AI-assisted optimization of multifunctional cement mortar incorporating banana peel-derived activated carbon

Information from the abstract

Multifunctional cement mortars combining mechanical performance with aqueous formaldehyde (HCHO) adsorption offer a promising route toward functional and sustainable construction materials, yet balancing these functions remains challenging. This study integrates banana peel-derived activated carbon (BPAC) with explainable machine learning to address this trade-off. BPAC (0–3 wt%) was incorporated as a partial cement replacement and evaluated at 7–28 days. The 1 wt% BPAC mortar achieved the highest 28-day compressive strength of 48.27 ± 3.22 MPa (18.7% above plain cement mortar) and the highest HCHO adsorption efficiency of 22.15 ± 1.15% at 7 days (90.13% improvement). Adsorption decreased with curing age, confirming a strength–adsorption trade-off. Among Decision Tree, Random Forest (RF), and Gradient Boosting Tree models evaluated by 10-fold cross-validation, RF provided a favorable overall predictive balance, with validation R² values of 0.88 ± 0.09 for strength and 0.88 ± 0.06 for adsorption. SHAP analysis identified curing age and BPAC content as the more influential variables for strength and adsorption, respectively. Equal-weight RF optimization identified 0.8–1.2 wt% BPAC as a model-predicted promising region for balancing both functions within the investigated design space. This study provides a data-driven framework for designing multifunctional cementitious materials that couple mechanical performance with aqueous HCHO adsorption while valorizing agricultural waste.

Why this record is monitored

This record has an Impact Signal of 71/100 based on recency, source, collaboration, and bibliographic signals. It prioritizes monitoring and is not a judgment of research quality.

Related topics: Concrete and Cement Materials Research · Innovative concrete reinforcement materials · Concrete Properties and Behavior