Due to factors such as imperfect construction techniques and inadequate maintenance, prestressed concrete structures exposed to chloride-rich environments for extended periods are highly susceptible to prestressing tendon corrosion. Prestressing tendons operate under high stress and possess small cross-sectional areas; consequently, corrosion often occurs without warning, posing a greater risk than in conventional reinforced concrete structures. The structural resistance of prestressed concrete members deteriorates progressively as tendon corrosion advances, thereby reducing the reliability of their safe service. Since structural resistance, loads, and environmental effects are all highly time-dependent, the reliability of prestressed concrete members in service also exhibits time-varying characteristics. Therefore, conducting in-depth research into the mechanisms and time-dependent patterns of resistance deterioration following tendon corrosion—and accurately assessing time-dependent reliability—is crucial for ensuring the safety of prestressed concrete structures during their service life. Integrating the fields of prestressed concrete durability and structural reliability, this study addresses the following aspects:

(1) Tensile tests were conducted to investigate the degradation patterns of mechanical properties (such as yield strength, ultimate strength, elastic modulus, and ultimate strain) of steel strands corroded in a chloride environment, and corresponding degradation models were proposed. Based on the corrosion characteristics of reinforcement in concrete under unidirectional chloride ingress, a localized corrosion model for the steel strand cross-section was established. Comparative analysis revealed that the flexural bearing capacity calculated using the localized corrosion model aligned well with measured values ​​from reinforced concrete beams subjected to unidirectional chloride ingress.

(2) The effects of corrosion degree, pull-out method, and the presence or absence of stirrups on the bond performance between corroded steel strands and concrete were analyzed. The study found that the average ultimate bond strength between the steel strand and concrete decreases as the corrosion rate of the strand increases. For specimens with stirrups, both the free-end slip and the ultimate bond strength were higher in eccentric pull-out specimens compared to concentric pull-out specimens. Compared to the non-corroded condition, stirrups are more effective at enhancing the bond strength between corroded steel strands and concrete. Based on the geometric and corrosion characteristics of the steel strands, a calculation model for the bond strength of corroded strands was proposed using the theory of thick-walled cylinders in elastic mechanics.

(3) The flexural performance of corroded prestressed concrete beams was analyzed, covering flexural failure modes, load-deflection curves, stress-strain relationships of concrete and corroded strands at the mid-span section, and the variation of flexural capacity with increasing strand corrosion levels. The study revealed that the beam’s flexural performance deteriorates significantly when the strand corrosion rate exceeds 2%. At a corrosion rate of 8.42%, wire breakage occurs, shifting the failure mode from ductile to brittle. Post-cracking stiffness decreases as the degree of strand corrosion increases. Corrosion accelerates the upward shift of the neutral axis, thereby reducing the height of the concrete compression zone. The ultimate load capacity of the strands at mid-span decreases as the corrosion rate rises. Both the cracking moment and ultimate moment of the test beams decrease with increasing corrosion of the prestressing tendons. Based on existing experimental data, a comparative study was conducted on the prediction of flexural capacity for corroded prestressed concrete beams according to Chinese, American, and European codes. Additionally, an empirical formula for the flexural capacity of corroded prestressed concrete beams was proposed based on the Chinese code, accounting for the effects of prestressing tendon corrosion.

(4) The influence of three different types of stirrups—HPB235, HRB335, and HRB400—on the shear performance of corroded reinforced concrete beams was analyzed. Beams with HRB335 stirrups exhibited the least sensitivity in shear capacity to stirrup corrosion, followed by those with HPB235 stirrups, while beams with HRB400 stirrups showed the highest sensitivity. An empirical model for the shear capacity of corroded reinforced concrete beams with web reinforcement was established by comparing 171 sets of experimental results (from domestic and international shear tests on such beams) with calculations from nine existing shear capacity models. Based on this, and considering the shear characteristics of prestressed members and the impact of prestressing tendon corrosion, a formula for calculating the shear capacity of corroded prestressed concrete beams was proposed.

(5) Based on experimental research and theoretical analysis regarding the mechanical degradation of corroded prestressing tendons and the deterioration of flexural and shear performance in corroded prestressed concrete beams, resistance degradation models were established for simply supported prestressed T-beams and box girders subjected to chloride corrosion. The higher-order moment method was employed to analyze the time-dependent reliability variations of flexural and shear capacities at the mid-span sections under ultimate limit states. The flexural reliability of the mid-span sections decreases as the corrosion level of longitudinal prestressing tendons increases; the rate of decline is fastest for localized corrosion, followed by pitting corrosion, while uniform corrosion results in the slowest decline. The rate and magnitude of the decline in time-dependent shear reliability at the mid-span sections are lower than those for time-dependent flexural reliability. When the probabilities of flexural and shear failure are comparable (within the same order of magnitude), the combined probability of failure (considering both modes) is significantly higher than the probability of failure when considering either flexure or shear in isolation.