Geology Reference
In-Depth Information
δ
δ
( )
t
had floor heights of 3.2 m and the same first floor
plan, comprising three 8m bays in each direction,
with a lateral force resisting system of MRFs in
the north-south direction and braced frames in the
east-west direction. While the regular building had
a uniform elevation (with section sizes reducing
with height), the irregular building had setbacks
at the first and sixth storys. A single MRF in the
north-south direction was designed, with gravity
loads of 4 kN/m
2
dead load and 2 kN/m
2
live load,
and an assumed 5% inherent damping. The MRFs
were designed using response spectrum analysis
and 0.3g PGA and Eurocode soil B site conditions.
A high behaviour factor of 6.5 was selected for the
regular building, and a reduced factor of 5.2 for
the irregular building, to account for vertical ir-
regularities. Basic details of the designs are shown
in Figure 2, with fuller information presented in
Whittle et al. (2012).
SAP2000 (CSI, 2009) was employed for mod-
elling the frames and aiding the design process.
Seismic performance levels selected include the
frequently occurring earthquake (FOE), which is
40% of the design basis earthquake (DBE), the
DBE (10% probability of exceedance in 50 years),
and the maximum considered earthquake (MCE)
(2% probability of exceedance in 50 years), which
is 150% of the DBE (Somerville et al., 1997). A
serviceability limit of 1% peak interstory drift
under the FOE was selected. This achieves the
Immediate Occupancy performance based design
level for the FOE and Life Safety for the DBE
(FEMA, 2000). Key building properties are pre-
sented in Table 1, where the peak drift is based
on the response spectrum analysis.
i
p
=
max
(5)
i
all i
,
A 'fully-stressed' design is achieved when
all
p
i
values tend to unity, i.e. the damping
distribution is such that the drift limit is just
achieved at each story. If this optimum is not
achieved in iteration (
k
), then the damping
levels are adjusted in step (
k
+1) according to:
(
)
1
/
q
( )
k
( )
k
C
p
i
i
(
k
+
1
)
C
=
C
(6)
i
t
n
1
/
q
(
)
∑
( )
k
( )
k
C
p
i
i
i
=
1
where
q
is a convergence parameter with
recommended values of 0.5 for linear
analysis and 5 for non-linear analysis. The
analysis-redistribution process is repeated
until all the constraint errors (differences of
p
i
values from unity) have been minimised
and the damping distribution does not change
significantly between iterations.
ANALYSIS METHODOLOGY
To assess the different damper placement methods,
they were applied to the retrofit of conventionally
designed steel moment-resisting frames. Perfor-
mance of the different schemes was then compared
by analysing the retrofitted frames under a suite
of earthquake ground motions. Both regular and
irregular frames were considered; results presented
here focus on the regular frame, with other results
given elsewhere (Whittle et al., 2012).
Damping Design
A strategic amount of added damping in the form of
linear viscous dampers was calculated to improve
the performance of the buildings. The objective
was to add dampers to achieve a linear elastic
building performance under the DBE, causing
Building Design
Two ten-story, steel MRF buildings, one regular
and another irregular in elevation, were designed
according to the Eurocodes 3 and 8. Both buildings
Search WWH ::
Custom Search