RMIT University
Browse
Guo.pdf (5.41 MB)

Analysis and Design of FRP-Reinforced Concrete Beams with Innovative Compression Yielding Mechanism

Download (5.41 MB)
thesis
posted on 2024-06-27, 04:52 authored by Bingcheng Guo
Steel reinforced concrete structures have been fundamental in civil engineering for centuries. However, they are susceptible to deterioration over time due to the corrosive effects of environmental factors such as chemical exposure, seawater and deicing salts. Corrosion of steel reinforcement significantly reduces the capacity and reliability of the beam. Fibre reinforced polymers (FRP) bars reinforced concrete beam is a fantastic alternative to solve the corrosion issue of steel reinforcement. The main drawback preventing FRP bars from widespread application is their poor ductility. Among the methods for improving the ductility of FRP-reinforced concrete beams, introducing a compression yielding (CY) mechanism has been proven to be the most effective. However, previous studies have mainly focused on the experimental testing of CY materials and determining the flexural capacity and ductility of CY beams. A comprehensive study on the analysis of CY beams, considering different cross-section shapes (rectangular and T sections), as well as strain-hardening and strain-softening behaviours of CY material, has not yet been conducted. Additionally, a systematic reliability analysis of CY beam is not available to facilitate the practical design of CY beams. In this study, a comprehensive investigation of the flexural capacity and ductility of CY beams with rectangular and T sections is conducted. CY materials with both strain-hardening and strain-softening behaviours are addressed. In particular, different failure modes of CY beams when CY material exhibits strain-softening behaviour are analysed, which have not been reported previously. A layered approach considering the location of CY block is then developed for section analysis. The importance of key design variables is determined using the element effect method. It is revealed that enhancing ductility performance is attainable by utilising a CY material with a post modulus ratio near 0, while an increase in the strength of the CY material significantly enhances moment capacity. The most effective strategy for concurrently improving both moment capacity and ductility performance is to increase the height of the CY block. As CY beam design conducted through layered section analysis is complex and time-consuming, this study proposes a machine learning integrated model to estimate flexural capacity and ductility of CY beams. Specifically, various activation functions are tested in artificial neural networks to enhance the accuracy of flexural capacity estimation, while different kernel functions are employed in support vector machines with Gaussian process regression to predict CY beam ductility. The results indicate that the integrated model can provide highly accurate predictions. Additionally, a genetic algorithm (GA) is developed to identify optimal design solutions for CY beam sections. The robustness of the proposed model in optimising the design of CY beams with rectangular and T sections is demonstrated through two numerical examples. To implement CY beams in engineering practice, it is essential to conduct reliability analyses under both the ultimate limit state (ULS) and the serviceability limit state (SLS). In this study, the model for CY beam long-term deflection, considering both creep and shrinkage, is proposed for the first time. This thesis employs rigorous statistical methods (i.e., Kolmogorov-Smirnov test and Monte Carlo simulation) to ascertain the uncertainties of variables associated with CY beams in reliability analysis. Furthermore, it utilises the First-Order-Reliability-Method (FORM) to conduct reliability analysis both under ULS and SLS. Results from the reliability analysis under ULS recommend using a flexural resistance reduction factor of 0.75 for building structural members and 0.70 for bridge members. The reliability analysis under SLS indicates that the long-term deflection of CY beams is less than that of conventional FRP-reinforced concrete beams. Furthermore, the degradation in reliability for CY beams is also less than that for conventional FRP-reinforced concrete beams, suggesting better reliability of CY beams over their service life.

History

Degree Type

Doctorate by Research

Copyright

© Bingcheng Guo 2024

School name

Engineering, RMIT University

Usage metrics

    Theses

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC