Professional Documents
Culture Documents
79-92, 1998
Abstract-Structural pounding occurs frequently during strong earthquakes between two buildings or
different parts of the same building. Structural pounding can also occur in bridges in the longitudinal
direction at the abutments or at expansion hinges, and laterally between narrowly separated superstruc-
tures. The dynamic behavior of a damped single-degree-of-freedom (SDF) structural system with one-
sided pounding during an earthquake is examined. The structural response of the SDF structure with
either elastic or inelastic structural behavior is analyzed. The pounding phenomenon is modeled as a
Hertz impact force, which represents the behavior of two colliding bodies during a completely elastic
impact. Artificial, as well as actual earthquake excitations, and realistic parameters for the pounding
model are used in numerical evaluations of the seismic response. The effects of separation distance and
inelastic structural behavior on the magnitude of the pounding force are examined. An increase in the
damping energy absorption capacity of the pounding structure results in the reduction of the pounding
forces. The present model and method of analysis can be used in investigations of pounding between
buildings or pounding which occurs in bridges during strong earthquakes. 0 1997 Elsevier Science Ltd
Key words-Elridges, Buildings, Earthquakes, Nonlinear pounding, Pounding force, Seismic gap
(a)
40
z 20
E
z
B O
ii
“a
.z -20
n
-40
-6O~,,,,,,,,,,,,,,‘,“,,,
0 5 10 15 20 25 30
Time (set)
(b)
0.6
0
,M
c -0.2
.z
rj -0.4
z
2 -0.6
4
-0.8
0 5 10 15 20 25
Time (set)
(c)
5.c
‘l-
4.5
4.0
z‘ 3.5
E
E 3.0
8 2.5
Et
s 2.0
s
0 1.5
a
1 .o
0.5
0 r
0
Time (set)
Fig. 3. Response of SDF elastic structure during artificial sinusodial earthquake: (a) displacement, (b)
acceleration, (c) pounding force; __- f, = 2 Hz, ---&=0.67 Hz.
Linear and nonlinear pounding of structural systems 83
(a)
0.40
-0.40 ,,......,,,.........,.........,.........,.........,
0 IO 20 30 40 50
Time (set)
(b)
1.20
._8 0.60
2
2
e 0
::
2
: -0.60
tj
-1.20
0 10 20 30 40 SO
(cl
Time (xc)
0.50
Gl
._8
0.25
5
2
s a
::
‘CJ
z
0 -0.;15
6
10 20 30 )
Time (set)
(d)
I .oo ,
I
G
z
._
O.!iO
5
If!
e 0
::
P
z -O.!iO
6
-I .oo
0 20 30 40 50
Time (set)
Fig. 4. Historical earthquakes: (a) Imperial Valley SOOEcomponent of 1940 El-Centro; (b) S16E com-
ponent of 1971 San Fernando; (c) Channel 8 of 1989 Loma Prieta; (d) 90” component of the Santa
Monica record of 1994 Northridge.
C. P. Pantelides and X. Ma
Time (set)
(b)
‘L -1
15 20 25 30
Time (set)
Fig. 5. Effect of the natural period on pounding of elastic structure: (a) displacement, (b) pounding
force; --- T = 1s, - TO=3 s.
agreement with more detailed studies using analyti- the maximum velocity over a separation distance
cal models of viscoelastic dampers [21]. ranging from 12.5 mm to the particular seismic sep-
A response spectrum for a SDF elastic structure aration distance, atim, for that structure’s period
with periods ranging from 0.1 to 3.0 s, at an inter- beyond which no pounding occurs. For stiff struc-
val of 0.1 s, was constructed using three damping tures with a period less than 0.3 s the response is
ratios of 2, 8 and 20% of critical for the 1940 El- affected only slightly with the increase in damping.
Centro earthquake. The impact stiffness parameter For flexible structures with a period longer than
was assumed as R = 80 kN rnrne3”. The results are 0.3 s the response is reduced significantly with the
shown in Fig. 8, where the horizontal axis indicates increase in damping. In addition, the increase from
the structure’s period and the vertical axis denotes 2 to 8% of critical damping seems to be more effec-
Linear and nonlinear pounding of structural systems 85
(a)
150
100
50
+2
& 0
z
l
m -50
2
6
-100
-150
-200
0 15 20 25 30
Time (set)
(b)
3
0
M
4
15 20 25
Time (set)
Fig. 6. Effect of damping on pounding of elastic structure: (a) displacement, (b) acceleration; ~
r = 2%, - 5 = 8%.
tive than the increase from 8 to 20% in reducing least (0.375 x R,) times the displacement due to
the structure’s response. seismic forces, where R, is a factor which defines
the lateral force resisting system. For bridges, the
Parametric study oJ separation distance
AASHTO 1995 Interim Provisions specify minimum
The value of the separation distance, a, between support length requirements at the expansion ends
two structures whilch is st&iciently large to prevent
of all girders [23]. However, even though the pro-
pounding in an earthquake is known as the seismic
visions encourage consideration of relative displace-
gap. Building Codes have recognized the existence
of a safe seismic separation distance. The Uniform ments between different segments of the bridge,
Building Code-UBC [22]-specifies that the separ- they do not provide specific guidelines for separ-
ation distance between two buildings shall be at ation distances to prevent pounding. Similarly, no
CAS 66/l-D
86 C. P. Pantelides and X. Ma
15 100 125
Fig. 7. Effect of damping on seperation distance and maximum level of pounding force: W 5 = 2%, 0
5 = 8%, A 5 = 20%.
provisions are given for prevention of lateral is the maximum pounding force during the earth-
pounding of the type shown in Fig. l(c). quake, and the horizontal axis is the separation dis-
A parametric study was undertaken to examine tance, a. Four earthquakes were examined as shown
the UBC code requirements for a structure modeled in Fig. 4: (1) the Imperial Valley SOOE component
as a SDF elastic structural system with period of the 1940 El-Centro; (2) the S16E component of
T = 1 s, damping ratio 4 = 2% of critical, and the 1971 San Fernando; (3) the Channel 8 of the
impact stiffness parameter R = 80 kN mmp3’*. A 1989 Loma Prieta; and (4) the 90” component of
series of separation distances were studied from the Santa Monica record of the 1994 Northridge
a = 25 mm up to where a was sufficiently large to earthquake [18]. As shown in Fig. 9, no pounding
prevent pounding, at an interval of 12.5 mm. The occurs when the separation distance is greater than
results are shown in Fig. 9 in which the vertical axis 175 mm for the El-Centro earthquake, 238 mm for
1.5
Period (set)
Fig. 8. Effect of damping on velocity response spectrum for SDF elastic structure: + 5 = 2%, n
5 = 8%, A < = 20%.
Linear and nonlinear pounding of structural systems 87
r
16
Fig. 9. Maximum pounding force as a function of seperation distance: + = 1940 El-Centro, n = 1971
San Fernando, A = 1989 Loma Prieta, l = 1994 Northridge.
the Northridge earthquake, 250 mm for the Loma A fourth-order Runge-Kutta integration algorithm
Prieta earthquaki: and 338 mm for the San is used to solve Equation (1). Comparisons are
Fernando earthquake. Thus, assuming that the de- made between the response of a SDF structural sys-
sign earthquake was the 1940 El-Centro earth- tem with elastic or inelastic structural behavior
quake, and the basic structural system was a under the condition of one-sided pounding in terms
bearing wall system with a concrete shearwall lat- of the separation distance.
eral force resisting system (R,= 6), the separation
distance required by the UBC would be equal to
Pounding of inelastic structure during the Imperial
0.375 x (6) x (175) = 394 mm; this separation dis-
Valley SOOE El-Centro earthquake of 18 May 1940
tance would be sufficient to prevent pounding from
occurring for any of the other three earthquakes. It In this section, the one-sided pounding of a SDF
should be noted that the above results are not very structural system simulating a reinforced concrete
sensitive with respect to the value of the impact building with inelastic behavior is studied for the
stiffness parameter, R. The UBC code values for 1940 El-Centro earthquake. The properties of the
the seismic separation distance to avoid pounding structure considered in this section are as follows:
are seen to be conservative, which is in agreement mass = 87.55 Mg, elastic stiffness = 3.5 kN mm-‘,
with previous studies [24,25]. plastic stiffness = 0, ductility = 4, damping = 2%
of critical; the period of the structure for elastic vi-
brations is T = 1 s. The value of the impact stiff-
ness parameter is assumed as R = 80 kN rnrne3’*.
POUNDING OF INELASTICSTRUCTURE
Two cases are considered for the structure
For an economical seismic design, the inelastic described above. In the first case, the separation dis-
behavior of the structure must be considered. In tance is sufficiently large so that pounding does not
this paper, the elastoplastic shear-displacement re- occur; in the second case, the separation distance
lationship [26] is used to model the inelastic beha- a = 25 mm for which one-sided pounding does
vior of a SDF structural system with one-sided occur. Figure 10 shows a comparison of the two
pounding during a strong earthquake. Equation (1) cases during the SOOE component of the 1940 El-
is used as the equation of motion and the inelastic Centro earthquake. The peak displacement in the
behavior of the structure is modeled by k(t) with no-pounding case is 127 mm, whereas for the one-
the elastic portion of the stiffness k, = 3.5 kN mm-‘, sided pounding case it is equal to -147 mm. The
and k,= 0 for the plastic portion. The behavior of one-sided nature of the pounding for the inelastic
the SDF inelastic structure with or without pound- SDF structure is evident. However, as Fig. 10(b)
ing is examined for appropriate levels of ductility. demonstrates, the peak acceleration when pounding
88 C. P. Pantelides and X. Ma
(a)
-150
Time (s&c)
(b)
2
e -0.5
._0
;;r
;
z
-1.0
::
d
-1.5
5 10 15 20 25 30
Time (set)
15 20 25 30
Time (see)
Fig. 10. Response of SDF inelastic structure during the 1940 El-Centro earthquake: (a) displacement,
(b) acceleration, (c) pounding force; --- with pounding, -- without pounding.
Linear and nonlinear pounding of structural systems 89
occurs is 1.6g, which is approximately four times tic structure with the same mass, an elastic
the peak acceleration of the no-pounding case. This stiffness = 3.5 kN mm-’ and a damping level of
result shows that pounding must be considered in 5 = 2% of critical. In both cases a separation dis-
seismic design. The pounding force is given in tance of = 25 mm is used. The time-histories of the
Fig. IO(c); pounding occurs only twice in the responses of the elastic and inelastic structures for
pounding case with a peak value of pounding force the four earthquakes used above are investigated.
equal to 1245 kN. The acceleration response for the elastic and inelas-
tic structure for the four earthquakes of Fig. 4 is
Comparison of pounding response for elastic and shown in Fig. 11. The inelastic structure has con-
inelastic structure siderably smaller accelerations as compared to the
The response of the inelastic structural system elastic structure. The peak responses of the elastic
described above 4.1, is compared to that of an elas- and inelastic structure for displacement, accelera-
’ ,
0
,M
z
4
lz
.EI
B
B
8
-1
-2
-3
P
-4
-5
-6 -I-
0 15
Time (set)
_ (b)
-12
5 10 15 20 25 30
Time (set)
Fig. ll(a, b).
90 C. P. Pantelides and X. Ma
(c)
1
-1
M
a -2
.o
5
2
2 -3
P
-4
-5
-6
0 5 10 15 20 25 30
Time (set)
(d)
-2
-a
-10
-12
15 20 25 30
Time (set)
Fig. 11. Comparison of accceleration response of SDF elastic and inelastic structure: (a) 1940 El-
Centro, (b) 1971 San Fernando, (c) 1989 Loma Prieta, (d) 1994 Northridge; --- elastic, -
inelastic.
tion, pounding force and number of pounding pounding occurrences are considerably less in the
occurrences are shown in Table 1. Even though the inelastic case as compared to the elastic case.
maximum displacement of the inelastic structure is These results indicate that the inelastic behavior
larger than that of the elastic structure, the maxi- of structures under pounding is less conservative
mum acceleration is considerably less. Moreover, than the elastic behavior assumption. This may help
the maximum pounding force and number of explain why adjoining buildings that have different
Linear and nonlinear pounding of structural systems 91
Table 1. Maximum elastic and inelastic response of SDF Structure with pounding
Earthquake Displacement (mm) Acceleration (g) Pounding force (kN) Number of poundings
Elastic Inelastic Elastic Inelastic Elastic Inelastic Elastic Inelastic
1940 El-Centro 145 145 5.8 1.8 4670 1120 35 2
1971 San Fernando 248 598 10.1 1.4 8500 1080 21 1
1989Loma Prieta 192 198 5.5 1.5 4680 1240 25 1
1994 Northridge 206 456 11.3 5.0 9060 4020 24 1
periods and are not separated by a distance less pounding, respond in a satisfactory manner during
than the seismic <gap, and thus would be expected earthquakes.
to experience damage due to pounding, have shown The seismic gap required to prevent pounding
satisfactory response in past earthquakes [8]. The was compared to the requirements of the Uniform
behavior of inelastic structures during pounding is Building Code using four earthquakes. It was found
an encouraging observation for efforts to consider that the code values for moderate damping are con-
an alternative to the seismic separation gap pro- servative as compared to the actual seismic separ-
vision in the design of structures to minimize the ation distance found through analysis of SDF
effect of pounding. Many investigators have pro- structural systems. The value of the seismic gap is
posed that building codes should consider alterna- reduced significantly as the damping capacity of the
tive regulations other than the seismic gap, such as pounding structure is increased. For flexible struc-
including in the design and detailing of adjacent tures, for which providing the appropriate separ-
buildings the possibility that pounding will occur ation distance is not possible, such as existing
[9,24,27], or by linking the adjacent structures buildings, an increase in the damping capacity of
through a link and beam system[28]. the pounding structure by means of passive or
active structural control and proper structural
detailing is an effective alternative for reducing
damage due to pounding. Building codes should
CONCLUSIONS consider specifying a less conservative seismic gap
in such cases.
A realistic pounding model was used for studying
the response of an elastic or inelastic SDF struc-
tural system under the condition of structural
REFERENCES
pounding during strong earthquakes. Numerical
simulations have shown that the pounding response 1. National Academy of Sciences, The Great Alaska
is not sensitive to the exact value of the impact stiff- Earthquake of 1964. Engineering, NAS Publication
ness parameter. For structures with different natural 1606, Washington, DC.
2. Hanson, R. D. and Degenkolb, H. J., The Venezuela
periods, the same earthquake excitation can pro- Earthquake-July 29, 1967. American Iron and Steel
duce different magnitudes of pounding force and Institute, New York, 1969.
resulting structural response. For stiff structures 3. Bertero, V. V. and Collins, R. Cl., Investigation of the
with a period less than 0.3 s the damping capacity failures of the Olive View stairtowers during the San
Fernando earthquake and their implications on seismic
of the SDF structure does not affect the response of design. Report no. EERC 73-26, Earthquake
the pounding structure considerably. For flexible Engineering Research Center, University of California,
structures with a period longer than 0.3 s the re- Berkeley, CA, 1973.
sponse of the pounding structure is reduced signifi- 4. Earthquake Engineering Research Institute, Managua,
Nicaragua Earthquake of December 23, 1972. Report
cantly with the irrcrease in effective damping, which
EP-12, Oakland, CA, 1973.
could be achieved by using passive or active struc- 5. Tezcan, S.S., Yerlici, V. and Durgunoglu, H.T. A
tural control devices. reconnaissance report for the Romanian earthquake
Comparison of the pounding behavior of elastic of 4 March 1977. Earthquake Engineering and
with inelastic structures for four earthquakes has Structural Dynamics, 1978, 6, 379-421.
6. Earthquake Engineering Research Institute,
shown that for m’oderate damping levels the displa- Thessaloniki, Greece earthquake of June 20, 1978.
cement response is larger in the inelastic case. Reconnaissance Report, Report EP-32, Oakland, CA,
However, the values for peak velocity, acceleration 1978.
and pounding force of the inelastic structure are 7. Earthquake Engineering Research Institute, The
Central Greece earthquakes of February-March 1981.
significantly less than those of the elastic structure;
Reconnaissance and Engineering Report, Report JP-
in addition, the number of pounding occurrences 05, Oakland, CA, 1982.
for the structure with inelastic behavior is less than 8. Bertero, V. V., Observations on structural pounding,
that of the elastic structure. This observation may Proc. Int. Conf. Mexico Earthquakes, ASCE, 1987,
help explain why adjoining buildings with different pp. 264-278.
9. Astaneh, A., Bertero, V. V., Bolt, B. A., Mahin, S. A.,
fundamental periods, which are not separated in ac- Moehle, J. P. and Seed, R. B., Preliminary report on
cordance with the seismic separation gap and thus the seismological and engineering aspects of the October
would be expected to experience damage due to 17, 1989 Santa Cruz (Loma Prieta) earthquake.
92 C. P. Pantelides and X. Ma
Report no. UCB/EERC-89/14, Earthquake practice in seismic energy dissipation, ATC 17-I.
Engineering Research Center, University of California, Proceedings of the Seminar on Seismic Isolation,
Berkeley, CA, 1989. Passive Energy Dissipation, and Active Control.
10. Earthquake Engineering Research Institute, Aonlied Technoloev Council. Redwood City, CA.
Northridge Earthquake of January 17, 1994- 1993, pp. 4499471.--
Reconnaissance Report. Vol. 1. ed. J. F. Hall. 20. Soong, T. T. and Reinhorn, A. M., Case studies of
Earthquake Spectra; 95-03, Oakland, CA, 1995. active control and implementational issues, ATC 17-1,
11. Earthquake Engineering Research Institute, The Proceedings of the Seminar on Seismic Isolation.
Hyogo-Ken Nonbu Earthquake of January 17, 1995- Passive Energy Dissipation, and Active Control.
Preliminary Reconnaissance Report, eds C. D. Applied Technology Council, Redwood City, CA,
Comartin, M. Greene and S. K. Tubessing. 95-04, 1993, pp. 701-713.
Oakland, CA, 1995. 21. Kasai, K., Munshi, J. A. and Maison, B. F.,
12. Matson, B.F. and Kasai, K. Dynamics of pounding Viscoelastic dampers for seismic pounding mitigation.
when two buildings collide. Earthquake Engineering Proceedings of the Structures Congress 1993. ASCE,
and Structural Dynamics, 1992, 21, 771-786. Irvine, CA, 1993, pp. 730-735.
13. Anagnostopoulos, S.A. Pounding of buildings in series 22. International Conference of Building Oficials, Uniform
during earthquakes. Earthquake Engineering and
Building Code. Vol. 2. Section 1631.2.11. Whittier.
Structural Dynamics, 1988, 16, 443456.
CA, 1994.
14. Liu, W. D., Nobari, S. F. and Imbsen, R. A.,
23. AASHTO, Standard Spectfications for Highway
Dynamic response prediction for earthquake resistance
Bridges-Division I-A Seismic Design, 15th edn.
design of bridge structures, Proc. Structures Congress
Washington, DC, 1992.
1989. ASCE, eds C. A. Kircher and A. K. Chopra.
24. Anagnostopoulos, S.A. and Spiliopoulos, K.V. An in-
San Francisco, CA, 1989, pp. I-10.
vestigation of earthquake induced pounding between
15. Maragakis, E., Douglas, B. and Vrontinos, S.,
adjacent buildings. Earthquake Engineering and
Analysis of the effects of the impact energy losses
Structural Dynamics, 1992, 21, 289-302.
occurring between the bridge deck and abutments,
Proc. Second Workshop on Bridge Engineering 25. Kasai, K. and Jagiasi, A. R., Building separation rules
Research in Progress. University of Nevada, Reno, to avoid seismic pounding, Proceedings of the
NV, 1990, pp. 201-204. Structures Congress 1993. ASCE, Irvine, CA, 1993,
16. Davis, R.O. Pounding of buildings modeled by an pp. 199-203.
impact oscillator. Earthquake Engineering and 26. Blume, J. A., Newmark, N. M. and Corning, L. H.,
Structural Dynamics, 1992, 21, 253-274. Design of Multistory Reinforced Concrete Buildings for
17. Van Mier, J.G.M., Pruijssers, A.F., Rienhardt, H.W. Earthquake Motions. Portland Cement Association,
Monnier, T. Load-time response of colliding concrete Skokie, IL, 1961.
bodies. Journal of Structural Engineering ASCE, 1991, 27. Stavroulakis, G.E. and Abdalla, K.M. Contact
117, 354-374. between adjacent structures. Journal of Structural
18. NISEE, Corrected Motion Records, Earthquake Engineering ASCE, 1991, 117, 2838-2850.
Engineering Research Center, Berkeley, CA, 1994. 28. Westermo, B.D. The dynamics of interstructural con-
19. Hanson, R. D., Aiken, I., Nims, D. K., Richter, P. J. nection to prevent pounding. Earthquake Engineering
and Bachman, R., State-of-the-art and state-of-the- and Structural Dynamics, 1989, 18, 687-699.