplasticizer
132 DUDZIAK, MECHTCHERINE: Enhancing Early-Age Cracking Resistance by Internal Curing Using SAP to DAfStb standards for self-compacting concrete were aimed for the reference mixture as well as the group of mixtures with SAP and extra water. More detailed information on the modifications of recipes with regard to IC as well as a statistical evaluation of the fresh mix properties for the mixes studied repeatedly (F-R, F-S.4, F-S.3.4, F.S.4.7) may be found in [Dud1]. 2.2 Properties in the fresh state Fig. 2.1 (a), (b), and (c) illustrate representative results regarding three main concrete properties in the fresh state for the reference mixture F-R (no SAP, no extra water), SAPenriched mixtures (varied SAP content with some or no extra water), and mixtures with equivalent w/c (no SAP but extra water). The results obtained for identical overall w/c ratios (incl. extra water) are connected by lines in order to demonstrate the trends in the value changes due to the addition of SAP. For the sake of a better overview, the names of the particular mixtures are indicated in Fig. 2.1 (a) (and partially in Fig. 2.1 (b)) only. 1 7 F-R.7-2 F-S1..16 F-R.4 6 9 5 8 F-R F-S.4.7 F-S.6.8 4 F-S.3.4 7 3 2 F-R (HIMvac) 6 1 F-R.4 (HIMvac) F-S.4 F-S.3.4 (HIMvac) 5..2.4.6.8 1...2.4.6.8 1. (a) SAP content [% by mass of (b) (c) SAP content [% by mass of cement] cement] slump flow spread [cm] referred value F-R referred value F-R, HIMvac extra w /c= extra w /c=, HIMvac extra w /c=.4 extra w /c=.4, HIMvac extra w /c=.7 extra w /c=.8 extra w /c=.16 air content [%] density x1 3 [kg/m³] 2.45 2.4 2.35 2.3 2.25 2.2 2.15..2.4.6.8 1. SAP content [% by mass of cement] Fig. 2.1. Slump flow spread (a), air content (b), and density of fresh concrete (c) depending on SAP content and extra w/c. Reference values of the mixture F-R (% SAP, extra w/c=) produced with HIM and HIMvac are indicated by the dashed black lines and grey dotted lines, respectively. Due to the absorptive nature of SAP, all mixtures with an IC agent demonstrated a loss of workability for the same entire water-to-cement ratio (incl. extra water). As a consequence the initial total w/c of the reference mix F-R had to be increased by approximately.15 for every.1% SAP used for any SAP-enriched mix to preserve the same range of slump flow spread values (75-83 cm [Dud1], average: 78 cm). A small overestimation of SAP absorption led to higher slump flow spread values (mix F-S1..16: 85 cm), while some underestimation resulted in noticeable changes in the reverse direction (e.g. mix F-S.6.8: 64 cm). As expected, the addition of extra water without the use of SAP increased the flowability of mixtures F-R.4 and F-R.7-2 (slump flow 89 and 96 cm, respectively) dramatically. At the same time the extra water led to lower air content and higher density of the mixtures in comparison to the reference mix F-R. Reverse effects were observed on implementation of the SAP: all of the mixtures enriched with SAP exhibited some increase in air content and a
International RILEM Conference on Material Science MATSCI, Aachen 21 Vol. III, AdIPoC 133 decrease in density in reference to F-R. In all mixtures produced by using the vacuum mixer HIMvac, the air content was very low (below 1 %). Subsequently, the density increased somewhat in comparison to the mixtures produced with HIM without applying vacuum. The phenomenon visually observed and not illustrated in the graphs was some loss of homogeneity in the mixtures with extra water and no SAP, especially apparent for the mix F- R.7-2 (% SAP, extra w/c=.7). For this reason, in the section on autogenous shrinkage only mixture F-R.7-1, which was produced with a reduced amount of superplasticizer and thus a somewhat lower total w/c, was taken into account. It showed only insignificantly smaller slump flow spread (72 cm) in comparison to the reference F-R, but possessed, in contrast to F-R.7-2, the required homogeneity. 3 EXPERIMENTAL METHODS The measurement of free plastic deformation was carried out using a test setup developed by Slowik [Slo8]. It consists of a set of two moulds, each of dimensions 3x3x1 mm³. The measurement units enable a continuous monitoring of free plastic shrinkage, settlement, and mass loss for the first mould, while capillary pressure and temperature developed in the sample are measured in the second mould. During each series of experiments, starting approximately 4 minutes from the end of mixing, the 6cm-thick samples placed in these moulds were subjected to constant air ventilation at a flow rate of an average of 3.7 m/s and at a temperature of 25±1 C. The measurement of free autogenous shrinkage was performed using the method developed by Jensen and Hansen [Jen95], where the special design of the measuring device (dilatometer) and the use of polyethylene corrugated, tube-shaped moulds enable continuous monitoring of the concrete s deformations beginning immediately after the filling and encapsulating of the tubes. The examination performed simultaneously on triple samples for one series of tests was carried out at a quasi-constant temperature of 2 C. The start of stress relevant deformations was evaluated from the final set, taken at time zero. Instrumented ring tests were performed to assess the magnitude of the tensile stresses developed due to restrained autogenous deformations both with and without SAP addition and hence to estimate quantitatively the tendency of these concretes to crack. The dimensions of the inner and outer steel rings used in casting the concrete annuli and adjusted to measure both the expansion and shrinkage of the mixtures was based on the test setup according to ASTM C1581-4. The strains measured on the inner steel ring were used for calculating tensile stresses in the concrete according to the equations proposed in [Hos4]. To maintain sealed conditions during measurement, the top surfaces of the concrete annuli were protected by a dual layer of self-gluing aluminium foil. The storing conditions during examination were 2 C and 65% RH. The Vicat needle penetration test in accordance with DIN EN 196-3 and DIN EN 48-2, ultrasonic transmission measurement and measurement of the temperature development in moulds of the same dimensions as in the ultrasonic tests, were used to estimate final setting and to monitor setting behaviour. The conditions during the Vicat tests were 2 C and over 9% RH. For ultrasonic and temperature measurements the probes were stored at 2 C and covered with plastic foil to avoid desiccation. Compression tests as well as bending tests were used to evaluate the mechanical performance of the concretes under investigation. First, three-point bend tests on prismatic beams with
134 DUDZIAK, MECHTCHERINE: Enhancing Early-Age Cracking Resistance by Internal Curing Using SAP dimensions 16x4x4 mm³ were performed. The halves of such specimens obtained were used subsequently in examining compressive strength. The storage conditions between demoulding (at 24 h) and examination (at an age of 3 or 7 days) included sealing with plastic foil and storing in a climatic room at 2 C and 65% RH. Only prisms examined at the age of 1 day were taken directly from the moulds. A helium pycnometer and a weighing method according to the Principle of Archimedes were used in measuring the true and the bulk density of materials, respectively. The material was taken from sealed beams at concrete ages of 1, 3, and 7 days, shattered and dried at 15 C until a constant mass was reached. For the estimation of true density, the milling process was applied in addition. 4 FREE PLASTIC DEFORMATION Fig. 4.1 shows the development of the key parameters in the measurement of free plastic deformation up to the hardened state in the first 24 hours. The addition of SAP and extra water to cement pastes led to a considerable reduction in their plastic shrinkage, i.e. from -5.9 to -2.8x1³ µm/m for the ultimate values measured (Fig. 4.1 (a)). The major portion of this improvement took place before both mixes had started to gain rigidity (see the characteristic increase in temperature developed in Fig. 4.1(b)) and before the loss of mass for the SAPenriched paste became higher than in the reference paste. The negative capillary pressure in the reference mix P-R reached its highest value, -47.2 kpa, at an age of approximately 5 hours. After the subsequent sudden drop, the capillary pressure remained nearly constant at a level of approximately -5 kpa. The paste P-S.6.87 containing SAP had the first peak of capillary pressure (-17.8 kpa) at about the same age as the reference mixture and a second, smaller peak approximately one hour later. The maximum capillary pressure reached by the reference paste P-R coincided fairly well with the maximum of the settlement deformations (-12.9x1³ µm/m), while the maximum of the settlement deformations for the SAP-enriched paste (-18.x1³ µm/m) appeared simultaneously with the second peak in capillary pressure. P-R P-S.6.87 P-R P-S.6.87 time [h] time [h] 5 4 8 12 16 2 24-72 4 8 12 16 2 24 5 shrinkage / settlement x1³ [mm/m] [µm/m] -5-1 -15-2 -25-3 (a) Fig. 4.1. settlement shrinkage 2x capilary pressure at depth 4 mm -6-48 -36-24 -12 capillary pressure [kpa] mass loss [g] 12-5 (b) Evolution over time of (a) plastic shrinkage, settlement and capillary pressure in comparison to (b) temperature development and loss of mass in the samples investigated. -1-2 -3-4 mass loss temperature in sample 4 3 2 1 developed temperature [ C]
International RILEM Conference on Material Science MATSCI, Aachen 21 Vol. III, AdIPoC 135 An increase in the settlement variable with the implementation of SAP and extra water could be attributed to two different phenomena. The first is the earlier and higher rise of the temperature due to the heat of hydration in the case of the reference mixture, cf. Fig. 4.1 (b). The corresponding thermal expansion makes the overall values for settlement deformation lower. The second phenomenon is the high positive autogenous deformations of the reference mix during the first few hours after casting, not apparent for the mix with SAP, cf. Fig. 4.2. 5 P-R P-S.6.87 75 autogenous strainx1³ [µm/m] 4 3 2 1 initial set: 6h final set: 7h initial set: 7h final set: 9h temperature in sample volumetrically equal to ultrasonic test 4 8 12 16 2 24 6 45 3 15 developed temperature [ C] time [h] Fig. 4.2. Evolution over time of autogenous shrinkage for cement paste with and without IC. Initial and final setting according to DIN EN 196-3 are indicated on shrinkage curves with hollow and full dots, respectively. 5 FREE AND RESTRAINED AUTOGENOUS DEFORMATION Fig. 5.1 summarizes the results of the development of free autogenous deformation over time obtained during the first week following the final set of UHPC mixes. Each curve in the graph is the average obtained from all test series performed on the particular mix. The maximum difference in the corresponding values between two series of one composition was approximately 1 µm/m, depending on the cement batch and the choice of the final setting time. All the mixtures containing SAP demonstrated pronounced reductions in deformation due to shrinkage. These reductions were particularly pronounced at a very early age, indeed in the first few hours after the final set. The addition of.3% SAP by mass of cement plus extra water (mix F-S.3.4) resulted in a decrease in autogenous shrinkage from approximately 1 μm/m (F-R) to approximately 3 μm/m in the first day of measurement. Thereafter the increase in volumetric changes of both mixes, F-R and F-S.3.4, tended to be alike. On further increasing the addition of IC agent, with a simultaneous increase in extra water, IC led to a more efficient reduction of deformations due to autogenous shrinkage and caused changes in their development over time. An extreme example is provided by the mix F- S1..16 with the maximum amounts of SAP and extra water (1.% SAP, extra water corresponding to an increase in w/c by.16). In this case, following the greatest reduction in the first day among all compositions investigated, only minor changes in autogenous shrinkage were observed with increasing age. Obviously, continuous internal curing could be achieved here. Interestingly enough, similar behaviour was observed also for the mix F-S.4
136 DUDZIAK, MECHTCHERINE: Enhancing Early-Age Cracking Resistance by Internal Curing Using SAP made with SAP addition but without extra water. An explanation for this phenomenon has still to be found. With respect to the result obtained for the mix F-R.7-1, the addition of water with a simultaneous reduction of superplasticizer, but without SAP, caused only a slight reduction in autogenous shrinkage (approximately 2 µm/m). In fact the efficiency of water addition without using SAP is about 4 times smaller than in the case when SAP and the equivalent amount of extra water were used. autogenous strain [µm/m] -2-4 -6-8 -1-12 F-S.6.8 F-S.4.7 F-S1..16 F-S.4 F-S.3.4 F-R.7-1 F-R 1 2 3 4 5 6 7 tensile stress [MPa] 5 4 3 2 1-1 -2 F-R F-S.4.7 F-S.6.8 1 2 3 4 5 6 7 time [d] time [d] Fig. 5.1. Evolution over time of free autogenous shrinkage after final set. Average curves from all investigations. Fig. 5.2. Residual tensile stresses derived from instrumented ring tests. Individual measurements. In good agreement with previous results, Fig. 5.2 demonstrates a substantial reduction in tensile stresses under restraint of autogenous shrinkage for UHPC mixtures with internal curing. The results obtained from the instrumented ring tests for SAP-enriched mixes containing extra water (F-S.4.7, F-S.6.8) are shown in comparison to those measured for the reference mix F-R up to a concrete age of 7 days. Considerable tensile stresses of approximately 3 to 4 MPa were recorded for the reference mix F-R at a concrete age of 24 hours, while stresses rose slightly over 1 MPa only in the mix F-S.4.7 and were much lower for the mix F-S.6.8, which had higher amount of SAP and extra water. Thus, the positive effect of internal curing could be observed at very early ages. In subsequent measurements, the mitigation of autogenous shrinkage using IC exhibited a continued trend towards the dramatic reduction in stresses caused by restraint. At a concrete age of 7 days the stresses induced in the specimens made of the mixes F-S.4.7 and F-S.6.8 did not show any tension while the corresponding tensile stresses in the reference concrete F-R were already in the range of 4 to 5 MPa at the same age. This demonstrates a clear, considerable reduction in the cracking potential of high-strength cement-based materials as a result of internal curing. Furthermore, as was reported in a previous publication by the authors [Dud1], a certain degree of internal curing remains beneficial for high-performance concrete up to ages of 1 days and probably even longer.
International RILEM Conference on Material Science MATSCI, Aachen 21 Vol. III, AdIPoC 137 6 DEVELOPMENT OF MECHANICAL PROPERTIES Table 6.1 gives the values of total porosity arrayed against the flexural and compressive strengths obtained. For those mixtures studied repeatedly the standard deviation is reported. The total porosity was calculated by comparing the values of true density, as obtained from helium pycnometer measurements, with the corresponding values of bulk density. Table 6.1. Porosity and flexural and compressive strengths for the UHPC mixes investigated. Standard deviations are given in parentheses. Mixture Porosity P [%] Flexural strength f cf and compressive strength f cm [MPa] P 1d P 3d P 7d f cf, 1d f cf, 3d f cf, 7d f cm, 1d f cm, 3d f cm, 7d 1 F-R 17.1 (.5) F-S.4 19.8 (1.8) F-R.4 19.5 F-S.3.4 21.4 (1.4) F-R.7-1 19. F-R.7-2 21.4 F-S.4.7 22.1 (1.1) F-S.6.8 23.3 F-S.1..16 28.4 14.8 (.4) 17.3 (1.) 15.5 17.3 (.8) 16.6 15. 18. (.3) 18.8 23.2 14.5 (3.2) 14.4 (1.4) 14.3 15.6 (.7) 14.1 14.5 17.3 (.1) 17.6 22. 1.4 (1.3) 9.8 (1.2) 9.7 (1.4) 6.5 (1.6) 11.1 (.7) 8.4 (.4) 6. (.9) 8.1 (.4) 5.4 (.3) 15.7 (1.5) 13.4 (.8) 15.2 (1.2) 12.5 (1.4) 13.8 (1.6) 13.2 (.9) 12.2 (.4) 11.7 (.8) 9.9 (.4) 16.5 1 (1.5) 14. (1.5) 14.7 1 (1.7) 14.8 1 (1.4) 15.8 (.2) 13. (.5) 13.9 (.8) 12.2 (.3) 11.2 (.2) 57 (1) corresponding flexural strength values for mixes F-R, F-R.4 and F-S.3.4 produced with HIMvac were: 17.8, 14.3 and 11.4 MPa, respectively. A clear tendency towards higher porosity and, as a consequence, lower flexural and compressive strength at early ages is apparent for those compositions containing SAP and extra water. The only exceptional case is the mix F-S.4 (.4% SAP, no extra water), which showed minor improvement in compressive strength at an age of 1 day (65 MPa) in comparison to the reference mixture F-R (57 MPa). Interestingly the lower air content of the mixtures with extra water but without the addition of SAP did not counterbalance completely the effect of a higher w/c in these mixes: the mechanical properties of F-R.4 and F-R.7-2 were in general affected negatively when compared to the reference mix F-R and slightly advantageously when compared to SAP-enriched mixes with equivalent w/c-ratios (F-S.3.4 and F-S.4.7, respectively). An exception to this trend was mix F-R.7-1, which, probably as a result of the reduction in superplasticizer, showed mechanical performance comparable to the reference mix F-R. With regard to the mechanical properties of the mixes produced with the vacuum mixer (HIMvac, see the corresponding values below Table 6.1), a reduction of air content in all mixes irrespective of the addition of SAP or extra water did not lead to new findings. Possible contributions to changes in mechanical performance have been partially discussed already in previous publications by the authors, e.g. [Dud1]. There it was suggested that some reduction in the strength of SAP-enriched mixes is mainly a result of the formation of 65 (8) 58 (6) 33 (9) 59 (1) 4 (1) 31 (1) 36 21 (1) 16 (5) 14 (5) 11 (6) 95 (6) 99 1 87 (3) 69 66 (1) 122 (7) 112 (5) 118 (6) 111 (6) 119 18 (3) 94 (4) 86 (5) 75
138 DUDZIAK, MECHTCHERINE: Enhancing Early-Age Cracking Resistance by Internal Curing Using SAP entraining pores, initially filled with curing water and subsequently dried out. However, at the standard testing age of 28 days, the strength values of material both with and without IC for some combinations of SAP and extra water were found to be nearly alike (e.g., compressive strength measured on prisms for F-R and F-S.3.4 according to [Dud1] was 141 MPa and 137 MPa, respectively). With regard to the results obtained for early concrete ages (Table 6.1), an additional explanation can be given. Firstly, as was shown in Fig. 4.2, internal curing retards somewhat the setting behaviour of cement paste. Indeed the paste with SAP and some extra water attained its final set approximately 2 hours after the reference paste. In Fig. 6.1 this phenomenon can be recognised in the shift in the development of the ultrasonic velocity curve over time toward later ages. Following the same tendency, a higher dosage of SAP and extra water added to UHPC forced an even more pronounced postponing of the final setting time (approximately 4 hours of difference between the mixes F-R and F-S.1..16): Fig. 6.2 shows a more pronounced shift in ultrasonic velocity over time towards later ages. This leads to a delay in the development of concrete strength. 4 ultrasonic velocity [m/s] 35 3 25 2 15 1 5 P-R P-S.6.87 signal velocity [m/s] F-R F-S.4.7 F-S.1..16 F-S.4 4 8 12 16 2 24 45 4 35 3 25 2 15 1 5 4 8 12 16 2 24 time [h] time [h] Fig. 6.1. Time evolution of ultrasonic velocity in cement paste without and with IC Fig. 6.2. Time evolution of ultrasonic velocity in UHPC without and with IC Secondly, concrete with a higher moisture content generally exhibits lower strength in comparison to the same concrete having a lower moisture content. In this context internal curing might have some negative effect on UHPC strength by providing additional water. This hypothesis is supported by reports of some researchers on the effect of SAP on mechanical performance having been reversed under two different curing conditions [Pai9]. 7 CONCLUSIONS It has been demonstrated that internal curing provided by addition of SAP can significantly reduce plastic and autogenous shrinkage. If sufficient amounts of SAP and extra water are used, the susceptibility to cracking can be minimised or even eliminated, which was shown in both free and restrained shrinkage tests. There is only minor improvement in autogenous shrinkage reduction on application of extra water with no SAP. Some extra water can
International RILEM Conference on Material Science MATSCI, Aachen 21 Vol. III, AdIPoC 139 preserve the workability of concretes with SAP addition without any loss of homogeneity. The supply of extra curing water carried by SAP causes some retardation in the setting behaviour of high-strength cement-based composites and leads to some increase in the porosity as well as some reduction in early age strength in comparison to the reference mixes. However, the decrease in strength is compensated largely at an age of 28 days. ACKNOWLEGEMENTS The investigations were conducted as part of the German Research Foundation (DFG) Priority Programme SPP 1182 on sustainable construction with UHPC. The authors would like to thank the DFG for its financial support. Furthermore, the technical support of Prof. Heinz and Mr. Gerlicher (cbm, TU Munich / tests with vacuum mixer) as well as the group of Prof. Slowik (Leipzig University of Applied Sciences / plastic shrinkage tests) is gratefully acknowledged. REFERENCES [Dud1] [Feh4] [Hen1] [Hos4] [Jen95] [Kov7] [Pai9] [Slo8] Dudziak, L., Mechtcherine, V., Reducing the cracking potential of Ultra-High- Performance Concrete by using Super Absorbent Polymers (SAP). In: Proceedings of the International Conference on Advanced Concrete Materials, 17-19 November 29, Stellenbosch, South Africa, Gideon P.A.G. Van Zijl and Billy P. Boshoff eds., CRC Press- Taylor & Francis Group, The Netherlands, p. 11-19, 21. Fehling, E., et. al., Entwicklung, Dauerfestigkeit und Berechnung Ultra-Hochfester Betone (UHPC), DFG Forschungsbericht (Kennwort: FE 497/1-1), 24. Henkensiefken, R., et. al., Plastic Shrinkage Cracking in Internally Cured Mixtures Made with Pre-wetted Lightweight Aggregate, Concr Int 32, p. 49-54, 21. Hossain, A. B., Weiss, W. J., Assessing residual stress development and stress relaxation in restrained concrete ring specimens, Cem Concr Comp 26, p. 531-54, 24. Jensen, O.M., Hansen, P.F., A dilatometer for measuring autogenous deformation in hardening Portland cement paste, Mat Struct 28 (181), p. 46-49, 1995. Kovler, K., Jensen, O.M., (eds.), Internal Curing of Concrete, State-of-the-Art Report of RILEM Technical Committee 196-ICC, RILEM Report 41, Bagneux, France, 27. Paiva, H., Esteves, L.P., et al., Rheology and hardened properties of single-coat render mortars with different types of retaining agents, Constr Build Mater 23, p. 1141-1146, 29. Slowik, V., Schmidt, M., Fritzsch, R., Capillary pressure in fresh cement-based materials and identification of the air entry value, Cem Concr Compos 3 (7), p. 557-565, 28.