Seismic Analysis of Structures: Improvements Due to Probabilistic Concepts - Dynamic Behavior of Concrete and Seismic Engineering - page 274

the structures, is presented in a quite assertive deterministic light, through the
seismic source data, through the whole approach for obtaining the response and the
conclusion as far as strength or collapse are concerned.
The current chapter aims to show that this deterministic approach can sometimes
lead to quite misleading results and that, in order to control things and to improve
the predictions, it is necessary to become aware of the probabilistic feature of the
problem raised and of the more or less justified hypotheses made. Thus, we are led
to ask ourselves questions such as:
- What model should be chosen to represent the motion of the soil?
- Is the conventionally-used “oscillator response spectrum” data still relevant?
- How can we preserve the simple feature of the new methods proposed while
keeping them consistent with regulations, especially in the case of non-linear
behavior?
8.2. The modal method
For further details see [DAV 85], [DAV 88] and [GIB 88].
8.2.1. Data about the seismic source
Open field motion is a rather complex acceleration time signal (Figure 8.1)
characterized by a level variable, the J smax maximum acceleration and frequency
content.
In the deterministic approach one or several earthquake levels (operation and
safety levels, for example) corresponding to different structure strength requirements
are associated with a site.
The frequency content results from various compilations and averages of
measured seismic signals. As we do not usually have any recordings of the
concerned site, we often just use general standardized data such as the “Regulatory
Guide 1-60” spectra (Figure 8.2), possibly corrected with the “site effects”.