Abstract:
Concrete, a foundational material in construction, boasts widespread availability and commendable attributes like high compressive strength, durability, and stiffness under typical conditions. However, its inherent brittleness poses challenges, notably in terms of low flexural strength and strain at fracture. To address these limitations, conventional practice often necessitates reinforcement through the inclusion of continuous deformed bars or prestressing tendons.
In contemporary construction, Glass .Fiber Reinforced Concrete (GFRC) has emerged as a revolutionary composite material, celebrated for its exceptional mechanical properties, durability, and architectural adaptability. Despite its myriad advantages, accurately. predicting GFRC's flexural strength remains a formidable task due to the complex interplay of various factors such as material composition, curing conditions, and fiber characteristics.
This project embarks on a comprehensive exploration into the realm of machine learning, seeking novel solutions to this .challenge by developing robust predictive models . .capable of accurately estimating GFRC's flexural strength across. diverse scenarios and applications. Through the. judicious application of various machine learning techniques, including supervised learning methodologies. such as Xtreme Gradient Boosted Decision Trees. (XG-Boost) and Multi-Expression Programming (MEP), our aim is to furnish architects, engineers, and researchers with indispensable tools for refining GFRC .mix designs, optimizing structural configurations, and enhancing its overall performance.
By serving as a bridge between conventional. concrete testing protocols and advanced predictive .modeling paradigms, this research endeavors to propel forward sustainable and resilient construction .practices, contributing to the evolution of the built environment towards greater efficiency, durability, and sustainability.
