This study investigates the effects of structural filtering on floor motions and floor response spectra (FRS) used in the seismic design of nonstructural components. While seismic design guidelines for buildings are well-established, nonstructural components often experience disproportionate damage due to amplified floor motions rather than direct ground shaking. To analyze this phenomenon, linear elastic time-history analyses were conducted on two types of school buildings: a regular single-story building and an irregular four-story building, both with rooftop structures. The study compares the frequency characteristics of ground and floor motions using Fourier Amplitude Spectra (FAS). The results show that buildings act as resonant filters, selectively amplifying harmonic components that are close to their natural frequencies. Additionally, the study evaluates the FRS by comparing numerical results with the displacement response factor and the ASCE 7-16 dynamic amplification factor (DAF). It was found that floor motions converge towards harmonic oscillations predominantly influenced by the building’s fundamental period, resulting in significant peak responses in the FRS. These findings suggest that the current design codes may require refinement to better account for these filtering effects, particularly for nonstructural components and structures that are sensitive to resonance.
This study proposes a numerical modeling approach to estimate the monotonic backbone curve of reinforced concrete (RC) columns based on cyclic loading tests. Backbone curves derived from cyclic loading tests tend to underestimate deformation capacity, leading to inaccurate predictions of post-peak behavior. To address this limitation, the IMKPinching model is adopted, with post-peak deformation capacity treated as a key parameter. A procedure is proposed to estimate the parameters of a monotonic backbone curve from cyclic data. The proposed approach was validated by comparing with monotonic test results, demonstrating a significant improvement in post-peak prediction accuracy.
In this study, we present a probabilistic simulation framework that combines a Pseudo-Dynamic Rupture Model with SPECFEM3D to estimate ground motions in areas with limited seismic observations. We generate multiple rupture scenarios that reflect the statistical characteristics of fault parameters, and we integrate these scenarios with a 3D velocity structure model for numerical wave propagation simulations. The numerical mesh is refined to accurately simulate ground motions in the frequency range of 0.1-3 Hz. Despite the limited resolution of currently available 3D velocity models, our numerical simulation results capture key features of ground motion, including the effects of fault rupture complexity, to some extent. We apply this method to the 2016 Gyeongju earthquake (Mw 5.4) and compute the probability density function of Peak Ground Velocity (PGV) at selected sites, allowing us to quantitatively assess the uncertainties associated with fault rupture.
The purpose of this study is to identify the optimal seismic Intensity Measures (IMs) that best reflect the seismic response characteristics of a submerged structure used as a data center. Structural modeling and analysis were performed using ANSYS, and a Nonlinear Time History Analysis (NTHA) was conducted with 84 ground motion records from the PEER database. The maximum acceleration and displacement responses at the top of the structure were extracted as engineering demand parameters. Correlation analyses with 23 IMs revealed that Peak Ground Displacement (PGD) exhibits the highest correlation with the peak displacement response. At the same time, spectral velocity at the fundamental period (Sv(T1)) correlates most strongly with the peak acceleration response. These results indicate that the PGD and Sv(T1) are suitable IMs for evaluating the seismic performance of submerged structures.
Response spectra used for seismic response evaluation are generally characterized by smooth shapes derived from ground motion prediction equations as a function of earthquake magnitude and distance. However, recorded earthquake ground motions exhibit peak and valley variations in response spectra, and this study investigated the effects of these characteristics on the nonlinear response of structures. To define the peak and valley variations in the spectrum, a smooth reference spectrum is required. For this purpose, a reference spectrum based on the design spectrum shape was derived for each ground motion spectrum. The variability of the structural response was evaluated for smooth input response spectra and response spectra with peak and valley characteristics, while varying the natural period and ductile behavior of the structure. The results showed that the peak and valley characteristics amplify the variability of the nonlinear response of structures. In addition, this variability was found to differ depending on whether the analysis was performed using the static capacity spectrum method or nonlinear dynamic analysis. Based on the results, the effects of the spectral peak and valley on the nonlinear response were discussed according to the analysis method.