Bond Behaviour between Reinforcing Steel Bars and Fly ash-based Geopolymer Concrete

Abstract

Ordinary Portland cement (OPC) has been criticised for a long time for its high energy consumption during the production process that releases quantities of carbon dioxide into the atmosphere. For decades, researchers have been searching for substitutive binders for manufacturing concrete. Of all the proposed alternative binders, geopolymers stand out because of their total environmental friendliness, good mechanical capabilities and inexpensive processing costs. A geopolymer is an inorganic polymer with an internal chemical composition very similar to that of natural zeolite. A commonly used raw material for producing geopolymers is low-calcium (class F) fly ash, which is a major by-product of power generation industries. Class F fly ash-based geopolymers are usually synthesized by activating the fly ash with alkaline activators. The reaction happening in synthesis is geopolymerisation which transforms fly ash into cement-like binder material. Previous research on class F fly ash-based geopolymer concrete (GPC) has proved its great potential as an alternative to OPC concrete in terms of outstanding chemical and comparable engineering characteristics. However, the structural performance of the reinforced concrete (RC) components relies on there being a sufficient bond between the concrete and the reinforcing bars. Before being utilised in any concrete structures, GPC must demonstrate that it possesses adequate bond strength and predictable bond behaviour with commercial steel reinforcements. Although the importance of investigating the bond characteristics of GPC has been recently acknowledged, nevertheless the bond performance between the reinforcement sand GPC matrices has not been sufficiently studied. The definitions about the bond mechanism and bond-slip rules for GPC were still not clear, which was an inevitable obstacle to its application. As part of the research committed to investigate the suitability of GPC in RC structures, this study aims to not only qualitatively but also quantitatively define the bond behaviour of GPC and, determine whether it could qualify as an alternative to traditional concrete in RC structures. This study used experimental tests, statistical analyses and finite element (FE) modelling to investigate the bond behaviour of GPC. The experimental program included manufacturing and testing plain and ribbed bar reinforced GPC cylinders in direct pull-out tests and beams in beam pull-out tests. Meanwhile, corresponding tests were conducted on equivalent reinforced OPC concrete cylinders and beams, in order to be used as control groups in comparison of bond quality. The experimental data were planned and obtained so as to be suitable for statistical regression and hypothesis testing. The empirical equations obtained from regression well predicted the bond-slip behaviour of GPC and were used in FE modelling. In the bond tests, the types of reinforcements were 16mm plain and ribbed bars. The samples used for direct pull-out tests were cylinders 100mm in diameter, and for beam pull-out tests were beams220mm wide, 450mm deep and 600mm long. In addition, standard 100mm × 200mm cylinders were cast together with the bond test specimens for mechanical testing on characteristics related to bond, including compressive strength, indirect tensile strength and elastic modulus. The mechanical test results illustrated that GPC possessed comparable compressive and indirect tensile strength but significantly lower elastic modulus than OPC concrete. Data obtained from mechanical tests were used in regression. The statistical correlations among these three crucial mechanical characteristics of GPC were examined and described by statistical predictions. Equations for predicting GPC’s indirect tensile strength and elastic modulus from its compressive strength were presented. The bond tests results illustrated that the failure modes observed in both the reinforced GPC and OPC concrete specimens were similar. All the ribbed bar reinforced samples failed in splitting while all the plain bar reinforced ones failed in pull-out. However, statistical significant differences had been shown in the bond strength and the bond-slip curves between GPC and OPC concrete samples. Specifically, test results demonstrated that the plain bar reinforced GPC samples had considerably higher bond strengths than those of OPC concrete. This was attributed to the relatively homogenous steel-GPC interface observed using scanning electron microscopy (SEM).It was observed that, unlike OPC, no particularly weak layer existed in the contact area between the steel and geopolymer. Meanwhile, the ribbed bar reinforced GPC samples had significantly better energy absorption capabilities than the ribbed OPC concrete ones. This was also attributed to the homogeneity of steel-GPC interface, as well as the low elastic modulus of GPC. Strain gauges were installed along the bars to obtain changes in the steel strain during pull-out procedure. The distribution of bond stress within the bond length was calculated from the recorded values of steel strain. The bond stress distribution curves illustrated that the bond stress distributed non-uniformly along the bond length of plain and ribbed bar reinforced GPC, with the maximum stress occurring near the loaded-end. The uniform stress distribution assumption is not true for either OPC concrete or GPC. The bond-slip behaviour of GPC was successfully modelled using ANSYS with empirical equations and material properties obtained from the experiments. The modelling results successfully reflected the bond-slip and bond distribution curves obtained from experiments. With the bond-slip effects, the FE analysis on flexural behaviour of GPC beams obtained more realistic results than the perfect bonding model reported in the literature. To sum up, GPC possesses eligible engineering properties, remarkable bond strength and predictable bond-slip behaviour. Therefore, it could qualify as an appropriate replacement for OPC concrete in steel bar RC structures.