Rheological Properties of Self-Consolidating Concrete Stabilized With Fillers or Admixtures

Transcription

1 Rheological Properties of Self-Consolidating Concrete Stabilized With Fillers or Admixtures H. Vikan 1, K. De Weerdt 1, and S. Smeplass 2 1 SINTEF Building and Infrastructure, Trondheim, Norway 2 SKANSKA, Norway ABSTRACT: The mix design of self-consolidating concrete (SCC) is crucial in order to obtain a robust SCC that is fluid enough to completely fill the form and simultaneously stable against segregation of aggregates and bleeding. This study investigates the effect of chemical stabilizers or additional filler on the castability and rheological properties of concrete and its matrix. Viscosity and thus thixotropy is of special interest since these parameters are believed to influence stability, form-filling ability, the migration and evacuation of entrapped air bubbles and thus the final surface quality of hardened concrete elements. The study includes three different types of filler: limestone powder, fines from crushed sand, and fines from sand which is both crushed and washed. Two types of chemical stabilizers and two types of superplasticizers with different stabilizing/splitting properties have, moreover, been used. Rheological properties have been measured both on concrete and concrete equivalent matrix. 1 BACKGROUND 1.1 Importance of robustness and stability for SCC Self-consolidating concrete (SCC) has been described as one of the most innovative developments in the field of concrete technology (De Schutter et al. (2008)). Improved placing of the concrete and and better work environment are two of the most important benefits of using SCC compared with traditional vibrated concrete. Unfortunately, SCC cast in-situ in Norway has stagnated at a low market share. Main reasons are probably low robustness against fluctuations of the concrete production, higher demand for quality control and that SCC over time has been too highly priced (Vikan (2008)). Robustness can de defined as the ability of fresh concrete to maintain its properties within narrow limits when the proportions of constituent materials change significantly (De Schutter et al. (2008)). Stability of fresh SCC defines the ability of a concrete mixture to retain its homogeneity through the fresh phase, both at rest and subject to loads due to transport, formfilling and compaction (Daczko (2002)). Dynamic stability refers to the characteristic of the concrete mixture to resist segregation during production, transport and casting. Static stability refers to the ability of the concrete mixture to resist bleeding, segregation and settlement once all placement and casting operations have been completed (Smeplass (2009)). 1.2 Principal objectives and scope The importance of mix design is crucial in order to obtain self-consolidating concrete (SCC) that is fluid enough to completely fill the form and simultaneously stable against segregation of

2 aggregates and bleeding. Stability of SCC can be achieved in at least three ways: By aid of fines, by aid of chemical stabilizers or a mixture of both fines and stabilizer (Takada et al. (1998), Kim et al. (1996), Corradi et al. (2003)). This study has been made with a practical approach based on scenarios that could occur on an actual building site for ensuring consistent stability and castability of fresh concrete, i.e. addition of fines or chemical stabilizers. Rheological properties of concretes and matrices stabilized by either of these methods are studied, bearing in mind that concretes stabilized with fines generally have a higher matrix content than concretes stabilized with chemical admixtures. Special attention is given to viscosity and thixotropy since these parameters are believed to influence stability, form-filling ability, the migration and evacuation of entrapped air bubbles and thus the final surface quality of hardened concrete elements. 2 EXPERIMENTAL 2.1 Materials EN CEM II/A-V 42.5 R Portland fly ash cement was used for all experiments. The cement had a Blaine fineness of 450 m 2 /kg and density of 3010 kg/m 3 Gneiss/Granite aggregates of fractions 0/8mm and 8/16 mm were used for all concretes. Filler (< µm) sieved from the 0/8 mm sand was used to prepare concrete equivalent matrix. Three powdered materials have been used as stabilizers: Sedigraph curves are given in Figure 1: Limestone powder (LS) of density 2700 kg/m 3 and Blaine 360 m 2 /kg Filler sieved from non-washed, crushed 0/8mm sand (C08). Density: 2730 kg/m 3. Filler produced from the same material as crushed 0/8 mm sand. The 0/8 mm sand was sieved to obtain the 0/2 mm fraction, washed and sieved once more to obtain the filler (W02). Density: 2730 kg/m 3. Figure 1: Sedigraph curves for powdered materials used in the study as stabilizers (max. sieve size 60μm). Two chemical stabilizers have been used: S1 is based on a polymer with high molecular weight. Normal is % of cement weight. Selected was 0.4%. S2 is based on cellulose derivate. Normal is 1-2% of cement weight. Selected was 1% Two types of acrylic superplasticizer were used. SP2 has lower molecular weight, longer side chains and higher charge density than SP1. These properties result in higher degree of sterical hindrance and rapid slump loss. A Gluconat based set retarder was added to all concretes at a of 0.4% in order to eliminate the effect of hydration on the rheological measurements.

3 2.2 Recipes Concrete The basis of the test matrix was a low-grade concrete with an aimed slump flow of 675 ± 15 mm. The reference concrete was designed in order to be on the verge of separation (i.e. instability). The mineral fillers were added in two s, namely 40 kg/m 3 (filler-cement ratio 0.12) and 80 kg/m 3 (filler-cement ratio 0.24) while adjusting the superplasticizer in order to keep the slump flow within a slump flow range of 675 ± 15 mm. The chemical stabilizers were added according to recommended given by the producer while adjusting the superplasticizer. The concrete recipes are given in Table 1. Table 1. Concrete mix design Reference 40 kg/m 3 Filler 80 kg/m 3 Filler w/c w/p f/c (%) Matrix (l/m 3 ) Paste (l/m 3 ) Cement (kg/m 3 ) /8 mm (kg/m 3 ) /16 mm (kg/m 3 ) A forced pan mixer with a volume of 50 litres from Eirich was used to prepare the concretes. The volume of the concrete batches was 40 litres. Slump flow and T 500 were measured according to EN : minutes after water addition. Simultaneously, the torque (T) was measured as a function of the rotational speed (f 0 ) by aid of a ConTec Rheometer-4SCC. The parameters G (A) and H (A s) were determined using linear regression according to the following equation: T = H f 0 + G. Due to the complicated geometry of the ConTec Rheometer-4SCC, there are currently no equations available to convert of G and H into other rheological parameters such as τ y and μ p Matrix The water-cement ratio was 0.45 for all mixes. All pastes were added 0.4% gluconate per cement weight. Total paste volume was 200 ml. The mixes were designed as concrete equivalent (abbreviated c.e.) matrices. Concrete equivalent superplasticizer s were used for the main test series. In order to eliminate the effect of variable superplasticizer, additional test series were made for which it was kept constant. The matrices were blended in a high shear mixer from Braun (MR5550CA). Rheological parameters of the matrix were recorded by a parallel plate (1 mm gap, upper and lower plate serrated) rheometer MCR 300 from Physica. The rheometer temperature was set to 20 o C. The following measurement sequence started 10 minutes after water addition: 1 minute pre-shearing with constant shear rate (γ ) of 60 s -1 1 minute rest without shearing Flow curve (hysteresis): Stress (τ) shear rate (γ ) curve with linear sweep of γ from 1 up to 100 s -1 in 30 points lasting 6 seconds each. Thereafter linear sweep of γ from 100 down to 1 s -1 in 30 points lasting 6 seconds each Thixotropy: o γ = 0.1 in 10 measuring points each lasting 12 seconds (level 1) o γ = 250 in 5 measuring points each lasting 6 seconds (level 2)

4 o γ = 0.1 in 50 measuring points each lasting 3.6 seconds (level 3) 10 seconds rest Shear rate (γ ) - stress (τ) curve with logarithmic sweep of τ from Pa in 28 points each lasting 5 seconds in order to measure the gel strength (data not discussed) 3 RESULTS 3.1 Correlations between concrete and matrix All viscosity parameters measured for concrete and concrete equivalent matrix, namely H, T 500 and plastic viscosity (µ) were interrelated as exemplified by Figure 1. Trends found in matrix can thus be expected to occur also in concrete and vice versa. Figure 1. Correlation of H measured on concrete and plastic viscosity (µ) measured on equivalent matrix. There were only vague correlations between yield stress parameters measured on matrix and concrete. This result can be explained by yield stress related parameters varying in a narrow range since all concrete were designed to obtain a slump flow within ± 15 mm. 3.2 Concretes and matrices plasticized with SP1 Rheological properties of concretes and matrices plasticized with SP1 are given in Table 2 and 3 respectively. G decreased with addition of either of the two chemical stabilizers, probably due to the increased superplasticizer s that were added in order to obtain constant slump flow. The matrix yield stress increased, however, in accordance with other reports (Khayat et al. (2010)). G and matrix yield stress decreased with addition of fillers. This effect may be caused by increased matrix content and dilution /dispersion of cement particles. Negative yield stress is an artefact of the Bingham model caused by curvature (shear thickening) of the flow curve. T 500, H and matrix plastic viscosity increased with the addition of chemical stabilizer or filler. The effect of the lowest filler addition was in the same range as found for the chemical stabilizers. The viscosity increased with increased filler due to increased particle concentration and adsorption of free water in the mix (Krieger-Dougherty (1959)). Both chemical stabilizers resulted in increased thixotropy values compared to the reference mix. The stabilizer of polymer type, S1, produced the strongest effect of all stabilizers tested within the study. The thixotropic effect of the chemical stabilizers increased with decreasing superplasticizer. Increased dispersion is linked to decreased structural buildup as illustrated by comparing matrices with concrete equivalent superplasticizer s with the 0.8% series. In the case wherefiller is used as stabilizer with concrete equivalent superplasticizer s, maximum thixotropy value was found for the lowest filler. These matrices had, however,

5 also the lowest superplasticizer s within the series. The lowest filler rendered lower thixotropic values than the reference when constant superplasticizer was applied. This effect may be caused by filler dispersing the paste. Increased filler will thereafter absorb more of the free water that again can be the cause of increased thixotropy. Table 2. Rheological properties of concrete plasticized with SP1 SP (%) (kg/m 3, %) Slump flow (mm) T 500 (sec) G (A) H (A s) S S LS C W LS C W Table 3. Rheological properties of matrices with concrete equivalent SP1 Stabilize r type (kg/m 3, %) Concrete equivalent superplasticizer s SP Yield Plastic Thix. stress viscosity value (%) (Pa) (Pa s) (Pa s) 0.8% superplasticizer Yield stress (Pa) Plastic viscosity (Pa s) Thix. value (Pa s) S S LS C W LS C W Concretes and matrices plasticized with SP2 The rheological parameters measured on concrete and matrix are given Table 4 and 5 respectively. No clear effect of chemical stabilizers on G could be found. Matrix yield stress increased by addition of S1 (polymer type) or S2 (cellulose type). The polymer type had the strongest effect on the yield stress. The effect which depended upon dispersion of the matrix (i.e.superplasticizer ). Addition of filler resulted in decreased G while no clear trends were found for the matrix yield stress. A slight increase of T 500, H and matrix viscosity was generally measured by S1 addition while no effect was found for S2. Filler addition resulted in stronger viscosity increase than addition of chemical stabilizers. The increase related to the filler fineness; limestone producing the highest values and filler from crushed 08 mm (C08) producing the lowest. Increased viscosity by increased volume fraction of solids is in line with the Krieger-Dougherty equation (1959).

6 Table 4. Fresh properties of concrete plasticized with SP2. SP (%) (kg/m 3, %) Slump flow (mm) T 500 (sec) G (A) H (A s) S S LS C W The thixotropy value increased by filler addition and increased filler fineness. As for matrices with SP1 the degree of structural buildup decreased with increasing superplasticizer (i.e. dispersion). The highest filler had a much stronger effect on thixotropy for these mixes contrary to the findings made for matrices with S1. The highest filler addition had thus a stronger effect on viscosity and thixotropy than the chemical stabilizers. Table 5. Rheological parameters matrices with SP2 Concrete equivalent 0.5% superplasticizer SP Yield Viscosity Thix. Yield Viscosity Thix. stress Value stress value (kg/m 3, %) [%] [Pa] [Pa s] [Pa s] [Pa] [Pa s] [Pa s] S S LS C W LS C W CONCLUSIONS The understanding of structural build-up is of importance not only for stability and robustness of fresh concrete, but also for the development of formwork pressure and aesthetic surface quality. The aim of this investigation was to compare two ways to stabilize concrete: by adding additional filler or chemical stabilizers. It was shown that the effect of the stabilizers depended on the plasticizer type and the : Increased dispersion is linked to decreased structural buildup. Of the two chemical stabilizers, the polymer type had stronger thixotropic effect than the cellulose type. However, for both superplasticizers, addition of filler gave a stronger viscosity increase than the chemical stabilizers. Matrix viscosity increased with increasing volume fraction of solids in line with the Krieger- Dougherty equation. Viscosity and thixotropy increase of the matrix dispersed with SP2 related to the filler fineness; limestone producing the highest values and filler from crushed 08 mm (C08) producing the lowest.

7 Futher research will be conducted on the effect of the combination of both additional filler and chemical stabilizers on the stability of both matrix and concrete. The final aim of the project is to develop robust recipees for SCC with materials available in Norway, and additionally to link the rheological properties with quality of the finished concrete surface. 5 ACKNOWLEDGEMENTS The paper is based on the work performed within COIN - Concrete Innovation Centre ( - which is a Centre for Research based Innovation, initiated by the Research Council of Norway (RCN) in COIN has an annual budget of NOK 25 mill, and is financed by RCN (approx. 40 %), industrial partners (approx 45 % of which ¼ is cash) and by SINTEF and NTNU (in all approx. 15 %). The Centre is directed by SINTEF, with NTNU as a research partner and with the present industrial partners: Aker Solutions, Norcem, Norwegian Public Roads Administration, Rescon Mapei, Skanska, Spenncon, Unicon, Veidekke and Weber Saint-Gobain. 6 REFERENCES Corradi, M., Khurana, R., Magarotto, R. (2003) User friendly self-compacting concrete in precast production, The Third International RILEM Symposium on Self-Compacting Concrete, Reykjavik: 457 Daczko, J.A. (2002) Stability of Self-Consolidating Concrete, Assumed or Ensured?, First North American Conference on the Design and Use of Self-consolidating Concrete De Schutter G.M., Bartos P.J., Domone P., Gibbs J.(2008) Self-Compacting Concrete, Whittles Publishing, Scotland, UK Khayat K.H., Hwang S.-D., Belaid K. (2010) Performance of Cast-in-Place Self-Consolidating Concrete Made with Various Types of Viscosity-Enhancing Admixtures, ACI Materials Journal 107 (4): Kim, J.K., Han, S.H., Park, Y.D., Noh, J.H., Park, C.L. (1996) Experimental research on the Material Properties of Super Flowing Concrete, Proceedings of the International RILEM Conference o Production and Workability Methods of Concrete, Paisley, June 3-5: Krieger I.M. and Doughert T.J. (1959) A Mechanism for Non-Newtonian Flow in Suspensions of Rigid Spheres, Trans. Soc. Rheol Smeplass S. (2009) Chapter 3 - Fresh concrete workability, TKT 4215 Concrete Technology 1, NTNU Takada, K., Pelova, G.I., Walraven, J.C. (1998) Self-Compacting Concrete Produced by Japanese Method with Dutch Materials, Proceedings of the 12 th European Ready Mixed Concrete Congress, June, Lisbon (2): Vikan H. (2008) Means of improving concrete construction productivity State of the art, COIN Project report 8, ISSN (online), ISBN (pdf)