Corrosion in tubular onshore and offshore structures poses significant safety and durability challenges, often necessitating retrofitting with materials like carbon fiber reinforced polymers (CFRP). Existing methods for assessing corrosion effects and repair efficiency often lack integration with nonlinear structural behaviors, such as plasticity and buckling, or rely on simplifying assumptions for seismic performance evaluation. This paper presents a new multiscale formulation based on Lumped Damage Mechanics (LDM) to model the coupled effects of corrosion, repair, and nonlinear structural response in complex trusses. The proposed model uniquely treats corrosion as a nodal degree of freedom, enabling direct imposition of measured corrosion states or driving forces, akin to mechanical loads or displacements, while accounting for damage progression. By leveraging state variables, the approach avoids assumptions about engineering demand parameter (EDP) distributions, instead using Monte Carlo analysis to derive fragility curves grounded in actual damage metrics. The method’s accuracy is validated against experimental data for a simple structure, and its practical utility is demonstrated through a case study on offshore tubular repair projects. Key contributions include: (1) a framework for quantifying earthquake resistance loss at observed corrosion states by comparing initial and degraded fragility curves, and (2) a systematic way to evaluate repair strategies by predicting their impact on seismic performance. The efficiency of the LDM model makes this high-fidelity analysis feasible for real-world applications.