Civil Engineering Reference
In-Depth Information
8
7
6
5
4
1
5
10
50
100
500
Closest distance to fault (km)
2.2
Magnitude-distance distribution of the database used for the
development of the Eastern North American GMPE of Atkinson &
Boore (2006).
in stable continental regions are more diffi cult to predict coupled with the
fact that the average inter-station spacing tends to be greater for stable
regions. Hence, even when a moderate-to-large magnitude event does
occur, it is relatively unlikely that a set of high-quality accelerograms will
be recorded at distances of primary engineering interest.
Because the strong-motion datasets are insuffi cient to constrain empiri-
cally derived ground-motion prediction equations in stable regions, the
approach taken for active crustal regions cannot be adopted. Instead, one
must look to make use of the data that exists from the relatively small
magnitude events in order to constrain a physically motivated model for
the scaling of spectral amplitudes. While there are detailed seismological
models that could be employed for this purpose, for engineering applica-
tions the current approach is to implement simplifi ed spectral representa-
tions of the earthquake process and wave propagation. These spectral
models, more commonly referred to as stochastic-based models (Boore,
2003), are derived by fi rst defi ning a model for the Fourier amplitude spec-
trum and then coupling this spectral representation with random vibration
theory to either directly estimate the peak response of a damped oscillator,
or to simulate time-series from which the peak can be computed.
An important point to note is that these spectral representations relate
to far-fi eld shear waves being radiated from a point source only. As a result,
the scaling that is commonly implied by these models for large magnitude
Search WWH ::
Custom Search