Recieved:

24/02/2026

Accepted:

16/03/2026

Page: 

doi:

http://dx.doi.org/10.17515/resm2026-1531ma0224rs

Views:

4

Performance assessment of sustainable one-part alkali-activated concrete: An experimental and machine learning approach

Sujeet Kumar1, Ajay Kumar Sinha1

1Department of Civil Engineering, National Institute of Technology, Patna, 800005, India

Abstract

One-part alkali-activated concrete (OPAAC) is regarded as a sustainable substitute for traditional concrete and demonstrates significant potential in practical field constructions. This paper explores the properties of OPAAC made using a binder system that consists of ground granulated blast furnace slag, sodium metasilicate (SMS), and calcium hydroxide. Three mix designs for grades M15, M20, and M25 were formulated. The performance evaluations revealed that higher SMS content and water-to-binder ratio favored the workability of mixes. The hardened-state tests suggested that OPAAC showed an enhancement in compressive strength (CS), flexural strength (FS), and split-tensile strength (STS) with an increase in SMS content. The highest CS, FS, and STS recorded were 32.72, 6.20, and 3.44 MPa, respectively. Various machine-learning models were tested to estimate the strength of developed concrete. The gradient boosting model showed the best performance among others with an average test R² of 0.9684 and was able to successfully predict the strength. OPAAC also exhibited crucial environmental and economic advantages, including carbon reduction by 55-59% and cost savings by 12-16% compared to conventional concrete. Therefore, the outcomes show that OPAAC does not only present a comparable material performance but also an eco-friendly and less expensive option for the field construction activities.

Keywords

One-part alkali-activated concrete; Workability; Strength; Machine learning prediction; Sustainability

Cite this article as: 

Share This Article
LinkedIn
X
Facebook
journal cover
News & Upcoming Events