Seismic Behavior of Long-Span Connected Structures Under Multi-Supported and Multi-Dimensional Earthquake Excitations
Seismic Behavior of Long-Span Connected Structures Under Multi-Supported and Multi-Dimensional Earthquake Excitations
Seismic Behavior of Long-Span Connected Structures Under Multi-Supported and Multi-Dimensional Earthquake Excitations
and X
= =
1
Y K K X X (2.4)
Where
ss sb
=
1
K K is the pseudo-static displacement matrix, it denotes the pseudo-static
displacement of the structure induced by unit displacement of the support nodes. Substitute Eqn. 2.4
into Eqn. 2.2
ss d ss d ss d ss s ss s sb b sb b
+ + = M Y C Y K Y M Y C Y M X C X
(2.5)
Assuming that the damping force is only related to the relative velocity, and thus in Eqn. 2.1,
s
b
`
)
X
X
can be replaced by
d
0
`
)
Y
, which yields
ss d ss d ss d ss s sb b
+ + = M Y C Y K Y M Y M X
(2.6)
Substituting Eqn. 2.4 into Eqn. 2.6
ss d ss d ss d ss sb b
( ) o + + = + M Y C Y K Y M M X
(2.7)
In which
sb
M is zero matrix when lumped mass matrix is applied, then
ss d ss d ss d ss b
o + + = M Y C Y K Y M X
(2.8)
where
b
X
can be obtained from the acceleration time-history of each support, relative dynamic
displacement
d
Y can then be solved from Eqn. 2.8.
3. NUMERICAL SIMULATIONS
3.1 Finite Element Model
A newly constructed long-span connected structure is selected. This symmetric building is composed
of two main towers and a corridor in between. The main towers are concrete frame-shear wall
structure, and the corridor is a steel structure which is connected between 39.85m and 55.85m height
with a span of 105.00m. As shown in Fig. 1, a finite element model was first built using ANSYS
software. Tianjin earthquake record (1976, NS) was used to excite the structure. In this numerical
example, when considering non-uniform earthquake excitation, it is assumed that the seismic wave
propagates along the long-axis of the structure.
Modal analysis was carried out using the ANSYS. The first 10 natural frequencies are listed in Table
3.1, and Fig. 2 shows the first 6 mode shapes. It is observed from Fig. 2 that unlike normal high-rise
structure, the long-span connected structure is characterized by its antisymmetric mode shapes of the
main towers which will easily result in torsional response of the corridor. Thus, this numerical study
will focus on the seismic responses of elements around the connection area.
Figure 1 Finite element model
Table 3.1 The first 10 natural frequencies
Mode 1 2 3 4 5
Frequency
(Hz)
0.790 0.831 1.015 1.158 1.306
Mode 6 7 8 9 10
Frequency
(Hz)
1.631 1.668 2.135 2.236 2.756
Figure 2 The first 6 mode shapes
3.2 Multi-dimensional Excitations
Seismic behavior of the model under single- and multi-dimensional uniform earthquake excitation is
examined. In single-dimensional excitation case, the peak acceleration is set to 0.15g. While in
multi-dimensional cases, the proportion of the peak acceleration of the three dimensions is 1:0.85:0.65.
Fig. 3 shows the time-history stress of horizontal bar 50001 and tilted bracing 60048 under single- and
multi-dimensional earthquake excitations. One can see that from these results, the peak stress
responses are increased by 43.7% and 69.5% respectively when subjected two horizontal earthquake
excitations. The seismic responses change little when vertical component is also considered.
0 2 4 6 8 10 12 14 16 18
-60
-40
-20
0
20
40
60
Time (s)
S
t
r
e
s
s
(
M
P
a
)
x-direction excitation
x- and y-direction excitation
multi-dimensional excitation
(a) Horizontal bar No. 50001
0 2 4 6 8 10 12 14 16 18
-20
-15
-10
-5
0
5
10
15
20
Time (s)
S
t
r
e
s
s
(
M
P
a
)
x-direction excitation
x- and y-direction excitation
multi-dimensional excitation
(b) Bracing No. 60048
Figure 3 Time-history of stress responses under single- and multi-dimensional excitation
3.3 Multi-dimensional and Multi-supported Excitations
The seismic behaviors of the structure under multi-dimensional and multi-supported earthquake
excitations are further studied. Three typical apparent velocities, 100m/s, 500m/s, 800m/s are
considered. Fig. 4 shows the comparison of time-history stress responses subjected to uniform and
travelling wave excitations with different apparent velocities, and the peak responses are further
compared in Table 3.2. It is conclude that the peak responses might be increased or decreased when
different apparent velocity is applied. For example, the peak stress of the horizontal bar 50001 is
decreased by 31.3% under 100m/s travelling-wave excitation, while the peak stress of bracing 60048
is increased by 37.4%.
0 2 4 6 8 10 12 14 16 18
-60
-40
-20
0
20
40
60
Time (s)
S
t
r
e
s
s
(
M
P
a
)
Uniform excitation
Travelling wave excitation 800m/s
Travelling wave excitation 500m/s
Travelling wave excitation 100m/s
(a) Horizontal bar No. 50001
0 2 4 6 8 10 12 14 16 18
-30
-20
-10
0
10
20
30
Time (s)
S
t
r
e
s
s
(
M
P
a
)
Uniform excitation
Travelling wave excitation 800m/s
Travelling wave excitation 500m/s
Travelling wave excitation 100m/s
(b) Bracing No. 60048
Figure 4 Time-history of stress responses under different travelling wave excitations
Table 3.2 Peak stress responses under different earthquake excitations (UnitMPa)
Cases
Uniform
excitation
Travelling-wave
excitation
(800m/s)
Travelling-wave
excitation
(500m/s)
travelling-wave
excitation
(100m/s)
Horizontal bar
No. 50001
x-direction excitation 12.1283 11.5569 9.9232 10.1421
x- and y-directions
excitation
17.6514 13.2651 14.6781 10.2178
Three dimensional
excitation
17.6618 16.5394 15.0370 12.1204
Bracing No.
60048
x-direction excitation 3.0643 2.4761 2.0757 3.1640
x- and y-directions
excitation
5.1931 4.8930 4.9908 3.9359
Three dimensional
excitation
6.1464 5.6714 4.4080 8.4459
The position and peak stress responses of the dominant element in different cases are list in Table 3.3.
The results show that both the dominant element and the peak value are changing with apparent
velocities, and dominant internal force of the corridor structure may be increased by 14.69% when
subjected to travelling-wave excitations.
Table 3.3 Dominant element and peak stress responses
Cases
Distance from the nearest
connectionm
Peak value
Mpa
Increase
(%)
Uniform 12.30 28.9315 -
Travelling-wave
excitation (800m/s)
12.30 29.4620 1.83
Travelling-wave
excitation (500m/s)
5.40 26.2510 -9.26
Travelling-wave
excitation (100m/s)
2.70 33.1828 14.69
5 CONCLUSIONS
Seismic response of long-span connected structure under multi-dimensional and multi-supported
earthquake excitation is investigated. The comparative performance is studied through a numerical
example, and the influence of wave-passage effect is investigated when different apparent velocities
are applied. The following conclusion can be drawn from the numerical results.
(i) Under uniform excitation, will increase the seismic responses will be increased greatly when
taking into account the multi-dimensional earthquake components;
(ii) Wave-passage effect will amplify the influence of vertical component; the internal force can
either be increased or decreased when different apparent velocities are considered. Different element
will be suffered from wave-passage effect differently, the impact of wave-passage effect on the seismic
response of bracings around the connection between main towers and corridor is particular significant.
(iii) The dominant internal force of the corridor structure may be increased by 14.69% under
travelling wave excitation. Therefore, it is necessary for aseismic design of long-span connected
structure to take into account multi-dimensional and multi-support earthquake excitations.
AKCNOWLEDGEMENT
The authors are grateful for financial support by National Natural Science Foundation of China (No. 51108089)
and National Natural Science Foundation of Fujian Province (No. 2011J05128).
REFERENCES
Hahn G. D., Liu X. (1994). Torsional response of unsymmetrical buildings to incoherent ground motions.
Journal of Structural Engineering 120: 4, 1158 - 1181.
Heredia-Zavoni E and Leyva A. (2003). Torsional response of symmetric buildings to incoherent and phase
delayed earthquake ground motion. Earthquake Engineering and Structural Dynamics 32: 1021- 1038.
Li B., Chan K., Chouw N. and Butterworth J. W. (2010). Seismic response of two-span scale bridge model due to
non-uniform ground excitation and varying subsoil conditions. Proceedings of 2010 NZSEE Conference.
LIN W.LI Z. X, DING Y. (2008). Trust-region based Instantaneous Optimal Semi-active Control on Long-span
Spatial Structure with MRF-04K Damper. Journal of Earthquake Engineering and Engineering Vibration. 7 :
4, 447-464.
Newmark N. M. (1969). Torsion in symmetrical buildings. Proceedings 4th World Conference Earthquake
Engineering. Santiago, Chile: 19-32.
Ohsaki M. (2001). Sensitivity of Optimum Designs for Spatially Varying Ground Motions. Journal of Structural
Engineering 127 : 11, 1324-1329.
Wang J., Cooke N., Moss P. J. (2009). The response of a 344 m long bridge to non-uniform earthquake ground
motions. Engineering Structures 31: 11, 2554-2567.
Zhang Y. H., Lin J.H., Williams F.W., Li Q.S. (2005). Wave passage effect of seismic ground motions on the
response of multiply supported structures. Structural Engineering and Mechanics 20: 6, 655-672.