Artificial Neural Networks for Evaluation & Prediction of the Mechanical Properties of Waste Ceramic Optimal Concrete Exposed to Elevated Temperature

Abstract

Resistance to concrete structures for rising temperatures is a major global concern today. So, this topic requires a research effort to arrive at a heat-resistant economic concrete. In this study, an Artificial Neural Network(ANN) model for compressive & tensile strength of Plain concrete (PC), waste ceramic optimal concretes (WOC) & WOC-Hybrid fibers concrete (PVA-CR-WOC) exposed to elevated temperature were devised. First, the authors developed these new types of concrete (WOC and hybrid-WOC) proprtes were used in the training of ANN. WOC was prepared with Natural Coarse Aggregates (NCA), Natural fine aggregate (NFA), Ordinary Portland Cement (OPC 43 grade) and ceramic waste tiles with 20% replacements for coarse aggregates, 10% replacements for fine aggregates, and 10% replacement for cement. In contrast, Hybrid- WOC is prepared with the addition of hybrid fiber (1% crimped steel fiber and 1% Polevenal alcohol fiber) in WOC. At 28 days, specimens were heated to elevated temperatures (100ºC, 200ºC, 300ºC). Tests were conducted to determine the loss in compressive and tensile strength. The results indicated that all concrete, particularly plain concretes, lost strength when exposed to elevated temperatures. Furthermore, based on the experimental results, an ANN based explicit formulation model was proposed to predict the loss in compressive & tensile strength of PC, WOC and WOC- Hybrid concrete for which is the temperature above 300ºC. ANN base model expressed in terms of the Portland Cement (PC), Water Cement Ratio (W/C), sand (S), Coarse Aggregate (20–10 mm) (CA), Ceramic Aggregate (20–10 mm) (CA), Ceramic Sand (CS) , Ceramic Cement (CC), Crimped steel fiber (CR), polevenal alcohol fiber (PVA ) and applied temperatures (T). It was found that the developed ANN empirical model seems to have a high prediction capability of the loss in compressive and tensile strength of PC, WOC and WOC-H

Keywords: Artificial Neural Network, Waste Ceramic Concrete, Hybrid Fibers, Elevated Temperature