Abrasion resistance behaviour of fly ash based geopolymer using nanoindentation and artificial neural network

Abstract

Abstract This study presents the investigation of the abrasion resistance of fly ash based geopolymer via the application of nanoindentation. Nine fly ash based geopolymer mixes with four variables: silica fume (SF), sand to cementitious ratio (s/c), liquid to solid ratio (l/s) and superplasticizer (SP) were investigated with statistical methods to obtain the plasticity index of the four fly ash based geopolymer phases. The results suggest that the pure fly ash based geopolymer paste with no SF, no SP, no sand and l/s ratio of 0.60 provides the highest abrasion resistance. Artificial neural network (ANN) is also adopted to predict the abrasion performance of the fly ash based geopolymer where good correlation is found. With the application of deconvolution method, it is reported that the plasticity index of N-A-S-H, partly activated slag and non-activated slag phases show a decreasing trend of plasticity index with increasing hardness or modulus. The non-activated compact slag phase shows no clear trend with changes in hardness and modulus. The relationship between degree of activation versus plasticity index (PI) is also studied, it was found that high degree of activation leads to formation of high volume fraction of N-A-S-H which this correlates with increased compressive strength and improved abrasion resistance. The application of nanoindentation with deconvolution method is suitable in quantifying the PI of the phases and abrasion resistance of the fly ash based geopolymer mortar.