Earthquake Engineering & Structural Dynamics最新文献

Accumulated debris may block post-earthquake roads and severely impact evacuation, emergency and recovery operations. It is of great importance to investigate the distribution laws of building debris. In this study, a probabilistic prediction model for post-earthquake debris extent and road blockage fragility estimation, conditioned on the intensity measurement of earthquakes, is presented. The main contributions of this paper are (1) highlighting the random distribution characteristics for post-earthquake debris extent including maximum debris width, debris length and debris area; (2) establishing an unbiased probabilistic prediction model for debris width and length based on the Bayes theorem with accumulated blocks from both the out-of-plane failure of infill walls and collapsed buildings; (3) developing a blockage fragility estimation method for post-earthquake road networks versus the intensities of earthquakes with Monte-Carlo simulation; and (4) achieving better applicability and less uncertainty for this presented model than those established with satellite images after an earthquake event. Finally, four buildings constructed as reinforced concrete frame (RCF) structures are used to generate a virtual debris extent data pool for developing the prediction model. The numerical results indicate considerable variability in the debris extent, with a coefficient of variation (COV) ranging from 3.0% to 18%. Earthquake intensity significantly affects the debris extent. Furthermore, the scenarios in which the road centerline is open or not open substantially affect the conditional blockage probability of the post-earthquake road network. This method is beneficial for designing rescue routes and evacuation plans for earthquake events.

Self-centering devices have been shown to be effective in reducing residual story drift after an earthquake. However, they typically result in high floor acceleration demand, especially in applications with rocking core systems where the stiff rocking core enhances the stiffness for higher-mode and induces acceleration amplification. To address this limitation, this study proposed a novel two-stage self-centering brace (TSSCB), providing an available alternative for balancing the trade-off between floor acceleration and post-earthquake residual drift. The TSSCB features a two-stage mechanism controlled by a gap configuration, enabling energy dissipation through slip friction under low-intensity earthquakes and reducing post-earthquake residual story drift using shape memory alloy (SMA) cables under high-intensity earthquakes. The paper first introduces the detailed configuration and working principles of the TSSCB. Cyclic tests on eight TSSCB specimens demonstrate stable hysteretic behavior, where the device behaves similarly to traditional friction dissipators under small displacements and exhibits delayed flag-shaped hysteretic loops under large displacements. The effects of SMA cable pre-tension levels and loading protocols on cyclic behavior are discussed, and the low-cycle-fatigue performance of the TSSCB is evaluated. A phenomenological hysteretic model is developed to describe the cyclic behavior of the TSSCB, followed by nonlinear time history analyses of five archetype rocking core-moment frame systems to validate the TSSCB's application. The results indicate that incorporating an appropriate value of gap into the TSSCBs can reduce floor acceleration without significantly increasing residual inter-story drift, highlighting the TSSCB's advantage for multi-objective seismic control and resilience enhancement of buildings.

Numerous case histories have demonstrated that earthquake-induced landslides can cause significant damage to infrastructure and pose serious threats to human lives. Therefore, realistic prediction of coseismic landslides is essential for effective disaster mitigation. The multiscale SEM-MPM method is a physics-based numerical tool for an integrated analysis of coseismic landslides. In this study, we use the SEM-MPM method to analyze the Aso Bridge landslide triggered by the M_w 7.0 Kumamoto earthquake. The Aso Bridge landslide is located on the hanging-wall side, with a rupture distance of less than 5 km, as one of the longest run-out coseismic landslides. Fault rupture, regional-wave propagation over complex topography, and local landslide failure and post-failure behaviors of the slope are simulated. The simulation demonstrates that areas around the Aso Bridge landslide experience high-intensity ground motions, resulting from the rupture directivity effect and topographic amplification over convex terrain. Moreover, significant velocity pulses in east-west and north-south directions are observed from the SEM-MPM simulation, which may cause a long run-out distance of the landslide. Finally, the simulation reproduces the Aso Bridge landslide in terms of run-out distance, size, and location, thereby verifying the accuracy of the SEM-MPM model.

This paper addresses the definition of the vertical seismic action for design, within the progress studies for the next generation of the European and Italian seismic norms, partly still in progress. Besides presenting the background of the vertical design spectra according to the second generation of the Eurocode 8, a period-dependent factor for the Vertical-to-Horizonal (V/H) response spectral ratios is introduced, following the findings from recent empirical studies. The proposed V/H factor is expressed in a simple format as a function of relatively few parameters (i.e., seismicity level and site conditions), which are typically adopted in a normative framework, being, at the same time, suitable to identify the main physical factors affecting the V/H ratios. This factor is found to explain reasonably well the results of probabilistic seismic hazard analyses for both H and V components conducted for selected test sites in Italy, with different approaches to account for the V component. A favorable comparison with probabilistic results is also obtained for the V/H spectral ratios according to the new Eurocode 8.

Seismic faulting can cause severe damage to buried ductile iron (DI) pipelines. Existing studies consider the scenarios of fault-pipe crossing at a pipe joint or the middle of a pipe segment only. Moreover, there is a lack of analytical assessment methods for maximum joint rotation under normal faults. In this study, full-scale experiments are conducted to analyse the influence of fault-pipe crossing position on the behaviour of DI pipelines under 90° normal faulting. New tape optical fibres and linear potentiometers under an improved instrumentation scheme are employed to monitor the pipe strains and joint kinematics. The results indicate that as the fault-pipe crossing position, rp, shifts from the joint (rp = 0) to three-quarters of the pipe barrel length (rp = 0.75), the pipe-soil interaction mechanism changes from the uplift capacity-controlled mode to the bearing capacity-controlled mode. In terms of joint kinematics (maximum joint rotation and axial displacement) and maximum bending strain, the most unfavourable fault-pipe crossing positions are rp = 0∼0.25 and 0.75, respectively. The joint rotation near the fault trace should be prioritized to avoid the loss of DI pipe serviceability. An analytical model that enables safe evaluation of peak joint rotation under normal faulting is proposed.

The Newmark sliding-block displacement (D) has been widely used to evaluate the seismic performance of slopes in both site-specific and regional scales. Current predictive models for D of slopes are generally developed without consideration of the specific slope orientation. This study presents the directionality analysis of preferential polarization for D and a model to predict the D considering the specific slope orientation in strike-slip earthquakes. Based on a database of 817 ground motion records with strike-slip faulting, it is shown that the orientation of maximum D tends to occur close to the transverse orientation (perpendicular to the orientation at a given site pointing to the epicenter). There is a significantly larger probability of exceeding the median displacement (D₅₀) over all orientations in the transverse orientation. The orientational D is strongly correlated with the angular difference with respect to the transverse orientation. The orientation-dependent model is developed to estimate D at specific orientations by combining with an existing D₅₀ predictive relationship. The model enables the transformation of D₅₀ to D in any specific orientation, and hence allowing for the site-specific and regional landslide hazard assessment with consideration of specific slope orientation in strike-slip earthquakes.