Investigating the Effect of Adding Nanosilica on Compressive Strength and Electrical Resistance of Concrete at Different Water-Cement Ratios and Predicting it Using Artificial Neural Networks

Abstract

Nanotechnology has brought about remarkable advancements in human knowledge and has attracted substantial attention from researchers across various scientific fields. Nanoparticles possess unique physical and chemical properties that have paved the way for the production of novel materials with exclusive capabilities. In civil engineering, one such application is the use of nanosilica, a nanotechnology product, as a highly reactive artificial pozzolan to partially substitute cement in concrete mix designs, potentially improving the properties of concrete.

In recent years, artificial neural networks, a branch of artificial intelligence, have been employed to address many civil engineering problems. The fundamental concept behind this method is inspired by the way the human brain's neural system processes data and information for learning and knowledge generation. A neural network comprises a desired number of processing units (neurons, cells, nodes) arranged in layers, which relate the input set to the output set. Essentially, neural network models can be regarded as complex linear or non-linear regressions, trained on input data and target outputs, capable of predicting conditions based on new inputs. The experimental results in this study demonstrate that the use of nanosilica additive and the reduction of the water-to-cement ratio increase the compressive strength and electrical resistance of concrete. Furthermore, by considering different amounts of nanosilica additive for various water-to-cement ratios at 7 and 28 days as the network input, the designed artificial neural network model performs satisfactorily in predicting the compressive strength and electrical resistance indices of concrete.

Keywords: Compressive Strength, Electrical Resistance, Nanosilica, Artificial Neural Network.