Prediction on Flexural strength of High Strength Hybrid Fiber Self Compacting Concrete by using Artificial Intelligence

Abstract

The primitive focus of this research work is about the artificial intelligence methods engaged for creating an outlook for flexural strength of High Strength Hybrid Fiber Self Compacted Concrete (HSHFSCC), which is considered to be a special concrete in order to tackle both workability and durability without disturbing the strength of the concrete. It possesses not only the good deformability during fresh state but also put forward high aversion to segregation resulting in superior homogeneity and increase in productivity by altering the period of construction. While incorporating various fibers like glass, steel, carbon, synthetic, and quartz powder in plain concrete, directs in the enhancement of post-cracking, toughness, ductility and limits the detrimental effect of shrinkage. The current work is classified into two stages. 1) Development of HSHFSCC and High Strength Self Compacting Concrete (HSSCC). 2) Engaging different Machine Learning (ML) models to divide the obtained information into Train, Test and Validation followed by 19 types of different ML regression models accompanied with Artificial Neural Network for engaging the function to appropriate the flexural strength of HSHFSCC and HSSCC. The boundary conditions discussed as input includes Setting time, percentage of quartz and river sand. Total 25 number of datasets are used for 5-fold cross validations by adopting MATLAB ML and Deep learning toolkit and Python is adopted to validate the optimized models. The evaluation factors like R-square and Root mean square show a great level of accuracy and reliability of the model.