8/13/2019 thesis.mj

1/76

8/13/2019 thesis.mj

2/76

6i)h uncertainty of the ductility desi)n strate)y is primarily attributed to7 Source7

8undamental of seismic base isolation by 9an), :en;o, Taiwan%

The desired stron) column wea( beam mechanism may not form due to existence

of walls.

Shear failure of columns due to inappropriate )eometrical proportion or short column

effect.

ase *solation by Sachin;andit, **T !elhi, "##$%

a.& 'i!mic rtro(itting )'tructural !trngt#ning&

*.& A!i!mic rtro(itting )+orc rduction&

1.1.1 'i!mic rtro(itting )'tructural 'trngt#ning&

This method is consistent with conventional seismic desi)n concept of structural

stren)thenin). *t involves stren)thenin) of the structure such that it can resist the lateral

loads. The stren)thenin) in the buildin)s can be achieved by additional bracin)s, shear walls,

wall panels, foundations etc. to the buildin) such that the lateral stiffness of the structure is

increased. *n the field of structural stren)thenin) the improvement in shear stren)th of the

structural elements is of prime importance, One ma/or )roup ,-ILTI" $ is there in the field

of shear connectors which are extensively bein) used in structural stren)thenin).

"

8/13/2019 thesis.mj

3/76

8i). 1.1a% Additional foundations.

8i).1.1b% Additional shear walls.

Source8i). 1.1a & 1.1b%7 ?ethods of seismic retrofittin) of structures

web.mit.edu@ist)roup@ist@documents@earth'ua(e@;art$.pdf , *ST )roup "##3%

8/13/2019 thesis.mj

4/76

8i). 1.1c% Bac(etin) of column. 8i).1.1d% Additional column.

8i). 1.1 c & d% 7

8/13/2019 thesis.mj

5/76

1.1. A!i!mic rtro(itting )+orc rduction&

This method is consistent with the aseismic desi)n philosophy. 6ere the ade'uacy of the

structure for lateral resistance is not important because it is aided by additional devices which

ta(e care of the expected seismic forces. This strate)y is classified under passiveand active

control of response of the structure.

C>ase isolators+, CDiscoelastic dampers+, Cfriction dampers+, Ctuned mass dampers+ are

examples of passive control. Supplemented dampin) systems are mechanical devices that can

be incorporated in the framed structure and dissipate ener)y at discrete locations throu)hout

the structure. These devices include either one of the yieldin) of mild steel, slidin) friction,

motion of pistons within fluids, orificin) of fluid or viscoelastic action of elastomeric

materials. !ampin) devices can provide also supplemental stiffenin) and stren)th tostructures that lac( such properties, in most cases without alterin) the existin) components.

All the above techni'ues have the flexibility to provide either more dampin), or stiffness, or

both, to better control the interaction with existin) components and reduce the seismic

demands without modification of the existin) structural components.

Active control can be achieved by usin) Caccentuated tunes mass dampers+, Cactuators+ and

Cactive tendons+. The principle of the active systems is to provide external corrective forces at

strate)ic points in the structure, to constrain the response within predetermined performance

limits. Active bracin) systems and active variable stiffness systems are systems built of

conventional structural components of structures enhanced with external forces that modify

either the effective dampin) , or the natural fre'uency of the system to produce more efficient

vibration suppression. An active control system is a dynamic system that comprises sensors,

controllers, control al)orithm and active control force )enerator which acts as an inte)ral

system.

$

8/13/2019 thesis.mj

6/76

1. Ba! I!olation

One of the most widely implemented and accepted seismic protection systems is base

isolation. Seismic base isolation is a techni'ue that miti)ates the effects of an earth'ua(e by

essentially isolating the structure and its contents from potentially dan)erous )round motion,

especially in the fre'uency ran)e where the buildin) is most affected.

*n recent years base isolation has become an increasin)ly applied structural desi)n techni'ue

for buildin)s and brid)es and especially for structures that must remain fully functional

durin) a ma/or earth'ua(e e.)., hospitals, fire stations, and emer)ency command centers.

?any types of structures have been built usin) this approach, and many others are in desi)n

phase or in construction.

*n *ndia we have >hu/ hospital buildin) which is the only existin) seismically

isolated buildin). Two more are in construction phase 7 i% A new ward bloc( at ET> Euru

Te) >ahadur% hospital Shadara !elhi, ii% Shimla hospital buildin) which is done with

desi)ns by faculty of **T 4anpur.

The ob/ective of base isolation systems is to decouple the buildin) structure from the

dama)in) components of the earth'ua(e input motion, i.e. to prevent the superstructure of the

buildin) from absorbin) the earth'ua(e ener)y. The entire superstructure must be supported

on discrete isolators whose dynamic characteristics are chosen to uncouple the )round

motion. Some isolators are also desi)ned to add substantial dampin). 8or e). F=> i.e.

Faminated =ubber >earin) with lead core provides substantial amount of dampin) by virtue

of ener)y dissipation in lead core.

!isplacement and yieldin) are concentrated at the level of the isolation devices,

and the superstructure behaves very much li(e a ri)id body. as shown in followin) fi)ures.

1." a & b%%

G

8/13/2019 thesis.mj

7/76

8i). 1."a% !eformation before base isolation

H

8/13/2019 thesis.mj

8/76

8i).1."b% !eformation after base isolation.

Source7 ?ethods of seismic retrofittin) of structures

web.mit.edu@ist)roup@ist@documents@earth'ua(e@;art$.pdf , *ST )roup "##3%

1..1 'uita*ilit/ o( *a! i!olation

arth'ua(e protection of structures usin) base isolation techni'ue is )enerally suitable if the

followin) conditions are fulfilled.as stated in a paper by Sa/al (anti deb **TE% seismic

isolation 7 An overview%

I The subsoil does not produce a predominance of lon) period )round motion.

I The structure is fairly s'uat with sufficiently hi)h column load.

I The site permits horizontal displacements at the base of the order of "## mm or more.

I Fateral loads due to wind are less than approximately 1#J of the wei)ht of the structure.

>ase isolation systems use a flexible layer at the base of the structure, which allowsrelative displacements between the foundation and the superstructure. !ue to the addition of

an isolation layer, the fundamental time period of the structure len)thens so as to move away

from the dominant time periods of )round motions, thereby reducin) the acceleration induced

in the structure.

K

8/13/2019 thesis.mj

9/76

Strate)ies to achieve seismic isolation includes7

;eriod shiftin) of structure. fi).1.%

8/13/2019 thesis.mj

10/76

8i). 1. % ;eriod shiftin) in case of base isolation.

source 7 !esi)n of seismic isolated structures by 0aeim & 4elly%

.

1.0 T/! o( I!olation '/!tm!

There are two basic types of isolation systems. They are

A. lastometric >earin)s

> .Slidin) System

1.0.1 Ela!tomtric Baring!

The system that has been adopted most widely in recent years is typified by the use of

elastomeric bearin)s, the elastomer is of either natural rubber or neoprene. *n this approach,

the buildin) or structure is decoupled from the horizontal components of the earth'ua(e

)round motion by interposin) a layer with low horizontal stiffness between the structure and

the foundation. This layer )ives the structure a fundamental fre'uency that is much lower

than its fixedbase fre'uency and also much lower than the predominant fre'uencies of the

)round motion. The first dynamic mode of the isolated structure involves deformation only in

the isolation system, the structure above bein) to all intents and purposes ri)id. The hi)her

modes that will produce deformation in the structure are ortho)onal to the first mode and

conse'uently also to the )round motion. These hi)her modes do not participate in the motion,so that if there is hi)h ener)y in the )round motion at these hi)her fre'uencies, this ener)y

cannot be transmitted into the structure. The isolation system does not absorb the earth'ua(e

ener)y, but rather deflects it throu)h the dynamics of the system. This type of isolation wor(s

when the system is linear and even when undamped- however, some dampin) is beneficial to

1#

8/13/2019 thesis.mj

11/76

suppress any possible resonance at the isolation fre'uency. Some examples of this system are

Faminated =ubber >earin)F=>%, 0ew Lealand isolation system0L%.

1.0.1.1 Laminatd Ru**r Baring

An F=> is made of alternatin) layers of rubber and steel with the rubber bein) vulcanized to

the steel plates. Therefore, the bearin) is rather flexible in the horizontal direction but 'uite

stiff in the vertical direction. 9ith its horizontal flexibility, the F=> provides protection

a)ainst earth'ua(es by shiftin) the fundamental fre'uency of vibration to a much lower value

and away from the ener)y containin) ran)e of the earth'ua(e )round motion. The horizontal

stiffness of the bearin) is also desi)ned in such a way that it can resist the wind forces with

little or no deformation. This baseisolation system has been used in a number of buildin)s inurope, Bapan and 0ew Lealand.

8i).1.3%. A tipical elastomeric bearin). source7 *ST Bournal, paper no.$3$,"##$ by Oliveto

& ?arletta, university of

8/13/2019 thesis.mj

12/76

isolator while the yieldin) property of the lead core serves as a mechanism for dissipatin)

ener)y and hence reducin) the lateral displacement of the isolator.

8i). 1.$ % Schematic dia)ram of a%F=> b% 0L System.

Source7 ?ultistory baseisolated buildin)s under a harmonic )round motion ;art *7 A

comparison of performances of various systems by 8aEun) 8an and Eoodarz Ahmadi %

1.0. 'liding '/!tm

The second basic type of isolation system is typified by the slidin) system. This wor(s by

limitin) the transfer of shear across the isolation interface. ?any slidin) systems have been

proposed and some have been used. *n ase isolators in which the only isolation mechanism is slidin) friction are classified as ;ure

8riction ;8% or Slidin)Boint baseisolation systems. *n this class of isolators, the horizontal

friction force offers resistance to motion and dissipates ener)y. These isolation devices have

no restorin) force and residual slip displacement between the structure and the foundation

will remain after each earth'ua(e. The examples of isolation devices in this system are ;ure

8riction ;8% or Slidin)Boint baseisolation system, =esilient8riction >ase*solation =

8>*% system, 8riction ;endulum system8;S%.

1.0..1 4ur +riction '/!tm! )45+&

>ase isolators in which the only isolation mechanism is slidin) friction are classified as ;ure

8riction ;8% or Slidin)Boint baseisolation systems. *n this class of isolators, the horizontal

1"

8/13/2019 thesis.mj

13/76

friction force offers resistance to motion and dissipates ener)y. These isolation devices have

no restorin) force and residual slip displacement between the structure and the foundation

will remain after each earth'ua(e.

8i).1.G% Fow friction bearin) device Source7 *ST Bournal, paper no.$3$,"##$ by Oliveto

& ?arletta, university of

8/13/2019 thesis.mj

14/76

8i). 1.H schematic dia)ram of a% ;8 & b% =8>* system.

Source 7 ,?ultistory baseisolated buildin)s under a harmonic )round motion ;art *7 A

comparison of performances of various systems by 8aEun) 8an and Eoodarz Ahmadi%

1.0..0 +riction 4ndulum '/!tm )+4'&

8riction pendulum 8;% isolators are special type of slidin) isolator that combines the ener)y

dissipation characteristics provided by friction with the restorin) force capability provided by

the spherical concave slidin) surface.

The isolator assembly consists of a polished stainless steel concave slidin) surface and an

articulated slider that is coated with a low friction composite material. The radius of

curvature of the spherical surface and the desired coefficient of friction between the slider

and slidin) surface are the properties of the 8; isolator that are specified by the desi)n

en)ineer. !urin) an earth'ua(e, the articulated slider moves within the spherical surface

followin) the curvature of the surface which results in pendulum motions for the supported

superstructure. As such, the period of the isolation system can be calculated based on

dynamics of a simple pendulum. An interestin) feature of the 8; isolator system is that the

isolation system period depends only on the radius of the slidin) surface and unli(e rubber

isolators, is independent of the buildin) wei)ht. !ue to the curvature of the slidin) surface, as

the slider moves up the surface durin) an earth'ua(e, a restorin) force is )enerated that

depends on the lateral displacement of the isolator.

13

8/13/2019 thesis.mj

15/76

8i). 1.K dia)ram depictin) mechanism of 8riction ;endulum System

Source7 fundamentals of seismic base isolation I 9an), :en;o%

1.0..6 Elctricid d +ranc )ED+&

An !8 baseisolator unit consists of a laminated steelreinforced% neoprene pad topped by

a leadbronze plate which is in frictional contact with a steel plate anchored to the base raft of

the structure. 9henever there is no slidin) in the friction plate, the !8 system behaves as an

F=>. *n this case, the flexibility of the neoprene pad provides isolation for the structure. Thepresence of the friction plate serves as an additional safety feature for the system. 9henever

the )round acceleration becomes very lar)e, slidin) occurs which dissipates ener)y and limits

the acceleration transmitted to the superstructure.

1.0..7 'liding R!ilint5+riction )'R5+&

The main feature of this system is the two frictional elements. 9henever there is no slidin) in

the upper friction plate, the S=8 base isolator behaves as an =8>* unit. 8or hi)h intensity

earth'ua(e )round accelerations, slidin) in the upper friction plate occurs which provides an

additional mechanism for ener)y dissipation and increases the effectiveness of the isolation

system.

1$

8/13/2019 thesis.mj

16/76

8i). 1.2 schematic dia)ram of a% !8 & b% S=8 system

Source 7 ,?ultistory baseisolated buildin)s under a harmonic )round motion ;art *7 A

comparison of performances of various systems by 8aEun) 8an and Eoodarz Ahmadi%

1.6 4rincil o( Ba! I!olation

The basic concept of base isolation is to reduce the fundamental fre'uency of structural

vibration to a value lower than the predominant ener)y containin) fre'uencies of earth'ua(e

)round motions. The other purpose of an isolation system is to provide means of ener)y

dissipation with which to reduce the transmitted acceleration to the superstructure.

Accordin)ly, by usin) base isolation devices in the foundations, the structure is essentially

uncoupled from the )round motion durin) earth'ua(es. Since a base isolator has a

fundamental fre'uency lower than both its fixed base fre'uency and the predominant

fre'uencies of )round motions, the first mode of isolated structure involves deformation only

in the isolation systems - the structure above remainin) almost ri)id. Thus, the hi)h ener)y in

the )round motion at the hi)her fre'uencies are deflected. *n this way, the isolation becomes

a very attractive approach where protection of expensive e'uipments and internal non

structural components is needed.

1G

8/13/2019 thesis.mj

17/76

Alon) with the elon)ation of time period of the structure, the forces in the structure are also

reduced due to the dampin) if present% of the base isolator. !urin) the earth'ua(e , the base

isolators )o into the inelastic zone and show hysteretic stress strain curve. The absorption of

the seismic ener)y is proportional to the area under the stress strain curve of the base

isolators.

The first desi)n )uide on base isolated structures was published by the Structural n)ineers

Association of est performin) parameters of linear and non

linear seismic baseisolator systems obtained by the power flow analysis.

?urat !icleli and Sri(anth >uddharam have done a

8/13/2019 thesis.mj

18/76

Bames S. >ailey and dmund 9. Allen have discussed The seismic isolation retrofittin) of

Salt Fa(e

8/13/2019 thesis.mj

19/76

U' alication! : Nuite a )ood number of base isolation wor(s r bein) done in Mnited

states 7 8oothill communities law and /ustice centre 8

8/13/2019 thesis.mj

20/76

=etrofittin) wor( usin) the approach of Ba! i!olation when carried out in the field should

follow the installation and construction se'uence )iven below 7

1.

8/13/2019 thesis.mj

21/76

8i). 1.1# *nstallation of temporary steel props.

>ench mar(s are to be introduced onto the column /ust above and below the final

position of the bearin), and measurements are to be ta(en, to enable subse'uent

chec(s to be performed of possible movements of the column.

Two horizontal cuts are to be made in the column usin) a diamond chain saw

8i)ure 1.11%. The bloc( of concrete in between is to be removed 8i)ure 1.1"%. The

"1

8/13/2019 thesis.mj

22/76

movements of the column above and below the cuts is then to be measured- in most

cases this is )enerally small, but can reach as much as Gmm. This is considered

acceptable. A bed of epoxy mortar is placed on the low half of the cut surface, and

the F=> is then rolled into place on steel ball bearin)s. The )ap above the bearin) is

then filled with epoxy mortar. The hydraulic /ac(s in the steel props are released and

the props are removed after curin) of the epoxy mortar

8i) 1.117 Saw cut throu)h the concrete column

""

8/13/2019 thesis.mj

23/76

8/13/2019 thesis.mj

24/76

8i) 1.1 7 Steel /ac(ets replacin) the discontinued reinforcin) bars.

"3

8/13/2019 thesis.mj

25/76

The bearin)s are wrapped in fire insulation, and brac(ets introduced to support

architectural finishes 8i)ure 1.13%.

8inal finishes are then applied.

8i) 1.137 8ireproofin) insulation

"$

8/13/2019 thesis.mj

26/76

1.

8/13/2019 thesis.mj

27/76

Submittal and review of the contractor+s detailed construction se'uence is re'uired

to ensure proper interpretation of the desi)n intent.

. OB8ECTIVE O+ 4RE'ENT WOR> :

To understand the effects of >ase isolation by analysis of fixed and isolated

buildin)s under consideration.

To study the effect of isolator stiffness asymmetry on a plan asymmetric Fshaped

H storey , bay , ! frame.

To study the effect of nonuniformity in isolator stiffness, on shear in columns of

different location at )round storey & torsional couplin) of superstructure, for 3

different cases viz 7 fixed, uniform isolated , with different isolator stiffness &

isolator stiffness in proportion of the mass ratio load comin) on each column%

"H

8/13/2019 thesis.mj

28/76

0. BA'E I'OLATION :

0.1 Ba! I!olator

>ase isolators are structural members that , li(e steel beams and columns , are part of a lateral

force resistin) system that enables a buildin) to respond acceptably to earth'ua(e

)round motion. A base isolator has a force versus lateral deflection curve as shown in fi)ure7

.1-

"K

Fig. 3.1) Idealized Force Deflection behaviour of an Isolator

8/13/2019 thesis.mj

29/76

Source fi)..1%7 A mathematical hysteretic model for elastomeric isolation bearin)s by

6wan),9u,;an,:an) "##"%7 Taiwan%

Two structural desi)n variables are obtained from this force versus deflection curve. The first

desi)n variable is the base isolator stiffness (b%, which is defined as

b

F Fk

+

+

=

The second desi)n variable is the base isolator viscous dampin) b%, which is

max max

1 area of loop"

bF

=

*n this e'uation , 8max and Pmax are the maximum absolute values of 8Q,8% and PQ,P%,

respectively. *f the dampin) of the isolator is very small , then the area of the loop is also

very small.

The chemical composition of the inner rubber layers used in base isolator determine the

lateral force versus lateral deflection characteristics of the base isolator. >ase isolator can

be cate)orized into two main cate)ories on the basis of the force deflection curve, ie linear

and 0on Finear .A base isolator which is desi)ned such that the line connectin) the

maximum force point in each cycle is linear is called as aLinear Base Isolator.

"2

8/13/2019 thesis.mj

30/76

Source fi).."% 7 >ase *solation, chapter K, !.E. 6art.%

>ase *solators can also be desi)ned to have envelope force versus deflection curve that are

not strai)ht lines but exhibit a non linear behavior. >ase isolators desi)ned to exhibit this

(ind of behavior are calledNon Linear Base Isolators.

#

Fig. 3.2)Force Deflection curve for a Linear Base Isolator

Fig. 3.3.)Force Deflection curve for non linear Isolator

8/13/2019 thesis.mj

31/76

Source 7 CC>ase *solation, chapter K, !.E. 6art.%

0. +i=d *a! !tructur

!esi)n starts with the specification of a $ J damped response spectrum for a desi)n basis

earth'ua(e. A Cfixed base+ structure is the structure that would exist if base isolators were not

used with the structure and elastic desi)n of this structure is performed for the desi)n basis

earth'ua(e where inelastic response parameters are reduced by a reduction factor. This factor

includes the ductility effect and the overstren)th of the structure. The seismic forces on Cfixed

base+ structure are calculated for desi)n response spectrum curvecorrespondin) to $J

dampin)%usin) response spectrum analysis. The stiffness (% and the mass m% of the Cfixed

base structure+ is (nown after the desi)n of the structure. The natural fre'uency, period of

vibration and critical dampin) ratio for the fundamental mode can be calculated for the Cfixed

base+ structure.

0.0 Trial D!ign o( I!olator

A structural dynamic analysis of a base isolated structure re'uires the properties of the base

isolator to be specified. 6ence, a trial desi)n of the base isolator is performed.

Two approaches can be used to develop a trial desi)n of isolator. >ase isolation ,

8/13/2019 thesis.mj

32/76

9here (b is the stiffness of the isolators under the column and Cm+is the

mass of the structure. The period of vibration is selected to provide a

good separation between fixed base period of vibration, Tn and base

isolated period of vibration, Tnb. Hence we consider a relation Tnb= n Tn

where n is 3 or greater. Using the value of T nb, the value of kb can be

calculated.

Alternativel the structural engineer can first set a value of k band then

use the e!uation to calculate the base isolated period of vibration.

bnb

k

m =

"t is important to recall that a positive benefit of using a base isolator isthe significant increase in damping of the structure. The damping of a

rigid structure sitting on base

isolators is effectivel the damping of base isolators. Therefore, a starting

value of damping #$%&'()* can be assumed for the base isolator, hence

the base isolated structure.

+ow using the estimated time period of base isolated structure and the

respective design response spectrum for assumed damping, the response

acceleration #a* and the response displacement #d* of the structure can

be calculated. The response parameters would be much lower than that of

the Cfixed base+structure, mainl due to higher time period and higher

damping.

"

8/13/2019 thesis.mj

33/76

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 2 4 6 8 10 12

Time (sec)

SpectralAccele

ration

5% damping

20% damping

9t#od :

"n method $. the intent was for the base isolated structure to have a

target natural period of vibration. As a result of this target natural period

of vibration selection, the magnitude of displacement of the base isolator

and the value of all desired structure response variables follow, using the

design basis earth!uake and method of structural analsis.

An alternative method can be used, which emphasi-es a desired response

parameter ofbase isolator displacement. The desired amplitude of base

isolator displacement ma be controlled b the open space called gap or

moat, around the base isolated building for architectural, mechanical,

electrical, or plumbing considerations. Also a variation in this method is to

set a value of absolute acceleration of the structure.

This criterion is )enerally re'uired for buildin)s, which house sensitive e'uipments and

machineries re'uirin) strin)ent control on the floor acceleration. Some times, acceleration

control may be re'uired for human comfort as well. After the parameter is decided

upon, the respective spectrum curve for an assumed high damping

Fig. 3.) !pectral "lot for different da#ping ratios$according to I! 1%&3'2((2)

8/13/2019 thesis.mj

34/76

#$%&'()* is considered for calculating the time period of the base isolated

structure.

0.6 Ba! I!olator 'lction

After initial trial desi)n of the base isolator, one needs to select a base isolator for the

structure. One can choose from the base isolators available, the one that exhibits similar

effective stiffness and assumed dampin), for the desi)n pseudo displacement, as considered

earlier. *t would not be possible to find an isolator that matches the exact re'uirement, but the

one that has closest parameters can be chosen for the structure.

0.7 4rliminar/ D!ign o( Ba! I!olatd 'tructur

>efore performin) a detailed analysis of the base isolated structure, a preliminary desi)n of

the base isolated structure is performed to modify the member size from those of the Cfixed

base+ structure, as the lateral seismic loads on the structure are expected to reduce

considerably.

The base isolated structure is desi)ned for the desi)n basis earth'ua(e reduced by the

response reduction factor. The response reduction factor is to include the effect of ductility

and overstren)th of the structure. *t is important to note that this response reduction factor is

different from the factor used in the desi)n of the fixed base structure. This is due to the fact

that the expected inelastic response of the Cfixed base+ structure and Cbase isolated+ structure

would be different.

3

8/13/2019 thesis.mj

35/76

Since, a preliminary calculation of time period and the dampin) of the base isolated structure

have been performed, an approximate value of the time period and the dampin) of the base

isolated structure is (nown. 6ence desi)n spectral acceleration can be found out for the base

isolated structure. The base shear of the base isolated structure can be calculated by

multiplyin) the spectral acceleration by the total mass of the structure. The fundamental

mode of the base isolated structure would be that of ri)id superstructure sittin) on flexible

base isolator, therefore, the base shear can be e'ually divided into all floors.

6ence the structure can be analyzed with fixed base and the lateral load applied e'ually at

each floor as calculated above. A preliminary sizin) of the sections of the structure can be

performed with this analysis.

0.< Anal/!i! o( t# Ba! I!olatd 'tructur

After selectin) the base isolator to be used in the structure, a detailed analysis of the base

isolated structure is performed to verify that the base isolators selected in the preliminary

phase are sufficient. *n other words, if the structure is not simplified to a ri)id box sittin) on

top of the base isolators, it will respond with the base isolators in a way that is within the

desi)n limit. 9e can perform a response spectrum analysis or a time history analysis to study

the behavior of the base isolated structure.

0.

8/13/2019 thesis.mj

36/76

Since the dampin) of the non linear isolator also depends upon the displacement of the base

isolator, the effective dampin) for the displacement calculated in the trial desi)n of the base

isolator is considered. 8or the detailed analysis it can not be assumed that the dampin) of the

structure is same as the dampin) of the base isolator. Therefore dampin) of the structure has

to be calculated. The effective dampin) for the different modes is calculated by formin) a

complete dampin) matrix

8/13/2019 thesis.mj

37/76

0.

8/13/2019 thesis.mj

38/76

The base isolators are modeled throu)h 0FFin( elements of SA; "###.A 0FFin( element is

a two/oint connectin) lin(. ach element is assumed to be composed of six separate

sprin)s, one for each of six deformational de)reesof freedom. All Finear@0onlinear

property sets contain linear properties that are used by the element for linear analyses, and for

other types of analyses if no other properties are defined.

Finear@0on linear property sets may have non linear properties that will be used for all non

linear analyses, and for linear analyses that continue from nonlinear analyses. 0on linear

behavior is only exhibited durin) time history analysis. 8or all other analysis, the lin(

element behaves linearly.

The non linar rorti! for each 0FFin( ;roperty must be of one of the various types

described below. The type determines which de)rees of freedom may be non linear and the

(inds of non linear force deformation relationships available for those de)rees of freedom.

8or each non linear type of 0FFin( ;roperty, there are six uncoupled linear effective

stiffness coefficients, >, one for each of the internal sprin)s. The linar ((cti? !ti((n!!

represents the total elastic stiffness for the 0FFin( element that is used for all linear analyses that start

from zero initial conditions.

8or each non lineartype of 0FFin( ;roperty, there are six uncoupled linear effective

dampin) coefficients, C, one for each of the internal sprin)s. >y default, each coefficient C

is e'ual to zero. The linar ((cti? daming represents the total viscous dampin) for the

0FFin( element that is used for responsespectrum analyses, for linear and periodic time

history analysis. ffective dampin) can be used to represent ener)y dissipation due to non

linear dampin), plasticity, or friction.

K

8/13/2019 thesis.mj

39/76

8or the non linear analysis, the plasticity property used for non linear analysis of the lin(

element is based on hysteretic behavior proposed by 9en 12HG%. 8i)..$%

Source7 SA; Analysis reference manual, pa)e no. "3", er(eley, Banuary "##H%

The non linear forcedeformation relationship is )iven by7

f = ratio k d +1 ratio% /ildz

where k is the elastic sprin) constant, Cd+ is deformation, /ild is the yield force, ratio is the

specified ratio of post yield stiffness to elastic stiffness k%, andz is an internal hysteretic

variable. This variable has a ran)e of W z W 1, with the yield surface represented by W z W =1.

The initial value ofz is zero, and it evolves accordin) to the differential e'uation7

9here Cexp+ is an exponent )reater than or e'ual to unity. Far)er values of this exponent

increases the sharpness of yieldin). The practical limit of exp is about "#.

2

Fig.3.) *Nonlinear Behaviour of NLlin+ ele#ent in !," 2(((

8/13/2019 thesis.mj

40/76

0.@ UBC Ba! I!olation D!ign 'ci(ication!

One of the few codes which provide the )uidelines for desi)n of base isolated structure is

Mniform >uildin) < is

)iven below.

The code recommends that the base isolated structure and the base isolator may be desi)ned

for the desi)n basis earth'ua(e, where the desi)n basis earth'ua(e is defined as that )round

motion that has a 1#J chance of bein) exceeded in $# years. 9hereas the stability of the

base isolator may be chec(ed for ?aximum

8/13/2019 thesis.mj

41/76

6. NU9ERICAL 'TUD :

6.1 nral !tud/ aramtr! :

The effect of nonuniformity in isolator stiffness on the torsional couplin) of superstructure

is bein) studied in this dissertation wor(. 8or that purpose an Fshape multistory frame is

bein) considered. Two parameters are bein) used as the measure of the torsional couplin)

that are namely the difference between the shear force values for )round storey columns of

particular location & the ratio of torsional fre'uency % to that of lateral fre'uency

% . ?atsa)ar Dasant & Ban)id =.S. >aseisolated >uildin) with Asymmetries due to

the *solator ;arameters Advances in Structural n)ineerin) Dol. K, no. G , !ec. "##$%%

*f there is lar)e difference between the shear force values for the different )round storey

columns then this indicates hi)her torsional effects in the superstructure i.e. structure will be

more torsionally coupled. Or it will vibrate in torsional directions. These shear force values

are bein) compared to have an idea about torsional couplin) in the structure considered.

The fre'uency ratio Y% i.e. the ratio of the torsional fre'uency % and the lateral

fre'uency % or Yy% )ives an idea as to what extent that structure is torsionally coupled.

8or example7

*f the torsional fre'uency is very low then the fre'uency ratio would be lower, i.e. Ome)a is

less. This shows us that the structure is torsionally very flexible, i.e. first fundamental% mode

of vibration itself could be in torsional direction. *n simple words, it will essentially vibrate in

its torsional direction .

And if the torsional fre'uency is very hi)h then the fre'uency ratio would be hi)her, i.e.

Ome)a is more. This shows us that the structure is torsionally very ri)id, i.e. vibration mode

in torsional direction may appear 'uite late almost absent%. *n simple words, it will not

vibrate at all in the torsional direction .

ccentricity i.e. the distance between the centre of )ravity

8/13/2019 thesis.mj

42/76

8or carryin) out of above analysis , an F shape H storey , bay, ! frame is bein) modeled

and analyzed. Structural members modeled are 7 beams , columns & slab. Then load on each

node is bein) calculated manually. Then coordinates for centre of )ravity @ centre of mass

are bein) calculated.

8or the isolated case the stiffness of isolators with is (nown. 0ow considerin) the

superstructure as ri)id body, the centre of stiffness is located for all the three cases for

isolated frame viz7 for uniform isolator stiffness , for different isolator stiffness , for isolator

stiffness in proportion of load comin) on individual column. *n this way the locations of

8/13/2019 thesis.mj

43/76

:i !istance of ith isolator alon) :axis from centre of ri)idity. as shown in fi). 3.1%

8i). 3.1 7 ;lan of the Fshaped frame showin) i & :i .

Thus the fre'uency ratios are bein) calculated & then the fre'uency ratio values for various

cases are bein) compared, to have an idea of the extent of torsional couplin) in the structure.

6. Numrical E=aml :

3

8/13/2019 thesis.mj

44/76

6..1 Dimn!ion! o( 0D (ram:

*sometric view, Top view & levation of reinforced cement concrete three bay, Fshape, H

storey frame are as shown in the 8i). 3." & 8i). 3. & 8i). 3.3 respectively.

storey hei)ht m is bein) modeled in SA;"###.

Sizes of the structural members are as follows7

>eam size 7 $## mm x $# mm.

Eround storey column size K## mm x K## mm.

=est all of the columns of G## x G## mm.

=oof slab 7 1$# mm thic(.

=est all floor slabs 7 "## mm thic( .

The beam & column sizes are (ept same for both fixed and isolated frame.

*sometric view, Top view & levation of reinforced cement concrete three bay, Fshape, H

storey frame are as shown in the fi)ures.

33

8/13/2019 thesis.mj

45/76

8i). 3." 7 *sometric view of Fshaped frame

3$

8/13/2019 thesis.mj

46/76

8i). 3. Top view of the Fshaped frame.

.

3G

8/13/2019 thesis.mj

47/76

8i). 3.3 7 levation of the fixed Fshaped frame. At Erid 1%

3H

8/13/2019 thesis.mj

48/76

8/13/2019 thesis.mj

49/76

*solator =M>% 7 ;laced at the base of column ">. 8i). 3.%

ffective stiffness in vertical direction 1K#$ 40@mm

ffective stiffness in horizontal direction #.2" 40@mm

The elevation of isolated frame is as shown in fi). 3.$% , an enlar)ed view of isolator is also

bein) shown.

8i). 3.$ 7 levation of *solated frame At Erid !%

32

8/13/2019 thesis.mj

50/76

6..0 'i!mic loading :

8or response spectrum analysis , the response spectrum )iven in *S 1K2"##" for !elhi

zone *D , L #."3% and medium soil Type **% is bein) used for seismic loadin). !ampin) in

analysis for both fixed base & base isolated structure is ta(en as $ J as the default value in

*S 1K2"##" response spectrum Finear isolator is bein) used so no additional dampin) will

be there due to the dampin) of isolator.

8or time history analysis , the time history of 0S component of l

8/13/2019 thesis.mj

51/76

6.0 Numrical 'tud/: 't! (ollo2d

The present study has been carried out to study the,E((ct o( non5uni(ormit/ in i!olator

!ti((n!! on tor!ional couling".

! Analysis of an F shape, H storey, frame with dimensions as )iven in 3.".1, is bein) done

in the present wor( for 3 different cases, to study the effect of isolator stiffness asymmetry on

the torsional parameters of superstructure.

1 8ixed base frame with no isolators.

". *solated frame with all isolators of uniform stiffness.

. *solated frame with isolators of randomly different stiffness.

3. *solated frame with isolators of different stiffness, *solator stiffness in

proportion of the load comin) on the individual column or in proportion to

mass ratio%.

The analysis is carried out, Then results are compared for the above 3 cases, basically the

base shear in the )round storey column of some particular locations are bein) compared to

observe the difference in the ratio of those base shear values, which indirectly represent the

extent of torsional couplin). Also the fre'uency ratio is bein) compared for all the isolated

cases. 8re'uency ratio is a direct measure of torsional couplin) in the structure.

*solators used for isolation are laminated rubber isolators, & they are of linear type. ?odelin)

and analysis of the frame is bein) done in SA;"### Dersion11 Advanced% usin) both

response spectrum *S 1K27"##"% and time history F

8/13/2019 thesis.mj

52/76

Analysis of fixed base ! =ase isolated frames for *S 1K2

"##" response spectrum seismic loadin).

0onlinear time history analysis of above all three base isolated ! frame usin)

F

8/13/2019 thesis.mj

53/76

As analysis results, the values of fundamental time

period for 1st mode% of the structure, as shown in

plan in fi). $.1% and base & top displacement i.e. absolute and relative displacements% are

obtained for all four cases, 1,",,3% usin) modal & both response spectrum and nonlinear

time history analysis. And after analysis the displacement time histories for top displacement

are obtained. All above results are obtained directly from SA;.

8i). $.1 7 ;lan specifyin) location of columns 1A,1!,">,"!,3A,3>.

Then

8/13/2019 thesis.mj

54/76

7.1 R!ult! o( Anal/!i! carrid out :

nral:

The four structural models under consideration are analyzed in SA; "### Der. 11% for

=esponse Spectrum *S 1K2"##"% & Time 6istory F

8/13/2019 thesis.mj

55/76

Table $. 7ase shear values in columns 1A,1!,">,"!,3A,3> location shown in fi) $.1% for both

response spectrum & time history is )iven in table $.. And ratio of extreme base shear

values i.e. maximum in ">% & minimum in 1!% value of shear in )round storey columns%

& stiffness values for isolators used in each case are also )iven in the same table $.%

olumn shear !alues("#)

Ratio o( =trm

column !#ar ?alu!

) BG1D&$solatorstiffness

ol.no. Response Spectrum Time %istor&

Res.Spec.

Time%is.

'('#mm)

. Fixed

!1 372.856 374.44

1.37 1.374

!2 372.343 374.313

!3 4"7.075 51".45.

!4 365.266 37".243

!5 365.81 382.6

!6 363.802 378.12

*. +niform isolated

!1 46.133 10".51

2.75 2.770.685

!2 46.03 10".275

!3 114.6 26".11

!4 41.0"8 "5.7

!5 42.05 "8.35

!6 41.683 "7.2

%ontd&&

$$

8/13/2019 thesis.mj

56/76

,. $solators with different stiffness

!1 40.813 "2.05

2.65 2.66

!2 40.706 "1.82'1 (0.275

!3 "".205 223.8725

'2(0.65!4 36.863 82."5

!5 36." 83.076 '3(0."75

!6 37.414 84.22

-. $solators withStiffness in mass ratio

!1 80.25 178.53

1.47 1.476

!2 80.203 178.65 '1 ( 0.45

!3 "2.16 204 '2(0.716

!4 5".45 131.685 '3(0."20

!5 6".513 154.104

!6 62.58 138.553

Table $.3 7 8re'uency ratio & ccentricity values7

8/13/2019 thesis.mj

57/76

Table $.$ 7 >ase and top displacementsin mm.%7

8/13/2019 thesis.mj

58/76

9ODE '-A4E' O+ +OUR +RA9E':

8irst mode shapes & deformed plan for fixed & all of three base isolated frames fi). $."

a,b,c,d,e,f% and The top displacement time histories fi). $. a,b,c,d% are obtained as shown in

the followin) fi)ures 7

8i). $." a% 1stmode shape for fixed base frame 7 ! view

$K

8/13/2019 thesis.mj

59/76

8i). $." b% 1stmode shape for fixed base frame 7 top view

8i). $." c% 1stmode shape for uniform isolated frame 7 top view

$2

8/13/2019 thesis.mj

60/76

8i). $." d% 1stmode shape for uniform isolated frame 7 ! view

G#

8/13/2019 thesis.mj

61/76

8i). $." e% 1stmode shape for different random% isolator stiffness frame 7 ! view

G1

8/13/2019 thesis.mj

62/76

8i). $." f% 1stmode shape for isolator stiffness in mass ratio frame 7 ! view

G"

8/13/2019 thesis.mj

63/76

8i). $. a% Top displacement mm% time history for frame restin) on fixed base

8i). $. b% Top displacement mm% time history for frame restin) on isolators of uniform

stiffness

G

8/13/2019 thesis.mj

64/76

8/13/2019 thesis.mj

65/76

Bnding momnt ?ariation!for fixed frame is as shown in fi). $.3 a% & b% for response

spectrum & time history analysis respectively. The values of maximum bendin) moments

obtained from analysis results for the structure usin) SA;% are )iven in Table $.H pa)e $%

G$

8/13/2019 thesis.mj

66/76

8i). $.3 a% >?! at Erid " for fixed frame sub/ected to response spectrum *S

1K27"##"% loadin).

GG

8/13/2019 thesis.mj

67/76

GH

8/13/2019 thesis.mj

68/76

8i). $.3 b% >?! at Erid " for fixed frame sub/ected to time history l

8/13/2019 thesis.mj

69/76

individual column. This ratio is one measure of torsional couplin) in the

superstructure , if the difference between these shear values is lar)e or in other words

if ">@1! is hi)h then this implies more torsional effects in the structure.

E++ECT ON +REUENC RATIO F ECCENTRICIT : The most important

parameter to have an idea of extent of torsional couplin) in the superstructure is the

fre'uency ratio % that is the ratio of torsional % & lateral fre'uencies

yxor % . 8or the same value of eccentricity or other parameters li(e ">@1!

here% , the torsional couplin) will be more if the fre'uency ratio is low i.e.

superstructure will vibrate in torsional direction in earlier modes.

8re'uency ratios for each case is bein) displayed in Table $.3 , is more for

case " 1.$"H% when the structure is restin) on isolators of uniform stiffness than

that for the case 3 1.31% when structure is restin) on the isolators of stiffness in

proportion of the load comin) on individual column. As per this observation in case "

the structure will be less torsionally flexible than in that of case 3% , this contradicts

the above ">@1! lo)ic. >ut actually the )overnin) factor for torsion is the fre'uency

ratio neither the ratio of shears nor the eccentricity values. Thus as far as torsion is

concerned uniform case is performin) better.

E++ECT ON RELATIVE TO4 DI'4LACE9ENT F INTER 'TORE DRI+T:

The results show that relative top displacement & inter storey drifts are reduced upto

1@Gth times & 1@Hth times respectively while movin) from fixedto the isolated fram.

*f top relative displacement table $.$% & maximum inter storey drift table $.G% is

considered then that is 'uite low in case 3 i.e. when structure is restin) upon the

isolators of stiffness in ratio of the mass comin) on the individual columns. >ut the

top relative displacement in case of isolated frame is in order of few mm+s , inter

storey drift 'uite low in isolated frames because of ma/or amount of displacement

occurs at the isolation level itself.%. So more important parameter of concern is the

fre'uency ratio as compared to this relative displacement & inter storey drift which

G2

8/13/2019 thesis.mj

70/76

are actually matter of few mm+s only in case of isolated frame, The reason bein)7 the

ma/or amount of lateral drift is accommodated at the isolation level, hence there is

substantial lateral deflection in isolators but the inter storey drifts are very low.

E++ECT ON 9AHI9U9 BENDIN 9O9ENT : 8urther when it comes to

bendin) moment values table $.H% it is clearly observed that 8or isolated cases the

maximum bendin) moment reduced upto 1@1# thof that of fixed case . The value of

maximum bendin) moment is very hi)h $$.$ 40m % in case when the structure is

restin) on the isolators of stiffness in proportion of the load comin) on individual

column compared to K"."% when the structure is restin) on the isolators of same

stiffness. The values of bendin) moment shown here is when F

8/13/2019 thesis.mj

71/76

On the basis of numerical study & discussions of results obtained after analysis, followin)

conclusions were drawn 7

>ase isolation substantially increases the time period of structure & hence

correspondin)ly reduces the base shear. As observed in the present study Table $.1

& $."% the time period is bein) increased upto ." times & base shear is reduced upto

1@Gthof that of fixed one.

The deformed shape in plan for fixed & isolated uniform stiffness% 7 fi). $." a% &

$." b% clearly indicates the reduced twistin) effect in case of isolated frame. This can

be ensured by comparin) the values of displacements in other " normal directions

than M1 i.e. M" & M. Thus seismic isolation considerably reduces the twistin)effect.

The study carried out here and results obtained su))ests the structure restin) on the

isolators of uniform stiffness as a better option than that restin) on the isolators of

stiffness in proportion of mass ratio as far as torsional couplin) because fre'uency

ratio is observed to be hi)hest1.$"H for uniformly isolated case amon)st the all 3

cases considered% & maximum bendin) moment is observed to be minimum12."H

40m for response spectrum & K". 40m for time history for uniformly isolated

case amon)st the all 3 cases considered% is concerned. 9hich is 'uite contradictory

from the eccentricity approach for torsional behavior.

As far as the relative top displacement & maximum inter storey drifts are concerned,

*n this present study, the results obtained su))est that the case 3 i.e. the when the

stiffness of isolators are in the ratio of the load comin) on individual column )ives

lower relative top displacement & maximum inter storey drifts as compared to case "

i.e. uniform isolator stiffness.

H1

8/13/2019 thesis.mj

72/76

The top displacement time histories for fixed & isolated cases conspicuously shows

the reduction in fre'uency & hence ma)nified time period in case of isolated frame.

And the base shear value will )et reduced correspondin) to the increased time period

of the structure.

>ased on the study carried out & discussion of results it is observed that

how effective seismic isolation is considerin) various aspects such as7 base shear, inter storey

drifts, maximum bendin) moments & column shears etc. *n case of torsion7 Analysis results

of the study su))est that isolators of uniform stiffness are better option as compared to

isolators of stiffness in ratio of column load, when torsional couplin) is concerned. The

reason bein) increase in fre'uency ratio in case of uniform isolation.

. 'CO4E +OR +UTURE RE'EARC- :

There is much to explore in this flourishin) field of seismic isolation. Specially in our

country a lot of research wor( can be done & needed to be done. *n this particular

dissertation wor( the study is bein) carried out only for 1stmode of vibration the effect of

hi)her modes was not bein) considered , the effect of hi)her modes on torsional

couplin) of superstructure can be a stuff to be explored . 8urther the superstructure was

assumed to be perfectly ri)id for this study, so the effect of superstructure flexibility can

also be investi)ated. 8urther the parametric studies can be conducted mi)ht be based on

extent of eccentricity or superstructure stiffness variation etc.

H"

8/13/2019 thesis.mj

73/76

RE+ERENCE'

1% Analysis =eference ?anual 7 SA; "###,

8/13/2019 thesis.mj

74/76

H.% *S 1K2"##" part *7 ase *solation

**T !elhi, "##$.

13.% Sa/al (anti deb **TE% Seismic isolation 7 an overview.

1$.% S(inner , =obinson & ?c Derry , An *ntroduction To Seismic *solation, 122, Bohn

9iley & Sons, *nc., 9ellin)ton, 0ew Lealand.

1G.% Tervor . 4elly, >ase *solation of Structures 7 !esi)n )uidelines 6olmes

8/13/2019 thesis.mj

75/76

1H.% Mniform >uildin)

8/13/2019 thesis.mj

76/76

http://www.csiberkeley.com/http://www.nicee.org/