A probabilistic approach for identification of compressive strength of fly ash bricks

Abstract

In the present industrial scenario, the production of fly ash being a waste from thermal power plants is increasing at an alarming rate. Fly ash not only creates a major problem at its disposal but also causes other environmental/health-related issues. To combat this waste is to find its application in the construction sector. In the past, fly ash is recognized as a promising replacement of clay in brick industry. The aim of the present work is to determine the characteristic compressive strength in a probabilistic approach for the Fal-G (fly ash–lime–gypsum) bricks constitutes the consumption of two waste products (fly ash and sludge lime) in brick manufacturing. For casting of Fal-G bricks, a semiautomatic, hydraulic operated brick manufacturing machine is fabricated and the compressive strength of the compacted bricks is observed. However, the compressive strength of the resulted bricks observed to be quite satisfied with the minimum value recommended as per IS 12894-1990. The study is further extended towards the fitting of randomly generated compressive strength values of Fal-G bricks to four different probability distribution functions and adopting Kolmogorov–Smirnov (KS) test to detect the best-fitted probability distribution of the present data. From the descriptive statistics of KS test results, it is inferred that the most common distribution, i.e. normal distribution, fits well to the data with the highest p value of 0.898. Normality test such as P–P and Q–Q plots is used to further confirm that the present data follow the normal distribution. Actual and expected z-scores are plotted to get all the data points on y = x line that follows the normal distribution. Determining the frequency factor of the normal distribution, the compressive strength is calculated at each probability level. At 95% probability level, the characteristic compressive strength of Fal-G bricks is obtained 4.8 N/mm². Further, a curve fitting between a probability of exceedance and compressive strength shows logarithmic relationship which holds between the two variables with R² = 0.932.