AI-Based Formulation for Mechanical and Workability Properties of Eco-Friendly Concrete Made by Waste Foundry Sand

Abstract

The casting process creates a significant amount of waste foundry sand (WFS). Using WFS as a concrete ingredient reduces the problems associated with the dumping process of these types of wastes, removes/reduces carbon dioxide, and is also considered economical in terms of overall concrete production cost. Besides, WFS has been reported to be a critical parameter affecting the mechanical properties and workability of concrete. Hence, predicting the behavior of concrete using the development of models based on artificial intelligence (AI) algorithms derived from the laboratory data can remarkably improve the project’s efficiency in terms of cost and time. This paper assessed the performance of artificial neural networks (ANNs) to predict the strength parameters of concrete containing WFS (CCWFS). In this regard, a comprehensive laboratory database consisting of 102, 397, 146, 346, and 169 data for the slump, compressive strength, elasticity modulus, splitting tensile strength, and flexural strength of CCWFS were collected from literature, respectively. Seven different variables including waste foundry sand to cement ratio (WFS/C), water to cement ratio (W/C), fine aggregate to the total aggregate ratio (FA/TA), coarse aggregate to cement (CA/C), waste foundry sand to the fine aggregate ratio (WFS/FA), 1,000 superplasticizer to cement ratio (1,000SP/C), and age (except for slump), were selected as input. The accuracy of the models was verified by examining 11 different performance indicators. Comparing the predicted and observed data’s overlap percentage indicated that the ANN models are less error-prone than those obtained using multiple linear regression (MLR). Also, the models’ validation and uncertainty analyses showed that all models were within the permissible range. Given the performance evaluation results, it can be concluded that ANNs can be used as a reliable and accurate method for predicting the mechanical properties of CCWFS.