PhD Oral Exam - Omid Habibi, Civil Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

The utilization of fibre-reinforced polymer (FRP) bars in concrete structures has seen a consistent increase in recent years. This growth is attributed to the properties of FRP bars, including corrosion resistance, low density, fatigue resistance, and high tensile capacity. Among various FRP bars, glass FRP (GFRP) bars have attracted notable interest because of their advantageous characteristics, such as exceptional chemical resistance, cost-effectiveness, and versatile applicability. The bond strength between GFRP bars and concrete is crucial for the serviceability and ultimate performance of GFRP-reinforced concrete (RC) structures. GFRP bars exhibit anisotropy, linear elasticity, and a lower modulus of elasticity compared to steel bars, rendering the design guidelines for steel inapplicable to GFRP. Therefore, a thorough investigation into the bond is essential to ensure effective force transfer between GFRP bars and concrete.

In this thesis, the bond strength and bond-slip behaviour of high-modulus GFRP bars were investigated using pullout and beam splice tests. In Phase 1 of the experimental study, 48 pullout specimens were constructed and tested to investigate the local bond-slip behaviour between single and bundled ribbed GFRP bar in concrete. The study considered the effects of bar embedded length, specimen dimensions, GFRP surface profiles, presence of steel transverse reinforcement, and bundling of the bars on the pullout strength. The experimental data were used to calibrate existing analytical bond-slip models for ribbed GFRP bars. The calibrated values could be utilized in numerical modelling for concrete elements internally reinforced with ribbed GFRP bar. In second Phase, lap splice tests have been conducted on 20 full-scale beams with a focus on investigating the effects of GFRP bar’s surface profile and GFRP transverse reinforcement on the bond strength. The study evaluated the influence of different surface profiles, including ribbed, sand-coated, and grooved, as well as GFRP stirrup spacing, bar splice length, and bar diameter, on splice strength. A comprehensive dataset was collected from both the literature and the present study to evaluate the reliability of existing bond models. A new model was proposed to enhance bond prediction accuracy. Additionally, the bond contributions of steel and GFRP transverse reinforcement were assessed and compared.

Furthermore, the lower modulus of elasticity in GFRP compared to steel bars leads to increased crack width and deflections in GFRP-RC structures. Consequently, the design of GFRP-RC structures is usually governed by the serviceability limit state. Effective control of crack width during the service stage of GFRP-RC structures is of paramount importance. To address this, the present research also investigated the capability of machine learning (ML) models to predict crack width in GFRP-RC structures, with the aim of enhancing prediction accuracy compared to existing code and literature models. Various ML models, including extreme gradient boosting (XGBoost), random forest (RF), K-nearest neighbours (KNN), light gradient boosting machine (LightGBM), gradient boosting (GB), adaptive boosting (AdaBoost), Ridge, and Lasso, were utilized to improve the prediction accuracy of crack width in GFRP-RC beams.