Age-dependent deflection analysis of high-volume fly ash RC beams: experimental and artificial neural network modelling

Abstract

This research examines the influence of high-volume fly ash (HVFA) on reinforced concrete (RC) beam deflections over 180 days of continuous loading. The study comprises eight fly ash-based concrete mixes with varying fly ash ratios (0–60%). Experimental assessments were conducted on two load levels, corresponding to 25% and 50% of the initial cracking load. Thirty-two full-scale RC beams, each measuring 100 × 150 × 1800 mm, were fabricated for both load levels. Long-term deflections were monitored for all concrete mixes through sustained four-point bending tests at ages ranging from 1 to 180 days. A comparative analysis of the overall mid-span deflection between conventional and fly ash concrete RC beams was carried out against established design codes. It was determined that high-volume fly ash RC beams exhibited less creep and shrinkage deflection compared with conventional RC beams. Moreover, artificial neural network (ANN) modelling was also carried out to propose a significant model helpful in predicting the total deflection of RC beams containing any fly ash content and at any age of concrete. The present study offers valuable insights for engineers and designers, facilitating the evaluation of age-dependent deflection patterns in RC beams constructed with reinforced high-volume fly ash concrete.