Page 189 - Dynamic Loading and Design of Structures

Page 164

relative movement of the adjacent parts during the earthquake, hence they are generally wider
than standard construction (expansion) joints.
Whereas the advantages of regular configurations have relatively long been recognized,
quantification of regularity requirements is a critical issue that can not be deemed as been
resolved so far. An important consideration is that most of the irregular buildings damaged by
previous earthquakes were designed on the basis of rough methods and the level of detailing
required by the then applicable codes was quite low, while poor construction practices often
made it even lower. Experimental studies involving irregular structures such as frames with
setbacks designed and detailed to modern codes (Wood, 1992), tested on the shaking table,
have clearly indicated that irregular R/C structures can indeed perform adequately, even when
subjected to earthquakes significantly stronger than the one they were designed for. It has to
be pointed out, though, that the foregoing tests involved application of unidirectional
earthquake input, hence they did not address the problem of torsion.
Code criteria for regularity tend to be conservative but the consequences of a building
being classified as irregular are typically not grave. In the UBC the presence of irregularities
affects the analysis procedure (compulsory use of multimo-dal analysis), but it does not affect
the value of the response modification factor R; in contrast, the EC8 q-factor is reduced by 20
per cent for irregular structures.

4.4.3 Failure mechanisms and capacity design
If the structure is allowed to behave inelastically during the design earthquake (Section 4.3.4),
it is obvious that it will respond even further into the inelastic range whenever a stronger
event (having a lower probability of exceedance) occurs. The requirement under such a rarer
event would normally be that the structure does not collapse and/or does not sustain damage
that would jeopardize human life. Given that it is very unlikely that the response of a structure
close to failure can be analysed (particularly in the framework of a practical design), it has
long been accepted that the main goal of seismic design should be to ensure that the collapse
(or failure) mechanism of the structure is a favourite one, so that the structure could displace
well into the inelastic range without falling down in part or entirely.
A typical illustration of the above concepts is made in Figure 4.17 that shows two generic
plastic mechanisms for a multistorey frame. In the first one (Figure 4.17a) the design strength
of all beams has been exceeded at a certain level of the lateral loading (roughly corresponding
to the design earthquake) and ‘plastic hinges’ have formed at the beam critical sections. This
is also the case at the base of the columns, but not in any other column section. In contrast, the
plastic mechanism shown in Figure 4.17b is characterized by column hinging both at the top
and bottom of the ground storey columns; this is commonly referred to as a ‘soft storey’ or
‘weak storey’ mechanism. It is clear from the kinematics of the two mechanisms that for the
same top displacement (δ u) the ratio of the required plastic