Civil Engineering Reference
In-Depth Information
Many factors need to be considered to completely solve a soil-structure
interaction problem. They include: the definition of the seismic unknown factor and
the resulting motion, soil behavior under cyclic stress, an estimate of soil response in
the open field and the response of structures under dynamic loading. In this chapter,
only those aspects dealing with the behavior of soils under cyclic loading will be
tackled, together with soil-structure interaction modeling.
The extensive literature dealing with the soil-structure interaction reflects both
the complexity of the phenomenon as well as the interest in the subject shown by the
scientific community. Two general syntheses have proposed a classification of the
methods used to take soil-structure into account ([LYS 78] and [IDR 80]). Both
publications stressed at the time that the study of interaction phenomena was
essentially limited to cases involving linear problems, approaches for non-linear
problems still being rare. However, in the last few years, important advances have
been made as far as our understanding of non-linear phenomena is concerned, which
makes broaching these problems possible.
4.1. Behavior of soils under seismic loading
4.1.1.
Influence of the nature of soils on seismic movements
Observations made on sites during real earthquakes clearly show the influence of
the nature of surface layers on the seismic motion recorded. This fact has
been acknowledged for about 20 years and has led to response spectra being
interpreted according to the nature of the soil.
The recent major earthquakes that have affected the world (Mexico, 1985; Loma
Prieta, 1989; Northridge, 1994; Kobe, 1995) confirmed the following facts: alluvial
soils tend to amplify incident motion, especially at low frequency. As an illustration,
recordings of the Loma Prieta earthquake obtained on rocky sites around San
Francisco showed a maximum acceleration of about 0.10 g; recordings of the same
earthquake made on alluvial sites showed maximum accelerations two to three times
as high (Table 4.1 [COL 90]), with spectra presenting important low frequency
peaks. As the epicenter was far from the recording sites (about 85 km), this could
not be a local effect due to the source (directivity). By the same token, as the
recording sites were near to one another (within a radius of a few kilometers), it
could not be an effect due to propagation between the source and the site inside the
Earth's crust. The only parameter that could have affected the nature of the recorded
motion was the geological nature of the sites, i.e. the mechanical characteristics of
the soil near the surface.
Search WWH ::
Custom Search