In seismic vulnerability and risk assessment applications, nonlinear static (pushover) analyses are often deemed as arguably the best compromise between accuracy and simplicity, as well as more suitable for engineering practice when compared to possibly more accurate, but certainly more complex and time-consuming, nonlinear dynamic (time-history) analyses. In these approaches, the seismic analysis of the structure is generally performed by comparing the capacity (i.e., the force-displacement capacity curve) and the seismic demand within a Acceleration Displacement Response Spectrum (ADRS) domain. Different methods are available in the literature (e.g., Capacity Spectrum Method, N2 method, Displacement Coefficient Method) and adopted in the main international seismic codes/guidelines. Among others, the recent Italian Building Code (NTC 2018) represents a particular case, as two alternative code-compliant spectrum-based methods are presented as alternative options for the seismic assessment of an existing building. However they can potentially lead to different results even when considering the same (pushover) capacity curve and seismic hazard. Yet, specific provisions/guidelines regarding the recommended approach to adopt for seismic response analyses in case of different expected seismic behaviour (in terms of ductility, stiffness and strength capacity) are still missing.    

Therefore, this paper investigates the main differences in evaluating the seismic performance of buildings according to different code-compliant approaches, with a specific focus on the two methods reported in the Italian Building Code. Specifically, the two alternative approaches, namely "Method A" and "Method B" (more specifically referring to the N2 Method and the Capacity Spectrum Method, CSM, respectively) are considered. Both methods allow assessing the performance point of the structure, considering the inelastic spectra ("Method A" / N2 method) or the elastic spectra with equivalent damping ("Method B" / CSM). An extensive parametric analysis is thus carried out by performing several non-linear static analyses on Multi-Degree of Freedom (MDoF) models of different Reinforced Concrete (RC) frame structures. A refined two-dimensional (2D) lumped plasticity model is implemented in the structural software Ruaumoko. The case-study structures are selected considering different geometric configurations (i.e., varying the number of storeys and spans), construction details and material properties. This allows considering a wide range of pushover curves, characterized by different strength/stiffness/ductility capacities, and different global inelastic mechanisms expected. Seismic assessment is then performed by applying the two spectrum-based methods, following the code provisions. Results are compared in terms of performance points (force and displacement/drift/ductility demand), safety index (IS-V) and economic Index EAL (Expected annual losses, or PAM, Perdita Annau Media). The influence of the adopted bilinearization methods is also investigated. Results highlight that stiffness and ductility capacity of the structure strongly affect the seismic performance assessment obtained by applying the two alternative methods, leading to higher differences when stiffer and more ductile structures are considered. Results presented in this work could be considered as a preliminary step to support the development of specific guidelines including provisions on the recommended simplified approach to adopt for seismic assessment of buildings (based on the observed/expected seismic behaviour) in practical applications.