Axial Capacity Prediction of Concrete Externally Strengthened by Low-Cost Natural Fiber Reinforced Polymer Composite Utilizing Machine Learning Intelligence

The concept of enhancing concrete's strength and sustainability with fiber reinforced polymer (FRP) composites has drawn the interest of engineering researchers. Previous studies on FRP composites with glass, carbon, and synthetic fibers as fiber material paved the way for the current applications of FRP composites in the construction industry. Despite its effectiveness, there are certain drawbacks, including high cost of fiber material and sustainability of production. A less expensive and more sustainable method of producing FRP can be explored by employing natural fibers as alternative fiber source for FRP composites. Pioneering investigations on the use of natural fiber reinforced polymer (NFRP) as external strengthening for concrete demanded several hands-on laboratory tests, resulting in prolonged preparation times and significant material waste. The study seeks to offer an alternative method for assessing NFRP's strengthening capabilities for concrete, decreasing physical laboratory testing costs, material wastage, and assuring time management. Creating a computational artificial neural network (ANN) model that uses the Levenberg-Marquardt algorithm to forecast the compressive strength of concrete confined with NFRP composites is the key objective of the study. Data that was utilized in the ANN modelling was derived from conducted compressive strength tests on concrete externally confined with NFRP featuring sisal, jute, flax, hemp, and cotton as fiber material. The model inputs were unconfined concrete compressive strength, NFRP tensile strength, NFRP elastic modulus, NFRP strain, and NFRP thickness. The resulting model had excellent predictive capabilities, with a total Pearson's correlation coefficient (R) of 0.9966 and mean square error (MSE) of 0.3826. Parametric analysis revealed that NFRP thickness had the greatest influence on enhancing the compressive strength of concrete.