Page 180 - Dynamic Loading and Design of Structures

Page 155

the base shear used in equivalent static analysis (eqns 4.27, 4.28) in the case of regular
structures, or 100 per cent this base shear in the case of irregular structures.
Modal combinations, torsional effects, and orthogonal (x and y) effects are treated in the same
fashion as in EC8.

4.3.7 Time history representations

Time history analysis is used for design purposes only as an exception (see Section 4.3.1), and
almost exclusively whenever non-linear effects are to be considered explicitly, rather than
through the R-factor approach. When acceleration time histories are used for design, it is
imperative that they actually correspond to the design earthquake for the site under
consideration, which means that the envelope of the response spectra of the accelerograms
used should reasonably match the elastic design spectrum for the site (no reduction through R-
factors).
Several options are available for selecting an appropriate set of design accelerograms:

●use of records from actual earthquakes, which generally have to be scaled to the design
earthquake intensity;
●use of artificial accelerograms generated so as to match the (target) elastic response
spectrum; this is sometimes referred to as the ‘engineering method’;
●use of simulated accelerograms generated by modelling the source and travel path
mechanisms of the design earthquake (‘seismological method’).

Each option has its own merits and limitations, as discussed in the following.

Selection of recorded accelerograms
This can be the ideal solution whenever an extensive database of acceleration time histories is
available, containing records from earthquakes with a large range of characteristics. Then, a
selection can be made of records matching the source parameters (focal mechanism and depth,
distance from source), travel path, magnitude, peak ground motion parameters (A, V, D), and
duration, for the site under consideration. Note that, with the possible exception of major
projects (such as the design of critical facilities) in areas where abundant data exist, such as
the US and Japan, the foregoing is a rather over-ambitious procedure, since not only an
adequate database of records is required, but also a complete charactema-tion of the seismic
hazard at the site.
A more pragmatic approach would involve the following main parameters to be considered
when selecting natural records:
●site conditions;
●magnitude;