The effects of physical and chemical properties of fly ash on the manufacture of geopolymer foam concretes

The development of sustainable construction and building materials with reduced environmental footprint in both manufacturing and operational phases of the material lifecycle is attracting increased interest in the housing and construction industry worldwide. Recent innovations have led to the development of geopolymer foam concretes (GFCs), which combine the performance benefits and operational energy savings achievable through the use of lightweight foam concrete, with the cradle-to-gate emissions reductions obtained through the use of a geopolymer binder derived from fly ash. Fly ash is a by–product of coal fired power stations, and has become a highly promising source material for geopolymer manufacture. Compared to clays, another type of usually used materials, fly ash is probably more technologically suitable as it requires less alkaline activator while providing good workability. However, fly ash particles are substantially heterogeneous in physical and chemical properties. The composition and mineralogy of fly ash have marked effects on the properties of geopolymers, such as setting behaviour. This will affect the pore structure of GFC. Unfortunately, there is very limited specification regarding feedstock utilisation in geopolymer manufacture at present. Understanding the effect of fly ash physics and chemistry on the manufacture of GFC is not only necessary for the development of commercially mature GFC technology but also important for the geopolymer technology as a whole section. Five fly ash samples sourced from different power plants around Australia were used to manufacture geopolymer binders, enabling investigation of the relationship between the physical and chemical properties of fly ash and the mechanical properties of geopolymer products. The results showed that fly ashes from different sources exhibit substantially different physical properties. One important property is the inter-particle volume of fly ash, which largely determines the liquid requirement. The liquid requirement furthermore affects the porosity of hardened binders and their production costs. Another factor is the reaction extent of fly ash, which determines the quantity and composition of gel phases. A general trend obtained is that fly ash with higher network-modifying cations seems to possess higher reactivity. Research by Rietveld quantitative XRD and XRF analysis found that the composition and chemistry of glassy phases play an equally important role as the quantity of these phases in affecting the reactivity of fly ash. In glassy phases, both FeO4 and AlO4 tend to randomly distribute and connect with SiO4 tetrahedra by sharing corners and this is due to the alkali/alkali earth cations, which act as charge compensators. A reactivity index (RI) was proposed in this thesis to quantify the reactivity of fly ash under geopolymerization conditions. If pentacoordinated Fe cations are regarded as network modifiers, in addition to alkali and alkali earth cations, and by considering the contribution of specific surface area, it was found that the RI order of the studied five ashes matched well with their reactivity order. Alkaline dissolution analysis under different liquid/solid ratios supported the RI results. Additionally, dissolution analysis also showed that the crystals such as mullite and quartz were also partially dissolved, particularly in the ‘impure’ fly ashes, which had relatively higher concentration of network modifying cations. The above stages of the works were very useful to understand and to obtain a strong geopolymer binder by selecting a reactive fly ash. However, GFC manufacture in the laboratory conditions showed that a fly ash suitable for making high strength solid geopolymers was not necessarily suitable for GFC manufacture. It appeared that fly ash physical properties played a more important role than fly ash chemistry in affecting the engineering performances of GFCs. Those fly ashes with lower particle density and irregular particle shape appeared best suited for the manufacture of foam geopolymers. For a foamed paste derived from a specific fly ash, quick setting was a key property to achieve fine pore size and a homogeneous microstructure. The orthogonal array study conducted showed that slag addition was an effective method to control, and shorten the setting time of the foamed paste. The pore structure and porosity were also changed significantly and contributed to an increase in compressive strength. Research of the characteristics of pore structure of a series of GFCs showed that the pore size distribution in GFC affected the compressive strength to a large extent, particularly for the large pores. Based on the statistical fitting and modelling, a new model was developed, called the ‘large void model’, which treated the porosity of critical size pores (>100 m) and total porosity separately. Two mathematical models relating the measured thermal conductivity with porosity and dry density were successfully developed. The mathematical models were proven to be able to predict the mechanical and thermal insulation properties precisely