Predicting the compressive strength of fly ash concrete with the Particle Model

Abstract

A novel approach called the Particle Model is used to predict the compressive strength of fly ash concrete. Thousands of fly ash particles are classified into the nine groups based on the chemical information of each particle. Predictive models for compression strength for 20% and 40% fly ash replacement levels are developed for seven different curing times between 3d and 180d. The R-squared value for the compressive strength prediction of the Particle Model is ≈ 0.99, while the Class C and F classification model is <0.50. Furthermore, the results show that the Particle Model can predict the compressive strength within ±10% for 95% of all measurements at 20% fly ash mass replacement and for 81% of all measurements at 40% mass replacement. The coefficients in the derived predictive models give important insights into the effectiveness of different fly ash particles over time.