Assessment of mechanical properties with machine learning modeling and durability, and microstructural characteristics of a biochar-cement mortar composite

Abstract

Rising CO2 emissions have become one of the biggest environmental challenges in recent years. Due to the rising carbon footprint of the building industry, CO2 emission regulation and mitigation have become perennial issues. The utilization of Biochar (BC) as a carbon-sequestering component in cement mortar is the novelty and main concern of this study. In this research, the effectiveness of BC in sequestering carbon was examined along with its effect on the mechanical, microstructural, and durability characteristics of the composite cement mortar. It includes a control and eight additional mixes prepared with 1%, 3%, 5%, and 8% BC by weight of cement added to mortar; the BC were prepared at two fixed temperatures of 300 °C and 500 °C. It also involves testing fresh properties, mechanical properties, durability properties, and microstructure analysis using scanning electron microscopy (SEM) and energy-dispersive X-ray analysis (EDX). A Universal Testing Machine (UTM) was used to determine the mechanical properties of the cement mortar, such as its compressive and tensile strengths. The water permeability and rapid chloride permeability tests (RCPT) were used to evaluate the specimens' long-term stability as a measure of their durability. Based on the test results, it has been found that the inclusion of BC enhanced the strength and durability of cement mortar through its pozzolanic action. In addition, BC is a filler material whose porous structure fills the voids within the cement particles, decreasing water absorption and improving workability. BC sequesters carbon by carbonizing biomass. Its presence in cement mortar stores carbon that would otherwise be emitted into the atmosphere, which helps to reduce the environmental effect. According to the conclusion of the study, BC has the potential to be a sustainable component of cement mortar. In addition, the two distinct algorithms built upon machine learning applied in the analysis using adaptive boosting (AdaBoost) and linear regression (LR); both these analyses demonstrate that it is feasible to predict the characteristics of cement mortar accurately. The AdaBoost methodology outperforms the LR methodology because of its strong correlation value (R2).