Multivariate Analysis for Prediction of Splitting Tensile Strength in Concrete Paving Blocks

Abstract

Paving blocks are concrete pieces exposed to the weather and subjected to loads and wear. Hence, quality control in their manufacture is essential to guarantee their properties and durability. In Ecuador, the requirements are described in the technical standard “NTE INEN 3040”, and tensile splitting strength is a fundamental requirement to guarantee product quality. The objective of the study is to predict the tensile splitting strength using two groups of predictor variables. The first group is the thickness in mm, width in mm, length in mm, mass of the fresh paving block in g, and percentage of water absorption; the second group of predictor variables is the density of the fresh paving block in kg/m3 and the percentage of water absorption. The data were obtained from a company that can produce 30,000 units per day of rectangular paving blocks with 6 cm thickness. The research involves sampling, analysis of outliers, descriptive and inferential statistics, and the analysis of multivariate models such as multiple linear regression, regression trees, random forests, and neural networks. It is concluded that the multiple linear regression method performs better in predicting the first group of predictor variables with a mean square error (MSE) of 0.110086, followed by the neural network without hidden layers, resulting in an MSE of 0.112198. The best method for the second set of predictors was the neural network without hidden layers, with a mean square error (MSE) of 0.112402, closely followed by the multiple linear regression model, with an MSE of 0.115044.