The corrosion of steel reinforcement can have detrimental effects on the long-term structural performance and functionality of reinforced concrete (RC) infrastructure. Non-metallic reinforcement materials, such as Glass Fibre-Reinforced Polymer (GFRP) bars, have proved a promising alternative to steel in aggressive environments due to their non-corrosive nature and high strength. Despite recent advancements in research, applications of the bars are currently limited to non-seismic zones and members under flexure and shear, such as bridge decks and slabs. This PhD study explores the feasibility of GFRP as a primary reinforcement in compression members subject to seismic loading. To achieve this, this research focuses on the behaviour of GFRP reinforcement under compression or cyclic loading, before characterising the response of bridge piers reinforced with GFRP transverse and/or longitudinal reinforcement.

The monotonic compression and cyclic tension-compression behaviour of GFRP bars, including the effect of buckling, was experimentally investigated using a modified RILEM beam test. For bars with a low slenderness ratio (< 5), the compressive strength corresponded to 81%, 90%, and 91% of the tensile strength for 12 mm, 16 mm, and 18 mm diameter bars, respectively. As the slenderness ratio increased to 8, 11, and 14, a transition in failure mode between crushing, buckling, and a combination of both was observed. This was accompanied by a 20%, 51%, and 64% average reduction in strength, respectively. Cyclic tension-compression loading led to a reduction in strength for stocky bars that were subject to several high-strain reversals. In contrast, cyclic loading had little effect on bars with a high slenderness ratio (> 11) where buckling occurred.

An existing analytical model for the buckling stability of steel reinforcement was modified to accommodate the unique confining behaviour of GFRP transverse reinforcement when restraining steel or GFRP longitudinal bars. The modified models showed good agreement with experimental observations from literature. Equations were proposed for the maximum buckling length to limit strength degradation and minimum GFRP transverse reinforcement diameter to restrict buckling to a single spacing. Comparisons with design guidelines showed current confinement provisions do not always satisfy anti-buckling requirements for GFRP transverse reinforcement.

The cyclic response of one-third-scale, concrete bridge piers reinforced with various configurations of steel or GFRP bars was experimentally investigated. Each specimen was tested under a constant axial gravity load, combined with reversed quasistatic cyclic lateral loading. The behaviour of the piers was assessed and compared in terms of strength, stiffness, deformability, energy dissipation, and damage. Using a hybrid combination of steel longitudinal bars and GFRP spirals offered a comparable cyclic response to conventional steel-RC, with the hybrid-RC piers achieving similar or enhanced levels of ductility and energy dissipation. Piers reinforced with GFRP longitudinal and transverse bars exhibited a stable and deformable cyclic response with little strength degradation and residual displacement. However, this was accompanied by a lower initial stiffness and a reduction in hysteretic energy dissipation.

A theoretical analysis was presented to capture the lateral strength and deformation response of GFRP-RC columns under combined axial and flexural loads. Parametric section analysis and existing integration techniques were used to develop simplified expressions for predicting the idealised bi-linear deformation curves for use in displacement-based design procedures. The equivalent yield curvature was found to depend on the dimension of the section, the yield strain of the concrete, and the axial load ratio. This parameter could be predicted within 15% error for any typical circular or rectangular GFRP-RC section. For the post-spalling branch, the ultimate lateral capacity was generally overestimated, whilst the ultimate deformation was conservative when compared to experimental data from this study and existing literature. Further research is required to refine confinement models, improve predictions of the ultimate rotation capacity of sections confined by GFRP transverse reinforcement, and capture strain penetration effects for GFRP longitudinal bars.

The Direct Displacement-Based Design (DDBD) procedure was modified to accommodate the unique response of RC members reinforced entirely with GFRP bars. Three key design parameters were developed to achieve this: a simplified function for the equivalent yield displacement, logical limit states and formulation for the design displacement, and an estimation of the equivalent viscous damping. The proposed procedure was presented in form of a design example for a case-study RC bridge pier, and verified using non-linear time history analysis (NLTHA).