Mathematical Modeling of Split Tensile Strength of Agrostone with Different Bio-fillers

Abstract

There has been a growing imbalance in the demand and supply of construction materials with increasing prices day by day at an alarming rate due to escalating cost of energy.  Quest for a low cost, environmental friendly material has resulted in the development of Agrostone made out of industrial and agricultural waste while saving the environment from solid waste disposal issues.  This also contributes for saving the energy required for production of conventional building materials up to 20 per cent thereby reducing the emissions. Agrostone made of different bio-fillers such as water hyacinth, bagasse and grass, satisfy the strength requirements as per Ethiopian construction material standard which is equivalent to a hollow block. The split tensile strength has been evolved through the compression test by adopting the standard test methods of applied for concrete. This paper presents the correlation of the experimental results split tensile strength with the values obtained using various empirical relations existing in the literature for unreinforced concrete. It is found that all these relations have underestimated the results while the experimental results of Agrostone with water hyacinth bio-filler have better correlation with results of relation given by Arioglu et al. This paper also presents mathematical models developed using linear and polynomial regression analysis to predict the split tensile strength of Agrostone. The correlation coefficients from the regression analysis are obtained using the Stats.blue software. It is found that the polynomial equations of degree five and the correlation indicators r and r2 show a strong relation between the independent compressive strength and the dependent split tensile strength. The linear regression also predicts the values with acceptable degree of accuracy. The results obtained by developed mathematical models are very close to the experimental within acceptable range of error.