Effect of expanded perlite aggregate on cyclic thermal loading of HSC and artificial neural network modeling

Abstract

This paper describes a laboratory investigation of the resistance to freezing and thawing of Expanded Perlite Aggregate (EPA) concrete, compared with that of natural aggregate concrete. The effects of EPA ratios on High Strength Concrete (HSC) properties were studied for 28 days. EPA replacements of fine aggregate (0–2 mm) were used: 10%, 20% and 30%. The properties examined included compressive strength, Ultrasound Pulse Velocity (UPV), porosity, microstructure and the Relative Dynamic Modulus of Elasticity (RDME) of HSC. Results showed that the compressive strength, UPV and RDME of samples were decreased with an increase in EPA ratios. Test results revealed that HSC was still durable after 100, 200 and 300 cycles of freezing and thawing in accordance with the ASTM C666. After 300 cycles, reduction in compressive strength and RDME ranged from 7% to 29% and 5% to 21%, respectively. In this paper, feed-forward Artificial Neural Network (ANNs) techniques were used to model the relative change in compressive strength and UPV in cyclic thermal loading. Genetic algorithms were applied in order to determine optimum mix proportions subjected to 300 thermal cycling. The best performance was obtained from HSC with about 10% EPA.