Development of Building-Specific Approach to City Seismic Response Analysis for Metro Manila

Abstract

Abstract – Information on the expected variability of maximum responses of buildings in cities of Metro Manila in the event of a strong earthquake can aid in disaster preparedness. A way to obtain such information is to combine the responses of individual buildings as influenced by different material properties and input ground motion. This study aims to develop a building-specific approach to city seismic response analysis for Metro Manila. Tools were developed for the tasks of processing feature data from GIS datasets, generation of building-specific input ground motion, automated MDOF model generation and analysis, and postprocessing of results. Input ground motion for each model was generated using the parameters, earthquake magnitude, distance to epicenter, and site/soil conditions. As an application example, a scenario earthquake analysis was conducted for five cities in Metro Manila considering C1-L, C1-M and C4-H building types (a total of 264,625 models). Results show that C1-M types for the city with Site Class D and located nearest to the fault obtained the highest mean and standard deviation of maximum story drift. Comparing the five cities, cities situated in Site Class D obtained higher mean of maximum story drift in all building types than in cities situated in Site Class C. The visualization of spatial distribution of buildings with varying story drifts allows for verifying the derived statistical information, as well as for direct comparison of the response of cities. Analysis of computation costs shows that for this application example, 6.5 hours of runtime (if running in single processor), 650 GB disk usage, and 50 GB of memory were required to simulate a scenario for a total duration of 18.4 seconds. These figures were estimated to be only 18% of the total cost if the whole Metro Manila is to be analyzed for one scenario earthquake using the building-specific approach.

Keywords—building-specific approach, city seismic response analysis, large scale computing