Assessment of Cement Mortar Strength Mixed with Waste Copper Mine Tailings (CT) by Applying Gradient Boosting Regressor and Grid Search Optimization Machine Learning Approach
DOI:
https://doi.org/10.52756/ijerr.2024.v42.016Keywords:
Cement mortar, copper tailings, gradient boosting regressor, grasshopper optimization algorithm, mechanical propertiesAbstract
Growing environmental concerns and resource scarcity, the construction industry must investigate sustainable materials that reduce waste while improving building material properties. This study investigates the viability of using waste copper mine tailings (CT) as a partial replacement for river sand in cement mortar, assessing the impact on mechanical strength and developing a predictive model using a Gradient Boosting Regressor (GBR) and Grid Search Optimization (GSO). Copper tailings mix designs ranging from 0% to 50% replacement by volume (river sand) were developed, with a constant cement quantity while varying the proportions of sand and tailings. The experiments were carried out at Poornima University in Jaipur, using standard protocols to prepare specimens and measure their compressive strength for 0% CT to 50% CT as volume percent in the mixture. The results showed that the addition of copper tailings up to 20% (3CT2) had highly increased the mortar strength, while mix design 6CT3 showed the best strength at 30% CT. Beyond this threshold, strength declined, thus indicating an optimal replacement level. In the final step of the research, the GBR-GSO-based machine learning approach was employed for developing the predictive model for compressive strength in mortar with different contents of copper tailing for mix types 3CT and 6CT. The predictions obtained by the developed model were in very good agreement with empirical data, thus supporting the potential of machine learning to predict material performance for guiding the use of unconventional additives in construction. It could be said that this work not only represents the material properties of copper-tailing-infused mortar but also showcases state-of-the-art machine learning techniques in the building materials science domain while opening new paths for more innovative and sustainable building practices.