Open AccessA Statistical Approach to Predicting Fresh State Properties of Sustainable Concrete
Author(s) -
Ruoming Jin,
Chen Qian,
Alfred Soboyejo
Publication year - 2021
Publication title -
epic series in built environment
Language(s) - English
Resource type - Conference proceedings
ISSN - 2632-881X
DOI - 10.29007/1h88
Subject(s) - slump , regression analysis , sustainability , multivariate statistics , variables , statistical analysis , logistic regression , raw material , environmental science , computer science , statistics , machine learning , mathematics , materials science , cement , ecology , chemistry , organic chemistry , metallurgy , biology
Using environmentally friendly concrete materials can help improve the sustainability of the concrete industry. However, the effects of such materials on concrete properties must be fully understood before sustainable concrete can be widely applied. Previous research showed limited applications of statistical methods in analyzing the effects of sustainable concrete materials on fresh concrete properties. This study applies multivariate regression analysis to modeling properties of fresh concrete (i.e., slump, air content, and density) made with multiple sustainable raw materials based on variables in mixture design. Different regression models were tested to explore the best-fit model(s) that can capture and predict how these variables affect fresh concrete properties. The regression analysis showed satisfactory results in predicting air content and density, but not in predicting slump. The regression analysis, as a statistical tool, can provide deep insights into how the selected independent variables affect fresh concrete properties and the degree of the effects.