Metakaolin concrete at a low water to binder ratio

https://doi.org/10.1016/j.conbuildmat.2007.06.003Get rights and content

Abstract

This paper investigates the performance of concrete containing metakaolin (MK) at a low water to binder ratio of 0.3. Portland cement (PC) was partially replaced with 0–20% MK. Testing included, compressive strength, ultrasonic pulse velocity (V), dynamic modulus of elasticity (Ed) and length change. Specimens were either cured in water or in air at 20 °C. The results indicate that the performance of MK concrete at low water to binder ratio is not different from that at higher water to binder ratios reported in a previous investigation. The maximum contribution of MK to strength occurs at 14 days of curing in that the relative strength of MK concrete shows a maximum value at that curing time as found in a previous investigation. The optimum replacement level of cement with MK is about 15%. Linear relationship exists between V and Ed for air cured and water cured specimens. A systematic increase in MK content of up to at least 20% leads to a decrease in shrinkage and an increase in expansion after 56 days of curing. Correlation between the various properties is also conducted.

Introduction

In concrete, properties including the microstructure of the cement paste at the immediate vicinity of the aggregate particles, normally known as the interfacial transition zone (ITZ), differ from those of the bulk of the cement paste. It is generally accepted that the ITZ is considered a zone of weakness in that it possesses higher porosity than that of the bulk of the paste [1], partly because of the increased water to cement ratio at the interface. There are various factors influencing the properties of the ITZ. These include the water to cement ratio and aggregate to cement ratio. The thickness of the ITZ increases with the increase in water to cement ratio [2]. In order to modify the performance of the ITZ, attempts were made to replace the cement with mineral admixtures such as fly ash and slag [3]. The partial replacement of cement with ultra-fine pozzolan such as silica fume or metakalolin plays an important role in densifying the ITZ because of the microfiller effect due to the relatively fine particles of these pozzolan [4], [5].

Due to its high pozzolanic activity, the inclusion of MK greatly improves the mechanical and durability properties of concrete. Wild et al. [6] found that the relative strength of concrete containing MK increases with the increase in curing time up to 14 days before it begins to decline. Bredy et al. [7] concluded that, when the MK content is below 20% (by mass of cement) the total porosity of paste decreases. Beyond 30%, an increase in porosity was found. The pore volume of mortar and the threshold diameter was found to decrease in the presence of MK [5]. Others [8] reported a slight increase in pore volume for pastes containing up to 15% MK as partial cement replacement and this increase is dependent upon the MK content. The presence of MK, however, causes refinement in pore structure, in that pastes containing MK decrease the threshold diameter and increase the percentage of small pores [8]. The presence of MK causes reduction in the calcium hydroxide (CH), especially during the early periods of hydration [9], [10], [11]. Sulphate resistance of mortars was found to increase with the increase in MK content [12].

The author has published a number of paper on the use of MK in concrete at water to binder ratios of 0.45 and 0.55 [6], [8], [11], [12], [13]. As mentioned earlier that the transition zone is affected by the water to cement ratio. Therefore, the present work investigates the performance of concrete containing varying amount of MK at a low water to binder (PC + MK) ratio of 0.30 in order to examine whether the low ratio affects the properties of MK concrete as compared with higher water to binder ratios. The concrete properties investigated comprised workability, compressive strength, dynamic modulus of elasticity, ultrasonic pulse velocity and length change.

Section snippets

Materials

The constituents of mixes were Portland cement (PC), metakaolin (MK), water, fine aggregate and coarse aggregate. The fine aggregate used complied with class M of BS 882:1992, and 20 mm nominal size crushed and washed was used as coarse aggregate. The superplasticiser (SP) used was naphthalene sulphonate based liquid with 60% water and a density of 1.19 g/cm3. The composition of PC and MK is given in Table 1.

Mix proportions

Six different mixes (mixes M1–M6) were employed to examine the influence of low water to

Workability

All mixes including the control (i.e. 0% MK) exhibited very low workability. Slump values were below 20 mm. Also Vebe times and compaction factors were more than 24 s and less than 0.8 respectively. The workability results are presented in Table 3. The presence of MK reduces the workability as reported elsewhere [6]. Special care was taken to compact the specimens well in their moulds and this is further supported by visual examination. Despite of low workability in concrete containing MK, full

Conclusions

  • 1.

    At a low water to binder ratio of 0.3, the optimum replacement level to give maximum strength enhancement is 15% MK. This optimum level is lower than that obtained at a higher water to binder ratio of 0.45.

  • 2.

    The maximum contribution of MK to strength occurs at 14 days of curing, which is similar to results obtained at higher water to binder ratio. As suggested previously that this is the result of an elementary change between the MK and calcium hydroxide reaction beyond 14 days.

  • 3.

    A systematic

Acknowledgements

The authors thank Mr. Plant for conducting the experimental programme and the concrete laboratory technical staff Mr. Skelton and Mr. Harwood for their assistance.

References (17)

There are more references available in the full text version of this article.

Cited by (96)

  • Calcined clay as hydraulic binder substitution

    2023, Advance Upcycling of By-products in Binder and Binder-Based Materials
View all citing articles on Scopus
View full text