Dynamic Response and Ground-Motion Effects of Building Clusters During Large Earthquakes

The objective of this study is to analyze the response of building clusters during earthquakes, the effect that they have on the ground motion, and how individual buildings interact with the surrounding soil and with each other. We conduct a series of large-scale, physics-based simulations that synthesize the earthquake source and the response of entire building inventories. The configuration of the clusters, defined by the total number of buildings, their number of stories, dynamic properties, and spatial distribution and separation, is varied for each simulation. In order to perform these simulations efficiently while recurrently modifying these characteristics without redoing the entire "source to building structure" simulation every time, we use the Domain Reduction Method (DRM). The DRM is a modular two-step finite-element methodology for modeling wave propagation problems in regions with localized features. It allows one to store and reuse the background motion excitation of subdomains without loss of information. Buildings are included in the second step of the DRM. Each building is represented by a block model composed of additional finite-elements in full contact with the ground. These models are adjusted to emulate the general geometric and dynamic properties of real buildings. We conduct our study in the greater Los Angeles basin, using the main shock of the 1994 Northridge earthquake for frequencies up to 5Hz. In the first step of the DRM we use a domain of 82 km x 82 km x 41 km. Then, for the second step, we use a smaller sub-domain of 5.12 km x 5.12 km x 1.28 km, with the buildings. The results suggest that site-city interaction effects are more prominent for building clusters in soft-soil areas. These effects consist in changes in the amplitude of the ground motion and dynamic response of the buildings. The simulations are done using Hercules, the parallel octree-based finite-element earthquake simulator developed by the Quake Group at Carnegie Mellon University.