Effect of porosity on predicting compressive and flexural strength of cement mortar containing micro and nano-silica by ANN and GEP

Abstract

The aim of this study is to evaluate the effect of porosity on mechanical properties of cement mortar containing micro and nano silica in two aspects of experimentation and modeling of prediction. For this purpose, 32 mix designs were considered with various replacement percentages of nano silica (Ns) and micro silica (Ms) in forms of alone and together. The microstructure effect of Ns and Ms on the mechanical properties of cement mortar was investigated by Field Emission Scanning Electron Microscopy (FE-SEM) analysis. Moreover, Artificial Neural Network (ANN) and Genetic Expression Program (GEP) models are presented to predict the compressive and flexural strengths of cement mortar by focusing on the effect of porosity in models. So, a comparative probe was carried out on two statuses. Once, porosity wasn't considered as input parameter, and, in the next step, it was considered as input parameter in developing ANN-I or GEP-I and ANN-II or GEP-II models, respectively, in order to specify the sensitivity of the models to select the proper input parameters for accurate prediction. The results showed that the use of simultaneous Ns and Ms led to a decrease in the porosity and an increase in the flexural and compressive strengths. This is due to the synergistic effect on the microstructure of cement paste. The current modeling results showed that the ANN-II and GEP-II models have higher accuracy in the prediction of mechanical properties considering porosity as an influential input parameter. Moreover, the validation of proposed models was evaluated with the help of a collection of previous literature.