Experimental investigation and prediction of compresive strength of concrete using soft computing techniques with different additives

Abstract

High Performance Concrete (HPC) is the latest development in concrete, But HPC not only demands High cement consumption, which pushes the natural resources towards depletion, but also increases C02 emission on a higher extent. In the recent year’s use of Supplementary Cementitious Materials (SCMs) is increased due to environment concerns, conservation of resource & economy because most of them are generally Industrial waste products such as fly ash, GGBS & micro silica. One of the costliest constituent of HPC is ultrafine material such as micro silica, alccofine. In recent years with the advancement in technology ultrafine fly ash is now being produced which is cheaper ultrafine material but, with less literature available on it. In available literature on Ternary blend concrete the level of replacement was restricted up to 30%-35%. In this Experimental Investigation an attempt was made to investigate compressive strength (100MPa) of concrete by replacing Cement on 40%, 45%, 50%, by incorporating P100 fly ash as an ultrafine material and GGBS. Each replacement was further divided into three sub parts (40%F.A-60%GGBS), (45%F.A- 55%GGBS), (50%F.A-50%GGBS). Among which 40% replacement of cement (50%F.a- 50%GGBS) gave maximum strength. Nominal mix was prepared with only OPC with w/c of 0.24.and all other ternary mixes was made on w/c of 0.2 to have an edge when compared with strength of nominal mix. Nowadays, soft computing techniques are used to predict the properties of concrete and hence reduce the experimental work. Thus, a neural network also known as a parallel distributed processing network, is used as computing paradigm that is loosely modeled after structures of the brain. It consists of interconnected processing elements called nodes or neurons that work together to produce an output function. This experimental investigation presents the application of Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) techniques for developing the model to predict the compressive strength of the concrete with SCMs. For this purpose, a systematic laboratory investigation was carried out. The compressive strength was evaluated on various mixes for 3 days, 7days, 14 days and 28 days of curing period. The data generated in the lab was used for development of the MLR and ANN model. The data used in the models are arranged in the format of four input parameters that cover the contents of OPC, FA, GGBS and w/c ratio respectively and one dependent variable as compressive strength of concrete for both MLR and ANN. Networks are trained and tested for various combinations input and output data sets. Keywords: High Performance Concrete (HPC), Supplementary Cementitious Materials (SCMs), Fly Ash (FA), Ground Granulated Blast Furnace Slag (GGBS), Artificial Neural Network (ANN), Multi Linear Regression (MLR).