Damage-State Threshold Definitions in Analytical Fragility Assessment of Non-Ductile and Code-Deficient RC Buildings: A Systematic Review of Framework Origins, Inter-Study Divergence, and Transferability

Abstract

Damage-state thresholds are the primary numerical input governing fragility curve position, yet their definition receives substantially less methodological scrutiny than ground-motion selection or structural modelling in seismic fragility assessment. This review systematically examines how damage-state thresholds are sourced, applied, and justified in analytical fragility studies of non-ductile and code-deficient reinforced concrete buildings. Five major threshold frameworks, HAZUS, EMS-98, ASCE 41, Eurocode 8 Part 3, and FEMA P-58, are classified by calibration basis and embedded structural assumptions; none was developed for a non-ductile, gravity-load-designed, or pre-code population as its primary target. Reported values for nominally equivalent damage states diverge substantially across structural classes, with the spread appearing to increase as damage severity increases. Transfer of thresholds from an external framework occurs in the majority of reviewed studies, yet full structural justification accompanies fewer than a third of these transfers. Controlled comparisons in which authorship and archetype basis are held constant demonstrate that divergence of comparable magnitude can arise from methodological choice alone. Detailing quality, governing failure mechanism, and demand-parameter compatibility are identified as the variables that determine whether a transferred threshold retains its physical validity. The evidence establishes that damage-state threshold definition functions as a primary determinant of fragility predictions and that current practice does not treat it accordingly, with direct consequences for seismic risk assessment of the non-ductile building stock that dominates seismically active regions worldwide.