Experimental study and development of machine learning model using random forest classifier on shear strength prediction of RC beam with externally bonded GFRP composites

Autor: Amna Hamed Salim Al Mamari, Rawan Said Humaid Hammad Al Ghafri, N. Aravind, Ragavesh Dhandapani, Eman Muhye Adeen Muhye Al Hatali, Rani Pandian
Rok vydání: 2022
Předmět:
DOI: 10.21203/rs.3.rs-1592560/v1
Popis: Life span of RC structures is reduced mainly because of steel reinforcement corrosion and there is a need to use alternate material. In the present study, GFRP composites are used as externally bonding shear reinforcement of the RC beams. For experimental analysis, RC beams were cast with C30 grade concrete with steel/ GFRP shear reinforcements. For the external bonding work, Chopped Strand Mat (CSM), Woven Roving (WR) GFRP composites were attached on the two sides as ‘U’ wrap on the shear zone of the RC beams. Two point loads were applied at 150 mm spacing on all the beam specimens and the experimental results of the RC beams with GFRP composites are compared with control beams. The major outcome of the present study is that, the RC beam externally bonded with GFRP WR ‘U’ wrap on full shear area performs well and the failure load is 20.93%, 6.57% and 18.94% more than that of beams externally bonded with normal ‘U’ wrap strips on shear area, sides of full shear area and strips on sides respectively. Also, theoretical analysis was carried out to determine the shear resistance of RC beams and the results are compared with experimental results. In addition to that, Random forest classifier was used for the development of Machine Learning (ML) model to predict the failure load, mid span deflection and type of failure. The predicted ML results are validated and it proved that the experimental results are well line with the developed ML model in terms of performance and accuracy.
Databáze: OpenAIRE