International Journal of Civil Engineering and Technology (IJCIET)

Volume 8, Issue 2, February 2017, pp. 570578 Article ID: IJCIET_08_02_059

Available online at http://www.iaeme.com/IJCIET/issues.asp?JType=IJCIET&VType=8&IType=2
ISSN Print: 0976-6308 and ISSN Online: 0976-6316

IAEME Publication Scopus Indexed

POUNDING MITIGATION USING PALL FRICTION

DAMPERS
Dr. A.K. Sinha
Professor, Department of Civil Engineering, NIT, Patna

Saket Kumar
Research Scholar, Structural Engineering, NIT Patna, Patna

ABSTRACT
Pounding is one of the major reasons of structural damages in buildings, constructed very near
to each other, when they are subjected to a strong ground motion during an earthquake. In
metropolitan cities where buildings have been constructed very near to each other due to high land
value and lack of awareness about pounding, pounding mitigation is very necessary for structural
safety and life safety.
This paper deals with the study of effect of pounding between adjacent buildings having
insufficient seismic gap after installation of friction dampers into taller building model. Time history
analysis has been done to evaluate the pounding effects for different models of building. The study
has shown the effect of pounding on global response of adjacent structures in terms of acceleration,
and impact force.
Key words: Pounding, Global response, Earthquake, Friction Dampers, Time History Analysis and
SAP 2000
Cite This Article: Dr. A.K. Sinha and Saket Kumar, Pounding Mitigation Using Pall Friction
Dampers. International Journal of Civil Engineering and Technology, 8(2), 2017, pp. 570578.
http://www.iaeme.com/IJCIET/issues.asp?JType=IJCIET&VType=8&IType=2

1. INTRODUCTION
Experiences of past and recent earthquake-damages have well established pounding as one of the major
causes of structural damages in buildings, constructed very near to each other or without any gap at all [1, 2,
7]. Pounding, which is a collision between adjacent buildings during an earthquake, commonly occurs due
to their different dynamic characteristics, adjacent buildings vibrate out of phase and there is insufficient
seismic gap between them. This situation can easily be seen in metropolitan cities where buildings have been
constructed very near to each other due to very high cost of land and lack of knowledge about pounding and
its consequences [2].
In case of pounding, during strong ground motion, the acceleration at pounding level considerably
increases and generates extra pounding force which causes structural damages or sometime results into
building collapse [10]. And to avoid this situation of pounding, regulations of minimum seismic gap between

http://www.iaeme.com/IJCIET/index.asp 570 editor@iaeme.com

 Pounding Mitigation Using Pall Friction Dampers

buildings have been formulated but it is often seen that these regulations are neither followed by landowners
nor strictly implemented by respective governments. This leads to a situation like Mexico City earthquake
(1985) where, 40% of 330 collapses or severely damaged buildings are cause of pounding only [7]. That is
the reason why proper mitigation for pounding in old adjacent buildings constructed very near to each other
is very necessary not just for structural safety, non-structural safety and life safety but for the making critical
buildings like hospital running operational.
Incorporating dampers into buildings are well established and reliable technique to control the response
of building, so Pall friction dampers have been used for the study. As the friction brake is widely used to stop
the motion of a moving body so by using the Pall friction dampers, the motion of a shaking building can be
slowed down by dissipating the energy in friction [11]. It is also the most effective, reliable and economical
way of dissipating kinetic energy in comparison to conventional methods like concrete shear walls or rigid
steel bracing.

1.1. SEISMIC GAP REQUIRED AVOIDING POUNDING

It is well established fact that providing a proper seismic gap between adjacent buildings is one of the best
methods to reduce the effect of pounding [10].

1.1.1. Square root of sum of squares (SRSS) rule

= + (1)

1.1.2. Absolute Sum (ABS) rule

S= + (2)
Where,
S = Seismic gap required
= Peak displacement response of building A
= Peak displacement response of building B

1.1.3. Double Difference Combination (DDC) method

= = + (3)
Where, , and are mean peak values of , ( ), ( ) and ( ) respectively. , is
correlation coefficient which depends upon the period ratio, = , as well as A and B, and is given by

(4)
! ! "! #$! %.$'.(
(')$ ) #*$! ! "'#$ %#*(! #! )$

Where, TA, A and TB, +B are natural periods and damping ratio of systems A and B, respectively.
Seismic gap calculated by using double difference combination (DDC) method was much more accurate
than that of square root of sum of squares (SRSS) method [4].

http://www.iaeme.com/IJCIET/index.asp 571 editor@iaeme.com

 Dr. A.K. Sinha and Saket Kumar

1.2. NON LINEAR DYNAMIC ANALYSIS

Non-linear dynamic analysis has been carried out considering various earthquake ground motion of different
PGA. The equation of motion for the structure when it subjected to ground motion is given as [13]:
,-./012 3 + ,4./053 + ,6.708 = ,-.,9./0:2 3 (5)
Where,
,;. is mass matrix, ,< . is damping matrix and ,= . is stiffness matrix of the building. 7>8 and 7>? 8 are
displacements of superstructure and base of the building. />2? 3 and />2@ 3 are base acceleration and acceleration
relative to ground. ,A. is the earthquake influence coefficient matrix.
Further, all non-linear properties are restricted to the non-linear link element i.e., gap element, only. The
above non-linear dynamic equation considering the superstructure as elastic and link as non-linear can be
written as:
,-. /02(B)3 + ,4. /05(B)3 + ,6C . 70(B)8 + $D (B) = $(B) ,$D (B) 6D0(B). (6)
Where,
[K] = [KL] + [KN] (7)
[M] is diagonal mass matrix; [C] is the proportional damping matrix; [KL] is stiffness matrix of all linear
elements; [KN] is stiffness matrix for all of the non-linear degrees of freedom; rN stands for the vector of
forces from non-linear degrees of freedom in the gap elements; r(t) in the equation is vector of applied load;
/>2( )3, />5( )3 and 7>( )8 are the relative acceleration, velocity and displacement with respect to ground,
respectively. The effective stiffness at non-linear degrees of freedom is arbitrary, but the value of it varies
between zero and the maximum stiffness of that degree of freedom.
The dynamic behavior of the structure installed with Friction Damper (FD) can be represented by
equation (8).
,-./012 3 + ,4./053 + ,6.708 + FG I J(B) = ,-.,9./0:2 3 (8)
Where,
,;. is mass matrix, ,< . is damping matrix and ,= . is stiffness matrix of the building. 7>8 and 7>? 8 are
displacements of superstructure and base of the building. />2? 3 and />2@ 3 acceleration relative to ground. KL
is the stiffness of damper brace system, M is the displacement of the damper brace system, ( ) is hysteretic
variable for the friction damper, ,A . is the earthquake influence coefficient matrix.

1.3. GAP ELEMENT

Gap element is a link element defined in SAP 2000. It is compression only member and is used to model the
collision between buildings and simulating the effect of pounding. The main purpose of providing gap
element between adjacent buildings is transmitting the force through the link only when contact occurs and
the gap is closed.

Figure 1 Gap element model from SAP 2000

http://www.iaeme.com/IJCIET/index.asp 572 editor@iaeme.com

 Pounding Mitigation Using Pall Friction Dampers

The force deformation relationship of gap element is given below.

F(Q + :RS), UO(Q + :RS) < W
O=P (9)
W, GBJ $XUY

Where, k is spring constant, gap is the initial opening which must be positive or zero and d is the relative
displacement across the spring.
Generally stiffness of gap element (k) is recommended as one or two orders of magnitude greater than
equivalent adjacent springs. Here stiffness of gap element has been taken 4.776 105 kN/m.

1.4. MATHEMATICAL MODELLING OF FRICTION DAMPERS

The idealized force displacement curve for a damper is shown (Figure 2). The equivalent stiffness = [ and
damping \ [ for a damper can be evaluated using the curve as [12]:
] # ^] )
= [ = # ^)
(10)
_ cd c Le LLf
\ [ = b j (11)
`a ]ghi ghi

Figure 2. A generalized force displacement curve for dampers to represent formulation of equation (10) and (11) for
linear dampers (dashed straight line) and for non-linear dampers (solid bilinear curve)
Where, F is the force and is the displacement and empirical expressions for = [ and \ [ depends upon
material properties and characteristics.

2. METHODOLOGY OF THE STUDY

The study has been carried out by assuming two different cases viz. (i) adjacent buildings of different heights
without dampers (ii) buildings with different height with dampers. Adjacent buildings are modeled in
SAP2000v17 and gap elements are introduced between buildings to analyse the effect of pounding between
them. Pall friction dampers have been used in one case for analysing the effect of dampers on pounding
between buildings. Non-linear dynamic analysis has been done by using response spectra method and time
history analysis. The gap between adjacent buildings has been kept 50 mm. Response envelopes of adjacent
buildings are provided in terms of acceleration of floor in first case. The impact forces achieved by
incorporation of gap elements are also shown in form of response envelop for different ground motions. The
reduction in impact force due to installation of dampers has been shown in form of global responses of the
building.

http://www.iaeme.com/IJCIET/index.asp 573 editor@iaeme.com

 Dr. A.K. Sinha and Saket Kumar

Four types of ground accelerations from 4 different earthquakes which are Petrolia (1992), Northridge
(1994), El Centro (1940), Loma Prieta(1989) have been applied to models of different cases to find response
of the models and to plot response envelope for acceleration and impact force of models. The peak ground
acceleration (PGA) of these ground motions varies from 0.22g to 0.883g.

3. MODELLING OF THE STRUTURES

Adjacent buildings are modeled in SAP2000v17 and gap elements are introduced between buildings to
analyse the effect of pounding between them.
In 1st case building of 7 stories (say building A) and building of 5 stories (say building B) are considered.
Live load of 3kN/m2 and 2.5kN/m2 has been used for building A and building B respectively. In this case no
any types of dampers are installed.
In 2nd case models of 7 stories (say building A) and building of 5 stories (say building B) are considered.
Live load of 3kN/m2 and 2.5kN/m2 has been used for building A and building B respectively. In this case,
Pall friction dampers are installed in building A in zigzag pattern. Following properties have been used in
modeling of friction dampers used for the analysis.

Table 1 Damper properties used in modeeling

Yield
Effective Effective Post yield Yielding
Link Type Mass Weight strength= slip
Stiffness* Damping stiffness ratio exponent
load
- (Kg) (KN) (KN/m) (=kl/n) (KN) - -
Plastic
429.32 4.2116 23772.853 0 700 0.0001 10
(Wen)
*Stiffness provided by damper brace system

(a)

(b)

Figure 3. (a) Plan view and (b) Elevation view of adjacent buildings, where floor levels are same, modeled in case I

http://www.iaeme.com/IJCIET/index.asp 574 editor@iaeme.com

 Pounding Mitigation Using Pall Friction Dampers

Figure 4 Elevation of buildings with friction dampers modeled in case II

4. NUMERICAL STUDIES
Global responses of buildings after pounding are given below in terms of acceleration for different cases
when buildings subjected to different ground motions.

CASE I: BUILDINGS OF DIFFERENT HEIGHT BUT SAME FLOOR LEVEL

For this case study, models of 7 storey building (say building A) and 5 storey building (say building B) have
been considered.

4.1.1. Response envelope for acceleration

BUILDING A
8
7 NORTHRIDGE (0.8g)
6
5 PETROLIA(0.6g)
Storey

4
3 ARRAY
2
1 Elcentro
0
-50 0 50 100 150
Acceleration

Figure 5. Response envelope of building A in terms of acceleration (case I )

http://www.iaeme.com/IJCIET/index.asp 575 editor@iaeme.com

 Dr. A.K. Sinha and Saket Kumar

6 BUILDING B
5 NORTHRIDGE (0.8g)
4
PETROLIA(0.6g)
Storey

3
2 ARRAY

1 Elcentro
0
-150 -100 -50 0 50 100
Acceleration

Figure 6. Response envelope of building A in terms of acceleration (case I)

4.1.2. Response envelope for impact force

Impact Force (link A-B)

6
5 Y=0
4 Y=3.5
iStorey

3
2 Y=7

1 Y=11.5
0
-2500 -2000 -1500 -1000 -500 0
Impact force

Figure 7 Response envelope of impact force of collisions between buildings during pounding for El Centro ground
motion.
Impact forces have been evaluated using gap elements, linking between adjacent buildings. These are
maximum at roof top of the building B i.e., building of lower height. Difference between the impact forces
at 1st floor and 2nd floor is very high. Impact forces at Y=3.5 and Y=7 are higher than impact forces at Y=0
and Y=11.5 plane of building throughout the whole length of the building B

4.2. CASE II: BUILDINGS WITH FRICTION DAMPERS INSTALLED IN TALLER

BUILDING
This is case type I and for this case, models of 7 storey building (say building A) and 5 storey building (say
building B have been considered. The floor levels of every building in the series have been kept at same
level. The elevation view of building model has shown in figure 4 and plan view is simillar as shown in figure
3 (a).

http://www.iaeme.com/IJCIET/index.asp 576 editor@iaeme.com

 Pounding Mitigation Using Pall Friction Dampers

5. COMPARATIVE STUDY OF IMPACT FORCES

Comparative study has been shown here in form of global responses of buildings, modeled in case I, (i.e.,
buildings without friction dampers) and modeled in case II, (i.e., buildings with friction dampers in higher
storey building) , in terms of impact force between buildings when they are subjected to different ground
motions. Here in figure 8 (a), (b), (c), (d) are global responses of impac forces when buildings of different
cases are subjected to different ground motions. In graphs, suffix 1 denotes the responses in terms of impact
force for case I and similarly suffix 2 stands for results in case 2. It is clearly evident from the graphs impact
force in case II are much lesser than what in case I

Elcentro Array
6 6

5 5

4 4

3 3
E1 A1
2 2
E2 A2
1 1

0 0
-3000 -2000 -1000 0 -2000 -1000 0
Impact Force (kN) Impact Force (kN)

(a) (b)

Petrolia Northridge
6 6

5 5

4 4

3 3
P1 N1
2 2
P2 N2
1 1

0 0
-3000 -2000 -1000 0 -4000 -2000 0
Impact Force (kN) Impact Force (kN)

(c) (d)

Figure 8 Global responses of impact force between two buildings of two different cases when subjected to different
ground motions viz. (a) Elcentro, (b) Array, (c) Petrolia, (d) Norrthridge.

http://www.iaeme.com/IJCIET/index.asp 577 editor@iaeme.com

 Dr. A.K. Sinha and Saket Kumar

6. CONCLUSIONS
Global responses of adjacent structures in terms of acceleration, amplify in case of pounding between them
when they subjected to strong ground motion than in no-pounding case. For pounding mitigation, dampers
are very suitable and convenient technique to use over conventional methods.
It is observed that there are considerable reductions in impact forces due to installation of friction damper
in adjacent buildings which is highly promising for pounding mitigation

REFRENCES
[1] Anagnostopoulos, S. A., Spilliopoulos, K V (1992), An Investigation of earthquake induced ponding
between adjacent buildings, earthquake Engineering & Structural Dynamics. Vol.21, No.4, and Pp.289-
302.
[2] Anagnostopoulos, S. A.,(1996), Building Pounding Re-examined : How Serious A problem is it? Paper
No. 2108, Eleventh World Conference on Earthquake Engineering, 1996.
[3] Arash Rezavani and A. S. Moghadam (2004), Using Shaking Table to Study Different Methods of
Reducing Effects of Building Pounding During Earthquake, 13th World Conference on earthquake
Engineering, paper No. 698, Canada.
[4] Bipin Shreshta (2013), Effects of Separation Distance and Nonlinearity on Pounding Response of
Adjacent Structures, International Journal of Civil and Structural Engineering Volume 3, No. 3.
[5] Chetan J. Chitte, Anand S. Jadhav, Hemraj R. Kumavat (2014), Seismic Pounding Between Adjacent
Building, IJRET; eISSN:2319-11631.
[6] Chenna Rajaram, Pradeep Kumar Ramancharla (2012), Comparison of Codal Provisions on Pounding
between Adjacent Buildings, International Journal of Earth Sciences and Engineering, February 2012
P.P. 72-82.
[7] Gregory L. Cole, Rajesh P. Dhakal and Fred M. Turner, Building Pounding Damage Observed in the
2011 Christchurch earthquake, NZSEE Conference, 2010.
[8] IS 4326: 1993, Indian Standard Code of Practice for Earthquake Resistant Design and Construction of
Buildings, New Delhi.
[9] IS 1893(part I): 2002, Indian Standard Criteria for Earthquake Resistant Design of Structures, part I
General provisions and buildings, New Delhi.
[10] K. Kasai, V.Jeng, P. C. Patel and J.A. Munshi (1992), Seismic Pounding Effects Survey and Analysis,
ISBN 9054100605, Earthquake Engineering, 10th World Conference, Balkerna, Rosterdam, 1992.
[11] A. Malhotra et all, Friction dampers for seismic upgrade of St. Vincent Hospital, Ottawa, 13 WCEE,
Canada, 2004, Paper no.- 1952.
[12] S. S. Sanghai and S. N. Khante, Seismic Response of Unsymmetric Building with Optimally Placed
Friction Dampers. International Journal of Civil Engineering and Technology, 8(2), 2017, pp. 7288.
[13] A.K. Sinha and Sharad Singh, Seismic Protection of RC Frames Using Friction Dampers. International
Journal of Civil Engineering and Technology, 8(2), 2017, pp. 289299.
[14] A.K. Sinha and Sharad Singh (2017), Seismic protection of RC frames using friction dampers,
International Journal of Civil Engineering and Technology, 8(2), 2017, pp. 289299.
[15] A.K. Sinha and Saket Kumar (2017), Effect of pounding on a global responses of buildings constructed
in a row, International Journal of Civil Engineering and Technology, 8(1), 2017, pp. 882890.

http://www.iaeme.com/IJCIET/index.asp 578 editor@iaeme.com