Documente Academic
Documente Profesional
Documente Cultură
1. Key Lab of Structures Dynamic Behavior and Control of the Ministry of Education, Harbin Institute of Technology, Harbin 150090, China
2. School of Civil Engineering, Harbin Institute of Technology, Harbin 150090, China
3. Key Laboratory of Earthquake Engineering and Engineering Vibration, Institute of Engineering Mechanics, China Earthquake
Administration, Harbin 150080, China
Abstract: Due to the randomness of earthquake wave magnitude and direction, and the uncertain direction of strong
axis and weak axis in the construction of engineering structures, the effect of the direction of ground motion on a structure
are studied herein. Ground motion records usually contain three vertical ground motion data, which are obtained by sensors
arranged in accordance with the EW (East -West) direction, NS (South- North) direction and perpendicular to the surface
(z) direction, referring to the construction standard of seismic stations. The seismic records in the EW and NS directions are
converted to Cartesian coordinates in accordance with the rotation of = 0180, and consequently, a countless group of new
ground motion time histories are obtained. Then, the characteristics of the ground motion time history and response spectrum
of each group were studied, resulting in the following observations: (1) the peak and phase of ground motion are changed
with the rotation of direction , so that the direction of the maximum peak ground motion can be determined; (2) response
spectrum values of each group of ground motions change along with the direction , and their peak, predominant period and
declining curve are also different as the changes occur; then, the angle in the direction of the maximum peak value or the
widest predominant period can be determined; and (3) the seismic response of structures with different directions of ground
motion inputs has been analyzed under the same earthquake record, and the results show the difference. For some ground
motion records, such as the Taft seismic wave, these differences are significant. Next, the Lushan middle school gymnasium
structure was analyzed and the calculation was checked using the proposed method, where the internal force of the upper
space truss varied from 25% to 28%. The research results presented herein can be used for reference in choosing the ground
motion when checking the actual damage to structures following earthquakes and explaining the seismic damage. Meanwhile,
it also provides a reference value for research into the most severe ground motion.
Key words: direction of ground motion; structure effect; coordinate transformation; response spectrum
has been presented, according to the parameters of 2 Ground motion transformation in different
ground motion and displacement ductility and damage coordinate systems
accumulation double parameter failure criterion (Zhai
and Xie, 2002). Therefore, some research for selecting In general, ground motions recorded by seismic
the most severe ground motion input for structures has stations are seismic time history results with three
been carried out by several scholars from the shell (Fan directions: EW (x direction), NS (y direction), and
et al., 2004), the bridge (Wang et al., 2007), the dam perpendicular to the surface (z direction). Among them,
(Wu, 2007) and so on. the peak ground motion obtained from the records in
Previous research on the most severe ground motion three directions is regarded as the maximum seismic
usually has focused on the NS and EW positive direction dynamic value of that point. However, if the horizontal
input of existing ground motion records to obtain the sensors (collectors) are placed in any two vertical angles
structural seismic response with regard to the most severe (but not in the EW and NS directions), then another
ground motion evaluation criterion. However, in practical different set of ground motion records can be obtained.
engineering applications, at the same epicenter distance The coordinate transformation of a set of mutually
and the same site conditions, for similar structures in perpendicular horizontal strong motion records, which
different construction angles with faults, the damage is are obtained by a station, is carried out at any angle.
also somewhat different; that is, the seismic response Figure 1 shows the rectangular coordinate conversion
and failure mechanism of similar structures are different schematic diagram of the ground motion with any angle
in the same ground motion. So far, some research has rotation (initially EW and NS directions). Equation
been carried out on large span bridge structure (Wang (1) is the new coordinate value after the coordinate
et al., 2007), large space structures (Yang et al., transformation.
2008), underground structures (Zhao et al., 2010),
space frame structures (Yang et al., 2013), and so on, to x(t ) x(t ) cos y (t ) sin
(1)
consider the influence of the input directions of ground y '(t ) y (t ) cos x(t ) sin
motion on the dynamic response of structures. And there
are some researchers who have carried out some studies In this study, the commonly used Taft seismic
on the angle rotation of the seismic inputs. For example, wave has been selected in calculation and analysis
the NGA-East project has selected rotated ground as an example. On July 21, 1952, a 7.3 magnitude
motion intensity measure called RotDnn described earthquake occurred in California of the western United
by Boore (Boore, 2010 and Boore et al., 2008). And States; the record was measured by No. 1095 seismic
some scholars have discussed the seismic motion station located in the Lincoln School of Kern County,
time, PGA, PGV and PGD and other ground motion which was located about 43.5 km from the epicenter.
parameters change for different angles of the 2008 The rectangular coordinate conversion of the record
Wenchuan ground motion (Ren et al., 2011). Chapter was obtained, the seismic record when = 0 (the
4 of reference 17 has investigated how this orientation- original seismic record without rectangular coordinate
dependence (directionality) should be considered in conversion) and the seismic record when = 40 (a new
the specification of seismic demand in codes of practice seismic record after rectangular coordinate conversion
(Damian et al., 2011). by the Eq. (1) ) is compared as shown in Fig. 2. The time
In this paper, the El Centro seismic wave, Taft seismic history direction of large (small) ground motions in the
wave, Wolong seismic wave in the Wenchuan Ms8.0 new coordinates is compared with the same ones in the
earthquake (DEDP, 2008; CNSMNC, 2014) and others original coordinates.
are used as the research object, and new ground motion Figures 2 (a), (b) are a comparison of new and
are generated by coordinate transformation. The direction original coordinates in seismic acceleration time history.
of the most severe ground motion input is determined From Fig. 2(a), when = 40, the maximum peak is t =3.76 s,
by comparing time history vibration characteristics and Amax= 212.77 Gal. On the contrary, when = 0o, the
of ground motions at different angles. Meanwhile, the
finite element model from the Lushan Middle Schools y
sports complex building has been established, and a
time history analysis of different ground motion input y'
directions was carried out to determine the influence of x'
x(t
t)
a(
checking actual earthquake damage, explanation of the y(t
damage mechanism, selection of the most severe ground ) cos
x(t
motion, and regression of response spectrum in the
seismic code. Fig. 1 Coordinate rotation diagram
No.4 Sun Menghan et al.: Study on the effect of ground motion direction on the response of engineering structure 651
Taft Lincoln School (Soil) (California earthquake Ms7.3) Taft Lincoln School (Soil) (California earthquake Ms7.3)
2 5
= 0 Max
Tn= 0.10 s
4
Acceleration (102 Gal)
Amax=425.3 Gal
1
= 40 Max
= 0 Min
3 Time history of ground motion with arbitrary
-1 time: 9.15 s angle conversion
Amax=-152.84 Gal
= 40 Min
-2 time: 4.25 s The Wolong seismic wave of the main shock from
Amax=137.86 Gal
the Ms 8.0 Wenchuan earthquake in China and the Taft
0 10 20 30 40 50 60 seismic wave from the California earthquake in the U.S.
Time (s)
are chosen as samples in this study. The time history
(b) Comparison of the ground motion in the direction y and y
changes of ground motion are obtained with the counter
Fig. 2 Comparison of the new coordinates rotating 40 and clockwise rotation of Cartesian coordinates, whose
the original coordinates in time history rotation angle s interval is 10o.
In the first 70 seconds of the Wolong seismic wave,
maximum peak is t = 3.76 s and Amax= 174 Gal. Similarly the x' coordinate rotates along = 0 90 and the y'
from Fig. 2(b), the peak magnitude of ground motion coordinate rotates along = 90 180, then the time
and the time of peak value are not the same between the history curves corresponding to each angle of the
new and the original coordinates. It can be presumed ground motion in 3D graphics are synthesized as shown
that, for any inputting of two new perpendicular in Fig. 4(a). It is seen that there is a significant change in
horizontal strong motions with different rotation angles, the time history of ground motion at different angles. As
the response analysis results of engineering structures the rotation angle = 110, the maximum peak value of
will also be different. the ground motion can reach Amax =1039 Gal. Figure
At the same time, two groups of ground motions are 4(b) shows a comparison of the time history when =
made into seismic response spectra with a damping ratio 20 and = 110, and the difference between them is
of = 5% , as shown in Fig. 3. The figure shows that the very large. They are the two vertical ground motion time
curve shape, peak value, and predominant period of the histories under the new Cartesian coordinates.
response spectra are also different between the original From Fig. 4, the maximum acceleration peak value
coordinates ( = 0o) and the new coordinates ( = 40o). of the Wolong seismic wave is generated at the angle =
If there is a continuous rotation of Cartesian coordinates 110o, when t = 33 s, and Amax = 1038.55 Gal.
of = 0 90, a series of successively different time Another example is the Taft seismic wave, which is
histories and response spectra groups of ground motions often used in structural analysis. As shown in Fig. 5, after
will be obtained. From Fig. 2, it can be seen that the the rotation of the ground motion at the angle = 0180,
ground motion records of different directions are the 3D graphics of the ground motion time history is
very different. Since the direction of the engineering obtained. The graph shows that the maximum peak value
structure location is uncertain, and the earthquake fault of ground motion is Amax = 212.78 Gal, when =170o.
is also uncertain for different types of faults and rupture And the maximum peak value of the original Taft ground
directions, the response analysis of the structure will be motion is Amax = 174.23 Gal in the x direction.
very different with the different directions of the ground Table 1 shows a comparison of the acceleration time
motion time history inputting load, which is selected histories of the two ground motions in angle, time and the
652 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.15
Wolong Strong Motion Station (Soil) (Wenchuan earthquake Ms8.0) Table 1 Comparison of the maximum peak in new and
time = 33.005 s original coordinates
= 110 Max
Amax= 1038.55 Gal
1000
Seismic Amax PD
800
Coordinate S
600 wave (Gal) (%)
Accelerate (Gal)
400
200
Wolong original 90o 33.005 957.7004 8.44
0
-200
(y)
-400 200 new 110o 33.005 1038.553
-600 150
100
-800 (Degree)
25 30 35 40 45 50 55 60 0
50
Taft original 0o 3.76 174.24 22.1
Time (s) (x)
(a) Ground motion time history of Wolong seismic wave in 3D new 40o 3.76 212.78
graphics as =0o180o
Note: is rotation angle; S is the time of peak value; Amaxis the
Wolong Strong Motion Station (Soil) (Wenchuan earthquake Ms8.0) peak value of acceleration; PD is phase difference
1.2
= 0 Max
time = 33.005s
0.8 Amax=957.70 Gal maximum acceleration peak value. For Wolong ground
Acceleration (103 Gal)
= 110 Max
time = 33.005s
motion, maximum acceleration peak value with rotation
0.4 Amax=1038.55 Gal is greater than the original value by about 8.44%. For
the Taft ground motion, the maximum acceleration peak
0 value with rotation is greater than the original value by
about 22.1%. Thus it can be seen that the uncertainties
-0.4 in ground motion direction has a significant influence
on engineering structures and further research is
0.8
needed, especially for post earthquake reconnaissance
20 30 50 4060 70 80 to determine the earthquake damage mechanism of long
Time (s) and large structures , and if the actual ground motion
Wolong Strong Motion Station (Soil) (Wenchuan earthquake Ms8.0) recorded in the vicinity is used, the effect of the direction
1.2 of the ground motion on structures.
= 0 Min
Time = 32.79 s
0.8 Amax= 652.851 Gal
Acceleration (103 Gal)
= 110 Max
Time = 54.09 s
4 Response spectrum of different ground
0.4 Amax=-568.911 Gal motion with arbitrary angle conversion
0 From Section 3, it is known that any earthquake
recorded at different stations will be changed and to be
-0.4 new after coordinate conversion with different angles.
Similarly, the new vibration response spectrum of
0.8
different angles through calculation also has a significant
20 30 40 50 60 70 80 difference in peak, width and other characteristics. As
Time (s) in the case study of the Wolong seismic wave in the
(b) Ground motion time history of Wolong seismic wave as =0o Wenchuan earthquake, Fig. 6 shows the 3D graphics
and =110o of the seismic response spectrum of the ground motion
Fig. 4 Wolong seismic wave time history 3D graphics and recorded at Wolong with an angle rotation of = 0180
comparison of the maximum time history and the mini- in the rectangular coordinate, and a damping ratio of
mum time history =5%.
As shown in Fig. 6, when =30, the response
Taft Lincoln School (Soil) (California earthquake Ms7.3)
spectrum peak value becomes the smallest one and
200 increases along with the angle , and the width of the
150 response spectrum also increases. When = 0 and
100 90o , the response spectrum correspond to the original
Accelerate (Gal)
50
0
coordinate x and y directions.
-50 Through the analysis of the Wenchuan earthquake
-100 Time = 3.76s
: 40 degree
(DEDP, 2008), Lushan earthquake (CNSMNC, 2014),
-150 Amax: -212.8 Gal
California earthquake and dozens of other seismic
-200 200
-250 100 records, it is seen that the angle max time history of
0
0 2 4 6 8 10
Time (s)
12 14 16 18 20 (Degree) the maximum seismic time history is not consistent
with the angle max response spectrum of the maximum
Fig. 5 Ground motion time history of Taft seismic wave in 3D
response spectrum. In other words, max time history
graphics as =0180o
No.4 Sun Menghan et al.: Study on the effect of ground motion direction on the response of engineering structure 653
max response spectrum. In Section 5, the influence of the which was destroyed in the Lushan earthquake on April
maximum time history input and the maximum response 20, 2013, is selected as an example to verify the proposed
spectrum time history input on structures is discussed. method. The Taft seismic wave and Lushan earthquake
Figure 7 shows the seismic response spectrum 3D Feixian seismic waves (Yang et al., 2008) were used with
graphics of the Taft earthquake wave with a rotation the original seismic records in the x, y direction, and the
= 0180 in the rectangular coordinate conversion, time history of the maximum ground motion direction
and a damping ratio of =5%. Along with the change or the time history of the maximum response spectrum
of coordinate rotation angle, the peak of the response direction as x', y', z input to calculate and analyze. In
spectrum and the width of the predominant period will the above calculation, the peak of the maximum seismic
also change. wave record of the original coordinate and the rotational
coordinate system were inputted along with the structure.
Figure 8 shows a finite element simplified model of
5 Practical example analysis the Lushan middle school gymnasium. The model unit
selection and material parameters are shown as in Table 2
In this study, the Lushan middle school gymnasium, and Table 3. In this study, the internal forces of weak
Table 2 Finite element model unit, material parameters and constitutive model
Unit
Model Unit type Material Material constitutive
(Num.)
Upper space truss 1080 Link180 Q235 BKIN follow-up intensified model
Lower support beam 171 Fiber beam element C50 Single axis and double segment
structure (concrete beam188, steel linear constitutive
bar reinf264)
column 122 Q235 BKIN follow-up intensified model
Response spectrum of Wolong Strong Motion Station (Wenchuan earthquake) [E104.0 N31.3] Response spectrum of Taft Lincoln School (soil) (California earthquake Ms7.3)
300
a (Gal)
a (Gal)
2000
200 200
200
1000
100 100
100
0 (Degree) 0 (Degree)
0 0.5 1.0 1.5 2.0 2.5 3.0 0 0 0.5 1.0 1.5 2.0 2.5 3.0 0
Tn (s) Tn (s)
Fig. 6 3D graphics response spectrum of Wolong seismic wave Fig.7 Seismic response spectrum of Taft earthquake wave in
of =0o180o 3D graphics of =0o180o
(a) Integrated finite element model (b) Finite element model of the upper space truss
Fig. 8 3D finite element modeling of Lushan middle school gymnasium
654 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.15
structural members, which are the internal forces of the compressive stress in the upper space truss unit of the
upper space truss structure, were selected for display. Lushan middle school gymnasium of =0 and =40
The Taft seismic wave selection is shown in Fig. 2, with three direction ground motion input. There are 1080
which are the ground motions in the x, y and z directions upper space truss units, and the stress contrast of each
in the original coordinates of = 0, and the ground unit in two working conditions is shown along with the
motion time history in x', y' and z directions of the direction of the horizontal axis. It can be seen from the
maximum ground motion of = 40 with three direction graph that, the tensile stress and compressive stress of the
input. unit are quite different in the two conditions. Comparing
Figure 9 shows a comparison of tensile stress and the maximum tensile stress of =0 in original coordinate
-1.0 -2
= 0
= 0
Bar number=1079
Bar number=338
-0.2
0
0 250 500 750 1000 0
0 500 1000
Bar number
Bar number
Fig. 9 Comparison of compressive stresses of each unit in the (a) Comparison of unit compressive stress of =50o and original
upper space truss structure seismic wave input with the biggest response spectrum
-2
Lushan Feixian Strong Motion Station (soil) (Yaan earthquake Ms7.0) = 0
Bar number=338
Compressive stress (108 Pa)
Tn: 0.15 s -1
Amax= 1481.80 Gal
= 70 Max
0.8 Tn: 0.15 s
Amax= 1424.44 Gal
0.4
0
0 500 1000
Bar number
0 (c) Comparison of unit compressive stress of =70o and =50o
0 1 2 3
Tn (s) with the biggest response spectrum
(b) Seismic time history Fig. 11 Comparison of unit compressive stress among = 0o/50o/70o
Fig. 10 Comparison of =0o/50o/70o from Feixian Station in Feixian Station (medium earthquake)
No.4 Sun Menghan et al.: Study on the effect of ground motion direction on the response of engineering structure 655
important reference value when checking the failure Mechanics, 27(12): 135140. (in Chinese)
mechanism of actual structures in earthquakes and Li Xiaozhen, Hong Qinye, Lei Hujun and Liu Zhenjie
selection of the maximum ground motion. Meanwhile, (2015), Effect of Input Directions of Seismic Ground
it also has a reference value on the regression of the Motion Seismic Responses of a Railway Extradosed
response spectrum curve in the seismic code. Bridge, Bridge Construction, 45(1): 2632. (in Chinese)
Ren Yefei, Dai Zhiyong and Wen Ruizhi (2011), A New
Reference Method for Computing the Orientation-independent
Parameters of Horizontal Ground Motion, World
Earthquake Engineering, 27(3): 134-141. (in Chinese)
Boore DM (2010), Orientation-independent, Wang Tao, Guo Endong, Zhang Lina and Hu Yuwen
Nongeometric-mean Measures of Seismic Intensity from (2007), The Critical Direction of the Earthquake Input
Two Horizontal Components of Motion, Bulletin of the for Dynamic Analysis of a Long Span Cable-stayed
Seismological Society of America, 100(4): 1830-1835. Bridge, World Earthquake Engineering, 23(4): 107
Boore DM and Atkinson GM (2008) Ground-motion 111. (in Chinese)
Prediction Equations for the Average Horizontal 99 Wu Zhaoying (2007), Study of the Worst Seismic
Component of PGA, PGV, and 5%-damped PSA at Motion Input for Earth-dam Seismic Stability under
Spectral Periods between 0.01 s and 10.0 s, Earthquake the Oblique Incidence Condition, Harbin: Institute of
Spectra, Earthquake Engineering Research Institute, Engineering Mechanics. (in Chinese)
24(1): 99138.
Xie Lili and Zhai Changhai (2003), Study on the
China National Strong Motion Network Center (2014), Severest Real Ground Motion for Seismic Design and
Report on Strong Earthquake Motion Records in China Analysis, Acta Seismologica Sinica, 25(3): 250261.
(2014), 17(1), Beijing: Seismological Press. (in Chinese) (in Chinese)
Grant DN, Padilla D and Greening PD (2011), Yang Hong, Yang Xiaoyi and Huang Hao (2013),
Orientation Dependence of Earthquake Ground Motion Effects of Seismic Wave Input Angles on Nonlinear
and Structural Response, Matja Dolek, Protection Response Characteristics of Spatial Frames, Industrial
of Built Environment Against Earthquake, Chapter 4, Construction, 43(7): 4351. (in Chinese)
Berlin: Springer Netherlands.
Yang Qingshan, Liu Wenhua and Tian Yuji (2008),
Department of Earthquake Disaster Prevention, China Response Analysis of National Stadium under Special
Earthquake Administration (2008), Report on Strong Variable Earthquake Ground Motions, China Civil
Earthquake Motion Records in China, 12(1), Beijing: Engineering Journal, 41(2): 3541. (in Chinese)
Seismological Press. (in Chinese)
Zhai Changhai and Xie Lili (2002), A Comprehensive
Fan Feng, Qian Hongliang and Xie Lili (2004), Method for Estimating and Comparing the Damage
Applications of the Severest Ground Motion to Potential of Strong Ground Motion, Earthquake
Anti-seismic Design for Reticulated Shells, World Engineering and Engineering Vibration, 22(5): 17. (in
Earthquake Engineering, 19(3): 1721. (in Chinese) Chinese)
Hu Yuxian (2005), Earthquake Engineering, 2nd ed, Zhao Baoyou, Ma Zhenyue and Ding Xiuli (2010),
Beijing: Seismological Press. (in Chinese) Seismic Response of a Large Underground Rock
Li Quanwang, Fan Jiansheng and Nie Jianguo (2010), Cavern Groups Considering Different Incident Angles of
Effect of Directional Uncertainty of Earthquake Earthquake Waves, Chinese Journal of Rock Mechanics
Ground Motion on Structural Responses, Engineering and Engineering, 29(Supp.1): 33953402. (in Chinese)