Civil Engineering Reference
In-Depth Information
REMARK 8.1.- The coherence of the method is quite obvious. An earthquake is
defined due to the response maxima of oscillators, with the maxima being precisely
the values that enable us to rule about the strength of the structure. As a matter of
fact, this implies that collapse is determined by a threshold that is exceeded once
(collapse from excessive strain and not from oligo-cyclic fatigue).
8.3. Criticism of the modal method
The drawbacks of the modal method are numerous and include:
- the simple quadratic combination hypothesis cannot always be verified;
- knowing the response maxima is not always sufficient. For example,
calculating a complex installation must be performed in several steps. The
calculation for buildings is effected, with the main elements represented through
their mass. To calculate the equipment, the motions of their anchoring points have to
be characterized. This is called the
floor spectrum
problem;
- collapse can be caused by combination of several cycles, which therefore
involve several relative maxima (oligo-cyclic fatigue, cumulative plastic strain);
- if we want to take a highly non-linear behavior into account, the modal method
is ineffective.
Therefore, other approaches have to be used. For instance, we might think of
resorting to time calculations. Carrying out shaking table tests implies a time
approach as well. The obvious question that must then be asked deals with
associating a temporal seismic signal with the ORS data.
In order to bring solutions to this set of questions, it is necessary to revert to the
definition of a seismic signal.
Observation of the seismic motions on a same site shows noticeable time rate
differences from one signal to the next. Such differences result from the complex
phenomena involved, and it would be mistaken to believe that it is possible to give a
detailed explanation of them.
The only solution involves representing the seismic motion of a site through a
random process that will be characterized by means of average values. The
previously defined ORS can be considered as such values. Are they relevant ones,
that is to say, is the ORS data enough to determine the average values associated
Search WWH ::
Custom Search