Assessing the potential of two methods for rapid first-order damage estimation using structural monitoring data

The increasing availability of structural seismic monitoring networks has significantly improved seismic monitoring capabilities within buildings, allowing for efficient structural response analysis and rapid damage estimation. However, such networks remain limited in number and spatial coverage, highlighting the need for approaches that rely on surface recordings to estimate the dynamic behavior of buildings using simple models, that allows to extend the estimation of building shaking for different building typologies over a reasonable area around the recording site.This study investigates two methods, the Z-transform and the Thomson-Haskell approach, for predicting top floor seismic motions by using a recording at the base of the structure as seismic input. The Z-transform applies a recursive filter based on seismic input and structural dynamic characteristics to predict drift and top floor motion, while the Thomson-Haskell method relies on physical parameters such as shear-wave velocity to compute a transfer function, later used to estimate top floor motion and drift under the assumption of a shear beam behaviour of the building. Model predictions are expressed as peak drift and peak top floor displacement, which are then used for damage assessment by using pre-defined thresholds for different damage states.The two approaches are validated at two test sites, a four-story building in Tolmezzo and a two-story building in Visso, Italy, where real earthquake records at the base and at the top of the structure are available. For the Tolmezzo building, the results of the simulations align well with observations of the top floor motions, though they vary depending on the choice of input parameter estimation. In the Visso building, both methods also show consistent performance, slightly overestimating the observation for Y-direction. A Sensitivity analysis carried out for the input parameters considering the Visso building shows that the results of the methods are highly susceptible to the damping factor choice, and to trade off between the damping ratio and fundamental frequency.To explore the applicability and scalability of the approach, a scenario-based test case is presented for the Municipality of Tolmezzo, where the building stock is characterized using empirical height-period relationships and 1D shear-wave resonance formulae. Results demonstrate that both methods are good candidates for estimating damage states.