A Time-Domain Method for Response-Spectrum-Compatible Ground Motions Using Autoregressive Models

Abstract eng:
Seismic codes require scaling the design ground motion to be compatible to the target spectrum for a specific region. Current methods of spectral matching are based on wavelet analysis and require a great deal of computations. This paper presents a simplified methodology of adjusting earthquake accelerograms to match target design spectra. A stochastic harmonic model is used to represent the ground motion as a stationary random process. To account for the non-stationarity of earthquake ground motions, the known ground motion is subdivided into a sequence of time windows to represent its temporal variation. Each window is idealized as an autoregressive (AR) model to establish its power spectrum, which is scaled according to the ratio of the spectral amplitude of the design spectrum to the spectrum of the unscaled ground motion. The harmonic model was then used to simulate the scaled ground motion, which was expanded for each window into a set of discrete frequency components and the approach is repeated for the sequence of windows representing the entire record. The response spectrum of the scaled record is then compared to the target spectrum and the scaling process is continued iteratively to achieve an optimal convergence. The procedure is illustrated by scaling a suite of ground motions that were selected from different regions of the world to design spectra representing eastern, central, and western United States.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket