Abstract
The study reports on the proposed gene expression programming (GEP) to predict the ability of CFRP wrapping to enhance the post-heated bond strength between steel rebar and concrete. For this purpose, the results of 80 pullout specimens (150×150×250 mm) were selected in two categories of compressive strength of concrete: C30 and C40. In the experiments, a 20 mm rebar was placed in the corner of each specimen with a 25- or 35-mm clear cover and 200 mm of bond length. The specimens were subjected to different levels of heat before using CFRP jackets. The exposure to heat reduced the concrete-rebar bond strength; however, confinement with CFRP jackets significantly improved the bond strength between steel rebar and concrete. The estimated bond strength could be predicted by the accuracy and high prediction ability of the proposed model. Based on the findings, machine learning can bring significant improvements and benefits to the civil engineering industry in terms of the estimation of the mechanical properties of the materials in various conditions.
TopThe Purpose Of The Chapter
In civil engineering, especially in reinforced concrete structures, establishing a connection between temperature and residual bond strength is one of the main factors in the failure mechanisms of elements particularly during seismic actions. The findings in the literature clearly represent that the binding strength dramatically reduces as the temperature rises. The key factors that have affected the results are the features of the materials in terms of concrete and steel rebar, mainly; the diameter and embedding length of the rebar and the cover of the concrete. The pull-out failure mode overtook the splitting failure as the confinement level increased. Therefore, it is important to switch the mode of failure to pull-out mode in concrete members where splitting failure is anticipated and to improve ductility performance by adding external jackets. Steel plate wrapping and fiber-reinforced polymer FRP sheets have all been employed as exterior jacketing in earlier research to strengthen the bond between rebar and concrete. In light of the above-mentioned factors, the study aims to have an improvement in the estimation of the bond strength between reinforcements and concrete in different conditions by employing Gene Expression Programming (GEP). For this purpose, the results of eighty pull-out specimens (150×150×250 mm) were selected in two categories of compressive strength of concrete; C30 and C40 with different covers as well as temperatures and compared with carbon fiber-reinforcement polymer CFRP jacketed specimens. The experimental results demonstrated that the CFRP retrofitted method could able to improve the bond strength significantly. Although the elevated temperature could reduce the concrete-rebar bond strength, employing CFRP jackets dramatically increased the bond strength between concrete and reinforcements. The estimated bond strength by employing Genetic engineering programming (GEP) could be predicted by the accuracy and high prediction ability of the proposed model. Based on the findings, machine learning can bring significant improvements and benefits to the civil engineering industry in terms of the estimation of the mechanical properties of the materials in various conditions.
Key Terms in this Chapter
Genetic Engineering: Genetic engineering, also called genetic modification is the modification and manipulation of an organism's genes using technology.
Artificial Neural Network: ANN is a branch of machine learning by employing the principles of neuronal organization discovered by connectionism in biological neural networks.
Fiber-Reinforced Polymer: FRP is a structure containing an arrangement of unidirectional fibers or woven fiber fabrics embedded within a thin layer of light polymer matrix material.
Machine Learning: Development and use of computer systems with the ability to learn and adapt without following explicit instructions, using various algorithms and statistical models to analyze and draw inferences from patterns in data.