Back to Search Start Over

Influence of treated recycled concrete aggregate and modified mixing approach on the mechanical properties of ternary blend geopolymer concrete: Experiments and machine learning algorithms.

Authors :
Singh, Paritosh Kumar
Rajhans, Puja
Source :
Journal of Cleaner Production. Mar2024, Vol. 443, pN.PAG-N.PAG. 1p.
Publication Year :
2024

Abstract

The variations in mechanical properties of geopolymer concrete (GPC) prepared with treated recycled concrete aggregate (RCA) in conjunction with a modified two-stage mixing approach (M-TSMA (sp-ggbs)) have been studied in the present research. Since GPC is prepared using fly ash (FA) and ground granulated blast slag (GGBS) as an alternative to cement, it is sustainable and eco-friendly. Further, FA is partially replaced with silica powder (SP) to evaluate the mechanical properties of GPC. Additionally, RCA is treated with sodium silicate to strengthen the adhered mortar. Moreover, based on experimental results, Bayesian optimization Support Vector Regression (SVR) and Gaussian Process Regression (GPR) models are developed to predict the mechanical strengths of GPC mixtures. Results show that treating RCA with sodium silicate improves RCA's physical and mechanical characteristics. Furthermore, treated RCA and M-TSMA (sp-ggbs) enhance the mechanical strengths of GPC mixes, and these strengths can be accurately predicted using SVR and GPR models. • Examining the impact of silica powder on the properties of fly ash/GGBS-based GPC. • Sodium silicate (45 %, 7 days) enhances RCA's physical and mechanical properties. • Treated RCA and M-TSMA (sp-ggbs) improve the properties of ternary blend GPC. • Bayesian optimization algorithm tunes the hyperparameters of SVR and GPR models. • BOA-SVR and BOA-GPR models accurately predict the mechanical properties of GPC. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
09596526
Volume :
443
Database :
Academic Search Index
Journal :
Journal of Cleaner Production
Publication Type :
Academic Journal
Accession number :
175569308
Full Text :
https://doi.org/10.1016/j.jclepro.2024.141007