PSI - Issue 78

Marco Martino Rosso et al. / Procedia Structural Integrity 78 (2026) 301–308

302

1. Introduction Earthquakes are one of the main natural disaster threats to the built environment and human life safety. Since this phenomenon acts under the dynamic regime, seismic performance-based assessment methods require a reliable and accurate identification of the buildings intrinsic dynamic characteristics, i.e., natural frequencies, damping ratios, and mode shapes. Considering traditional and well-established approaches, modal characterization of buildings can be achieved using either Experimental Modal Analysis (EMA), where both input and output are measured, or Output Only Modal Analysis (OMA), where the input is assumed to be stochastic and broadband [2]. OMA is a more attractive suite of tools to achieve this because it could provide dynamic identification of structural dynamical properties by recording the structural response under natural ambient vibration under in-service operating conditions. In Rainieri (2008), OMA methods for seismic protection of structures are presented, also focusing on their short term impact due to natural hazards and their long-term effectiveness considering structural deterioration processes, and even for evaluating the effectiveness of manmade activities such as seismic retrofitting interventions. Indeed, structural damage typically results in reduced stiffness, thus affecting the modal data changes with expected natural frequency decrease. To perform this kind of assessment, modal parameter estimates are required for phenomenological and numerical model updating calibration procedures, especially in the current digital era, where high-fidelity smart digital twins of structures are sought. For instance, Gara et al (2021) adopted OMA techniques with finite element modeling (FEM) for assessing seismic retrofit of a school building. Combey et al (2025) adopted OMA to study seismic microtremors effects on a historical Colonial church in Peru, leveraging modal parameters also to infer site soil-structural interactions. Rainieri et al (2012) adopted OMA also for capturing spurious modal components of surrounding structures and environment, which can be useful for fast post-earthquake emergency support. Nowadays, another promising trend is the adoption of automatic OMA procedures denoted as AOMA integrated with continuous structural health monitoring (SHM) systems for a continuous evaluation of modal properties, i.e., providing a proper and continuously updated digital twin, even depurated by seasonal trends and with reduced false alarm rate in the standpoint of an effective early warning system. Moreover, artificial intelligence and machine learning supported AOMA strategies are nowadays fostered to provide regional-scale smart monitoring solutions, see e.g., Rosso et al (2023), Zhou & Li (2022). Laboratory testing provides controlled input conditions to characterize building frames and to benchmark identification procedures against known excitation. Therefore, OMA is typically adopted for global evaluation of dynamic properties and non-destructive techniques. For instance, in Liao (2025), OMA dynamic identification was performed before and after introducing damage, analyzing the rate of changes. Typically, in a laboratory environment, the structure can be excited directly using real or synthetic earthquake waveforms or using random signals (e.g., white noise), simulating in this latter case the main hypothesis of OMA input nature, and theoretically avoiding the use of force transducer measurement devices or precisely knowing the input. For instance, in Astroza et al., a full-scale five-story reinforced concrete (RC) building was tested on shaking table with six earthquake motions that progressively damaged the structure, and using AOMA to assess modal parameters and providing statistical analysis on them. In this study, two RC three-story frame buildings placed onto the same unidirectional shaking table were tested under an alternate sequence of scaled earthquake motions and white noise time histories, inducing progressive accumulated damage over the structure. The main goal is assessing the effectiveness of the retrofitting system implemented on the building 2, comparing the variation of their modal parameters with respect to the un-retrofitted one. The preliminary results herein reported are referred to the OMA conducted over the white noise case studies, evidencing an increase of structural damage, i.e., the stiffness reduction, testified by the overall reduction of global modal parameters. In Section 2, a description of the case study under investigation is provided, whilst in Section 3, the preliminary results obtained with OMA tools are reported and validated by comparing them with a FEM model of the case study. 2. Shaking table laboratory tests description In current laboratory tests, two three-story reinforced concrete (RC) frame buildings with perimetral brick infill were tested. Each frame has a rectangular shape plan with dimensions of 5.00 m x 2.00 m, and floor levels are