Strength Model for FRP-Confined Concrete: Comprehensive Database and Reliability-Based Partial Factor Calibration

Despite several formulations available to predict the compressive strength of fiber-reinforced polymer (FRP)–confined concrete columns, predictive models for cases with a rectangular cross section, high/ultrahigh-strength concrete, or discontinuous confinement systems are still under development. Furthermore, due to the lack of a systematic approach to data gathering and model recalibration, the reliability of these models remains limited and is validated only in certain cases covered by their employed data set. This study addresses these limitations by compiling a comprehensive database of 4,639 experimental results and presenting a recalibrated version of a previously developed unified strength model. The revised model incorporates key influencing parameters such as unconfined concrete strength, cross-sectional shape, confinement continuity, and size effects, leading to superior predictive performance across a wide range of scenarios. The comparative evaluation of the proposed strength model with existing ones evidence its superior performance according to the obtained statistical metrics of prediction-to-test value ratios. A reliability-based design framework was implemented to account for uncertainties in material, geometry, and load. A partial safety factor was calibrated through a least-squares optimization process to ensure alignment with target reliability indices under diverse design conditions. This integration of reliability into the predictive model supports safe and cost-efficient structural design, providing a robust tool for the design of FRP-confined concrete columns.