Strengthening performance of reinforced concrete beam using combined externally bonded and near surface mounted techniques / Kh. Mahfuz Ud Darain

Kh. Mahfuz , Ud Darain (2016) Strengthening performance of reinforced concrete beam using combined externally bonded and near surface mounted techniques / Kh. Mahfuz Ud Darain. PhD thesis, University of Malaya.

	PDF (The Candidate's Agreement) Restricted to Repository staff only Download (190Kb)
Preview	PDF (Thesis PhD) Download (7Mb) | Preview

Abstract

Structural strengthening is a technique to upgrade and improve existing structural systems to carry additional loads and prolong design life. Various structural strengthening techniques are now being used in the construction industry. Among these techniques, the externally bonded reinforcement (EBR) technique is the most common. However, it has a tendency to fail by debonding failure at the plate curtailment location. This is due to interfacial shear stress, which increases with increasing plate thickness. The near surface mounted (NSM) technique is comparatively new and more efficient. Nevertheless, it exhibits premature failure when multiple grooves are used within a narrow cross-sectional width. The aim of this research was to develop a strengthening solution to improve the structural performance of RC beams while avoiding premature failure. To achieve this objective, this study proposed a combination of the EBR and NSM techniques, calling it the combined externally bonded and near surface mounted (CEBNSM) technique. In this study, the performance of RC beams strengthened with the CEBNSM technique were investigated experimentally. An artificial intelligence technique was used to predict the serviceability behavior of these strengthened beams. The finite element method (FEM) was also used to simulate the structural behavior of the strengthened beams. The experimental test matrix consisted of a total of twenty-seven RC beams divided into four groups. Round steel or CFRP bars were used in the NSM grooves for the CEBNSM-B beams, whereas rectangular CFRP strips were inserted into the CEBNSM-S beams. In both cases, CFRP fabric was externally bonded at the tension face of the beam soffit. NSM and EBR strengthened beams were also tested in order to compare results with the CEBNSM technique. Fuzzy logic expert system (FLES) was used as an artificial intelligence (AI) tool to predict the serviceability behavior of the strengthened beams. Incremental static load and variable NSM bar length were the input parameters and the outputs were deflection and crack width of the strengthened beams. Applying expert knowledge using if-then rules, the input and output variables were expressed linguistically as well as in numeric values. FEM was applied to develop a numerical model to verify the experimental results of the strengthened RC beams. The plastic damage behavior of concrete, elasto-plastic behavior of steel reinforcement and material nonlinearity were considered in developing the FEM model. Beams strengthened with NSM CFRP showed greater increment in strength, although they failed prematurely, in contrast to the flexural failure of steel bar NSM beams whose stiffness and cracking behavior were superior as well. For CEBNSM beams, the ultimate capacity increased from 32% to 176%, depending on the variation in strengthening reinforcement ratio. The failure mode, serviceability and stiffness of the beams also improved considerably. The output of the AI models excellently predicted the deflection and crack width of the strengthened beams. In evaluating the FLES prediction model, it was found that the relative error of the predicted deflection and crack width values were within the acceptable limit (5%) and the goodness of fit of the predicted values was close to 1.0. The results simulated by the FEM model satisfactorily agreed with the load-deflection and strain values of the CEBNSM strengthened RC beams. The simulated damage pattern of the beams also matched well with the experimental beams.

Actions (For repository staff only : Login required)