Page 187 - Dynamic Loading and Design of Structures

Page 162

‘life safety’ with a 10 per cent/50 year probability earthquake (the usual ‘design’ earthquake
in current codes), and ‘collapse prevention’ with an earthquake having a 2 per cent/50 year
probability of exceedance.
Explicit checks of performance can also be made at local level (i.e. for each member in the
structure), in which case appropriate limits for local deformation quantities, such as rotations
of plastic hinge zones are required; tabulated values of such quantities for various types of
members can be found in FEMA (1997b), but it has to be stressed that these issues are still the
subject of current research. Local ductility of members (e.g. plastic rotation capacity of R/C or
steel beams and columns) is dependent on appropriate design and detailing, but also on
quality control, particularly at the construction phase. Moreover, exceeding the available
ductility capacity at one or even a few critical regions does not necessarily mean (incipient)
collapse of the structure, particularly when the latter is characterized by high redundancy and
ability to redistribute loading.
Recent approaches, such as those briefly outlined previously, are essentially deterministic
procedures, since uncertainty is explicitly accounted for only in the case of the seismic input
(spectral accelerations are adjusted to the target probability of exceedance selected for the
performance level that is being checked). However, the real issue is the reliability (or the
probability of failure) of the structure when subjected to a particular earthquake. This is only
marginally addressed in EC8, where it is stated that target reliabilities for the ‘no collapse
requirement’ and the ‘damage limitation requirement’ should be established by national
authorities for different types of structures, on the basis of the consequences of failure.
Unfortunately, these target reliabilities are not given, even as ‘boxed’ (i.e. indicative) values
in EC8 or other codes. In a recent study (Wen et al. 1996) addressing this issue, suggested
target 50-year probabilities range from 30 per cent to 50 per cent for the SLS and from 4 per
cent to 6 per cent for the ULS; both values refer to a specific performance criterion, i.e.
exceeding a drift limit.

4.4.2 Configuration issues
The selection of the configuration of the structure (i.e. of the arrangement of the structural
system as well as of the non-structural elements and their connection to the former) is
arguably the most critical step in the seismic design procedure. It is clear from the distribution
of damage in earthquake struck regions that structures with a reasonably regular and
symmetric configuration perform consistently better than structures with irregular
configuration (Arnold and Reitherman, 1982). An irregular configuration is characterized by
one or more of the following problems:

●The plan of the building includes large re-entrant angles (L-shaped, C-shaped, H-shaped
plans; see Figure 4.16a).
●The distribution in plan of stiff members, such as walls, is not symmetric with respect to the
mass centre (M in Figure 4.16b).