Correlating dynamic segregation of self-consolidating concrete to the slump-flow test

https://doi.org/10.1016/j.conbuildmat.2011.08.052Get rights and content

Abstract

The three key characteristics of self-consolidating concrete are flowability, segregation resistance and passing ability. Quality control of flowability is typically predicted by the final diameter (DF) of the slump-flow test. In this study, the time required to reach the final diameter (TF) of the slump-flow test is correlated to dynamic segregation for mixes with a constant DF and aggregate-to-binder ratio. Segregation was determined by measuring the radial aggregate distribution from the slump-flow test. It was demonstrated that increasing the TF improved dynamic segregation resistance. It was also found that the TF was more indicative of viscosity than the time to reach a diameter of 50 cm (T50).

Highlights

► We correlate results from the slump-flow test including final diameter, time to 50 cm and time to final diameter to dynamic segregation. ► Segregation was determined by measuring the radial aggregate distribution. ► At constant aggregate content and final diameter, increasing time to final diameter improved segregation resistance. ► Time to final diameter was more indicative of viscosity than time to 50 cm.

Introduction

The three key characteristics of self-consolidating concrete (SCC) are flowability, segregation resistance and passing ability. Quality control of flowability is typically predicted by the final diameter of a slump-flow test; a larger diameter indicates higher flowability. Segregation resistance concerns the ability to retain homogenous distribution of aggregates; segregation can occur both during and after casting. The ability to keep the homogeneity of the aggregate distribution is governed by the volume fraction, distribution and physical properties of the aggregates, as well as the rheological properties of the suspending matrix [1], [2], [3], which can be considered the cement paste or mortar for concrete systems [4]. Cementitious materials such as cement paste or mortar are generally regarded as yield stress fluids which can be characterized by the Bingham, Herschel–Bulkley, Casson or other rheological models [5], [6]. For concrete at rest, static segregation occurs when the yield stress of the suspending matrix is insufficient to support the weight of the aggregate minus its buoyancy [7], [8], [9]. If an aggregate particle settles within the concrete, the stress in the fluid along the surface of the particle exceeds the yield stress (although Beris et al. [7] clearly demonstrate that certain regions above and below the particle remain rigid due to stagnation points). As the distance from the surface of the particle increases, the stress in the fluid decreases to zero. As such, at a certain point, the stress in the fluid will no longer exceed the yield stress, and the fluid will remain rigid [7]. If the yield stress is not exceeded anywhere along the particle surface, the particle will remain suspended. Stability criteria have been developed which depend on gravity, yield stress of the suspending fluid, particle size, difference in density between the suspending matrix and particle and a particular constant (see for example, Roussel [8]). This constant depends on the size of the yielded region around the particle and has been determined theoretically, experimentally and numerically [8]. In order to avoid static segregation, larger particle sizes require higher yield stresses of the suspending fluid. The viscosity of the fluid can also help mitigate segregation by decreasing the speed at which the aggregates settle before the concrete begins to set [3]. However, for typical SCC mixes, the inherently low viscosity plays a small role compared to the yield stress in mitigating static segregation until concrete sets (see the example given in [8]).

Still, for dynamic segregation, where the fluid is in motion, the viscosity may have an important role. During motion, the fluid structure breaks down which may allow aggregates to settle if the yield stress is reduced sufficiently [10]. Thus, dynamic segregation is due to the movement of the fluid. Higher viscosities will help drive aggregates along with the flow and also reduce the rate of settlement until the concrete comes to rest. At this point, the fluid structure can rebuild, restoring the yield stress and preventing further static segregation.

Several simple field tests have been developed that take advantage of the relationship between segregation and rheology, such as the falling ball viscometer [11], the penetration probe [12], the rapid segregation test [13], the flow trough test [14] and the V-funnel, L-Box, U-Box and J-ring tests [15], [16]. The first three tests determine the depth at which an object penetrates concrete at rest, thus predicting static segregation. The last four do not predict segregation directly, but have been correlated to both static and dynamic segregation resistance by considering the rheological properties of the concrete ascertained by these tests. Only the flow trough test measures dynamic segregation directly using a channel device to compare concrete that has flowed through the channel (undergoing potential segregation) to concrete not subjected to flow [14]. This test requires wet sieving on site.

For the slump-flow test, it has been shown that the final diameter can be calculated based on the yield stress of the material [17]. Furthermore, for the minislump-flow test, it has been demonstrated for cement pastes that the time it takes for the flow to reach the final diameter is related to the ratio between the viscosity and yield stress [18] as shown in Fig. 1.

In the present study, this relationship between time to final spread and viscosity is extended to concrete so that a relation between TF and segregation can be highlighted for SCC. This was demonstrated in a preliminary study by Tregger et al. [19] Two sets of data were obtained for flow diameters of 65 cm and 70 cm, both with a constant water-to-binder ratio (w/b) and aggregate-to-binder ratio (agg/b). Within each set, the viscosity of the matrix was systematically varied with the use of a viscosity-modifying agent (VMA), while the final diameter (and thus yield stress) was kept constant by adjusting the high-range water-reducing (HRWR) dosage. From the slump-flow test, the final diameter, time to 50 cm, time to final diameter and segregation resistance of SCC were recorded. The results show that information about dynamic segregation can be obtained from rheological properties inferred by the slump-flow test.

Section snippets

Materials, proportioning and mixing

In this study, an ASTM type I Portland cement was used in addition to pea gravel (maximum size 10 mm, 0.4 in), and river sand (maximum size 5 mm, 0.2 inches). A polycarboxylate-based HRWR was used along with a biopolymer-based VMA. Two sets of compositions were used: Set A was designed to have a final diameter of 65 cm (25.6 in) while Set B a final diameter of 70 cm (27.6 in). Within each set, mixes were designed with VMA additions of 0–5% by weight of binder while the HRWR was adjusted to keep the

Studies with cement paste

Compositions in each of the two sets were designed to have constant yield stress but increasing viscosity. Increased dosages of VMA were used to increase the viscosity, while HRWR additions were adjusted in order to keep the yield stress constant. In order to verify these trends, rheological and minislump-flow properties for the cement mixes are plotted against the addition of VMA in Fig. 6 and Fig. 7.

It can be seen that increases in VMA led to increases in both the viscosity and TF values. On

Conclusions

From this study, the slump-flow test has been shown to be capable of indicating dynamic segregation resistances in addition to flowability. This strengthens the slump-flow test as a more complete quality control test for SCC. It was found that slump-flows with higher TF values result in less dynamic segregation. Higher TF values were achieved by adding VMA, while at the same time keeping the flow diameter constant with adjustments in the HRWR dosage. This increase in TF is due to an increase in

Acknowledgments

The research presented in this paper was funded by the Infrastructure Technology Institute of Northwestern University, the Center for Advanced Cement-Based Materials and the National Science Foundation (Award CMS 0625606). Their financial support is gratefully acknowledged. The first author would also like to recognize the support of Lombardia regional council for research conducted at the Politecnico di Milano (code AZ#2). The third author wishes to acknowledge the financial support of the

References (23)

  • Song T, Chiew YM. Settling characteristics of sediments in moving Bingham fluid. J Hydraul Eng...
  • Cited by (0)

    1

    Present address: Department of Structural Engineering, Università degli Studi dell’Aquila, Nucleo Industriale di Bazzano Sud, 67100 Monticchio, Italy.

    2

    Present address: Department of Structural Engineering, Politecnico di Milano, Piazza Leonardo da Vinci, 20133 Milano, Italy.

    View full text