OF STRUCTURES
EARTHQUAKE
RESISTANT DESIGN
OF STRUCTURES
PANKAJ AGARWAL
Assistant Professor
Department of Earthquake Engineering
Indian Institute of Technology Roorkee
and
HANISH SHRIKHANDE
Assistant Professor
Department of Earthquake Engineering
Indian Institute of Technology Roorkee
,t.
Rs. 425.00
It! 2006 by Prenti(",.;;Hal! of India Private Limited, New Delhi. All rights reserved. No part of this book
may be repfOdoced in any iorm, by mimeograph or any other means:, without permission in writing
from the publisher,
'SBN812032S922
T'le export rights of thiS boCll\ are vested solely with the publisher.
L
Puhlished by Asoke K Ghosh, PrenticeHall of lndia Private limited, Mw97, Connaughl Circus,
New Delhi110001 and Printed by Jay Print Pack Private Limited, New Delhi"110015,
Contents
Contributors .................. ,., .... ,......... ,............... ,....... ,... , .. ,...................... ,...... ..................... ,nv
Part I
EARTIlQUAKE GROUND MOTION
1. Engineering Seismology ._...._................................_..................... 344
1.1 Introduction ... ,... ., ............. ,................ ,............ ,............. ., ....... ,....... ." ...... ." .. . ............... 3
1.2 Reid's Elastic Rebound Theory ...... " ....... ",., ................ ", ..... _., ............. " ......................... 3
1.3 Theory of Plate Tectonics ................................................................................................. 4
1.3.1 Lithospheric Plates ........................... ., ........... ., .................. " ............... ,.................. 6
1,3.2 Plate Margins and E.a.rtllquake Occun:ences .. ,. ... ,....... _..... ,... " .............. , .. __ .... " .... 7
1.3.3 The Movement of Indian Plate"."... .. ............... """ ........... ,, ................... , 9
1.4 Seismic Waves .............. " .. " ................ " ... ,..... .. .................... 10
1.4.1 Body Waves." .............. ., .......,., ................. 10
1.4.2 Surface Waves .. " .................... . ................................. I J
1,5 Ear[hquake Size .............. __ ... .. ...... " .......... ,." .. ,.......... " .......... ,. .................. ,.... , ] 3
1.5.1 Intensity .... " .......................................... ,. ...... ,. ................. " ............. ,." ................. 13
1.5.2 lsoseismal Map ...................................................................................................... 18
1.5.3 Earthquake Magnitude ................................................................................... 18
1.5.4 Energy Released in an Earthquake ..... ,............... " .......... "," ............................... 23
1.5.5 EarthqUake Frequency........... ......................... ................... 23
1.6 Local Site Effects ............................................ .................... 24
1.6.1 BasinlSoil Effects." ............................................................................................ 24
J ,6.2 Lateral Discontinuity Effects ....... ,... ,.... ,... . .27
1.6.3 Effect of the Surface Topograpby ....... . .......................... "., ................. 28
i.
L
444.... (=CQ~n~'~.n~t~s==================____________________
~=========~
Contents )
r.7 lnternaI Structure of the Earth
4.2 Terminology of Strong Motion Seismology .............................................................. 73
1.7.1 Crust ................................ ::::::::::: .. ::::::::::::.::::::::::.::............................................. 30
L7.2 Upper Mantle .............................................. 30 4.2.1 Amplitude Parameters ........................................................................................... 73
:.:~fFt~;;;~7i:i=.;;~:~:;;;:;::;;:;:: 11
4.2.7 Spatial Variation of Earthquake Ground Motion .............................................. 80
4.2.8 Damage Potential of Earthquakes ....................................................................... 81
Summary ................................................................................................................................ 86
References "", ................... ,......................._... ".... ,.... "............................................................... 86
Glossary Of EarthquakelSeis;uology ... "...... 38
References ................................................................................................................................................................:..::.:.. :..:. ::..::..::.:: ::::. 38
Ii. Evaluation of Seismic Design P ....ameters ................ m ............ 88107
41 5.1 Introduction ., ................................................ ,", .................. ,........... " ................................. &8
2. Seismic Zoning Map of India .._. _.._.... __ ..._.... 4558 5.2 Types of Earthquakes ...................................................................................................... 88
~ f[f~2~;;;:(;~S~;~j2j~_~H
5.3 Fault Rupture Parameters .................................................................................................. 90
504 Earthquake Ground Motion Charaeteristics ..... ",., ..... ,"", ...... ,', ..................................... 91
5.4.1 AmplitUde Properties .......................................................................................... 91
5.4.2 Duration ................................................................................................................. 93
5.4.3 Effect of Distance .................................................................................................. 93
2.6 Seismic Zone Map of 2002 ................................................................... 52 5.4.4 Ground Motion LeveL ...................................................................................... 96
2.7 Epilogue.. .. ............................................................................................ 56
, .................... , ... . 5.4.5 Geographical. Geophysics and Geotechnical Data ............................................ 96
'" ........ ,', .............. ., .......... ,., .. ,. .. , .... .. ,.......... " .. 56
SummorJ' 5.5 Detenninistic Approach .................................................................................................... 97
Rejeren(es .................................... ,', ..... ,.. ,.. ,.... ,.......................... ,.................. " ........ ,", .... ." 58
5.6 Probabilistic Approach ..................................................................................................... 98
58
5.6.1 Example ............................................................................................................... 100
Strong
3. 3 Motion Studies in India ......................... m ......................... 5969 5.7 Response Speetra ............................................................................................................. 101
.1 Introduction 5.8 Design Spectrum .............................................................................................................. 101
3.2 vnderSlandin~.. ili~N;~~;~ ~fG~undM~~~~s .. ........... "............"...........................
59
Summary ,.. ,', .... ,.... ,', ... ," .. " ... .,., ......... " .......................... " ....... "... ..,., .. ,", .. ,.. " ...... 105
3.2.1 Source Effect .................... ,..... .. ............ 60
References ....................................... ................................. ................... .. ........... 106
3. 2.2 Path EffccL.::::::::.:::::: .... ...... ....... ...... ............... ........ 60
3 .2 .3 Site Effect ................................................................ 62
Part II
3.3 Esrimation of Ground Moo P 63
3.4 Tbe Indian PeTS tive on arameters .......... " ............... "" .................................... " 64 STRUCTURAL DYNAMICS
3.5 Utilizalion of S=g M~ti~~.,D~~~"' ...... "..........'....." ...........,..................................
65
Summary ................ .. ...................... " ........ """ .............. ", ......... " ....... ,,, ... 65 6. Initiation into Structural Dynamics ......................................... 111114
.............. ,,'
Rcfcrettces ................. . .......................................... 66 6.1 Introduction .............................. ,.... .,., ...... ""', .... ,'.", .. ,' ................................................... , 111
........................................................................................... ......... 66 6.2 Mathematical Modelling ................................................................................................ 112
4. Strong Motion Chal'llctel'istics .......................................................................................................................... 114
4. 1
Introdu:::tlOn .................. .. ............................
U.h........H. . . ...... ~ u ~ . . . . . . . . . o.a_ .. 7087
Summary
References .................. " ...... " ......... ,........................ _"." ...... ,"" .... " ................... u, ........ " ......... 114
.. .... .. .. .. .... .. .. ...... .. ................... 70
L
_ (Contents
..................................... 146
10.3 Response Speerra ............................................................. ..
7. Dynamics of Single Degree of Freedom Systems ................ 1l5128
7.1 Introduction ..................................................................................................................... 115 10.3.1 ~~:~~~~~iti~ii;;;. R~~;::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: :151~~
. : . : . .: : : : :~ ~ .: ~: :~ :~: :~: :~: ~: : ~ ~ ~ ~ :~ 1~~
7.2 Free Vibration of ViscousDamped SooF SyStems ..................................................... I I 6 10.3.2 ..
7.2.1
7.2.2
Underdamped Case (~< 1) ................................................................................. 118
Criticallydamped Case (~= 1) .......................................................................... 118 19'~:; ~::~n~;:;::~~::.::.:.:::::
7.2.3 Overdamped Case (~> I) ................................................................................... 118 Summary ............................................~............................. 155
References .................. ,., .. ,.... ..
.................................................................................
7.3 Forced Vibrations of SOOF Systems ............................................................................. 120
7.3.1 Response of SDOF Systems to Harmonic Excitations ..................................... 120 Dynamics of Mnlti_DegreeofFreedom Systems .................. 157188
7.3.2 Excitation by Base Motion ................................................................................ 122 11. 157
7.3.3 Response of SDOF Systems to a FInite Duration Excitation ......................... 122 11.1 Introduction .................... ~ .................................................................................................. 158
11.2 System Property Matrices ............................................................................................. 159
7.3.4 Response of SDOF Systems to a Short Duration Impulse .............................. 124
113 Dynamics of Two Degree of Freedom Systems .......... ,. ... " .................. ,., ., ............. ,. .... 162
7.3.5 Response of SooF Systems to General Dynamic Excitation .............. ,""" ...... J 25 J 1A Free Vibration Analysis of MDOF Systems ........... ,.. " ................ "P. . ........... " ....... .
7.4 Vibration Isolation ..................................................................... "" .............. ", .. ,", ........ 126 11.4.1 Orthogonality Conditions ............................................................................... 163
Summary .............................................................................................................................. 128 165
11.5 Determination of Fundamental Frequency ................................................................. , 165
References ............................................................................................................................. 128
11.5.1 Rayleigh Quotient ........................ ...... ...... .. ................................................. 165
~:;:n;es ~ 87
8.5.2 Calibration of Accelerometers ........................................................................... 136
Summary .............................................................................................................................. 136 :::::::::::::::::::'::::::::::::::::::::::.:::::::::::::::::::::::::::::::::::::::.:::::::::::::::.:::::
References ............................................................................................................................. 136
Pari m
9. Numerical Evaluation of Dynamic Response ........................ 137143 CONCEPTS OF EARTHQUAKE RESISTANT DESIGN OF
9.1 Kumerical Solution Based on Interpolation of Excitation ..,.......... ..... _137 REINFORCED CONCRETE BUILDING
9.2 Numerical Solution Based on Approximation of Derivatives ,.. _................. " ............. 139
Stability and Accuracy Consider.tions ......................................................................... 141
12. Earthquake and Vibration Effect on structut:es:
9.3 06
Summary ............................ .. ...................................................................................... 143 El ts f Earthquake Resistant DeSIgn ................ 1912
BaSIC emen 0
Ref<'rences ............................................................................ ~ ................................................ 143 .................... ...................... .................... 191
12.1 Introduction ........................ " ................. ,.. .... ,.......... ,"'" ... 192
12.2 Static and Dynamic Equilibrium ....................... ........ .... .... .. .......... 194
10. Response Spectra ............................................................................ 144156 12.3 Structural Modelling . 194
10.1 Introduction .................... ~.~ .............................. ~ ............................................................... 144 12.3.1 Structural Models for Frame Building ..................... ...... .......... .... ........ .. .
10.2 Fourier Spectra ................. ~ .............................................................................................. 144
lc
ii$'ii (~C~o~n~t~en=t~8__________________________________________________~ ==::========::========::==================~____~C~o~n~t.~n~'s~) ""."
14.2 Vertical Irregularities ..................................................................................................... 221
12.4 Seismic Methods of Analysis.. .................................................................................... 196
14.2.1 Vertical Discontinuities in Load Path .............. "............................................ 227
12.4.1 Codebased Procedure for Seismic Analysis .................................................. 191
14.2.2 Irregularity in Strength and Stiffness ............................................................. 230
12.5 Seismic Design Methods ............................................................................................... 198
12.5.1 Codebased Methods for Seismic Design ....................................................... 198 :::;:! ~:~rre~~:~ .i;;:;;;;;j;;;;~.:::: .:::::::::::::::::::::::::::.::::::::::::::::::::::.:::::: ;~;
12,6 Response Control Concepts ........... ,.............. ,......................................... ., ...... ,.... " ... .", 199 14.2.5 Proximity of Adjacent Buildings .................................................................... B3
12.6.1 Earthquake Protective Systems ....................................................................... 200 14.3 Plan Configuration Problems ........................................................................................ 234
12.7 Seismic Evaluation and Retrofitting .................... "" ................................... ,", ...... _." .. 201
12.7.J Methods for Seismic Evaluation ..... ,............. ,', ........ ",.", ....... ,,, ........ ,, ............ 202
12.7.2 Methods for Seismic Retrofitting ................................................................... 203 :!.~:~ ~::EJ~7:::::.:.:.:.::: . : : .: .:.: : : ::.: .: : . . .:.:. . . . .: :. .: .: . :. . .:.:. .: ~~~
12.8 Seismi{; Test Methods ., ..... " ..... "''' ..... "''" .......... ",,............. ,.204 14.3.4 Diaphragm Discontinuity ............................................................................... 237
J 2.S,! Methods for Seismic Testing ... ,.................... .,.",,, ....... ,,,, ......... ,, .... ". 204 ]4.4 Recommendations ........................................................................................................ 237
Summar,v
........................................................... .. .. ............... 205 SUI1IJ1Ulry .... ,., ... " ................. ,., ............................................................................ 238
Referen.ces
................. ...................................................................................... 205 References "" ....... " .......... ,." .. ..................................................................... 238
13A Damage to Structural Elements .................................................................................... 217 15.4.1 Mode Shapes and Fundamental Period ......................................................... 243
J3.5 Damage to !'1on~Structural Panel Elements ..... ,....... , .. ,................ ,....... ,.", ................... 2J9 15.4.2 Building Frequency and Ground Period ...................................................... 244
D . ................................ 244
,I
L
Part IV 18.4
SEISMIC ANALYSIS AND MODEIJJNG OF
REINFORCED CONCRETE Bun.DING
16. Code Based Procedure Cor Determination oC Design
lteral wads u .................u . u u,. . , . _...._ . . . . . ._ ............_ ............"". . .~...... H ......... _. 251281
J6.J Jntroduction .................................................................................................................... 25J
16.2 Seismic Design Philosophy ........................................................................................... 251
16.3 Detenninmion .of Design Lateral Forces ...................................................................... 252
16.3.1 Equivalent La~ral Foree Procedure ............................................................... 253
16.3.2 Dynamic Analysis Procedure ........................................................................... 259
SwrtmJry ........................................................................................................................... 280
References .... """,." ........... ,..... ,.,.,.,.,., ................. ".~.... ,.................................................... '" 280
21.5
21.4.3 Earthquake Load Analysis ............................................................................... 376 , 24. Identification of Damages and NonDamages in
21.6
~:~nC:~~~~~:..:k::~. ::.:::. .::: . . :: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .:. . . :. ::::::::::::::::::::::::: ;~~ 1
f
.. Masonry Buildjngs from Past Indian Earthquakes ........_.427448
21.6. J
21.6.2
Design of a Flexure Member ........................................................................... 382
Design of Exterior Columns ............................................................................ 385 ;::~ ~:~tO~~i~n~~~~~.::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::;;
24.3 Features of Damages and Nondamages ....................................................................... 429
II
21.6.3 Design of Interior Columns ............................................................................. 387
21.6.4 Detailing of Reinforcements ............ " .................. " ....................... ,.............. ,,"' 389 24.3.1 Bhuj Earthquake, January 26. 2001 ............................................................... 4429
Summary 24.3.2 Chamoli Earthquake, March 29. 1999 ........................................................... 31
............................................................................................................................ 390
References 24.3.3 Jabalpur E3J!hquake. May 22. 1997 ............................................................... :33
............................................................................................................................ 391
24.3.4 Kiliarl Earthqnal:e. September 30.1993 ........................................................ 36
22. 24.3.5 Uttarkashi Earthquake. October 20. 1991 ...................................................... 4/~
Earthquake Resistant Design of Shear Wall as per
24.3.6 BiharNepal Earthquake. August 21. 1988.................................................... 4
i!; I
IS 13920: 1993 ......................................... _ ...................... 392403
;; ~ ~~~~~~~i~~n~;:B~;i~i;~:::.::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::':::::::::::::::::.:: ,
;::~ ~~:~~:::~~~;...::::.::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Summary
:::
........................................................................................ 445
22.3 DeterminatIon of Design Lateral Forces ................................................. " .................. , 393 ...................... "",, .................... ,....... 446
R ,e'r'nees ....... " ...... " ,.......... "
;i:: ~!r::n~f!h:!~!~~~~~;.::.::::::::::::::::::::::::::::::::::::::::::.:::::::::.:. . : ..........
5;um.mary
397
............. 402
403
'GJ~.. .." .. ,"", ....... , , ............................................ .
Re/I?rrnccs
403
25. Elastic Properties of Structural Masonry .............................. 449462
23. Capacity Based DesignAn Approach for Earthquake 25.1 Intro~uctinn .................................................................................................................... 449
25.2 Materials for Masonry Construction ............................................................................ 449
Resistant Design of Soft Storey RC Bulldin 044
gill .... ........... 4 24 449
26'2'4 step 4: eterm;nation of Rigidity of Shear Wall ......................................... 470 29.3.1 StrUctural Damage due to Discontinuou. Load Path .................................... 529
. _ tep: Deternunation of Direct $hear Forces and Torsional 29.3.2
structural Damage due to Lack of Deformation ............................................ 529
l
31. Seismic .Provisions for Improving the Performance of
Nonengtneered Masonry Construetion with
Experimental Verifications  . __._ _. ___..... _...... _. __576590
; ~.~ ~troduction .. . H H " " " " ' H " ' H ' " H ' ' ' ' ' ' ' ""H'''''HH'''''''''HHH'H 576
31', S nl~enaFfor Eartbqualre Resistant Provisions ""'''''''''HHH''''''''''H'''''HH''HH'HH''''H"." 577
.... a lent eatures of Earthquake Resistant Provisions
31.4 Seismic Siren thenin F "H""H"''''''''H'''''''''''''''''''''''''H 577
References ............. '._.. ..... " ..................... ,,, ............. ,................... .. ....................... 590
...... "".""""''' .... " .... " ....... " ... " .......... " .. "" .... "''''" .......... ,,'''',, .. 590 The vast devastation of engineered systems and facilities during the past few earthquakes has
exposed serious deficiencies in the prevalent design and construction practices. These
32. RetrOfitting of Masonry Buildings.. ......................... 591624 disasters have created a new awareness about the disaster preparedness and mitigation. With
32.' Introduction ,., ....... ,... ,........... ,.. ,..... ,', .. ,.. ,.... ,.".,., ... ".... "" .... ,,. .. ,., .... ,................. '........... 59 J increased awareness carne the demand of learning resource material which directly address
32,2 Failure Mode of Masonry BUildings ." ... '" .... " ...... "."."" ........ "''''"'',, ... ,,'''''',, ...,''''''' 592 the requirements of professionals without any circumlocution. While the recommended codes
32.2.1 Out~ofplane Failure """'""",,,,.., . of practice fOT earthquake resistant design do exist but those only specify a set of criteria for
~fHg~~4~:;:;i;it;;;~~;t~;jl~
compliance. These design codes throw little light on the basic issue of how to design. The
problem becomes more acu~ as students graduate with degrees in civil/structural engineering
without any exposure to earthquake engineering in most of the universities/institutes. The
shortterm refresher courses routinely offered by various institutes and universities for
professionals achieve little more than mere familiarization with the subject matter. Any
32,3 Methods for Retrofitting of Masonry Bulldings .... '''''''''''''''' .. "''''''".,,''''''''''',,.... ,,'' 595 sMrttenn approach to the learning process, which requires the relevant ideas and concepts
to be assimilated. is doomed to faiL Realizing the practical difficulties of professionals in
;;:~~ ~~~~~;;;t;;R~~;~fi;~i~"""''''''''''''''''''''""""'"'''' 596
32,3,3 StructuraUGiObaI Retrofi~~''''''''''''' .. ' .. ''''''''''''''' .... '''' .. '''''' .... '''''''' .. '''''''''''' 596
attending any longterm direct contact academic programme on earthquake engineering, a
sixmonth modular courile in distance education mode was offered by lIT Roorkee in 2004,
. . g """"""" ...... """".""" .. ""'''''''''"."."'''',,.,, .. ,,'' 596
32.4 Rep:nrmg Techniques of Masonry ..... " .................................. " The course was well~received and culminated in a lwoday workshop at Roorkee which was
. """""."''''"""",,,.,,. 596 attended by a large number of participants, providing valuable feedback, This book derives
~2.4. J Masonry Cracking "".""" ...... """ .... "."".",,. ."" ............. " ................ ,,"" ....... . 596
."12.4.2 Masonry Deterioration .. its origin from the set of lecture notes prepared for the participants with later additions to
... ............................ ...... " .......... " ........ 60]
incorporate some of the suggestions made in the feedback workshop.
32.5 Member Retrofitting ..... ,., .. .,................... ................ ';"02
'2 ,U Tbe guiding principle in developing the content of this book has been to provide
 .5. I Relrofitting Techniques .......................... ............. ,' enough materialat one place(o develop the basic understanding of the issues as required
'" 6 S ...... 602
d_. tructurd! Level Retrofitting Methods.. ............ " ........ , .. ,",.,., ......... , for correctly interpreting and using the standard codes of practices for earthquake resistant
",,,,,,,,.,,,,,,605
32.6.1 Retrofitting Techniques .. " ................... ,. ...... ",,,,, .. design. The objective is to present the essentials in a clear and concise manner with adequate
32.7 ~eismjC Evalu~tion of Retrofitting Measures in Slone' ~~~~~"~.~~.;~'::::~:~~:~.:,...:. :~~ illustrations~ whiie intentionally steering clear of some of the advanced topics which require
,,2.7.1 Beh{tVlOUr of Retrofitted Model;; ...... more rigorous mathematical treatment.
32.7.2 Findings.. ............. " . . . " .... ,. ........................ .. '",,'''',,''' 620 This book is divided into seven parts, each dealing with a specific aspect of earthquake
Summary "",621 engineering. We slart with the discussion of the physics of the earthquake generation~ the
Re,fprenct's
....... ,..., .............. ......... " ... < ...... H .......... 621
''''., .....
evolution of the seismic zoning map of India. characteristics of the earthquake strong ground
""'"'''''''''', 622 motions, and determination of seismic design parameters in the first part on Earthquake
Index ............... ,. ... ,.. Ground Motions. The second part on Structural Dynamics is concerned with the study of
"."""" .....,.. ""...''''''''''''" .. "''"."."'', ....,,'''_.,,'''',, .. ,,',, .. ,,'""""""'''''''''''' 625634 f
L
_____________~ 'Pi>:,..;:;iif.;;';;'e) 4g'"
( Preface (~..~
bI . earthquake resistant design and construction, Only the problems
2nalytical treatment of vibration problems. Starting with an introductory chapter on enl~~'::t~:~qu= :::Sistant design of buildings have been addressed in. this book to
Mathematical Modelling for Structural Dynamics Problems, the theory of structural re. ble ize It is planned to address the problems concerrung earthquake
dynamics is developed gradually to the level of dynamics of complex structural syStems restrict It to a reasona s . . b I th t SO e errors
. . of other structural types in another volume. it IS ut natura a m
including multisupport excitation and dynamic soilstructure interaction analysis. The :~~~~~~!Pt into the text of such volume, We will appreciate if such errOrs are brought
treatment is intentionally focused on deterministic problems in time domain as most of the to our notice. Suggestions for improvement of the book are also welcome.
professional engineers do not fee) comfortable wjth the probabilistic framework and
frequency domain methods. The basic philosophy of the earthquake resistant design is
discussed along with the deficiencies in the prevalent design and construction practices with PANKAJ AGARWAL
the help of several case studies in the third pan on Concepts of FArtJ.4uake Resistant Design MANISH SHRIKHANDE
of Reinforced Concrete Buildings. Simple architectural considerations that go a long way in
improving the seismic performance of reinforced concrete (RC) buildings are also discussed.
The modelling issues, including the modelling of infill panels, and seismic analysis of RC
framed buildings are el.borated through several workedout e.amples in the fourth part on
Seismic Analysis and Modelling of Reinforced Concrete Buildings. The actual design
caleul.hons as per relevant IS codes are presented for the seismic design of fourstorey RC
framed buildings and RC shear walls are described in the fifth pan on Eanhquake Resistant
Design of Reinforced Concrete Buildings. A detailed example on the capacity desigo method
to handle the softstorey problem in RC framed buildings has also been presented. The
modelling, analysis and design of masonry bUildings to resist earthquake load forms the I
thrust of pan six: entitled &T1hquake Resistant Design of Masonry Buildings. Finally, the
seventh part on Seismic Evaluation and Retrofitting of Reinforced Concrete and Masonry
Buildings elaborates upon the very challenging problem of seismic evaluation and
I
retroftttingfstrengthening of eXisting buildings. A stateoftheart compilation of methnds and
materials has been presented along with experimental verification in some case studies. Thus
a gamut of earthquake engineering starting from seismology and seismic hazard analysis to
analytical study of dynamic behaviour to design and retrofit of RC aod masonry buildings
has been presented in single volume.
I
This book is the result of team work. We have received tremendous support and
cooperation from our colleagues and students in bringing it to this form and are greatly
indebted to them, in particular, to Prof. Susanta Basu and Prof. S,K, Thakkar who read
early drafts and offered useful suggestions for improvement in addition to contribUting
some chapters for the book. Dr. J.P, Narayan pitched in with his expertise in the
I!
engineering seismology to contribute a chapter introducing me basic seismological concepts.
Mr, V.V.S_ Dadi belped with the calculations and Mr, J.P. Singh and Mr. Hemant Venayak
helped with the figures. We greatly appreciate the kind support extended to us by the staff
of PrenticeHal! of India, New Delhi. We particularly admire the seemingly infinite patience
of Ms. Seema Zahir, who readily accepted numerous revisions/corrections till the last
moment. Finally. we are grateful to our wives, Mahima and Ashwini. for their suppon
during the period when time was at a premium.
Although this book is primarily designed to serve as a textbook for undergraduate and
postgraduate students of civil engineering, it can also be used as a reference book for regular
academic courses on design of reinforced concrete and masonry buildings. The book will
also serve the needs of structural designers as a ready reckoner for most of the commonly
!
l
j
Contributors
I
I
I Jay Prakash Narayan
Assistant Professor, Department of Earthquake Engineering
I
Shashikant Thakkar
Rmlway Bridge Professor
Department of Civil Engineering
Indian Institute of Technology Roorkee
SlISanm Basu
Professor, Department of Earthquake Engineering
Indian Institute of T""hnology Roorkee
,
t
L
PART
_ _ _I _Cl=~""== .!
!
I
!
I
I
Il Chapter 1
Engineering Seismology
I
,
1.1 INTRODUCTION
) Seismology is the study of the generation, propagation and recording of elastic waves in the
earth, and the sources that produce them (Table I.l). An earrhquake is a sudden tremor or
movement of the earrb's crust, which originates naturally at or below the surface. The word
natural is important here, since jt excludes shock waves caused by nuclear tests. manmade
explosions, etc. About 90010 of all earthquakes result from tectonic events, primarily movements
on the faults. The remaining is related to volcanism, collapse of subterran:ean cavities or man
made effects. Tectonic earthquakes are triggered when the accumulated strain exceeds the
shearing strength of rocks. Elastic rehowtd theory gives the physics behind earrbquake genesis.
This chapter includes elastic rebewtd theory, plate tectonics, earrhquake size, earrbquake
frequency and energy~ seismic waves, local site effects on the ground motion characteristics.
interior of the earrb and seismicity of India.
3
( Earthquake Resistant Design oj Structures C  __ _ _ _ _ _ _ _ _ _ _ _ _ _._'C'hG"':P~er
_,_ j Engineering Seismology) +.
nobody had yet been able to explain what was happening within lhe earth to cause earth uakes ranges also found in these zones, too? An explanation to these questions can be found in plate
;rom an .examlnatIO? of the displacement of the ground surface which accompanied th~ 1906 tectonics, a concept which has revolutionized thinking in the Earth Sciences in the last few
e:~qu~e, H.F. ~eld, Professor of Geology at lohns Hopkins University concluded that the
decades. The epicentres of 99% earthquakes are distributed along narrow zones of interplate
19l1i ;h
u
es m~sll ave lnvol~ed an "'elastic rebound" of previously stored ~elastic stress (Reid. seismic activity. The remainder of the earth is considered to be aseismic. However, no region
of the earth can be regarded as completely earthquakefree. About I % of the global seismicity
rebou~d t:;~~'?~/~~~~:~:~~~ and release of stress and strain is now referred to as the "eJastic is due to intraplate earthquakes, which occur away from the major seismic zones. The seismicity
" 1},(bFut the blocks are d,storted So thaL lines originally straight across the fault have become
obi 'que 'gure l.lb).
,
(al ,, (b) ,, ,,
,, ,, (e)
,,
. :
,,,
, ~
, ~
,,
,, "~
,,, ,, ,,

Itt=
~{:
= ...
~, ,,
,
,, t 'i'l
1i~
1tJ
!
1
 __ iiI
O) ,
: Rupture
",,
,, ,,,
,
,,, ~
, ,, , 1i~,,
,,, ~
, ~ FIGURE 1.2 Geographical distribution of epicentrcs of 30,000 earthquakes occurred during
,
No strain Slr.1ined eanhquake 19611967 illustrates the tectonicaJly active regions of the earth (after Barazangi and Dorman,
FIGUR El.1 SclJ I%g),
2(00). ema~c representation of elastic reboond theory (afler Mussett and Khan,
The pioneering work was done by Alfred Wegener, a German meteoroJogist and
The weakest part the faulL slips s dd I h th . geophysicist, towards the development of the theory of plate tectonics. He presented his
can support. The ru ture fro . u en y w en e steal.n becomes more than what the fault
the blocks on either~ide of i~:~e ::akest p;rt
exre~ds rnpldly along the fault plane, allowing cominental drift theory in his 1915 book JOn the Origin of Continents and Oceans'. He
proposed that at one time all the continents were joined into one huge super continent. which
fault jn Fi ure I Jer In:o a ess stramed condition. The half arrows beside the
accumuhU!d en~::yS~:;~h~~~~:~ ~~l~~su~den di.spJacement. called th.e elastic rebound. The
he named Pangaea and thar at a later date the continents split apart, moving slowly to their
present positions on the globe. Wegener's theory was not accepted since he could not
waves and a pan lS converted ," to h
n
0 h
rock
IS suddenly reJeased to the fonn of seismic
eat or ot er forms.
f satisfactorily answer the most fundamental question raised by his critics, i.e, what kind of forces
~:
earthquake in Indonesia which had triggered a massive tsunami. [A~~
FIGt:RE 1.6 Schematic representation of continentalcontinental convergence.
!,
Oceaniccontinental convergence
If by magic we could pull a plug and drain the Pacific Ocean. We would see the most amazing About 40 to 50 million years ago the boundary
sight, a number of long narrow, curving trenches thousands of kilometres long and 8 to 10 km
deep cutting into the ocean floor. Trenches are the deepest parts of the ocean floor and are
created by subduction. At the oceaniccontinent boundaries, oceanic plate subducts due to higher
between Indian plate and the Eurasian plate was
oceaniccontinental in nature and later on it became
continental~continental convergence after consump~
t
t,
density (Figure J ,5), Strong, destructive earthquakes and tbe rapid uplift of mountain ranges lion of the Tethys Sea,
towards the side of ovcrriding pJate are common at the convergenee boundaries. Oc~.\njc
continental convergence also sustains many of the earth's active volcanoes on the side of Transform boundaries
overriding plate.
The ZOne between two plates sliding horizontally
past one another is called a trans/orm.faull boundory. FIGt7RF.. 1.7 Scbematic H'presentati\Hl
or simply a transform boundary (Figure 1.7), ~e of transform boundary.
concept of transform fault originated with CanadIan ,
t geophysicist J, Tuzo Wilson, who proposed that these large faults orjracfure zones ,~Jn~'<,;c~ ~~;:
spreading centres (divergent plate boundaries) or; less commonly, trenches ,(c~m,_ft6.~n: ~,!::,';
boundaries), Most transform faults are found on the ocean floor. Howevcr, a lew occm ~t. 1",,,1,
I
Litho!iphere :;=:>
r.or exampe.
I th " Andreas Fault zone in California, Along the tmnsform boundan"" the
e ,.an Th f" h 1
ASthenospbere ' earthquakes 'occur at shallow depth. unaccompanied ?y
volcanic activity, e n~:lOn e.:~~~
the plates can be so great that very large strains can bUild up before they are pCT10dtcaJly !Cll~. ! ... ",
f
f by large earthquakes,
L~
proCe~:S}l j,,~ljch i" f{'Ieme;), 'rIl{, MarJ:,n~~'s Tn~n(:h (J1<l111Ikling the :\.itlri:!n:l hl.1nd,,), for exmnp!c:, which :tretcbcs 2,900 km along the border between lndia (lDO Tibet Alt.er spilHIl1L', (If ,lqn;,),CY
... CEarthquake Rerutont St1"Ucture.~ M'M
Design of
~==~==~~ c Chapter 1 Engineering Seismoio!l!l )
about 200 minion years ago, India began to SVwave and SH~wavef respecti,,'dy. They are sometimes called secondary waves because they
forge northward. Abollt 225 million years ago. EURASIAN PLATE travel more slowly than P~waves in the same material. Swaves do not change the instantaneous
India was a large island still situated off the volume of tbe materials through which they pass, but as they pass through materials, they distort
Australian coast, and Tethys Sea separated India tbe instantaneous shape of those materials, The velo;;ity of Swave is directly related to the shear
from the Eurasian continent About 80 million strength of materials. Swaves do not propagate through fluids as those do not have any shear
years ago, India was located roughly 6,400 Jan
strength.
south of the Eurasian continent. moving north
ward at a rate of about 9 m a century. By study
ing the history and ultimately the closing of
1.4,2 Surface Waves
the Tethys Sea, scientists have reconstructed A disturbance at the free swface of a medium propagates away from its source partly as seismic
India's no!1hwardjourney (Figure 1.8). Immense surface waves. SUlface waVes, sometimes known as Lwaves., are subdivided into Rayleigh (1<)
HimaJayan mountain range began to form and Love waves (LQ ). These surface waves are distinguished from each other by the type of
between 40 and 50 million years ago, when two motion of particles on their wavefronts.
Equator /.
large landrnas:.es, India and Eurasia, drjven by
plate movement, ~ollided. Both these continen Rayleigh waves
tal Jandmass~s have same rock density. so one
plate could not be subducted under the other. ~:l!~:+ Lord Rayleigh (1885) described the propagation of Rayleigh wave along the free surface
of semiinfinite elastic halfspace. In the homogeneous halfspace, vertical and horizontal
Further, the rate of northward movement of i
i F~"
INDIAN
OCEAN components of particle motion are 90* out of phase in such a way that as the wave propagates.
India reduced to about 45 m a century after
collision. Thc coUision and associated decrease (j~~~l the particle motion describes a retrograde ellipse in the ve..rtical plane. with its major axis vertICal
in the rate of plate movement are interpreted to L/ and minor axis in the direction of wave propagation, The resulting partjcle motion can be
mark the beginning of the rapid uplift of the regarded as a combination of P and SVvibr.tions (Figure 1.9), In the case of a layered and
Himalayas.
L
~ l~.~.~~qv.o.ke Resistant Design of Structure.., )
dissipative medium, the path is always elliptical but not necessarily retrograde, Further, the axis 1.5 EARTHQUAKE SIZE
("If the ellipse may not be vertical and horizontaJ since the phase difference between vertical and
Q
l'orizonta] displacement can be different from 90 The velocity of Rayleigh wave is very much
1.5.1 Intensity
dependent on the Poisson's ratio and it is equal 00,9194 times to that of Swavein the Poisson's Seismic intensity scale is a way of measuring or rating the effects .of an e~hquake at different
,olid (poisson's ratio = 0.25). The particle displacement is not confined entirely to the surface site . The assignment of intensity of an earthquake does not requtre any mstrumentaJ records.
of the medium but the passes of the Rayleigh waves also displace the particle below the free It d~pendS very much on the acuity of the observer, and is in .principle sub~ective. l~tensi~y to
",;face up to a depth equal to the wavelength. In a uniform half space, the amplitude of particle different places of an affected area can be assigned based on vlsuai observa?ons and mterVJews
'.;ispl.{ccmenl decreases exponentiaUy with depth, of residents or based on evaluation of questionnaires completed by res~dents: of tha~ area.
Intensity data are very much useful for the development of s,eismic risk map of ~ regIOn or
CQuntry, Seismic risk maps are useful in planning safe sites ~or Important stru~tures hke ~udear
I'.J:.H, Love (191]) explained the mechanism of generation of Love waves in horizontal soil power plants Or large dams. Ris~ maps are a~so v~u~b~e to msura~ce compames. lntenslty data
ltl.ye.r ht;\rlying the halfspace (Figure 1.10), When the angle of reflection at the base of soil is also important in determjnatlOn of htstonc se1smIcity of a regIon.
;s more than the critical angle, SHwaves are trapped in the soU layer. The constructive The RossiFord intensity scale, developed in the late 19th century, have ten stages to
'!i"'t'('!\'nce of reflected SHwaves from the top and bottom of the soil layer generate describe the earthquake effects in increasing order. Mercalli (1902) propos~ an i~ten!\.jty fcale
',:,r:':':ouiOiHy travelling Love waves. The particle motion is in horizontal plane and transverse in which earthquake severity was classified in twe1ve stages. The Mercalh mte~slty ~cal~ w~s
~;~ the direction of wave propagation. The velocity of Love wave lie.<; between the velocity of modified in 1931 to suit the building conditions in the United States, The moddied versJOn 15
::;. w;;.ve in the soil layer and in the halfspace, The veJocity of Love wave with short wavelength widely known as Modified MereaHi Intensity (MMI) scale. The MedvedevSpoonheuerKarnik
ic.: close 10 the velOCIty Swave in soil 1ayer and velocity of longer wavelength Love wave is close (MSK) intensity scale introduced in 1964 also has twelve stages and differs from the MMI scale
to the Swave velocity in halfspace, This dependence of velocity on wavelength is termed mainly in details. The MSK (Table 1.5) and MMI (Table 1.6) intensity scales are commonly used
dispersion. Love waves are always dispersive, because they can only propagate in a velocity to seek information on the severity of effects of an earthquake. lntensuy raungs are expressed
:".;'ered medium. as Roman numerals. The intensity scale differs from the magnitude scale in that the effects of
anyone earthquake vary greatly from place to place, so there may be many intensity values fOI
one earthquake. On the other hand~ there is only one magnitude value for an earthquake.
1IIIIIIII
in fieldstone, rur<il structures, unburntbtkk house'S. J)OU/C';.
L
Canld, Con/d.
TABLE 1.5 Conld.
TABLE 1.5 Conld.
Intensity
Descriptions ,,
Intensity Descriptions
cm. Water in lakes becomes turbid, New reservoirs
investigated specially.
come inlo existence, Dry wells refiil and existing
wells become dry. In many cases change in flow and )<11 landscape changes (a) PracticalIy aU structures above and below the
ground are greatly damaged or destroyed.
IX
~~______~~~I~e:v~el==O~f~~~'a~t~er~iS:;O~bs~e~rv~e~d~,~____~__~____
(a) Ge~eral panic; considerable damage to furniture.
(b) The surface of the ground is radically changed.
General damage
to buildings A I Considerable ground cracks. with extensive vertical
Olma S run to and fro in confusion and cry. and horizontal movements are ohserved. Falls of
(b) Many buildings of'lYpe C suffer damage of Grade 3, rock and slumping of river~b:anks oVer wide areas,
and a few of Grade 4. Many buildings of Type B lakes are dammed: waterfalls appear, and rivers are
show damage of Grade 4, and a few of Grade 5. deflected. The intensity of [he eanhquake rcquires
Many buildings of 1Ype A suffer damage of Grade 5. to be investigaled specially.
Monument~ and columns fall. Considerable damage ..~~ ... ~.. ..~~.~. .. ~~.~
~ ~e.servolrs; underground pipes parlly broken. In The details of Modified MercaHi Intersity (MMI) scale is given in Table 1.6 llnd i:;, also
mdlvlduaJ cases raHway lines are bent and roadway
damaged. used to seek information on the effects of an earthquake ljke MSK intensity scale.
(c) On flat land over11ow of water. sand and mud is
TABLE 1.6 Details of MMI intensity scale
often observed. Ground cracks to widrhs of up to ~~~~~~ .. ~~~ ~,,
IO em, On slopes and river banks more than 10 cm; Intensity Descriptions
furthermore a Jarge number of slight cracks in
ground; faUs of rock, many landslides and earth Vibrations are recorded by instruments. People do nO( feel any earlh movement.
flows; large waves in water. Dry wells renew their n People at rest upstairs notice shaking. A few peopJe might notice movement if they
   ,_ flow and eXisting wells dry up. are at rest andlor on the upper floors of tall bUildings,
X General destructio'::n;~;(:):M;::=7~::::=~':::=~:!:..2:'_.~
of buildings
a ~ any
d
buildings of Type C suffer damal1e of Gr"de 4,
& .. ill Shaking felt indoors; hanging objects swing. Many people indoors feel movement.
an a lew of Grade 5. Many buildings of Type B Hanging objects SWing back and forth. People outdoors might not realize thaI an
show d~mage of Grade 5; most of Jype A have earthquake is occurring.
desuucUon of Grade 5; critical damah"e to dams and ""  '  .  ' .. ''~'~
dykes and severe damage to bridges. Railway lines IV Dishes rattle; standing cars rock:; uees shake. Most people indoors fee! movement.
are bent slightly. Underground pipes are broken or Hanging objects swing. Dishes, windows. and doors raule. The earthquake feels
bent Road paving and asphajt show waves. like a heavy truck hitting the walls. A few peopJe outdoor~ may feel movement.
(b) I~ ground, cracks up to widths of several Cm, some~  Parked cars rock.
..~ .. ~~~~~
times up to 1 m. Parallel to water Course OCCur broad V Doors swing; liquid spills from glasses: sleepers awake. Almost everyone feels
~ssures. Loose ground slides from steep slopes. From movement. Sleeping people are awakened. Doors swing open or close. Dishes are
rIverbank and steep coasts, considerable lands.lides broken. Pictures on the wall move. Small Objects move or are turned over. Trees
are possible. In coastal areas, displacement of i'iand might shake. Liquids might spi1l out of open conlainers.
and mud; change of water level in weJls; water from
canals, lakes, rivers, eIC., thrown on Jand. New lakes VI People walk unsteadily: windows break; pictures fall off walJs. Everyone feels
occur. movement. People have trouble Walking, Objects falJ from shelves. Pictures fall
XI Destruction off walts. Purniture moves. Plaster in walls might crack. Trees and hushes shake.
(3) Severe damage even to weH built buildings, bridges.
Damage is slight in poorly built buildings. No structural damage.
water dams and raHway lines; highways become
useless; underground pipes destroyed. "11 Difficult to stand; plaster, bricks, and tiles faU: jarge beBs ring. People have difii M
(b) Ground considerably dj,[ened by broad crack, and culties in standing. Drivers feel their cars shaking. Some furnilUteS break.. Loose
slighlto~m{)derate
I 
flss~es, ~s welt as by movement in horizontal and bricks faB from buildings. Damage is in wellbuilt huildings;
vMIca} dUcctions; nUmerous landslips and falls of considerable in poorly buHt buitdjngs.
rock. The intensity of the earthquake requires to be Vm Car steering affected; chimneys faU; branches break: cracks in wet ground. Drivers
.. ~~. .. ~.. ..~~~.~~'. ~~~
Contd. Conlrl.
L
8M ( Earthquake Re.sistant Design oj Structures C
.~~rC;;;"a;'p;;te;'r::j,EE;;;,,:;;g<inn.~~~i'tlg Seismology) p
seismological observatories for an event may var~. The u~certajnt! in an estimate ?f the
TABLE 1.6 Details of MMI intensity scale magnitude is about O.3 unit. Seismologists often reVIse magmtude estimates as they obtam and
Intensity Descriptions analyze additional data.
have trouble steering. Houses that are not bolted down might shift on their
foundations. Tan structures such as towers and chimneys mjght twist and fall, lUchter magnitude (Hz)
Wellbuilt buildings suffer slight damage, Poorly built structures suffer severe One of Dr. Charles F. Richter's most valuable contributions was ~o reco~nize tha~ the seismic
damage. Tree branches break;, HHlsides might crack; if the ground is wet. Water lev~ waves radiated by earthquakes could provide good estimates of thetr magmtudes. Richter (1935)
els in wells might change.
' collected the recordings of seismic waves from a large number of earthquak~ and constructed
IX General panic; damage to foundations; sand and mud bubble from ground. WeU d' am of peak ground motion versus distance (Figure 1.11). The loganthm of recorded
built buildings suffer considerable dam.age. Houses that are not bolted down m.ove :m~~~de was used due to enormOUS variability in amplitude. Richter inferred ~at the ger
the intrinsic energy of the earthquake, the larger the amplitude of ground mouon at a glven
ta:
off their foundations. Some underground pipes are broken. The ground cracks.
Reservoirs suffer seriou$ damage.
distance.
x Most buildings destroyed; large landslides; water thrown out of rivers. Most build
ings and their foundations are destroyed, Some brjdges are destroyed. Dams are
seriously damaged. Large Jandslides occur. Water is thrown on tbe banks of canals,
rivers, lakes. The ground cracks in large areas. Railroad tracks are bent slightly.
XI Railway tracks bend; roads break up; large cracks appear in ground; rocks f.ll.
Most buildings ooUapse. Some bridges are destroyed. Large cracks appear in the
ground. Underground pipelines are destroyed, Railroad tracks are badly bent.
XII Total destruction; "waves" seen on ground surface; river courses altered; vision
distorted. Almost everything is destroyed. Objects are thrown mto the air. The
_ _ _~gr~.~nd ~~~.:s in waves or ripples. Large amounts of rock ma.y move,
I rapidly with magnitude, For example, when the magnitudes of two earthquakes differ by I, their
correspooding energies differ by a factor 28 (=101.44) according to Bath's equation, or 32 (101.')
according to the GutenbergRichter fonnula,
I
More recently, Kanamori came up with a relationship between seismic moment and seismic
wave energy. It gives:
Energy = (Moment)I2O,OOO (1.11)
For this relation moment is in units of dynecm, and energy is in units of erg.
~~~ ! ~~t/"/;trained On this globe, tbe annual frequency of smaU earthquakes is very large and that of large
L.==::~:
 ... ~..!../
volume
just before rupture
_ _ _ _...J.
I earthquakes is very small (Table 1,8). According to a compilation published by Gutenberg and
Richter in 1954, the mean annual number of earthquakes in the years 19181945 with
FIGURE; 1,12 Schematic diagram Co seism' magnitudes 44.9 was around 6000, while there were only on an average about 100 earthquakes
2000). r JC moment determination (after Mussett and Khan, per year with magnitudes 66.9. The relationship between annual frequency (N) and magnitude
FjnaJIy. moment can be computed as TABLE 1.8 Earthquake frequencies since 1900 (based on data from the USGS! NEIC)
and Ibe estimated mean annUliI energy release based on Batb (l!l66)
Mo = J1 A d (1.7)
The moment magnitude M'. b b ' d ' , Earthquake magnitude Number per year Annual energy {l015 Joule yr I)
and Kanamori, 1979). w L n eo tame usmg followmg relation (Kanamori, 1977; Hanks ~~~~==~~~==~~=~~~=
;, 8.0 01 0600
2 77,9 18 200
M. ~ 3' [log" Mo(dynecm)  16,OJ 66.9 120 43
0.8)
'The rigidity 'J!' is measured using samples of rock . " 55.9 800 12
In the area. AHershocks are believed to reveal the or IS estnnated from knowledge of the rocks 44.9 6,200 3
plane. The simplest way to measure the len th' , rupture area t:ccause most of them 1ie on a 33,9 49,000
look at the newly faulted surface Or fault br:ak ThL and av~ge dISplacement 'd' of a fault is to 22.9 ~350,OQ() 0,2
th I ' ' . e seIsmiC moment ean also be estimated from
e Dog penod components of seismograms (Bullen and Boll, 1985),
L
11.9 ~3.000,Q()0 0.1
 ..__.=.:.:..._.
r
L  .           
Chapf.er 1 _ Engineering Seismology)
(Ms) is logarithmic and is given by an equation of the form damping are neglected, the conservation of elastic wave energy requires thallhe flow of energy
(energy flux, P VsV') from depth to the ground surface be constant. Therefore. with decrease in
10gN= a bMs (1.12) density (p) and Swave velocity (Vs) of the medium, as waves approach the ground surface, th.e
The value of '0' varies between about 8 and 9 from one region to another~ while 'b' is particle velocity (v), must increase. Thus, shalcing tends to be stronger al Sites WIth softer soli
approximately unity for regional and global seismicity. Most of the time' /J' is assumed to be layers.
equal to I; 'b'>1 in an area generally means that small earthquakes occur frequently; 'b'<l
indicates an area that is more prone for a larger earthquake. In volcanic areas where there is lots Resonance
of earthquake swarm, 'b' >1. Along subduction zones and continental rifts the value of 'b'<1. Tremendous increase in ground motion amplification occurs when there is resonance of signal
The mean annual numbers of earthquakes in different magnitude ranges are listed in Table 1.8. frequency with the fundamental frequency or higher harmonics of the s.oillay~r. Various spectral
peaks characterize resonance patterns. For onelayer ID structures, thiS relation IS very sImple:
1.6 LOCAL SITE EFFECTS io ~ V,,/4h (fundamental mode) andi, ~ (2n + I)io (harmonics)
where V is the Swave velocity in the surficial soil layer, and h is the thickness. The amplitudes
Significant differences in structuraJ damage ed' d .
in basin as compared with the surrounding
TABLE 1.9 Classificallon or local geology SI
of these spectral peaks are related mainly to the impedance contrast and S Iment ampmg.
in different category
exposed rocks~ or even in the basin itself from
place to place, have been observed during LOCAL SITE EFFECTS Damping in soil
earthquakes. The amplitude of shaking in A~.~B~o~SI~nI~S~O~ilr::::====~B~.~~~~~= Absorption of energy occurs due to imperfect elastic properties of medium in which the collision
basin can be more than 10 times stronger ..: between neighbouring particles of the medium is not perfectly elastic and a part of the energy
Q. Ridge
than the surrounding rocks. Other geological a. Impedance contrast in the wave is lost instead of being transfenred through the medium. This type of attenuation
conditions, which affect amplitude and signal b. Reson(lnce h. Volley of the seismic waves is referred to as anelastic damping. The damping of seismic waves is
duration, are topography (ridge, valley and c. Trapping c. Slope/slope described by a parameter called as quality factor (Q). It is defined as ille fractional loss of
slope variation) and the lateral discontinuities. d. Focusing variation energy per cycle, 2;r;1Q = llEIE. where llE is the energy lost in one. cycle and E is the total
The historical references regarding earthquake e. Basin~edge C, Strong Lateral elastic energy stored in the wave. If we consider the dampmg of a selSIlUC wave as a functJon
ti_"u_i"ti"e"s
damage due to 10caJ site condition extend back =f_D_o_m'p_in...:8'_ _ _ _ _ _d_is_c_o_"e. of the distance and the amplitude of seismic wave, we have
to nearly 200 years (Wood, 1908; Reid, 1910).
Mac Murdo (1924) noted that the buildings situated on the rock were not much affected as A = A" exp (  ~~) =A"exp(ar) (1.13)
those situated on the soil cover during Kutch earthquake (1819). Recent examples regarding the
intense effects of local site conditions include Michoacan earthquake (J985) which caused only where IX ai2QV is absorption coefficient. This relation implies that higher frequencies will
moderate damage in the vicinity of its epicenter but caused heavy damage some 400 km away he absorbed at a faster rate.
in the Mexico city (Dobry and Vacetie, 1987), damage caused by ille Lorna Prieta, California
earthquake (1989) in the city of San Francisco and Oakland (USGS, 1990) and damage pattern Basin edge
observed during Bhuj earthquake of January 26, 2001 (Narayan et al., 20(2).
Intense concentrations of damage parallel to the basin~edge had been observ~ due to strong
generation of surface waves near the edge, during recent earthquakes (Nonhndge earthquake,
1.6.1 Basin/SoU Effects 1994; Kobe earthquake, 1995 and Dinar earthquake, 1995). The conclusion that basinedge
Study of different aspects of ba"iin effects on the ground motion characteristics needs special induces strong surface waves had been drawn in many studies by examining the phase and group
attention since most of urbanized areas are generally settled along river valleys over young, soft. velocities, polarity and arrival azimuth (Bard and Bouchan, 19l1O " b 1985: Hatayama et aJ.,
surficial soil deposits. 1995; Kawase, 1996; Pitarka et al., 1998; Narayan, 2003a, 2004, 2005). Suriace waves start
generating near the edge of the basin when frequency content in the body wave exceeds the
lm,proanc.e c.llutrast fundamental frequency of the soil and their amplitudes decrease with increase of edgeslope
(Narayan, 2004, 2005). .' .
Sei~mic waves travel faster in hard rocks than in softer rocks and sediments. As the seismic Figure 1.13. shows the vertically exaggerated basinedge models haVIng dIfferent thIckness
waves pass from hard medium to soft medium, their celerity decrease, so they must get bigger of single soil layer over the bedrock. Figure J.l3b dep~cls the vertical component of groun~
in amplitude to {;any the same amount of energy. If the effects of' scattering and material motion, computed for thickness of soill.yer as 195 m uSlllg a doublecouple source (dIp ~ 45 .
_n. ( Eart~.t.
,..... _e
D
n.eSJ8tant DesJg1l 01 Strueture.s
( Chapter J Engineering SeismfiiOiiiJ Mi[.
rake = 60 and strike = 90) just below the edge at a depth of 13 7 km w,th. do' fr
J 0 Hz Th P dS .. . mman! equency The major conclusions drawn in papers of Bard and Bouchan (I980 a. b). Hatayarna
. ' . e . an waves velocl!Jes and densities were taken as 13965 mis 400 0 mi d
1.9 g/cm" for soil and 3464.! mis. 2000.0 mls and 2.5 glcm' for half s~ace <hard ~ock s;;:.,
et .J. (1995). Kawase (1996) Pitarka, et al. (1998) and Narayan (2oo3a. 2004, 2005) are listed
ground respo~se was computed at 26 eqUidistant (105 m apart) receiver points. Fi ure i.13b
J below.
:~eals f~ur weBseparated wavelets at receIver points some distance away from the edge lbe
8
Basinedge induces Strong surface waVes near the edge.
1 erentta] ground motion in northsouth direction cl I d" . Edgeinduced surface waves propagate normal to edge and towards the basin.
surface waVes since vertjcally travelling bod ear y eplcts hOrIzontally travelling
y waves are more or less removed (Figure 1.13c). Surface waves start generating near the edge of the basin when frequency content in
the body wave exceeds the fundamental frequency of the soil deposit.
+~N Surface wave amplitude deereases with increase of edgeslope.
p. A, /1'; 0
Rl Receivers is RlO Vacu Damage caused by edgeinduced surface waves is confined in a narrow zone (width
" \~~Y~"~I~6:i=m~Ygt105m
"~Y~t~Y~Y~t ~~um~
2.53.5 km) parallel to the edge, and at some distance (0.51.0 km).
Surface wave amplitude increases with the decrease of propagation velocity in soiL
150m Soil  Further, their characteristics are highly variable with change in propagation velocity
14.5" 195 m and thickness of soil deposit.
The characteristics of edge~induced surface waves are also very much dependent on the
angle of incidence of body waves.
Hard rock Edgeinduced surface waves develop significant differential ground, the main cause
of damage during earthquakes. in addition to amplification and prolongation of the
signal.
I'V" ..
. , Basement topography
=
<lAW
1U'
The focusing and defocusing effects caused by basement topography are strongly dependent
# II or,'" on the azimuth and angle of incidence of waves. Seismic waves traveling upward from depth
1/'" ... may be redirected by subtle irregularities at geological interfaces. particularly the basement
topography. The effects of focusing and defocusing are maximum for normal incidence of
waves and it decreases with increase of angle of incidence. Similarly. azimuth also affects the
focusing and defocuslng effects. This effect reveals the importance of considering not only the
s.urficial soil layer but also the basement topography for seismic mJcrozonation.
'Ifapping of waves
The fundamental phenomenon responsible for the increase of duration of motion over soft
sediments is the trapping and multiple reflections of seismic waves due to the large impedance
contras.t between soft sediments and underlying bedrock. Sometimes. when a wave cnters a basin
~ ~ 9 /2~; /82\ *2'4 ~ 6 ~ 12 115j';~'Tt~ through its edge, it can become trapped within the basin if postcritical incidence angles develop+
TIme (sec) Time (SL'C) causing total iutemal reflection at the base of the Jayer. Waves that become trapped in deep
(b) (e) sedimentary basins can therefore be potentially very damaging.
HGt;RE 1.13 (a) Vertically exaggerated basined e .
displacement,. and (c) the differential ground displa~e model, (b) ve.rU.cal component of ground
ment 1.6.2 Lateral Discontinuity Effects
Dent of ground motion at 26 reiver 'points (aft N correspondmg to the nrtieal wmpo
. er arayan, 2Dil5).
There are numerous consistent macroseismic observations showing a significant increase in
The generation of surface waVes near the ed e was confi
coherence in recording stations increase of tr ' J Iff "
Inned on the baSiS of the Jarge intensity of damage in narrow zones located along lateral discontimlities, Le. areas where a
j
tb ed . d ' a\e tme 0 atcr phases as we move away from softer material lies besides a more rigid one. An amplitude amplification and local surface wave
(Neara::~.e~~:t B~~r velocity ~~ater phases and the analysis of differential ground motion generation in the softer medium and large differential motion caused by shorter wavelength
wave an wave have caused generation of Rayleigh waves. of the suIface wave can explaln the observed damage. In past, .a number of field observations
L
.1. ( Earthquake Resiqtant Design 0/ Structures
(Narayan and Rai, 2(01) and theoretical studies have reported significant increase of damage
c Deamplification (2ltl<\> rimes) occurs in valley relative to the top.of .the valley.
Topographic amplification decreases with increase of angle of mCldenee of body
in the narrow zone located along strong latetal discontinuities (Moczo and Bard, 1993),
waves. 5)
Ridge amplification increases with elevation (Pigure l.l .
1.6.3 Effect of the Surface Topography
:r
It has often been reported after destructive earthquakes in hilly areas that buildings located at
bi)) tops suffer much more damage than those located at the base: examples of such observations I
may be found in Levre! et al., 1986 (Lambesc, France, 1909 earthquake), Siro, 1982 (lIpini.,
Italy, 1980 earthquake), Celebi. 1987 (Chile, 1985 earthquake) and Narayan and Rai, 2001
~cre~
350m'
(Chamoli, 1999 earthquake).
There are also very strong instrumental evidences that surface topography considerably S ~' Base
affects the amplitude and frequency contents of ground motion (Pedensen et al., 1994). A review
I Ridge profile /
of such instromental studies and results may be found in Geli et al. (1988), Aki (1988) and more
recently in Facdoli (1991). The theoretical and numerical models have also predicted a
, III
, I
I
I '
systematic amplification of ground motion at ridge crest (convex part) and deamplification in 2 .J
valley (concave parts) of the surface topography (Kawase and Ald. 1990; SanchezSesma, 1990;
Faccioli, 1991; Narayan and Rao, 2003; Narayan, 2oo3b).
Narayan and Rao (2003) reported sorface wave generation near the top of the ridge and
their propagation towards the base of the ridge, in addltion to amplification of ground motion
J
with elevation and slope of the ridge using 2.5 D model (Narayan, 2(01). Narayan (2oo3b)
reported strong generation of surface waves for weathering thickness more than one~eighth
of wavelength. He reported on the basis of the simulated results that damage to the built
environment may be maximum on the top of the ridge, if it is not weathered. But, if the velocity 0.2
of the weathered material is very Jess as compared to the underlying rock formation. maximum
damage may be more near the base of the ridge, due to large amplitude and duration of the
generated surface waves. Some of the findings of the abovementioned studies are listed below,
o o 50 100 150
Ground motion amplification increases with ridgeslope. Elevation (m)
Maximum amplification (2n:'$ times, where <\> is the crest angle) occurs at the crest of Variation of average amplification factor with elevation (after Jibson, 1987).
the triangular wedge type topography relative to the base for wavelength comparable FIGURE 1.15
to width of the base (Figure 1.14). Sumcc waves are generated near the top of the topography. .
The presence of neighbouring' ridges accentuates the topographIC effects,
x 1 terference between the incident waves and outgoing diffracted waves rroduces
" Crest
~
r:pidly varying amplitude and phase, thereby causing differential ground motion along
Trough
~
Z~I the slope of the topography. . . bal 'f
M~_~
.... . G The amplitude of ridge_weatheringinduced ~urface wave lOc~eases towards Its se,l
~ ~ thickness of weathering is more than oneelgbth of the wavele~gth ,
Decrease of weathering velocity increases the amp1itude of rIdgemduced surface
>
{a) (b) ~~:;iicat.eddamage pattern occurs on hills with variable slopes. Generally, houses
situated on or near the slopechange suffer more d~mage. .
fiGURE 1.1~ CharlIcterization 01' simple topographic irregularities; (a) notation for a trion " In general, theoretical studies predict lower amphficatIo n than those obtaIned by
gular wedge; (b) approximation of actual ground surface (soiid line) at trough and crest by
wedges, (after FaccioJi, 1991). analysis of recorded motion.
L
i. ( Earthquake Re8i6to.nt
increase of velocity occurred was 54 km, Now. this seismic discontinuity between crust and
~:r:~~~eation ~f i:ernal ~t~c,tun~ of ~e Earth, different discontinuities and nature of material mantle, where there is sudden increase of seismic wave ve1ocity~ is called as Mohorovicic
discontinulry, o~ simply Moho.
refracted wo major sconlmultles IS mamly based on the analysis of the recorded reflected and
. seJSmIC waves. In broad sense, the internal structure of the earth is divided int th V. Conrad (1925) found faster Pwave (P*) and Swave (S*) as compared to Pg and Sg
concentnc ceHs namely crust mantl d ' 0 fee waves during the analysis of T.uem earthquake of 1923 (Ea.stem Alps) in upper crnstallayer.
materials (Figu;'; LJ6). Purthe;' crust i: ~~Vid:e accordrng to the chemical property of the The estimated velocities of P* and S* waves (6,29lan1s and 3.57 kmis, respeclively) were lesser
lower mantle and core as inner core as upper and lower crust, mantle as upper and than the velocities of Pn and Sn waves refracted from the Moho, Conrad inferred the ex.istence
classjfied as lithosphere asthenos here~ outer core. The crust and mantIe together are also of a lower crustal layer with higher velocity as compared to the upper crustal layer. The interface
materials, Following s~bheadin:' deSCri~:esos~here. ~n the basis of physical property of the separating the crustal mass into upper and lower crust is caUed as Conrad discontinuiry, in
discovery of major discontinuities. crus mant e and core in bnef along with the honour of V. Conrad.
Worldwide analysis of recorded reflected and refracted seismic waves reveals that the
structures of the crust and upper mantle are ve..y complex. The thickness of crust is highly
laterally variable, It is 510 km in oceanic region, below the mean waterdepth of about
4,5 km. The vertical' structure of continentaf crust is more complicated than that of oceanic crust.
, The thickness of continental crust varies from 35 to 40 km under stable continental areas and
50 to 60 kIn under young mountain ranges,
'
mantle together with crust forms the lil/w"phere, The lithosphere play an important role in plate
tectonics.
Ouler There is an abrupt increase of seismic wave velocity (3  4%) at depth of around 220
'\ core
h
30 km. This interface is called as the Lehmann discontinuity. Between the base of lithosphere and
:j'
the Lehmann discontinuity, there is low velocity layer (LVL) with negative velocity gradients,
'.'..~'=============='
\1; The average thickness of LVL is around ISO km. This LVL is known as asthenosphere, which
also plays an important roJe in plate tectonics. Asthenosphere behaves as viscous fluid in long
term and thus decouples the lithosphere from the deeper mantle.
The traveltime versus ePicentraJdistance curves of body wave show a distinct change in
slope at epicentraJ distance of about 20"'. This is attributed to a discontinuity in mantle velocities
FIGURE 1.16 Inlrnal structure of !be eartb based .
Kennett and Engdahl, 1991). on p~ and S~waves velocity variations (after at a depth of around 400 kID. This is interpreted as due to a petrological change from an olivine
type lanice to a more closely packed spineltype lattice, A further seismic discontinuity occurs
at a depth of 650670 km, This is a major feature of mantle structure that has been observed
1.7.1 Crust world~wide. In the transition zone between the 400 kID and 670 km discontinuitie& there is a
further change in structure from J3spinel to ,),spinel, but this is not accompanjed by appreciable
Alldrija Mohorovicic (1909) found only direct P.wav ( . .
lhe analysis of an earthquake in Croatia B t b d 1~ .Pg) amvals near the ep1centre during chilnges in yhysical properties.
and direct Pwave was overtaken bv;h u ey~n p km two P~wave arrivals were recorded
possible when Pn has travelJed t ~ e secon wav~ (Pn). He conduded that it is only 1.7.3 Lower Mantle
from the upper mantle Accord' a gre~ter speed,. MohoroVIcic identified Pn as a refracted wave
Pwave was 5.6 km/; and 7 ~n:~7 hIS calculaMns, the velocity of direct Pwave and refracted The fower mantle lies just below the important seismic discontinuity at 670 km, lts composition
" 5, respectIvely; and the estimated depth, at which sudden is ralher poorly known, but it is thought to be consisting of oxides of iron and magnesium as
eM ( Earthquake Resistant Design oj Structures
weH as iron~magnesium silicates with a perovskite structure. The uppermost pan of the lower
mantle between 670 and 770 km depth has a high positive velocity gradient. Beneath it, there
is great thickness of normal mantle, characterized by smooth velocity gradients and the absence
of seismic discontinuities. Just above the coremantle boundary an anomalous layer approxi~
j
mately 150200 km thick, has been identified in which bodywave velocity gradients are very
small and may even be negative.
1.7.4 Core
R.D, Oldham first detected the fluid nature of the outer core seismologicaJly in 1906, He
observed that, if the traveltimes of P~waves observed at eplcenrrat distances of less than l00()
were extrapolared to greater distances. the expected travel~times were Jess than those observed.
This meant that the P~waves recorded at large epicentral distances were delayed in their path.
Oldham Inferred from thIS the existence of a central core in whieh the Pwave velocity was
reduced. Gutenberg (1914) verified the existence of a shadow zone for Pwave!> in the epicentral
range between lOS and 143", Gutenberg also located the depth of top of outer core at about
2900 km, A modern estimate for this depth is 2889 km, It is characterized by very large seismic
velocity change and is the most sharply defined seismic discontinuity, In honour of Gutenberg.
the core~mant1e boundary is known as the Gutenberg seismic discontinuity.
lnga Lehmann (1936), a Danish seismologist, reported weak Pwave antival, within the
shadow zone. She interpreted this in terms of a rigid inner core with higher seismic velocity
at depth of around 5154 km, Thus core has a radius of 3480 km and consists of a solid inner
core surrounded by a liquid outer core.
t
o (,<11<><' (',S
o
,
J.'hh,d~
.
'I(JV1"";\.\
MYsw''
o
0
o
o
7OIu"'7.5
15,.,""jL1
Mortc!h;m~\)
I.Ul CLASSIFICATION OF EARTHQUAKES
I (104m)
@ D:v jh.\J~ slltn::::' 1, Ba~ed on location 2. Based on eplccntral distances
X" r 0 ':'h"''''''1,"",1I'" 0' i'I'''''liwlm" ltl 'l..:4(,,)
I Pu,.~m I'rcm:: the Sl,me rn'g~1 la) lnterplale (a) Local earthquake
t_~.?.~_~ .. . .~:::~7::.. .~=;;.c= (b) Intraplate (b) Regional earthquake
I'll
(c) Teleseismic earthquake
3. Based on focal depth 4. Based on magnitude
1'I(:lJRE 1.18 Seismicity map of India (.fter, 15:1893 (Pt. 1): 2002).
{al Shallow depth ()..71 km (a) Mircocarthquake < 3.0
(h) hHcrmcdiate. depth 71300 km (b) [ntannediate earthquake :14
~ _ _ _ _ _ _ _ _ _ _ _ ~. _ _ _ . . ~c_
_ _ ~ _ _ ~c
ChapU:1' 1 Englneel"'tn~
, s l ..... ' i _
eUfflO 0,:<, " __ ~
)
L
. . l.Earth9uake Re,'JiJJtant De$igJ:,:oT~::c:;tu=",,=.    .===. ) ; 
C;;hc.;p"C"':';E;;"",::::;:::,,:::,,=,,:::.~:::g:C<S"':::i,=mo"logy="":"1) .M.~
water depth and the coastal co~figuratjon. the waves may undergo extensive refraction., another
p~cess that ,may c~nverge therr energy to particular areas on the shore and thus increase the Body wave. A seismic wave that moves through the interior of the earth, as opposed
heIghts and Inundauon even more. FieJd surveys are carried out after every larcre tsunam' f to the surface wavt;.S that travel near the earth's surface. p~ and Swaves are body waves.
assessment of :unups and inundation limits and to coUect associated data fro;; ey~witn~ or Crust. The outermost major layer of the earth, ranging from about 10 to 65 km in
such as the number of waves, arrival time of waves and the largest wave. ses thickness worldwide. The uppermost 15 to 35 km of crust is brittle enough to produce
earthquakes.
Core. The innermost part of the earth. The outer core extends from 2500 to 3500 miles
SUMMARY below the earth's surface and is liquid meta!. The inner core i. the central 500 miles
and is solid metal.
Seismology is the study of generation pronagation and r'"~rdl'ng of l ' . th Earthquake. This tenn is used to describe both sudden slip on a fault, and the resulting
e aSHe waves In e canh
and of so.urces Y:at.. 'to" ....
produce them. An earthquake is a sudden tremor or movement of eanh's ground shaking and radiated seismic energy caused by the slip, or by volcanic or
crust. WhJc~ ongmates ~atu~dHy at or below the surface. About 90% of aJ1 earthquakes re~mI~ magmatic activity, or other sudden stress changes in the earth.
from tecu:mc events, pnmanly movements of the faults. The remaining proportion is reJated Earthquake hazard. Anything associated with an earthquake that may affect the
to volcafl)sm, collapse of subterranean cavities, Or manmade effects. The epicenters of normal activities of people. This includes surface faulting, ground shaking. landslides.
earthquakes are nol randorrJy distributed over the earth's surface They tend [0 b d liquefaction. tectonic deformation, tsunamis, and seiches.
. ) T " . e conCentrdte
l~ narro~ lOnes. hIS cha~ter descnbes in detail the elastic rebound theory, seismic waves local Earthquake risk. The probable building damage, and number of people that are
SHe effe~ts ~n g~und motIo,n characteristics, interior of the earth and movement of India~ plate expected 10 be hurt or killed if a likely earthquake on a particular fault occors.
along WIth ItS selsmolectomc features. Earthquake risk and earthquake hazard are occasionally used interchangeably.
Epicentre. The point on the earth's surface vertically above the pOint in the crost where
seismic rupture begins
GLOSSARY OF EARTHQUAKE/SEISMOLOGY Fault. A fracture along which the blocks of crust on either side have moved relative
to one another parallel to the fracture. Strikeslip faults are vertIcal (or nearly
Active fault. A fault that is likely to have another earthquake some time in the futur vertical) fractures where the blocks have mostly moved horizontally. If the block
~auJts are commonly considered to be active if they have moved one or more tim:~ opposite to an observer looking across the fault moves to the right, the slip style
In the past.
is termed righllateral; jf the block moves to the left, the motion is termed left lateral.
Aftershocks. Earthquakes thal follow the largest shock of an earthquake sequence. Dipslip faults are inclined fractures where the blocks have mostly shifted vertically.
They are smaJJer than the mamshock and continue over a period of weeks months If the rock mass above an incUned fault moves down~ the fault is termed normal,
or years. In general. the huger the mainshock. the larger and more num~rous th~ whereas if the rock ahove the fault moves up, the fault is termed revetse (or thrust).
aftershocks, and the longer they will continue Obliqueslip faults have significant components of both slip styles.
AIl~viu.m. le:0se gravel, sand, siJt, or day de~sjted by streams. Foreshocks. Foreshocks are relatively smaller earthquakes that precede the largest
AseismIC: ThIS term describes a fault on which no earthquakes have been observed. earthquake in a series, which is termed the mainshock. Not all mainshocks have
Attenuabon~ "'hen you throw a pebble in a pond, it makes waves 'on the surface that foreshocks.
move out from the place where the pebble entered the water. The waves are largest Hypocenlre. The point within the earth where an earthquake rupture starts. Also
,,:here they a~e funned and gradually get smaller as they move away. TIDs decrease in commonly tenned the focus.
s.lze, Of amphtude, of the waves is ca1ied attenuation Intensity. A number (written as a Roman numeral) describing the severity of an
" B;~ment. Harder and usually older igneous and me~orphic rocks that underlie the earthquake in tenns of its effects on the earth's surface and on humans and their
mam sedImentary rock sequences (softer and usuaUy younger) of a region and extend structures. There are many intensity values for an earthquake, depending on where you
I
downward to the base of the crus!. are~ unlike the magnitude, which is one number for each earthquake.
~ ~~r~~~.~elaliveJy hard, soHd rock that commonly underlies softer rock, ~ediment, Intraplale and interplale. Intraplate pertains to process within the earth's crustal
or so~l. '" .)UbSCf of the basement
plates. Interplate pertains to process between the plates.
/{ .sen~on: zone~ A dipping planar (flat) zone of earthquakes that is produced bv the lsoseismal A eontour or line on a map bounding points of equal intensity for a particular
lfllcwcllOfl oj u downlJoino oceanic crust'l I t <th 
I
h uk l:> 0 a p a e Wl a contInental plate. The<;e earthquake.
eart_ qu. c.s can be prod~ced by slip along the subduction thJUst fault or by slip on Leftlateral. If you were to stand on the fault and look along its length, this is a type
fault:.. wnhm the downgoJna plate as a result of bend' . d . . . of strikeslip fault where the left block moves toward YOll and the right block moves
. J' he'. 1l1g an extenslOn as the plate JS
;)0 Jlce Into t e manrle. Also known as the W~adatiBenjoff zone,
L
away.
WliM (Earthquake Re,ststc:nt "Design of Structures Chapter 1 Engineering Seismology J . , .
Lithosphere, The outer solid part of the earth, including the crust and uppermo't Ring or Fire. The zone of earthquakes surrounding the Paciftc Ocean which is called
mantie, The lithosphere is about 100 krn thick, although its thickness is age dependent the CircumPacific belt about 90% of the world's earthquakes occur there, The .next most
(older lithosphere is thicker). The lithosphere below the crust is brittle enough at some seismic region (56% of earthquakes) is the Alpide belt (extends from MedIterranean
locations to produce earthquakes. by faulting, such as within a subducted oceanic plate. region, eastward through Turkey, Iran, and northern India). .
Love wave. A type of seismic surface wave having a horIzontal motion that is Swal'e. A seismic body wave that shakes the ground back and forth perpendicular to
iIansverse (or perpendicular) to the direction the wave is travelling. the direction the wave is moving, also caned a shear wave.
Magnitude. A number that characterizes the relative size of an earthquake. Magnitude Sand boil. Sand and water that come out onto the ground surface dunng an earthquake
is based on measurement of the maximum motion recorded by a seismograph. Several as a result of liquefaction at shallow depth. .
,cales have been defined, but tbe most commonly used are (1) local magnitude (MJ, Seismic gap. A section of a fault that has produced earthquakes in ~e ~ast but 1S ~o~
commonly referred to as "Richter magnitude," (2) surfacewave magnitude (Ms), (3) quiet. For some seismic gaps, nO earthquakes have been observed hlstoncaUy, but It .1S
bodywave magnitude (me), and (4) moment magnitude (Mw). believed that the fault segment is capable of producing earthquakes on some other basls,
Mainshock. The largest earthquake in a sequence. sometimes preceded by one or more such as platemotion infonnation or strain measurements.
foreshocks, and almost always followed by many aftershocks. Seismicity, The geographic and historical distribution of earthquakes.
~Mantle. The part of the earth's interior between the metallic outer core and the crust Seismic moment. A measure of the size of an earthquake based on the Mea of fault
Moho. The boundary between the crus.t and the mantle in the earth. The boundary is rupture the average amount of slip. and the force that was required to overcome the
between 25 and 60 krn deep beneath the continents and between 5 lUld 10 km deep friction' sticking the rocks together that were offset by faulting. Seismic moment can
be.neath the ocean floor. also be calculated from the amplitude spectra of seismic waves.
Oeeanic spreading ridge, A fracture zone along the ocean bottom where molten Seismic zone. An area of seismicity probably sharing a common cause. Example: "'The
mande material comes to the surface) thus creating new crust. This fracture can be seen
Himalayan Zone."
beneath the ocean as a line of ridges that form as molten rock reaches the ocean bottom
Seismogenk Capable of generating earthquakes, .
and solidifies. Seismngram. A record written by a seismograph in resp~nse to ground motIOns
Oceanic trencb. A linear depression of the sea floor caused by the subduction of one produced bv an earthquake, explosion. or other groundmouon sources.
plate under another. Seismolo!lY. The srudy of earthquakes and the structure of the earth, by both naturally
Pwave, A seismic body wave that shakes the ground back and forth in the same and artificially generated seismic waves.
direction and the opposite direction as the wave is moving.
Plate tectonics. A theory supported by a wide range of evidence that considers the
earths ernst and upper mantle to be composed of several large, thin. relatively rigid REFERENCES
plates that move relative to one another, Slip on faults that define the plate boundaries
commonly results in earthquakes. Severa] styJes of faults bound the plates. including [IJ Abe, K., "Tsunami and Mechanism of Great Earthquakes", Physics of the Earth Planer
thrust faults along which plate material is subducted or consumed in the mantle, oceanic Interiors, 7: 143153, 1973. .
spreading ridges along which new crustal material is produced, and transform faults [2J Aki, K. lUld Chouet, B., "Origin of Coda Waves: Source, Attenuation and Scattenng
that accommodate horizontal slip (strike slip) between adjoining plates. Effects", Journal of Geophysical Research, 80: 3322, 1975. , .
Rayleigb wave. A seismic surface wave causing the ground to shake in an elliptical [3] Aki, K., "Local Site EffecL, on Strong Ground Motion':' In EarlhquakL Engmeermg
motion, wjth no transverse, or perpendicular, motion, and Soil Dynamics IIRecent Advances in Ground MOlton Eval~atlOn, J .~. Von ~u~
Recurrence intervaL The average time span between large earthquakes at a particular (Ed,), Geotechnical Special Publication No. 20, 103155, Amencan Soc,ety of CIvIl
she. Also termed rerum period. Engineering, New York, 1988. , '
Reflection. The energy or wave from an earthquake that has been returned (reDected) [4] Barazangi, M. and Dorman, J" "World Seismicity Map Complied from ESSA, Coat
from a boundary between two different malerials wiibin the earth, jU!>i as " mirror and Geodetic Survey, Epicenter Data, 1961J961", Bulletin of the SClSmologlCal
reflecL" light. Sodetv of America, 59: 369380, 1969.
Retraction. The deflection, or bending 1 of the ray path of a s.eismic wave cau!ied by 15] Bath, ].,1" "Earthquake Energy and Magnitude", Physics and Chemistry of the Earth,
its passage from one material to another having djfferent elastic properties. Bending > Ahren, LH. Press. 115165, 1966.
of a tsunami wave front owing to variations in the water depth along Ii coastlille.
.. Rightlateral. If you were to stand on the fault and look along its length, this is a type I [6] Bard, P,Y and B~uchon. M., 'The Seismic Response of Sedimentfilled ValIeys~
Pnrt 1: The Case of Incident SH Waves", Bulletin ~f the Setsmologlcal SocJety of
of strike~slip fault where the right block moves toward you and the left block mnves
,rv.'fiy. I! America, 70: 12631286, 1980.
L.
   ) [26) Kawase, H. and Aki. K., "Topography Effect at lhe Critical SV Wave Incidence:
[7] Bard, P.Y. and Bouchon, M., "The Seismic Response of Sedimentfilled Valleys Possible Explanation of Damage Pattern by the Whitter Narrow", Earthquake of I
Pan 2: The Case of Incident P and SV Waves", Bulletin of rhe Seismological Sodety October 1987, Bulletin of the Seismological Society of America, 80: 122, California.
of America, 70: 19211941. 1980b. 1990.
IS] Bard, P.Y. and Bouchon, M .. "The Twodimensional Resonance of Sedimentfilled 127] Kennett, B.L.N, and Engdahl, E.R., "Travel Times for Global Earthquake Loeation
Valleys", Bullelin of the Seismological Society of America, 75: 519541, 1985. and Phase Identification", International Journal of Geophysics, 105: 429465, 1991.
[9] Bullen, K.E. and Bolt, B.A., An Introduction to the Theory of Seismology, Cambridge [28] Khan~ P.K . "Recent Seismicity Trend in India and Adjoining Regions" ~ ISET~ New
University Press, Cambridge, 1985. Lett.. October 2003July 2004, 1014, 2004.
[10] Celebi, M., ''Topographical and Geological Amplifications Determined from Strong [29] Kim, W.Y., "The ML Scale in Eastern North America", Bulletin of 'he Seismological
mmion and Aftershock Reconds of the 3 March 1985 Chile Earthquake", Bulletin of Society of America, 88(4): 935951, 1998.
Ihe Seismological SocielV qfAmmca, 77: 11471167, 1987. [30] Langston, c.A., Brazier, R., Nyblade, A.A., and Owens, TJ.. "Local Magllitude Scale
f111 Conrad, Y., "LaufzeitkuTvcn Des Tauernbebens", vom 28: 59; 123. MiH.Erdb. and Seismicity Rate for Tanzania. East Africa:' Bulletin of the Selsmo[ogtcal SOCiety
Komm. Wien. 1925. of America, 88(3): 712721, 1998.
112] DeMel', el aJ., Current Pia" Mnlions, 101: 425478, 1990. [311 LeV!et, A., Loup, C., and Goula, X., "The Provence Eanhquake of June II, 1909
[13] Faccioli. E.. "Seismic Amplification in the Presence of Geological and Topographic (France): New Assessment of Near Field Effects", Proceedings of the 8th European
Irregularities", Proceedings of 2nd lmernolional Conference on Recent AdvQn<es Conference of Earthquake Engineering, Lisbon, 2. p. 4.2.79, 1986.
in Geotechnical Earthquake Engineering and Soil DY1Ulmics. SL Louis, Missouri, 2: [32] Love, A.E.H., Some problems of Geodynamics, Cambridge University Pres" 191 L
17791797, 1991. [33] Lowrie W. Fwufamenla/s of Geophysics, Cambridge University Press, 1997.
114] Geli, L., Bard, P.Y. and Jullien, B .. "The Eff""t of Topography on Earthqu.ke Ground [34] MacM~rdo, J., "Papers Relating to the Earthquake which Occurred in India in 1819",
MOtlon: A Review and New Results", Bulletin a/the Seismological Society of America, Philosophical Magazine, 63: 105177,1824. ..,
7K: 4263, [988. [35] Moczo. P. and Bard, P.Y., "Wave Diffraction, Amplification and Dlfferenttal Mollon
[15] Gutenberg, B., Th,' Energy of Eartlu/uakes, 112: 114, 1945. Near Strong Lateral Diseontinuities", Bulletin of the Seismological Society of America,
116] Gutenberg, B. and Richter, c.F., Seismicity of Earth OM Related Phenomenon, 83: 85106, 1993.
Princeton University Press, Prineeton, New Jersey, 1945. [31\1 Mohorovicic, A., "Das Beben Vom 8 x 1909", lb. Met. Obs. Zagreb, 9, 163, 1909.
117] Gutenberg, B., "Magnitude Determination for Deep Foeus Eanhquakes", Bulletin of [37J Musselt, A.E. and Khan, M.A., Looking into the Earth: An Introduction to Geological
the Seismological Society of America, 35: 117130, 1956. Geophysics. Cambridge University Press, 2000. .
11S] Hatayam., K., M.atsunami, K .. Iwata, T. and Irikur., K., "Basininduced Love Wave [38] Narayan, J.P., "Site Specific Strong Ground Motion Prediction Using 2.5D Modelling",
in tbe Eastern Part of the Osaka Basin", foumal of Physics of the Earth, 43: 131 155, Geophysical Journal Imernalianal. 146: 269281, 2001. . .
1995. [39] Narayan, J.P. and Rai, D.C., "An Observational Study of Local SIte Effects In th~
[19] HUllon, L.K. and Boore, D.M., "The Me Seale in Southern California", Bull<tin of the Chamo1; Earthquake", proceedings of Workshop on Recent Earthquakes of Chama/.
Seismological Society of Americo, 77: 6: 20742094, 1987. and Bhuj', 273279, 2001.
[20] Jibson, R., "Summary on Research on the Effects of Topographic Amplification of 140] Narayan. J.P. Sharma, M.L. and Ashwani Kumar, 'A Seismological Report on the
Earthquake Shaking on Slope Stability", OpenFileReporl87268, USGS. California. January 26, 2001 Bhuj, India Earthquake', Seismological Research Letters, 73:
1987. 343355, 2002.
[21] Kanamo,;, H .. "Mechanism of Tsunami Earthquake", Physics of ,he Earth Pinnet. (41) Narayan, J.P. and Prasad Rao, P.V., "Two and Half DimensionalSimulation of Ridge
Interiors, 6: 246259, 1972. Effect' on the Ground Motion Characteristics". Pure and Applted Geophys.cs, 160:
[22] Kanamo';. H, "The Energy Release in Great Earthquakes", Tectonophysics, 93: 185 15571571,2003. .
199. J977. f~2] Naravan, J.P., "2.5D Simulation of Basinedge Effects on the Ground MotIon
[23] Hank, TC. and Kanamo]'i. H" "A Moment Magnitude Seale", fGR, 84: 2343..2350, Char~cterj5tics" > Proceedings oj Indian. Academy of Sciert("es (Science of the Earth
1979. Planet), 112: 463469, 2003a. .
1241 IS: lH93. indian Standard Criteria for Earthquake Resistant Dt::,rgn of Structures, 143] Narayan, J.P., "Simulation of Ridge Weathering Effects on the Ground MotIon
Part 1, BIS, Ncw Delhi. 2002. Char1teterislics". Journal of Earlhquake Engineer/nR, 7: 447461, 2003b. .
I Narayan, J.P., "3D Simulation of BaSinedge Effects on the Ground Monon
[25J Kawase, H.. 'The Cause of Damage Belt in Kohe: 'The Basinedge Effect', 1M]
Constructive Interference of the Direct SWaves with the Basin Induced Diffracted!
Rnylcigh W3\,(';';". Seismologicm ReselJfrh Lel1ers, 67: 2534. 1996.
I
I
Characteristics", 13WCEE, August 16. P'dper No. 3333. Vancouver, Canada, 2004.
L
. , . ( Eart~ Resistant Design oJ Stf"UetUW$ )
[45) Narayan, J.P., "Study of Basinedge Effects on the Ground Motion Characteristics
Using 2.5D Modelling", Pure and Applied. Geophysics, 162: 273289, 2005.
[46] Oldham, R.D., "A Catalog of Indian Earthquakes from the Earliest Time, to the End
of A.D., J 869", Memoir X, Geological Survey of India, J 883.
Chapter 2
[47] Oldham, R.,?., "The Constitution of the Interior of the Eanh, as Revealed by
Eanhquakes , Quarterly Journal of Geological Society of London, 62: 45675, 1906.
[48] Pedersen,H., ~tzfeJd, D., CampiIJo, M., and Bard, P.Y., "Ground Motion Amplitude Seismic Zoning Map of India
Across RIdges , Bulletin of rhe Seismological Soceiry 0' America 84: J7861800
1994. ""
[49] Pitarka, A" Inkum, K., Iwata. T., and Sekiguchi, H.~ '''Threc dimensiona1 Simulatlon
R
of the Near Fault Ground Motion for 1995", Hyogoken Nanbu (Kobe). Japan
earthquake, Bullerm of rhe SelSmoiogicai Soceicv of America, 88: 428440, 1998.
[50] Pl~SInger A., Zm~s~al, M . and Zednik, J., Automau~d Pre~processing of Digital
Smmogroms: PrlflClples and Software. Version 2.2, E. Bergman (Ed.i, Prague and
Golden, 1996. 2.1 INTRODUCTION
[511 Rayleigh, Lord, "Onwave Propagated Along the Plane Surface of an Elastic Solid",
The goal of seismic roning is to delineate regions of similar probable intensity of grou~d moti~n
Proceedmgs of the London Mathematical Society, 17: 411, 1885.
in a country, for providing a guideline for provision of an adequate earthqua~ reslstan~ in
[52] Reid, H.F., The California Earthquake ofApril 18, 1906, Publication 87, 21, Carnegie
constructed facilities, as a step to disaster mitigation. Various social, economIC, and pohtIcal
Institute of Washington, Washington, D.C., 1910.
considerations govern the prescription of a minimum standard of safeguard against earthquake
[53] Reid, H.F., "The E1a,<tic Rebound Thenry of Earthquakes", Bul.ierin of DepanlMn! of
in the design of a structure. These are 0) economic concept of, 'acce,ptable ~isk', and (ii) ans~er
Geology, 6: 413444, university of Berkeley, 1911. to social question 'How safe is safe enough;. The strongest mtenslty of likely ground motIOn
[54] Richter, c.P., "An Instrumental Earthquake Magnitude Scale" Bulletin of the
is based on the anSwers to the above tWO posers, In terms of pure economic theory. earthquake
Seismological Soceity of America, 25: 132, 1935. '
causes two types of losses known as primary and secondary losses. A primary. loss is
[55] Richter, c.F., Elementary Seismology, W.H. Freeman, San Francisco, 1958.
irrecoverable loss, which results in the losS of human life in earthquake. All other losse5 incurred
due to earthquake that can be recouped are termed as secondary losses. Thus minimum standa~d
[56] SanchezSesma: FJ., "Elementary Solutions for the Response of a Wedgeshaped
MedlUm to InCIdent SH and SV Waves'\ Bulletin of the Seismological Soceirv of in a code to withstand earthquake is prescribed such that complete collapse of structure t$
America, 80: 737742, 1990. ..
prevented which ensures that no human life is lost. This requires a forecast of the strongest
[571 Sa~ake. K.. Tsunamis, International Handbook of Earlhquake and Engineering
intensity of likely ground motion at a particular site during the service life Of. struclure. Thm';
Sel1mwlogyPan B, Lee, et .1. (Eds.), 437451, 2002,
estimate of acceleration. velocity. displacement, frequency content and duratlOn of expected
[58] Siro, L. "Southern Italy November 23, 1980 Earthquake". Proceedings of 7th
maximum strong motion is required for a site. Seismic zoning map of a country segregates
European Conference on Earthquake Engineering, Athens~ Greece, 1982.
country in various areas of similar probable maximum intensity of ground m~tion. ~e
[59] Wood, RO., "Distribution of Apparent Intensity jn San Francisco. in the California
maximum il1lensity is fixed in such a way that the lifelineJcritica~ struet~res WIll re:n~m
Earthquake or April 18, 1906", Report ofthe Slate Earthquake investigation Commission, functional and there is low possibility of colJapse for structures deSIgned WIth thc provlsiOn
1: 220245, Carnegie Institute of Washington, Washington. D.C., 1908. 
provided in the code even for the occurrence of earthquake with higher intensity. Thus (\
structure designed with the provision of code can suffer damage of both structural and non~
Struchmil type. The damage is repairable but its economic viability is not warranted.
Very heavy
Damage zone
Light damage
Bombay S",ble
plateau
Madras
,,
.,
,; : '1;'
Scale
100 0 100 4<)0, km
FIGUR~; 2.1 ~;.rl.l!qu.ke Damage Zone oflndi. by West (1937). F1GURE 2.3 Seismic Zoning Map of lodia by Jai Krishna (1958 and 1959).
L
eM ( Earl/Uluake .Resistant Design of Stt'"Uct'Uf'CS ) ' _..
.
_...
.. 'I
Chaptc1' 2 Sei$mic Zoning Map of India) U
than V' V VI vn VIII IX and 'X and above' in twelve point MM
Mithal and Srivastava (1959) classified occurrence of earthquakes in India on thickness of produced mtenslues ess ".., 1
continental shelf using geophysical data and based on Assam (J 897 and 1950), Kang'" (1905) scale. Enve10pes for various MM intensities stated as above were drawn. ,These enve opes were
and BiharNepal (1934) earthquakes. Three seismic zones (Figure 2A) named as belt of odifi d to take into account of past seismic activity of smaller magmtude earthquakes, the
frequent. occasional and rare earthquakes are of continental shelf thickness of more than 1500 J.1l d Je '  al tectonic features and local ground condition (lithological groups),
f
tren ThO p~ncl~ t round Deihl' were modified in conformitv to the trend of Aravalli Hills.
m, less than 1500 m and sh~eld blocks of marginal overburden. respectively. e lSOselsmal a ~  h
DeUary isoseismal was also modified in the activity trend of mmor ~hocks that occurr~d ]n ~ e
re 'on ~tending from Chennai in Tamil Nadu to Thiruvananthapuram In KeraJ~. A SelSmlC ~n~ng
Zone 1Frequent earthquakes zone Ill. (F' 25) was, adopted in IS: 18931962. In thIS recommendatIon
Zone 2...()ccasiQnal earthquakes zone map havmg reven zone~ 19ure . ,
Zone 3FeW Or rare earthquakes zone
SEISMIC ZONES
(1962)
7"'1'..,:11
7.."",;V
7o:r~V
and mitigation efforts 8.'isociated with the earthquake hazard. In view of s.canty data of pas.t 7<1o:\I(
earthquakes in the country, the committee agreed that evolution of zoning map based only on \
statistical approach is. not likely to provjdl! a representative sei~mic bazard appraisal. Broad I
seismotectonics framework of the country was considered to be the basis of the ~ismk zoning. I
This wa, augmented by preliminary tectonic map prepared by GS] and map of epicentres of past i
canhquakcs: prepared by the india Meteorological Department (IMD), These maps wl!re
included in the Indian Standard Recommendation for Earthquake Resistant Design of Structure
IS: 18931962. The committee utilized the average intensity attenuatjon (intcnsitymagnitude
l'IGURE 2.5 Seismic Zuning Map .f IS: 18931962,
distance) relationship to draw idealized isos.ci'>mal of twentythree major earthquakes. that
.,j
l
the ZOne corres~onding to MM intensity of less than five is termed as Zone 0 (zero), to sug est
) _ _ _ _~Chapter 2 Seumic Zoning M(Jp of India.) .i.
map included in the published vo1ume was so small that the regions bet\Veen MM intensity
that Iatera~ loadmg On .the stnJcture was envisaged to be s.mall so that the design of structu! to vVI (seismic zone 111) boundaries were not visible at the southern part from the east of
carry verueal loads wIth proper factor of safety was considered to be adequate. Vadodara in Gujarat. Moreover. IMD had assigned magnitudes to many historical earthquakes
. Guha (1962) prepared a seismic regionlisation map (Figure 2.6) with the premise that past using correlation between magnitude and felt area. Tectonic Map was published by GS} in
earthquake data (eplcentres) contams all the information pertaining to seismotectonic setup of 1964 at the time of International Geological Congress. The BIS committee revised the seismic
~~~=u~tr~I I~~~l~~:S l~ere drs'NIl and five seismic regions corresponding to MM intensity zoning map to account for new available information of the historic data and to provide
d "~~. X and Xl were termed as no damage. moderate damage heavY additionaJ emphasis on ge{)iogy and tectonics in demarcating zones. The committee also decided
. a~~g.e ~nd very heav~ damage. The presumption of availability of historic seismicity'data is that the number of zones and methodology need no changes. The major modifications can be
Imp tCIl In the assumpUOn made to prepare the seismic regionalization map.
grouped as:
L
._ , . M
_ (BC1'''"'l
,,_....~ke Resistant Design of Structures
.~~ I c_
to one and termed as MM intensity 'IX and above' zone. Epicentre location, steep geothermal
gradient, movem.ent of crust, geomorphic evidence of offset of small landfonns. strati.graphic
evidence. gravity anomalies, were utilised in demarcating seismotectonic setup of India.
SEISMIC ZONES
(1966)
I Krishnaswamy (in Srivastava, 1969) proposed five principal seismotectonjc units (Figure 2,8
and Table 2.1) based on these. The proposed seismotectonic units are as follows:
I 140 lW ?
Stale
12.0 21" km
I
I
....= "
z....,l

""" zJ
=
E3
_m
,w
7_Y
of l' .
_ , ec.;lSIon t at a single zone of
data indicated that zones correspon:E~gl~~ Me MreV,lsed ~.e~swln. The availability of strong motion
In ten,.."ttes X and 'X and' Po
pro( ucmg comparable ground cu;celeraf10 n, Th us, these two zooes were decided 3 ve were capabJe
10 be merged
i, zone and their extension corresponding to MM ;ntensity VII wlth lslands of MM
intensity VIII.
l
TABLE 2.! Generalised tectonic units of India with decreasing magnitude and frC4{Uency of earth(IUukc occurrence (after
Krishnaswam3' (in Srivastava, 1969).
.~~ ... 
I
Tectonic unir Eorthquake occrtrnmce Seismic
~'S:
ZOlles
<
Name De5uipliol/
~.~.. ~
Orogenic Orogenic unil of Cenozoic folding Common shocks of magnitude 5~6.5 IV and V 111
I~
and faulting. The Sbillong Massif, with a number of shocks of magnitUde
which has been greatly affected by
this faulting. has been included in
6.57.5, a few shocks of magnitude 7.5 . .,
8 and occasional shocks greater than 8
this zone. originating on some of the major Hima ~
layan thrust and faults (Satlitta thrust,
Pinjal thrust, Central Himalayan Ihrust.
Dauki faults, etc,)
i
iIl.'"
2. foredeep and marginal Duit of Himalayan foredeep and mar~ Common shocks of magnitude 5,6 with III and IV with
depression ginal depression {where the boundary a few shocks of magnitude 67 and islets of V
is llot positively established, sOlUe of occasioual shocks of magnitude 7.58
the marginal parts of the shield may originating along active faults in the
~
really be included in this zone. The basement (patna fault? Of other base
Tectonic Map provisionally defines ment faults, Kutch faults),
the boundary at 200~ 1000 m contour of
r
r
"<
~a
Vi o
~O'S'f;f~
ot::!:tQ..o
"~"'1i~\l~'"
" 0 _. _. t;; " ~ ;>
8 it '~p;
"
....
g~~~O!:::T'~::f
::;:l P
I;::l .... ..0 ~ 01
_o9a::;:l~(;U)
~ ~ 8. ~ ~ c. ~, g ('\l ('\I ~ tn ae.
"O~oE=~~e
v> ..,::::l1
o 'g ::t (") ~ 0 0 o 0""'"'
"'""'10" :;::,  ' <
e'j;lo~.e.."""~N ""::::!
&;<1'9 g ~ e:;;;.g ~'l!!. [:::: S'
__ . 0 ~ ~ El " _<~. r;l 511.
s fl~;;S.~SOio~ ~i1qo!?~
l:1'a::J~O:::l*7' =~a~p.
,;q' ~ II (l 0 '" :::, 0 <
~
.. ~ ~~ ~~ e:.
:r..o:::1Il:
~
~0 ~0
~_.
:!'gfl'O &:0
t:! E '<: Q g (il e::
~ ~~. < ()'
'"
~'~g"ag.gg~ E3 a.
" , , 0 , . , .  0 ""
a
""'" ::I S '" '" <l  _. 
""g0CJ..~::r'fi;('J')
t') no til til .,
8.~e!l
I:l;:;;
i< 1,1>
h. R3'
cT o
tl'
5' .....
0;;
'Os<\:
~ g ~. ~ I:l ~
;; o~s.:::s,
'0
V>/:I)
(l8.a,;e II
..~".
!::'
.... til 0 
c:: "1
'7l;:::J.c .
~ ()'S
:::,1 ....
,,< '" :::: (l~ o
t
:!II"> El" ~.
~~ oil' "
e;;6:
11 <!i ;;: 0"
0' ~ 0.. f,:I. a
flI 0. s 'a.
~ t z :::t> ~ ~~~g i!:.
&~ g:;' oS t;;<~
..... ,<; N ...... ;!i:!
~< 0: "
.'>C
;>
 \:l
s
.o~
PE..
1lo~a."..,
03
09
2
::so~ .... 
9 a~ 5_.
,,< '"
::;:l
o()QV:
So ~~
oil
.
Il.
~i g~g~~
5'
;o~R
" e: " il ~
....; < g ~ . . . :T
0" og.o.t;:::;
:::r 0", e;...........
"'g ~ ~" !l ..,
~ .a. I!Il
=~
(") ""
a ~ gE.
0.....
~~
~~o
~?~~.o
IJ.> 0
;.;l~i'ig.
oo[!~
0 i
"""
!? ~ ~~
:::l.... t::I
'"
I
(
 ..... ~
mm~~ _____   
" , I, ~., ,"
( .,IfI]' ...
S,',m'" Zoning _M
.....".p of j:n~
~.~)
(iii) around western Nepal and Pithoragarh ofUttaranchal was associated with extension of
Himalayan thrust and
(iv) epicentral tract of 1934 BiharNepal earthquake in seismic zone IV.
An islet of zone IV encompassing epicentral region of ] 967 Koyna earthquake was also
introduced in the new Western coast seismic zone III. Seismic status of entire Assam, Mizo hills,
Tripura and northeast part of Kashmir were enhanced by one based an tectonics and seismicity,
In recognition of total tectonic framework entire Andaman island was demarcated as zone V.
Two new seismic zones of leveJ III were introduced along Western coast and along marginal
depression jn Rajasthan. Seismic status of parts of IndoGangetk trough was raised due to
fedemarcation in accordance to tectonic features. Narmada~SonTapti zone was merged with
Damodar graben in the east. The fsosei::.mals fur BeHar)' and Coimbatore earthquakes were
redrawn due to revi'>ed estimates of magnitudes. The islet around Coimbatore was nonexistent
due to introduction of costa) belt of seismic zone III in the Western coast. The entire
Lakshadweep group of islands was in seismic zone ill in confrumity to Western coast belt. The
seismic zone I was limited on the basis of extent of platform cover and past earthquake data
resuhlng in curtailment of extent of zone 1.

Srivastava (1974) indicated that the correlation of likely maximum Richter magnitude with
the seismic zones IV was 'less than 5', 56,66.5,6.57 and 'greater than 7'. The maximum _ ,""",11
The proposed hazard maps prior to IS: 18931962 presumed that earthquake processes in the
~~~~~~~~ ~~~~;~~~' ~:' ::;~:l ~:::;e~~e~"'~~fnclude :~ClTIh~r (;r~~:~;i:;dc~:::~'~;~nr~~:
fim1llec/l.:ommefce ministry to e..;;;tablish the m.IOJmum reqUJfcmC'Jl ..oh '" h dance; of
recent past in aU likelihood would be the same a'i those in the near future. These maps were . . h hI~l n' q. ha!,;ls man lor l C cOllnlt'r, Qlln "
Moreover. prior lO cmbarkmg on pro a I I.,) h... 1.0 lllz:,' ,l,." "
primarily based on the effect::: of four Himala.yan earthquakes. and Kutch earthquake. Milhal and diFferent scismo~ledonic block of the entire country should lit> fhaOl:.
C!?arthquake ResIstant Dliul'ign of StMlctUF'etl
)
SUMMARY
:::'e d~scription. of perceived threat from ea~~quakes in different pans of the country in the form Chapter 3
n1 a map began Jfl 1935. The most recent reViSIOn of this map was taken up 'tn 2002 Th' l'
f " . ' JS~~oo
o selSm.c zomng map of Ind.an subcontinent is described in this chapter, The basis and data
u~ed .In 1~~ pre~aratJon of these zO~Jng maps are discussed. The future trends in the preparation
of selsmIC zonmg map on the basts of the probabiiistic hazard analysis is also discussed,
Strong Motion Studies
REFERENCES
in India
Il] AUden~ J.B.,."Earthquake in Relation to D~modor Valley Project", EarrhqlUlke Engi
n~trmo ,Sem;,nar, H.L. Sally (Ed.), UmversJty of Roorkee, India, pp. 212216, 1959.
[2] ~eS!, WD .. Earthquake Tn Ind.a ,Presidential Address)" 24 bu:1' S" C
pp. 199227, 1937. ' wn "enee ongress,
r3) Kri;;~na, J., "~arthquake Engineering Problems in lndia", Journal of Institution. 0/
Enf!,tneers, IndIa. 1958. 3.1 INTRODUCTION
(4] ~ris:lI1a, ~., "S~ismic Zorl:ing of India", Earthquake Engineering Seminar, H.L Sally,
(Ed.), UmversJlY of Roorke<:, India, PI', 2431, 1959. The UNESCO had dec1ared the nineties as the international Decade for Natural Disaster
[5] Mirhal, KS. and Srivastava. LS., "Geotectonic Position and Earthquakes f G Reduction (lDNDR). Ironically, this decade also witnessed four earthquakes of magnitude 6.0
Brahma t R . " . HL 0 anga and above occurring in the Himalayas. centra] India, and the peninsular region resu1ting into
, pu fa eglOn. 111 , . SalJy~ (Ed.). Earthquake Engineering Seminar,
UniverSIty of Roarke<, India, Pl'. 217233, 1959. approximately 10000 casuaJties and many more injuries. These figures would have been much
[6] IS: 18931962, lmiian Srandard RecommendaJions for Earthquake Resistance of higher, if these earthquakes had occurred in the neighbourhood of large urban centres, as was
Stru('tures. Indian Standards 1nstitute, New Delhi. 1962. seen in the case of the Bhuj earthquake of January 2001. This earthquake alone accounted for
[7] Guh~, S.K. "Sei~miC Regionalisation of India", Secami Symposium on Eanhquake more than 15000 Jives in addition to having a crippling effe<:t on the eeonomy of the region.
Engmecnng, UnlVerSlty of Roorkce, India, 1962. These disastrous consequences eouJd have been avoided had the systems been designed to
IS] IS: 18931966, Indian Standard Recommemialions for Earrhquake Resistance of withstand earthquake ground motions, Quoting from the declaration of the United Nations
S,rU(/ures (FITS! RevlS.on), IndIan Standards Institute, New Delhi 1967, Inremational Decade for Natural Disaster Reduction (UNlDNDR) Yokohama Convention 1994:
[9] Knshnaswamy VK" "The Evolution of Seismic Zoning Map of India", Souvenir The Impact of natural disasters in tenns of human and economie losses has risen in recent
Volume, RM. Choudhury, (Ed.), Surh World Conference of Eanhquole Engineering years, and SOciety in general has become more vulnerable to natural disasters. Those
Prabhat Press, Uttar Pradesh, pp. 7789, 1977. ' usually most affected by narural and other disasters are the poor and soeially disadvantaged
[lO] IS: 18931970, Indian Standard Criteria for &rthquake Resistance of Structures groups in developing countries as they are least equipped to eope with them.
(S~ond Revlslon), Indian Standards Institute! New De1hi, 1971. Disaster prevention. mitigation, preparedness .and relief are four elements which contribute
Ill] Snvastava LS., "A Note on the Seismic Zoning Map of India", Bulletin of the Indian to and gain from the implementation of sustainable development policies. These elements,
.I0c/CI)' of Earthqunke Techn%ilY, 6(4): pp. 185194, 1969, along with environmental protection and sustainable development, are closely interrelated.
! 121 Snvaslava, LS., "Seismic Zoning Map of India" Earthquake Eng' . J' Therefore. nations should incorporate them in their development plans and ensure efficient
K "hn S' f ' ' . " l f u : e T m g  al
r~~ a IX, tetft BErth AlUuversary Commemoration Volume, Sarita Prakashan Me rut followup measures at the community, sUbregional, regional, national and international
lna.a, pp. 4965, 1974. ' e , levels.
I cq 1~: 1893. /lIdi~~ SI.aruiard Crileria far Earthquake Resi'ito17Ce of StructuresPart l: Disaster response alone is not sufficient. as it yields only temporary results at a very
(Jeneral Pr,()VlJum.~ and Buildings (Fifth Revision) Bureau of Ind' St d high cost. We have followed Utis limited approach for too long. This has been further
New DeihL 2002. ' lan anUm s.,
demonstrated by the recent focus On res.ponse to complex emergencies which, although
compelling shouJd not divert from pursuing a eomprehensive approach.
The facts are clearwe cannot prevent big~ destructive earthquakes from occurring.
These pose a continUing threat to lives and property in more than 55% of the area of this country.
HOwever, it is possible to avoid the disastrous consequences of an earthquake and that precisely
59
B1 (EQ,l'thqwzke Resistal'tt De8ign of Structure8 L....
, I e to " velocl'ru of shear waves in the rock mass near the source. The
is the objective of every sejsrruc design code of practice. The seismic design codes are framed locity of rupture is C os Ule OJ . . '
ve. .c waves observed at a site in the direction of fault rupture WIll have hIgher frequency )0
primarily with the objective of prevention of loss of life, In order to meet this objective it is m1
selS . th aves observed at an equally spaced site in the direction away from the
companson to e w , II 'J" 3 I J
essential that the structures/constructed facilities res.pond to the expected earthquake ground . .on of fault rupture. This phenomenon is illustrated schemauca y Ill. ~lgure ". n
dlre~u 1 the rupture directivity effect results in a large amplitude, short duratlOn pu.ls .at lhe
motions at the site in a designated manner which in turn depends on the nature of ground motion
j
exciting the structure. Thus the reHability of achieving the life safety penormanee objective of
any constructed facility is governed by the mos.t uncertain element in the chainexpecled
::7: e
:~ direction of rupture and a small amplitude, long duration pulse at the SIte In the
direction oppos.ite to the direction of fau1t rupture.
ground motion. The strong motion studies are aimed at reducing the uncertainties in the
specification of expected earthquake ground motions for designing any structure.
ncarfield strong motion data from s.everal recent earthquakes, Analyses oftbe nearfield strong
mNion data pointi'> to three d~stinct effect,.:; described as folJows: velocity pulse. which can lead to one Yleid reversal Wit11 a large ducuhty ~e~d, ~n th.e .~th~r
hand, fling ste'p affects the peak velocity and displacement of groun,d mollons, rhes~ n:a! a~ I
(i) Rupture direclivity "ffect Source effects which comprise of brief and impulsive ground mOflons. can nol be <ldcqua~e y
described in fj:e uency domain whIch characterizes iJ unifor',fl distribu~ion of energy t~rou~ out
The rupture direcLivjty dYed i:; essentjally a manifeslatioll of Doppler's effect in seismic wave the duration of ~ollon, Thus the t;onvel1tional charactenzatHln or
deSIgn grOlmd motion 10 the
propagation and affects the 1',rOllnd molion in fault parallel direction. ThiS occurs when the
CEarthqu.ake ReHiittant Design of Structures
M
STRIKE SUP
DIP SUP
J c Chapter 3 Strong Motion Studies in India)
(Plain view) attenuation wIn cause the amplitude of the body waves to decrease at a rate of l/r, where r
Fault
,c,, {Cross Section)
..... represent<; the radius of the wavefront. It has also been shown that the geometric attenuation of
surface waves occurs at a rate of 11,,"[7], Thus the surface waves attenuate much more slowly
Fiing St~"~~
than the body waves, which also explains the fact that the ground motion at large epicenttal
Ground Surface
distances predominantly comprises surface wave. In addition to lhese geometric attenllllUOl\
I \ Directivity Pulse
\ I' q Directivity Pulse effects, the seismic waves also experience an attenuation of amplitudes on account of the
; Fling Step ~
Fault,\:::
dissipation of seismic energy due to material damping in the soH. A combination of both these
effects represents the influence of path effects on the nature of ground motion at a site.
,,t,,
~'l~UR~ 3.2 S~h~~Hic illustration of the orientations of fling step and directivity I I 3.2.3 Site Effect
,trike.shp and dlpsbp faults (after [32]). pu50 n
Incorporation of the site effects in ground motion estimation procedures is aimed at reducing
form of reBponse spectra need!> to be augmented with a simplified description f th the uncertaiuty in the ground motion estimates as mea$ured by the standard error of the
pulses in tim d . A . I 0 e nearsource regression anaJysis and also to remove the potentia) bias in median estimates. The site effects
H ." e ,omam., lHmp e characterization is indeed possible with the use of Peak
.onzontal Vel<>C1ty (PHV), approximate period of the dominant pulse (T,), and lbe number of represent the local influences on the nature of ground motion and include the local ground
"gmficant halfcycles of motion in the larger, faultnormal direction [16J, response. basin effects, and the surface topography. The local grou.nd response refers to the
response of shallow geological deposits to the vertically propagating body waves, The modelling
(iii) Hanging wall effect should ideally involve lbe full soil profile up to Ihe bed rock level at the site but for deep alluvial
deposits, reasonable estimates can usually be obtained by considering the soil profile only
~: ~anging v.:an effect is primarily due to the proximity of much of the fault to the sites on up to a depth of 100120 m. The basin effects correspond to the influence of two or lbree
gmg wall Side, It has been observed to have lbe most pronounced effect for periods shorter dimensional extent of the sedimentary basin structures on ground motions, including critical
~": about J s, and at locations away from the topedge of lbe fault on the hanging wall side, body wave reflections and surface wave generation at basin edges. The dividing Hne between
rupture d1fec~vtty effect, on the other hand, is due to rupture propagation and radiation the local ground response effects and the basin effects is rather arbitrary and USually the local
pattern effects, It IS more pronounced for periods longer than I s, and is concentrated over the ground response effects refer to the onedimensional response of soil column and basin effects
top edge of the fault Tbe relationship between lbe rupture directivity effect and lbe han in are considered account for the observed ground motion characteristics that deviate from the
wall effeci IS th~ complementary both In the region of influence and lbe affected period ~.eg predictions of onedimensional analysjs, The presence of ridges and valleys at the site can also
!herle~y mcreasmg the degree of spatial variation of strong ground motion around dippi;g' have some influence on the nature of ground motion, An approximate estimate of these effects
.au ts (2J,
for certain cases of ridgevalley terrain is possible by using analytical solutions for some
'ha Sites on lbe hanging wall of a dipping fault have closer proxintitv to the fault as a whole idealized geometries [24J,
' ~ do the SItes at t~e same closest distance on the foot wall side. ca~sing larger short period The effect of soil layer on the nature of surface ground motion can be divided into two
motIOns on the hangmg wall than on lbe foot waiL The hanging wall effect is observed to be principal components: (i) the predominant period of surf... ground motion. and (ii) spectral
lhe greatest In the closest distance range of 8 to 18 km for pen' ods of 0 t 06 dd amplification with respect to the bed rock motion, These effects may be studied by analyzing
to UOlty at 5 s I J 1. 0 . s, an ecreases
the recorded strong motion data after eliminating the source and path effects from the strong
motion recordings. Two different approaches have been adopted to achieve this end: (i} by
3.2.2 Path Effect comparing the strong motion recorded on soil site with motions from a reference site (usually
rock) 15, 31J. and (ii) without using any reference site recordings [11, 19J,
Although the lotal energy released during an earthquake is a constant parameter for a particular
'lclsmlC event the specific energy, which is defined as the seismic energy per unit volume (i) kll'rence Site Approach
\ ecre2SC':: du'" to advancin fT'  ,
'. < v g wave r(lIlL hIS decrease is purely geometrical jn nature as the
vo I ume 01 the medium over wh' h th I" , If a reference site can be found in the dose vicinity of the soil site then the motions at both sites
. f lC e tota seIsmIC energy IS distributed Increases as the
W<:lVe ront advances When the earth
, yf. I '.
k .
qua e energy IS released from a fault below the ground are cxpe<:led to have similar source and path effects. Thus a comparison of the two motions
su. ace. )~dy waves travel away from the source in all directions, ]fwe assume the ru ture zone prOVide!>; an estimate of the local site effect. In practice ratios of either response spectr'd, or
to he D pOint .'.;(JUree, the wavet'ronts wiH be spherical and it can be proved that the ~eometric (smoothed) Fourier spectra of the motions recorded at soil site and the reference site is taken
to be a representative of the transfer function. of the soil column at the site. The predominant period
Chapter S Strong Motion Studies in India} M$_
MtfM ( Earth.qu;J.;e Re~i.stant Deaign of Stf"UCturea )
the ground motion parameter in question. Th~ uncertainties ~s~ociated .with this ~stimale are
and amplification factor can be estimated from the location and amplitude of peaks of derived reflected in the error term (fJI) y of the regresslOn. Such predlctive relatIons. have mdeed been
transfer functions. developed for the Fourier and pseudospectral velocity spectra [36, 37, 39J, peak ground
acceleration [8, 10, 12, 13) and strong motion duration [21, 22, 43], etc.
(ii) Nonreference Site Approach
Since the availability of a rocky outcrop in the dose vicinity of the site can not always be
~teed., a f~~ approaches have been proposed for estimation of site effects, which do not 3.4 THE INDIAN PERSPECTIVE
reqUIre avadabllity of reference site data, A very popular approach is to take the ratio of the The strong motion studies were taken up in India in 1976 on the recommendations of the
spectrum of horizon~l component of motion with respect to the spectrum of the vertical planning commission. Two types of strong motion recording instruments namely. Roorkee
component of the mouon (termed as HIV ratio) at the same site. This spectral ratlo is considered Earthquake School Accelerograph (RESA) and the Structural Response Recorders (SRR)
as thde tranhsfer funetion ?f the site. The ease of applicability of this approach has lured many were designed and fabricated at the Department of Earthquake EngIneering, Indian Institute of
to a opt t IS approach In slte characterization studies. However, it has been shown that the Technology, Roorkee (then, University of Roorkee). Today, the department maintains a network
estimates (predominant period and amplification factors) predicted by HAl ratios are not stable of over 200 acceJerographs and 350 SRRs in the HimaJayan and subHimalayan bell as shown
for me ,sa~e site using differe~t records [25], Yet another nonreference site approach involves in Figure 3,3, Seventeen earthquakes have been recorded by the instruments of this network
nonnahzatlOn of the spectra ot recorded motions by a reference spectrum for rocky site obtained tiU dale with the most recenl one being the Kutch earthquake of January 26, 2001 i9J. The dala
from some attenuation relationships {30], ~ recorded by this network of strong motion instnlments has been of immense help in develo~rnent
. Since the data from large, strong earthquakes in nearfield region of intense motions is of the standard spectral shape in the recent revision of the seismic design code IS1893 (Part
qUIte sparse, the use of microtremor data and/or records from small, frequent earthquakes has
I): 2002 [23J. Currently work IS in progress for the installalmn of 3dimensional array of strong
been proposed to study the site effects [14, 17]. Unfortunately, the behaviour of soil column motion acceJerographs In Tehri to study the effects of soil column response at variolls elevations,
du~ng strong motions generated by large, strong earthquakes diffeTS substantially from that
The analysis of data recorded by this network will aid in the development of a more rational
dunng small earthquakes (weak motions). In particular, the large strain levels associared with basis for design of underground facilities The strong motion data recorded by this network, will
the str?ng ~orions during lar~e earthquakes force the soil ro respond non~linearly as against an also allow studies on the effect of topography, and basin structure. In addition to the freefield
~se?t1allY bnea: response dunng sman strain weak motions. It has been shown in several studies
earthquake recording instrUmentS, the department also maintains networks of accelerometers in
m dif~erent regIons of the world that site effects estimated by using weak motion records~ or
,everal highrise buildings in different parts of the country.
the mlcrotremor data do not correlate well with the observed behaviour of the soil during strong
earthquakes [3, 6, 18,25,38,40, 41. 42, 44J.
......
~ 
............
,.
where t' rep~esents the desired ground motion parameter, (assumed to be distributed log
3.S UTILIZATION OF STRONG MOTION DATA
lJormal~y), M IS a measure of the size of the earthquake, R is a me<tsurc of sourcetosite distance There can he nO sholt~uts for developing specifications for design earthquake ground motIons,
fl, and 12 ~e some suitable functions of sollrce parameters (e.g., type of faulting, etc") and locai whicb have lO be derived on the basis of analyses of recorded !:trong motJOo data. The commoo1y
slfe condltJon:>: (soil/rock), respecfively. UsuaIly, such relations provide the median estimate of
L
) C=~_~~~"________,C",M=p!:.:t~r S Strong Motion Studie$...~!1 India)  .
recommended approach of using easily available weak motion data instead of data from less [5] Borcherdt, R.D., "Effects of Local Geology on Ground Motion ~ear San Francisco
frequent stron.g earthquakes may Jead to gross errors in ianduse planning in addition to leadin Bay". Bulletin of the Seismological Society of America, 60: 29"'{)1, 1970.
to unsafe desIgns [251. Strong motion data helps in developing a better under<tanding of (i~ [61 Bresnev, l.A. and Wen, K.L., "Nonlinear Soil ResponseA Reality?", Bullerin of the
g~ound response near fault ruptures of large earthquakes, (ii) effects of severe shaking on
Seismological Sociel)! of America, 86: 19641978, 1996.
dIfferent subsurfaee structures and geologic materials, and (iii) ground response in areas that' [71 Bullen, K.E., An Introduction to !he Theory of Seismology, Cambridge University Press,
~re prone to llquefacuon, Ana.)ysls ~f s:rong motion records also leads to the development of .
London, 1953,
Improved methods for gcneraung artificIal earthquake motions for regions where the data fro [8J Campbell, KW., "Near Source Anenuation of Peak Horizontal Acceleration", Bullelin
~eal "eart~quak~s ~re not aValIable 127, 28), The data recorded by the network of instrumen: of the Seismological Society of America, 71: 20392070, 1981.
In hIghn.se .bUlJdmgs ca~ be used for deriving infonnation for remote monitoring of the health [9) Chandra, B., Thakkar S.K., Basu S., Kumar A., Shrikh.nde M., Dos J., Agarwal P.,
of the ,bulldmg, the locatrons and extent of repair works required. and to verify adequacy of the and Bansal M.IL, "Strong Motion Records", Eanhquake Spectra, Supplement A to
analytical modeling and design guidelines.
Volume 18: 53"'{)6, 2002.
. The strong motion datu as recorded by the strong motion instruments. however, are not (10) Crouse, C.B., "Groundmotion Attenuation Equations for Earthquakes on the Cascadia
dlfectl~ useful ~or strong motion studies. The raw data has to be first processed and corrected
Subduction Zone". Eanhquake Spectra, 7(2): 201~236, 1991.
for var~ous P?SslbI~ sources of er:ors which might have crept in during the proeess of recording. [111 Field, E.H. and Jacob, K.H., "A Comparison and Test of Various Site Response
A detaIled. dis~ussJon of the varIOUS is!'ues of strong motion data processing may be found in Estimation Techniques, Including Three that are Non Referencesite Dependent".
other publIcatIOns [4, 20, 26. 33, 34. 35]. Bulletin of the Seismological Society of America, 85: 1127~1 143, 1995.
[121 Joyner, WB. and Boore, D.M., "Method for Regression Analysis of Strong Motion
Data", Bulletin of the Seismological Society of America, 83: 469487, 1993 (Errata in
SUMMARY
1994).
[131 Joyner, W.B. and Boore, D.M., "Errata: Method for Regression Analysis of Strong
T~i~ c~apter. cont~ns discussion of the various issues involved in strong motion studies
Motion Data". Bu/letin of the Seismological Society of America, 84: 955~956, J994.
prImarIly a dlscussJOn of w~a:. why, ~d how. The emphasis is on developing an understanding
[l41 Konno, K. and Ohmachi, T., "Grouodmotion Characteristics Estimated from Spectral
?f the problem of charactenzmg deSIgn earthquake ground motions. This will help readers to Ratio between Horizontal and Vertical Components of Microtremor". Bulletin of the
~nterpr~t carrectly the relevant clauses in design codes and also in making a judicious decision
10 SpecIal case:, requiring special attention beyond the scope of the standard codes of practice.
Seismological Sodel)! of America, 88(1): 228241, 1998
[l5] Kramer, S.L., Geotechnical Eanhquak Engineering, Pean;on Education, Singapore,
1996 (Indian reprint 2003).
REFERENCES [16] Krawinkler, H. and Alavi, B., "Development of Improved Design Proeedures for Near
Fault Ground Motions". In SMIP98. Seminar on Utilization of Strong MOlion Data,
[11 Abrahamson, N.A. and Silva. WJ., "Empirical Response Spectral Attenuation Oakland, California. 1998.
Relations for Shallow Crustal Eanbquakes". Seismological Research Letters 68: 94 [17J Lermo. J. and ChavezGarcia, F.J., "Are Microtremors Useful in Site Response
127, 1997. ' Evaluation?", Bulletin of the Seismological Society of America, 84: 135()...1364, 1994.
i21 Abrahamson, KA. and Somerville, P.G., "Effects of the Hanging Wall and Foot Wall [18] Mohammadioun, B., "Nonlinear Response of Soils to Horizontal and Vertical Bedrock
on Ground r..1olious Recorded during the Northridge Eanhquake", Bulletin of the Earthquake Motioo", Journal of Eanhquake Engineering, 1(1): 93119, 1997.
Seismological Sodctv (~f America, 86: S93S99, 1996. . [191 Nakamura, Y. "Clear Identification of Fundamental Idea of Nakrunura's Technique and
1,_~J Ak'1, K.. C'hin. B.!'!.. and Kate, K., "Seismological and Geotechnical Studies of Local its Applications", In Proceedings of rhe 12th World Conference on Earthquak
Sitc Effects on Strong and Weak Motions t " In Proceedings of tlte International Engineering, VoL 5, Paper # 2656, Auckland, New Zealand, 2000.
Symposium on !he Effects a/Surface Geology on Seismic MOlion, ESG1992, Odawara, (201 Novikova, E.l. and Trifun.c, M.D., "Digital Instrument Response Correction for the
Japan. Pa~e, I: 97~1 10, lASPElIlAEE Joint Working Group on ESG. Association for Force Balance Accelerometer". Earthquake Speclra, 8(3): 429442, 1992.
Earthquake Dlsasier Prevention, Tokyo. Japan, 1992. . [211 Novikova, EJ. and Trifun.c, M.D .. "Modified Mercalli lnstensity Scaling of the
14~. Boore, D. IvI., Stephens. CD" and Joyner, \V.B., "Comments on Baseline Correction Frequency Dependent Duration of Strong Ground Motion", Soil Dynamics and
O! Digi1.aJ Strong :"v1otion Data: Examples from the 1999 Hector Mine CaIifornia Eanhquake Engineering, 12: 309322, 1993.
Earthquake", Bulletin oj th('" Seismological Socie1v of America 9'1(4)' i5431560 [22] Novilwva, E.!. and Trifunac, M.D., "Duration of Strong Ground Motion in Terms
2001. ' . ' '" . of Earthquake Magnitude, Epicenrral Dislance, Site Conditions, and Site Geometry",
Eanhquak Engineering and Structural Dynamics, 23: 1023~1043. (1994),
)
:1 Strong Motion Studies in India m
'f M D and TodofOvska, M.I., "Nonlinear Soil Response as a Natural Passive
[40] T n unac, .  ;' S ./ D . d
[23] 1S1893, Indian Standard Criteria for Eanhquake Resistant Design of St",cturesPan IsoIatlOn . .c.
Mech aDl:smUA: 1994 Northridge California. Earthquake, 01 ynamlc' an
1; General Provisions and Buildings. Bureau ofIndian Standards, New Delhi, 2002. Earlhquake Engineering, 17: 4151. 1998. . . . .
[24] SanchezSe8ma, E, "Elementary Solutions for Response of a Wedgeshaped Medium 'f M D and Todorovska. M.l., "Can Aftershock Studtes Pred,ct SIte Amphfl
[41] Trl u n a c , . . . HS"D . d
to Incident SH and SVWaves", Bulletin of the Seismological Society of America, 80: cation Factors: Northridge. CA, Earthquake of 17 January 1994, ou yn.amu::s an
737742, 1990. Earthquake Engineering, 19: 233251, 2<JOO. . .
[25J Shrikbande, M. and Basu. S., "Strong Motion versus Weak Motion: Implications for T 'fu M D and Todorovska M.I . "Long Period MicrotremofS, MlcroselSms and
[42]
Microzonation Studies", Journal of Eanhquake Engineering, 8(1): 159173,2004. E~~~~~ D';"age: Northridge, CA, Earthquake of J7 January J994", Soil Dynamics
[26] Shnkhande, M . Basu, S., and Kumar, A., "Earthquake Strong Motion Data Process and Earthquake Engineering, 19: 253267, 2000. .
ing", In Atlas of Indian Strong Motion Records. M. Shrikhande (Ed.), Department of .f M .D, and Westermo , B .D ., "A Note on the Correlanon of Frequency
[431 T n unac, . Goo] ,
Earthquake Engineering, Indian Institute of Technology Rooritee, 2001. De dent Duration of Strong Eatthquake Ground Motion With the MMI and ~gtc
[27] Shrikhande, M. and GuPta, V.K" "Synthesizing Ensembles of Spatially Correlated Co~~tion at the Recording Stations". Bulletin 0,1 lhe Seismological Society of Amenc(}.
Accelerograms", Journal of Engineering Mechanics, ASCE. 124(11): 11851192, 67(3): 917927, 1977. .
1998. . d' FE and Trifunac. M.D .. "Comparison of Earthquake and MIcrotremor
[441 Udwa la, . . . IS' .r
[28] Shrikhande, M. and Gupta, V.K., "On the Chardeterization of the Phase Spectrum for Ground MotIons in El Centro, California", Bulletin of the Seismologu..a oCiery OJ
Strong Motion Synthesis", JourlUJ.1 of Earthquake Engineering, 5(4): 465482, 2001. America, 63: 12271253, 1973.
[29] Singh, I.P" "Earthquake Ground Motions: Implications for DeSigning Structures and
Reconciling Structur.1 Damage". Eanhquake Spectra, 1(2): 239270, 1985.
[30] Sokolov. Y.Y., Loh, c.R., and Wen, K.L., "Empirical Study of Sedimentftlled Basin
Response: TI10 case of Taipei city". Earthquake Spectra, 16(3): 681707, 2000.
[31] Steidl, H.I., Tumarkin, A.G. and Archuleta. R.I., "What is a reference site?'' Bulletin
of the Seismological Society of America. 86: 17331748, 1996.
[32] Stewart, J.P" Chiou S.I" Bray J.D., Graves KW" Somerville P.G., and Abrahamson
N.A~ "'Ground Motion Eva]uarion Procedures for Performance Based Design", PEER
Report 2001109, Pacific Earthquake Engineering Research Centre, Berkeley, 2001.
[33] Sunder, S. and Connor, J., "A New Procedure for Processing Strongmotion Earthquake
Signals". Bulletin of the Seismological Society of America, 72(2): 643661, 1982.
134] Sunder, S. and Schumacker, B" "Earthquake Motions Using.Jl New Data Processing
Scheme", Journal of Engineering Mechanics, ASCE, 108(EM6): 13131329, 1982.
135] Trifunac, M.D., "A Note on Correction of Strongmotion Accelerograms for Instrument
Response'" Bulletin of the Seismological Society of America, 62(1): 4014{)9. (1972).
[36] Trifunac; MD., "Influence of Local S'oil and Geologic Site Conditions on Fourier
Spectrum Amplitudes of Recorded Strong M(Jtion Accelerations"; Technical Report
8704, Department of Civil Engineering, University of Southern California, Los
Angeles, 1987.
1371 Tlifunac, M.D., "Fourier Amplitude Spectra of Strong !\1otion Acceleration: Extension
to High and ww Frequencies", Eanhquake Engineering and Structural Dynamics, 23:
38'1411, 1994.
138] Trifunac, M.D., HaD, T.Y.. and Todorovska, Ml., "On the Recurrence of Site Specific
Resjl<mse". Soil Dynamics and Earthquake Engineering, 18: 569592, 1999.
[391 Tlifun.c, M.D. and Lee, V.w., "Preliminary Empirical Model fol' Scaling Fourier
Amplitude Spectra of Strong Ground Ae<:cleration in Terms of Earthquake Magnirude,
Source 10 Station Distance, Site Intensity and Recording Site Condjtions". Technical
Report 85D3, Department of Civi) Engineering, Univer~ity of Southern California,
Lo~ Angeles. I ()g:').
c Cil4lpter 4 Strong Motion Characteristics J 4'M
(i) peak acceleration by using acceleration trace and it.:; calibration data.
(ii) duration of strong sbaking.
(iii) frequency of predominant wave and rough idea of frequency range,
Chapter 4 (iv) amplitude and frequency relation between horizontal and vertical accelerogram. and
(V) approximate distance of hypocentre from the recording station.
 ...",.V\';\J~V~l\;~V!t 15"W
CFixedtrnce Q 10 15 20
~~~!*,(Vflk~'fIfif(~~.,."" qt,." PI lit * 9ertt~ Time (s)
FIGURE 4.1 Traces of analog records of Uttarka,bi earthquake. engineering purposes~ the ground acceleration is the most significant parameter of strong
motion. being directly proportional 10 the inertia force imposed on tile structure<. The ground
70
. . . ( Earthquake ,Q~si ..tant _!J:.:e;::":J!gn=0:11'..::;St""'=:ctu""re~,:._ _ _ _ _ _ _ _ _ __
c
3 0.4 r...r...........,.......,..~=:::71 velocity is better correlated with the intensity of damage and it is a]so directly related with the
.8 0.2 energy transmitted to the structures, The ground displacement, however. is imponanr for design
~ 0.0 of underground pipelines and is also an indicator of the amOUnt of s.train the foundation of a
ilo 0.2 large structure will be subjected to.
." O.4~~ ____ ~._. __ ____
~ ~ __ ~ ____ ~ __ ~ __ ~
15.0
:'1 " :. I
20.0 25.0
:
30.0
: 1
35.0 40.0
The amplitude parameters of the ground motion were the earliest of the strong motion
parameters to be proposed. Typically, the peak values of the ground acceleration, velocity and
displacement give an idea of the severity of shaking at a site, predominant period of vibration,
etc.
Time (8)
FIGURE 4.3 Th~e components of the motion recorded at Uttarkashi. Peak Acceleration
flE~''':l
The Peak Horizontal Acceleration (PHA) is the most commonly used measure of the intensity
of shaking at a site and is taken to be the Jargest absolute value of the horizontal acceleration
recorded at a site. It is also possible to extract the maximum oftbe vector sum of two orthogonal
components of the horizontal ground acceleration recorded at a site. Ground motions with high
."  0 . 1 0 _ . _
........_ _'._ _ _ _'_
peak accelerations are usually, but not always. more damaging than those with lower peak
0.0 5.0 10.0 15.0 20.0 25.0 acceleration, However, a short duration stray pulse 'kith large amplitUde may not Cause any
Tune (s) significant damage as there is very litde time available for the system to respond to such
lJE~~"w,~'l
excitation. Therefore, the duration of the excitation is al,.:;o an important consideration in
estimating the damage potential of a ground motion.
via ratio
0.0 5.0 10.0 15.0 20.0 25.0 As the peak accelerations (aJ and peak velocities (v) are usually associated with the motion of
Time (s)
:3 D.lOr , , different frequencies, the ratlo via can be related to the frequency content of the motion. For
eanhquake motions that include several frequencies, the parameter 21rVla can be interpreted as
j
Vert.
..E ~ 0.2
."
(J[i(t)]'dl) takes place [22J. Usually the time interval between 5% and 95% contributions is .~ 0.2
taken as the strong motion duration.
0.1 ~0 0.1
0 0.0 Z 12 14
Z 0 0.2 0.4 0.6 0.8
Frequency (Hz)
Frequency (Hz)
4.2.3 Fourier Specl:TJJ.m
E
The frequency content (distribution of energy with respect to frequencies) of an accelerogram .sg 1.0
0.9
is represented by Fourier Spectrum. The Fourier transform of an accelerogram x(t) is given by. 5! 0.8
~
(4.1) 11 0.7
~ 0.6
E
Assummg ground acceleration as nonzero in t E (0, T] the Equation (4.1) can be written as, .g 0.5
+ ~'ll i .(r) e 't~,,{/f"! sin Ci) (t 1') dr: rteorded at Uttarkashi for different damping rahos.
F7 0 ~ d
(4.11)
Chapter" Streng Motion Charocteri6tica J. _
It may be seen that for damping ratio i; E (0,0. 0,20)
For an undamped linear elastic SDOF system, substitution of x(t) by the Equation (4,6)
S,(I;, ll\,) = W,S",<i;, w,) S",,(I;, ll\,) and ):(t) by tbe Equation (4,11), the Equation (4.21) reduces to,
(4.17)
~ 0.04
J ::: \
~
!t 0,002 0
"" 0.02
0,00 ':::::",_~'~
0,01 O,} 1
",_~_~ '" 0,00 L'~~~~~
10 100 5678910 5 10 15 20
Period (,) 15 20
Period (s) Frequency (Hz)
(aJ
(bl
FIGURE 4,8 Comparison of pseudospectral velocity spectrum ror 0% damping and tbe
FlGURE 4.7 Comparison of spectral and pseudospectral acceleration of longitudinal compo Fourier spectrum of longitudinal component rKorded at Uttarkashi.
nent recorded at Uttarkashi for S% damping.
(4.23)
where Yjk(iCO) is known as the coherency function and is a measure of correlation between the
given pair of timehistories, Sjk(iW) is the crosspower spectral density function for the pair
of motions recorded at stations} and k. Sjj(})) and Su(w) are the respective autopower spectral
density functions of the motions at stations} and k. riJ represents the projected horizontal
separatIon between stations j and k. and Yapp denotes the surface apparent velocity of
propagation of the wave at frequency fJ).
Although seismic waves of different frequencies, in genera], travel with different speeds
o.os 0.10 0.15 0.20 0.25 it is common to prescribe a constant value for vapp for aU frequencies because of inherent
Spectral displaccmelll (m) difficulties in the estimation of this parameter. The first factor, on the right hand of Equation
(4.23), represents the effect of incoherence and can be either derived from entirely theoretical
considerations [12, 25], or is empirically obtained from the anaJysis of strong motion array data
[4,5], and the second factor represents the effect of propagation delay due to finite velocities.
where. Pi/= ,U!/I, T) Efa/t)oj(t .... r))) represents the covariance between two orthogonal Pi) ~ (b:U)bj(/
components (Ji(t) and a.(t) and E[] represents the mathematical expeetation (ensemble average) where, the superscript r denore.s the rth sample from the ensembJe of the process and angular
operator. As a fir'::l app~oximatjon. rear earthquake accelerograms can be represented by shot or braCkets represent hme averagmg over the duration of motion.
white noise processes [71. In such 2 situation, the random variables ai(t) and ait + 1) would be It has been reported [17 J that there exists a strong correlation between the direction of
statjstically uncorrelated for non~zero values of time difference T. Hence, the elements of the one of ,the pdncipal axis (most often the major principal axis) and the general direction to the
covariance matrix of the ground acceleration proces::: may approximated by Pi} =: E[ar(t)a;{t}]. f.uJ'shp zone. However. analysis of data from San Fernando earthquake of February 9, 1971
Substituting from Equation (4.24) into Equation (4.25), the covariance matrix can be written
~~dlc~tes t~atthis eorrel.atlon is ~ot very strong. Further, it was observed that one of the principal
lCectlOns l~ usually aligned WIth the vertical direction [13]. For the Indian earthquakes also,
(4.26) no COrrelatIOn could be established between the directions of principal components of the
recorded motions and the direction of the rupture of fault plane [l8, 19J.
where. f.tq(!) = Elo,(t)a;Cti1 .and (;r;(= E[bi(t)blt)]) js the time invariant covariallce of stationary
processes bi(t) and b/t), fOf i; j = x, J, z, Furlher, the components of motion along an arbitrary Measures of severity
::.ti of orthogOlllJI ax~s x', y', Z can be transformed to components along orthogollal axes x, y.
F
1 hy " 2implc ir:m~;ff:rlTlatic" G:';, Various parameters have heen defined to cbaracterjze severjty of strong Shaking. Tht; Peak
~round Acceleration (PGA) is the most widely used parameter to measure severity of eaJthquake.
o~ever, the PGA js a rather poor parameter for mea<;uring severity of strong motion due to
(4.27) vanous. reasons such as, its possible association with a pulse of very high frequency, ampHfication
due to J~guJar loca1 topography, interaction of large structure at the site of recording, etc. The
Populanty of PGA as a measure of severity of ground motion is partly because of it being the
Chapter /; Strong Motion Characteristics) _ _
c
only parameter that can be directly measured by an instrumentall other strong motion This measure is defined as spectral intensHy and is given by
1"
parameters are derived from the processing of strong motion data, Further, the PGA measure (4.33)
is intuitively appealing to the engineers as it is proportional to the maximum inertia force Sl(O = S,,(s, T)dT
0.1
imposed on a structure during an earthquake. Several parameters that have been proposed as , (' n re re.'l.ents. the relative velocitv spectrum. Although
replacements for PGA as a measure of severity of ground shaking are discussed below. wbere, T is the period of SOOF, and S~, . P f Y _ 0 0' but presently it is customary to
Arias defined earthquake intensity as sum of the total energy per unit weight, stored in .. 1 . sed for dampmg ratiO 0 , ,  . , r
ongmal y tt was propo . d T [01' 5] and for lightlv damped structure S,("
undamped oscillators uniformly distributed with respect to their frequencies at the end of the cilcula te SI(0.05). In the range of pene E . , .
earthquake [2]; ot) ~ Sp,(" %l and the Equation (4.33) reduces to
2" 1 r2.5 " (4.34)
(4.29) SI( 0 ~ f,;'Sp,,(', T) aT = 2" J01 S'" (S, TiT dT
. f th seudovelocity and pseudo~acceleration spectra. The
where, x(t) refers to the ground acceleration, g is acceieration due to gravity and TD represents Here, Sir' and Spa. resp.ect)ve1y re er to , ee~sit is higher for strong motIon with ricber content
the earthquake duration. Thus duration and amplitude is implicitly considered in the definition Equation (~J4) unpbes that spe~r:u l~n of ~I as a measure of damage potential parameter of
of IA Since x(l) = 0 for t > TD , the Equation (4.29) can also be written as, of long penod waveforms. The hmltati f ~rtra The ef:&ect of dt1.fation of strong
.. . f th d finition 0 response S r   ' l' .
earthquake IS tnhented rorr: eel tra Thus spectral intensity SI value of strong motIon
motion is not accounted form response.spec ld b p"red for anv meaningful conclusion.
(4.30) . I ' '1 duration shou e com '" J ,
records. of appr?Xlmate Y slml ar _ rd could be due to large amplitude stray pulse, It IS
Smce a hIgh value of PGA In a reeo . fundo motion An alternative parameter
where, IX(w)1 is the Fourier amplitude spectrum of x(I). Thus IA will be large for strong motion not a reliable parameter to measure the severd't~ 0 dgrb W. tabe and Tohdo [23]. The EPA is
with significant amount of high frequency components, high amplitude and long duration. . A I . (EPA) has been e,me Y a .
Effective Peak cce eratlon ed d acceleration time history for which the
Hausner used the mean square acceleration during the rise tjme of strong motion to define defmed as the pe.ak ~alue of amplitude uuncat .. gr~Ut~me history. This way lhe effect of any
earthquake average power [9]. Let spectrum inlenslty IS 90% of that for the ~ngl.na.
. . h ded time history is elimInated. ,
spunous peak 10 t e reco: . ccounted for the effect of maximum amplitude,
(4.31) Araya and Saragom slmuJtaneously a , . escribing earthquake destructiveness
duration and frequency content of strong motton lTI pr
be the total energy in a strong motion. The earthquake average power is defined as, potential factor as [I]:
(4.35)
(4.32)
_ ..' . th 'ntensity of zero crossing defined as No/TD' No is ~he
where, to.05 and .0.95 are the time t at which J has 5% and 95% value respectively and the rise where IA IS Anas mtensl~ an~ J1fJ is e 1 am in lotal duration TD with positive and negallve
tolal number of zero crossmg ill an accelerogr. I t . the Equation (4 35) is most rational
time is defined as (Ts = 10.95 10.0,) the duration of strong motion part of an accelerogram [22J. ed d' age potentia parame ers ,. .
The foot mean square of P(1 is the measure of average rate of input energy to an elastic system slope. Among all t he propos am. not consider effect of inelastic deformations wbJch
for linear elastic structure, However, It does
and is denoted as rmsa ::: Fa. The larger valu&'of PQ is obtained for an accelerogram which is primarily responsible for damage. 'd 1 (N)based on inelastic deformatl0ns
is of short duration and impulsive in nature. The value of earthquake power and Arias intensity A criterion of equivalent number of Yle~ eyc es . d by Zahrah and HaH 124J. This
is comparable. Both lA and Pa are fairly good indicator of damage potential for brittle structure. . fi d d tility ratlo was propose
of SnOF systems for a speCl e ue d. . ted bv yieJdin'' \E/,\ in a structure
The elastic response spectra indicate directly how a linear elastic single degree of freedom .' . f the lOtal energy lsslpa J ,,,j
parameter IS defmed as the rat~o 0 under the resistancedispJacement curve tor the
(SDOF) system responds to strong ground motion. It aiso indicates maximum elastic deformation
When sUb.jeeted to ground motton to. the area' 't reache(' the same maximum displacement it
produced in structures having periods in the range of computation. However, it cannot be a good structure when it is loaded monotomcally unt)l 1 ~
predictor of damage potential as the damage is primarily an inelastic phenomenon. For ductile expeci~;nces during Ule excitation, i,e"
structure tne daIDllge depends on duration of strong motion, number of 'tress reversals and
(4.36)
amplitude of vibration excursions. Number of stress reversal and inelastic defonnatjon are
largely dependent on strong motion duration. To measure intensity of ground shaking from the
elastic response of structure Hausner {S] proposed an average response in a range of periodS.
where, (J) denotes the natural frequency of the SDOF structure, U y represents the yield IllJ Hudson, DE, "Some Problems in the Application of Spectrum Technique to Strong
deformati~n, and f1 is the specified ductility. The smallest value N can have js 1; in this case, Motion Earthquake Analysis", Bullelin of the Seismological Society ofAmerica, 52(2):
the structure yields only in one direction and reaehes its maximum displacement. 417430, (1962).
[12J Kiureghian, A. Der, "A Cohereney Model for Spatially Varrying Ground MQUQru;",
Earthquake Engineering and Structural Dynamics, 25: 99111, 1996.
SUMMARY [13J Kubo, T, and Penzien. J., "Analysis of Threedimensional Strong Ground Motions
Along Principal Axes, San Fernando Earthquake", Earthquake Engineering and
A discussion of various issues involved in the engineering interpretation of strong motion Structural Dynamics, 7: 265278, 1979.
data is presented. Starting with the explanation of basic tenninology used. jn ~trong motion 114J Luco, J,E. and Wong, H.L., "Response of a Rigid Foundation to a Spatially Random
seismology the reader is guided through the different forms of charactenunon of ~round Ground Motion", Earthquake Engineering and Structural Dynamics, 14(6): 891908,
motions. The chapter concludes with a diseussion of the various parameters used to quantIfy the 1986.
damage potential of the earthquake ground motion recorded at a site. This wili help in [15J Nigam, N.C., "Phase Properties of a Class of Random Processes", Earthquake
developing an understanding aboul the ground motion characterization and the parameters used Engineering and Structural Dynamics, 10: 711717, 1982.
to indicate the severity of the motion at a site. [16J Ohsaki, Y, "On the Significance of Phase Content in Earthquake Ground Motions",
Earthquake Engineering and Structural Dynamics, 7: 427439, 1979.
[l7J Penzicn, J. and Wat.be, M" "Characteristics of 3dimensional Earthquake Ground
REFERENCES Motions", Earthquake Engineering and Srructural DYlUlmics, 3: 365373, 1975.
[18J Shrikhande, M" Das, 1.D., Bansal, M.K., Kumar, A .. Basu, S., and Chandra, B"
llJ Araya, R and Swagoni, GK, "Earthquake Accelerogram Destructiveness POlential
"Analysis of Strong Motion Records from Dhannsala Earthquake of April 26, 1986",
Factor". In Proceedings of the Eighth World Conference on Earthquake Engineering,
In Proceedings Of the l:.1eventh Symposium on Eanhquake Engineering, Department of
San Francisco, California, U.S.A., Pl'. 1I: 835842, Prentice Hall Inc., New Jersey,
Earthquake Engineering, University of Roorkee, India, Dec. 1719, pp. 281285,
1984,
1998,
[2] Arias, A . "A Measure of Earthquake Intensityll. In Seismic Design/or Nuclear Power
[1~l Shrikhande, M., Das, J.D., Bansal, M.K., Kumar, A., Basu, S., and Chandra, B.,
Piant", R.l, Hansen, (Ed.), pp. 438469. MIT Press, Cambridge, Mass.chusetts,
"Strong Motion Charaetertisties of Uttarkashl Earthquake of October 20, 1991 and its
1970. Engineering Significance", In Research Highlights in Earth System Science; Volume 2:
13J Benda! 1.S. and Piersol A.G" Random Data, 2nd nd" John Wiley and Sons, 1986,
Seismicity, O.P, Vanna, (Ed.), Indian Geological Congress, Roorkee, India, pp. 337
14J Hao, H" Oliviera, C.S" and Penzien, J., "MultipleStation Ground Motion Processing
342, 200!.
and Simulalion Based on SMARTI Anay Data", Nuclear Engineering and Design,
120J Shrikhande; M. and Gupta, V.K., "Synthesizing Ensembles of Spatially Correlated
Ill: 293310, 1989.
Aceelerogr.ms", Journal of Engineering Mechanics, ASCE, 124(11): 11851192,
[5J Harichandran, R.S. and Vanmareke, E,H., "Stochastic Variation of Earthquake Ground
1998.
Motion in Space and Time", Journal of Engineering Mechanics, ASCE, 112: 154174,
[21] Shrikhande, M. and Gupta, VK" "On the Characterization of the Phase Spectrum for
1986.
Strong Motion Synthesis, Journal of Earthquake Engineering, 5(4): 465482, 2001.
(6J Hausner. G:W., '"Calculating the Response of an OscilJator to Arbitrary Ground
122J Trifun.c, M.D. and Brady, A,G., "A Study on the Duration of Strong Earthquake
Motion", Bulletin of lhe Seismological Society of America, 31: (43149, I94L
Ground Motion", Bulletin of the Seismological Society ofAmerica, 65: 581626, 1975.
[7] Housner, G.w., "Characteristics of Strong Motinn Earthquakes", Bulletin ~f the
[23J \Vatabe, M. and Tohdo, M., "Analyses on Various Parameters for the Simulation of
SeL,mological Society "j'America, 37(1): 1931, 1947,
Threedimensional Earthquake Ground Motions", In Transaction afthe 51h Imernational
fSl Hausner, G.W., "Spectrum Intensities of Strong Motion Earthquakes", In Proceedings
Conference on Structural Mechanics in Reactor Technology, 13~17 August 1979,
of the Symposium of Earlhquake and Blast Effects on Structure", Eartllquake
",umber Klfl in K(a), pp. 111, Berlin, Germany, 1979.
Engineering Research Institute, Los Angeles, California, pp. 2136, 1952.
1241 Zahl1lh, TE and Hall, W.J" "Earthquake Energy Absorption in SDOF Structures",
19] Housner, G.W., "Measures of Severity of Earthquake Ground Shaldng'. In Procedings
Journal of Structural Engineering, ASCE, 110(8): 17571772,1984.
of the US National Conference on Earthquake Engineering, Earthquake Engineering
[251 lerva, A. and Shinozuka, M.~ "Stochastic Differential Ground Motion", Structural
Research Institute, Ann Arhor. Michigan, pp. 2533, 1975.
Sq{ety, 10: 129143, 1991.
l Wi Hudson, D.E., Response Spectrum Techniques in Engineering Seismology", In
Proceeding." of the Firs! World Conference on Eartr.quake Engineering, Earthquake
Engineering Reseatch Institute, Los Angeles, California, Vol. 4. pp. 112, 1956.
5.2,1 Intensity
Intensity is a qualitative measure of the strength of an earthquake. It gives a gradat~on of
Chapter 5 t gth of earthquake using observed damage to structures and/or ground and reaction of
~:':ans to the earthquake shaking. An earthquake has many intensities, the highest near the
maximum fault displacement and progressively to lower grade at further away. Smce the
Evaluation of Seismic measure IS not instrumental, intensity can be assigned to historical earthquakes also. The popular
intensity scale is the Modified Mercalli (jJMl) scale with twelve gradation denoted by Roman
Design Parameters numerals from I to Xli. Another intensity scale deveJoped for central and eastern European
states is known as Medvedev.Sponheuer~Kamik (MSK) intensity scale. The twe]ve gradation
MSK scaJe differs with MMI in details only. Like many other countries, IS 1893 (Part 1), the
Indian Standard: 2002, also refers to the MSK scale [14J. An isoseismal map shows intensities
of a past earthquake in a contoured form of line of equal intensities. Note that the defined. scale
is subjective in nature and depend on social and prevailing c?nstructi~n pra~tJces. ReVISIon of
intensity scale from time to time is made so that gradatIOns of mtenslty as per current
construction practices are made,
of this moment as time [ ~ OCt IS known as the seismic moment. However. only geodetic data (i) L is equal to 75% of the subsurface rupture length,
can provjde MfJ as t ~ 00. Further. estimate of Mo is also made from Jow frequency cnd of the {Ii) the average surface displacement per event is about 50% of the maximum surface
seismic spectrum (period much larger than 20 s). This farfield seismic parameter is a direct displacement per even~ and
measu:e of the extent of faulting and lS used for comparison with ncarfield geodetic and {iii) the average subsurfaee displacement on the fault plane is bounded by the average
geo1ogical measurements. Jt may also be noted that surface wave. magnitude A1s is an energy surface displacement and (he maximum surface displacement.
measure and is determined by seismic wave amplitude at a period approximately in the range
of 18 s to 22 s. The moment magnitude Mw as defined by Hanks and Kanamori [11] is given
by. 5.4 EARTHQUAKE GROUND MOTION
Mw = 213 log Mo  6.7
CHARACTERISTICS
(5.1 )
Mw is in~in~ically related to seismic moment Mo (Nm). For values at about 6.5 the mb and Ms The earthquake ground motion of sufficient strength that affect human and their environment
scales eOlnClde. The small earthquakes 6.5) are better represented by mb scaJe and 114S scale (strong ground motion) arc of interest for earthquake resistant dcsjgn. The strong motions are
underestimates the same. The magnimde scales (M L . mb and Ms) salur.ite at Home upper bound. measured by occelnographs and its reeord is time history of acceleration (accelerogram). The
ML and mb saturate at about 6.5 and 7 respectively, Upper bound of Ms is about 8.5, Since, MLl temporal evolution of an accelerogram is composed of three parts viz. rise. strong motion and
mb and Ms are determined from seismic wave of particular period and wavelength that is mueh decay. It is compos.ed of nonperiodic sequences of aeceieration pulses of various dUT'dtions.
shorter than the earthquake source size of great earthquakes (magnitude eight or larger). The Thus not only peak of amplitude but also frequency content of record is necessary to characterize
A1w scale adequately measures the size of the ~ource since the scale is independent of particular accelerogram. The characteristic of strong motion in the vicinity of causative fault (near field)
wave type is strongly dependent on the nature of faulting. The motion depends on source parameters such
as fault shape. its: area, maximum fauh dislocation, and complexity of slipping process, stress
drop. and the distance of fault plane from ground surface. The elastic properties of the material
5.3 FAULT RUPTURE PARAMETERS through which the generated seismie waves travel also influence the strong motion. The effect
of ground shaking is mostly dependent on duration of strong motion part. The earthquake
Tocher 129J, Slemmons [281 and WclJs and Coppersmith 132] among others studied the corre ground acceleration is generally broadband in frequency composition. It is rieh in high
lations of fault rupture parameters (e.g. length and displacement) to assess the future earthquake frequencies in the near fields. The high frequency components attenuate faster than the low
pote?tjal ~n a region. Based on 216 worldwide past earthquakc WeBs and Coppersmith gave frequency components, therefore the contribution of high frequency components is reduced in
relattonshIp between moment magnitude Mw and fault rupture parameters. For all styles of the accelemgrams recorded at large distances from the fault. Further, the amplitude of ground
faulting. some of the relations are acceleration decreases with increasing distance from the causative faults in general. Moreover.
Mw = 1.J6 log(L)+ 5.08 0.28; in general the vertical component of the ground acceleration is richer in high frequencies than
log(L) = 0.69 Mw  3.22 0.22
the two horizontal components.
M", 2.25 log(W) + 4.06 0.41; log(W) = 0.32 Mw  1.01 0.15 (5.2)
Mw (j.9g log(A) + 4.07 0.24; log(A) = 0.91 Mw  3.49 0.24 5.4. I Amplitude Properties
where, L~ Wand A are surface mpture length (km), downdip rupture width (km) and mpture
Horizontal component of acceleration i~ primarily used to report ground motion as structureS
area (km) respectIvely. Simj~ar relationships between moment magnitude Mw and displacement
are also reported. The);e are, are designed for vertical loads and margin of safety in the vertical direction are usually adequate
for ear1hquake induced vertical load. TIle common amplitude measure of a ground motion is
Mw 0,74 !og(Dm) + 6.69 OA~); log(Dm) ~ 0.82 Mw 5.46 (l42 the largest horizontal acceleration and known as horizontal peak ground acceleration (PGA).
Mw O})2 log(D,,} + 6.93 0.39; The largest dynamic forces induced in very stiff structures are closely related PGA. Historic
lag(Da) ~ 0.69 Mw  4.80 0.36 (5.3)
earthquakc~ can have only intensity information. Various authors have attempted to propose
r\~hcre, D;II i~ maximum ~urfaee" displac0mc.nt (m) and D(I is avcmge surface di!l."piaccment (m). reJation between l'tfMl and PGA.. Ambraseys f2] propo~ed such a relation using ~outhern
I he maXlT),,11ll ~urface: displacement provides the farge~';t slip at a point along a rupture and turopean earthquakes as log a : : : ; 0.36 I MM  0.16. Later, using 187 records of 57 western USA
average surfa~ displacement gives the mean displacement along the length of rupture, 'They also eanhquake with MMI between III and X. Trifunac and Brady [30J gave a relation as log a ~
presented relations for different styles of faulting but concluded that difference is insignificant. 0.30 IMM + 0.014. Murphy and O'Brien [20], using worldwide data of 1465 records having more
'." :.
) Charlier 5 E'lJQluQtion af Seismic DeMgn Parumeters MS_
than 900 records with pe.k horizontal ground .""eleration greater than 10 cmfs2 and MMJ in 5.4.2 DuratiDn
=
the range I to X, also proposed correlation between MMI and PGA as log a O.25IMM + 0.25. Strong motion duration is related to the time required to release of the accumulated strain energy
These relationships are shown in Figure 5.1. Horizontal ground velocity is derived from in the causative fault. Thus, duration of strong motion increases with increasing earthquake
accelerogram and have less contribution from high frequency component tban acceleration magnitude, As stated earlier It has a strong influence on damage. HOlisner 113 j uses the mean
record. Most buildings are in the range of the frequency content of ground velocity. Thus peak square acceleration during the rise time of s.trong motion to define earthquake average power.
ground velocity (PGII) v is a better indicator of damage potentiaL Trifunac and Brady [301 also Let
proposed a correlation between MMI a.~d PGV. The ratio of PGV and PGA is a representation
of the frequency content of the motion. This ratio ean be jnterpreted as the period of vibration
of an equivalent harmonic wave and thus provides an indication of the significant periods of
J= f a'(/)dt (5.4)
the ground motion [26]. The peak ground displacement PGD d is the most inaccurate ground
be the total energy in a strong motion. The earthquake avemge power is defined as
motion information because of long period noise in the record and errors in filtering and
integration of accelerograms. The displacement record is associated with lower frequency
component of ground motion. Statistical analysis (Mohraz [19], Newmark and Hall [23]) on Po ~
1 r111
t; JI:H6 a
'1$ 2:
(t}dt (5.5)
ground motion was carried to estimate ground motion properties. These studies suggest uSe
of both vIa and adlv 2 to estimate ground motion parameters, According to Newmark and where, to.l1'i and 10.95 are the time t at which 1 has 5% and 95% value respectively. The former
Rosenbleutll [241 for earthquake of engineering interest the ratio ad/V' ranges between 5 and 15. is known as the rise time and tbe later is known as the decay time respectively, Trifunac and
This ratio can be considered as a measure for the sharpness of response spectrum in the velocity Brady [31] defined t, as the duration of strong motion part of an accelerogram and is defined
region. Small ratjo indicates a sharp response spectrum while a Jarge value eorresponds to a flat as ts = to.95  10.05
spectrum in the velocity region. Note that all the amplitude parameterS discussed are peak in
a single cycle of ground motion. The damage in structure is essentially cumulative damage and 5.4.3 Effect of DistJuu:e
it requires repeated cycle of high amplitudes. The use peak amplitudes for design purpose are
generally questioned on this ground. This leads to the concept of effective peak acceleration Attenuation law gives the effect of distant earthquake to the site and is expressed as peak ground
(EPA) and effective velocity related acceleration A,. The pseudo velocity corresponding to A, motion. The attenuatjon relationships in the literature can be broadJy classified into three
is termed as effective peak velocity (EPII). Let Sa be the mean pseudo acceleration value in the different types. The first one corresponds to those proposed by Bolt and Abrahamson [4]. These
period T E [0.1, 0.5] sand S, be the pseudovelocity value at about I s. for 5% critical damping relations are in the form of Pearson family of probability curves as functions of source to site
NEHRP, 1997 [21] speeifles the EPA is defined as Aa = Sa/2.5 and the EPV is similarly defined distanee jn various ranges of moment magnitude. They concluded thal the data does not indicate
as V,. = S,/2.5. increase of PGA systematically with the increase of magnitude in the near~source region. The
second group of attenuation relation i, of the form presented by Campbell [7]. These relations
IOrrrrrr, indicate that influence on PGA of site to source distance and magnitude is nonseparable, The
Ambrasey, (1974)  
Trifunac and Brady (1975) ~~~~. last group of attenuarion relationships are popularly known as Joyner and Boore [15J type. Basic
Mu!]>hy and O'Brien (1977) . feature of this relationship is separability of influence of magnitude and site to source distance
on PGA. Using we,tern USA earthquake data. relations for larger PGA, which are valid for
3 similar range of closest distance to surface rupture from site d:5 370 km. were given by Bolt
.g and Abrahamson [4]. These relations are specified for different r.anges of moment magnitude
I~ 0.1
Mw and are given by
L20 [R' + 1]0.033 exp (O.066R) for M wE [5.0, 6.0). R d + 23 and 0' A = 0.06 g
A ~ 1.20 [R'+ IJo.o42 exp (0.044R) for MwE[6.0, 7.0),R=d+25 andO',,=O.lOg (5.6)
{ O.24lR2 +1]10 exp ( Q.022R) for Mw E [7.0,7.7], R = d + 15 and 0' A = 0.05 g
Where, (Ttl. ifi the standard error of one observation (i,e. standard deviation of the prediction
Modified Mercalli Intensity (MMI)
error). These relations show similar trend in attenuation as predicted by the equation (5.6) in
FIGURE 5.1 PGAMM! ..I.tions. Mw E [5.0. 7.()). bur exhibir an entirely different trend in the range of Mw E [7.0,7.7]. This
MM QarthqttCtke Resistant Design oj StMlCture.(l ( CMp~tr 5 Evaluation oj Seismic Design Parometers
deviation in trend is mainly due to the use of nonlinear regression in this study. This study also The source to site distance R is defined as
attempts to define a significant (effective) peak acceleration. On the basis of worldwide
earthquake data Campbell [7] proposed attenuation relationships for peak value of both R = Jh'+d' (5.12)
horizontal and vertical components of ground acceleration. These relationships are defined for
moment magnitude Mw and the shortest distance R from site to the zone of seismogenic rupture where, d is the closet distance from site to surface projection of the rupture rotle ltl km.
on the fault. For peak horizontal ground acceleration (PGA), defined as the geometric mean of Campbell [7] recommends this attenuation relationship for Mw ~ 5 and source to site distance
the PG,4 of two horizontal components, in units of g (= 981 Cml52), the relation is given by, R ,; 60 km. This study .1'0 reports fitted equations for peak vertical acceleration. peak ground
velocity and pseudoabsolute acceleration response spectra. An equation of Joyner and Boore
InA 3.512 + 0.904Mw L3281nJR2 +[OJ4gexp(0.647 M w )]' [15) type for PGA in units of g was proposed by Boore etal. [5] using moment magoitude MOl'
distance d in kID from site to the surface projection of fault rupture and average shearwave
+ [1.125  0.112 In R  0.0957Mwj f + [0.440  0.171 In Rjs, (5.7)
velocity V. in units of mfs. This relation is given by,
+ [0.405 0.222 In R]s" + E
(5.13)
where, the following are recommended:
ofor strike~slip faulting; f = 0.5 for normal faulting:! ;;; 1 for reverse, thrust. reverseoblique where, b;. hz, b3 b s, b v are empirical constants~ h denGte~ a fictitious depth parameter, VA is a
and thrusloblique fauiling: 5, = $, = 0 for alluvium or flIm soil (Quaternary deposit with depth fictitious normalising shearwave velocity determined by regression analysis, and R = ~ d + h
2 2
> 10 m); s~ 1, Sh Z: 0 for ~oft rock (Tertiary Sedimentary deposits and soft volcanic deposits);
represents the source to site distance. The geometric mean of two horizontal component of
SJ = O. Sh :::::: 1 for hard rock (Cretaceous and older Sedimentary deposits, metamorphic rock,
ground acceleration is used as peak acceleration in this study. The standard deviation of the
crystalline rock. and hard volcanic deposits like basalt); and E is error of the regression relation
regression is given by
having mean zero and standard deviation <1. In this study magnitude data was postulated as
M .. = Ms for Ms;;' 6.0 and Mw = ML for ML <: 6.0. The standard deviation O'is correlated with (5,14)
In A as,
i:2 :2 (5.15)
0.55 if A <: 0.068 g <1IDA == ...J<1 e + (JT
0' = 0.173 0.140 In ,4, if 0.068 g :s A ,; 0.21 g (5.8) where, <1 and (J, are the standard deviation of earthquake to earthquake variability. determined
10,39 otherwise in the s:cond stage of regresslon~ and aU other components of variability respectively. In
equation (5.14), <11 is the standard deviation of first stage of regression, and <1c is the needed
However. a correlation between <1 and Mw is also reported in this study as. correction in standard deviation for randomly oriented horizontal component of ground
acceleration. The shearwave velocity, VI is used in the proposed relation to define site
r0.889  0.691 Mw if Mw <: 7.4 conditions. A time weighted average shearwave velocity value js used in the analysis and is
<i=<; (5,9)
,0.38 otherwise recommended as 30 m divided by shearwave, travel time from ,urface to 30 m helow. The
authors Tecommend b1 value according to the type of fault mechanism as
The equation (5.8) is found to be more robust th.n equation (5.9) by rsquared value.
Further, the shortest dis.tance from ~ite to rupture zone depends on the average depth h to the r 0.313 for strike.lip mechanism
lOp of [he sei~mogenic rupture wne of a presumed earthquake. In absence of any infonnation (5,16)
hI == 1: OJ 17 fOT reverseslip mechanism
it is recommended as
 0.242 for unspecified mechanism
h {O.5[h,+h b WSina+ ... ] ifn:!!h,
(5.10) Other recommended smooth parameters are b, = 0.527, h, ~ 0, b, = ~.778, bv ~ ~.371,
hf otherwise
VA = 1396 mis, h =5.57 km, 0', = 0.43' 0', 0.226, 0', =0.486, <r, =0.184 and 0'" A = 0.520,
where, hI and hI> re~pectively are the depth to the top and bottom of the seismogenie crust in The authors recommend this attenuation relation for moment magnitude Mw E [5.5, 75) and
km, a is Lhe angie of dip of the fault plane, and W is the downdip rupture wjdth in km, Down d 'S gO km, This study aho provides empirical relations for pseudoacceleration response
dip rupture width HI can he estimated using fcilawing empirical relation obtained by Wells and spectra in unlls of g at 5% damping for randomly oriented horizontal component of ground
Coppersmitll 132] acceleration.
Jog W~~I.OI, 0.32 Mw 0.15 (5.11 )
.i. (rE;;:::ar:;,"i:k=qu=.:i..
information about the type of rocks te be found and indicates faults in a geographic region. It
is necessary to detennine evidence of any motion in recent times. Locations of epicentres of all
The important consjderation in evaluation of design forces is the consequences of damage of available seismic data (instrumental and historical) in the region are to be plotted to determine
a particular type of structure from shaking. from ground failure, etc, Welldesigned and possible trend to indicate active faulting. A fault that does not extend to. earth ~urface and
constructed structure will be less susceptible to damage than old, poorly constructed structure. normally terminates upward at the ax.ial regIon of convex upward fold (AntLClme) IS known as
Very important structures. such as dams and nuclear power plants. whose failure would lead to blind fault. This fault cannot be examined and is thus associated with very high uncertainties.
disaster from secondary phenomena. should be provjded with least susceptibility to damage than But this type of fault has a very great effect on 1evel of ground motion estimation and is often
ordinary masonry buildings, which may be permitted to undergo repairable non~structura1 verv difficult to incorporate in the evaluation. Different faults have different degree of activity.
damage but prevented from structural failure and collapse. Thus the design criteria for very S~cific definition of fault activity is given for regulatory purpose. For quan.titative assess~e~t
Important structures will be different from that of ordinary andlor conventional buildings. The of activity, some time a capable fault is defined as a faull that had surface dIsplacement wlt~)m
enginee~ing project site with such structure, therefore, requires special consideration in defining past 10,000 years (Holocene active). Generally faults originate as sman. fracture, successl~e
appropnate levels of severity of ground motion at a given site to permjt analysis of the behaviour earthquakes propagate (lengthen) it. If a site does not belong to a ~ery ~ctrve complex tectonic
of the structure to remain functional' during and after the earthquake, One of these levels can rone. chances of new fault breaking are almost none for the servIce life of a structure.
be considered to be the maximum ground motion that reasonably can be expected to occur at
the site Dnce during lifetime of the :<:tructure. The earthquake corresponding to this level of
ground motion is often called as Design Basis Earthquake (DBE). The other level may 5.5 DETERMINISTIC APPROACH
correspond directly to ultimate safety requirements. This level of ground motion has a very low
probabiliry of being exceeded and represents the maximum level of ground motion on the basis When the causative fault cannot be jdentified. it is very difficult to avoid arbitrariness in
of estimates of upper threshold magnitude of seismic sources, The earthquake corresponding to specifying MC. The standard practice is to determine tbe intensity of the site (from the
the ultimate safety requirements is often called as Maximum Credible Earthquake (MCE), The available hwseismaI map of the region) from the strongest earthquake that has ever occurred
DBE is derived on the basis of historical earthquakes that have affected the site, expressed as around the site. Lacking strong motion accelerogram record, it is customary to increase the scale
ground motion having a definl!d probability of not being exceeded during the service life factor by one for the specification of MCE. Many earthquake intensityacceleration relationships
ro the facility and may be derivcd using probabilistic approach or the approach may include are available in the literature. These data suggest that the median value of the maximum ground
seismotectonic consideration (combined probabilistic and seismotectonic approach), An velocity is approximately 20 cmis for MMI intensity VIII and cbanges by a factor of 2 for each
alternative to rigorous probabilistic analysis for evaluation of DBE. when data on earthquake unit of change of intensity. The peak acceleration of 0.167 g thus correspond to MMI mtenslty
is meagre or not available, DBE is taken as a fraction (e,g, OA for the bridge site) of MCE, VIII, since a velocity of 122 cmls corresponds to a peak acceleration of g. The correlation
where MCE is determined by rigorous application of seismotectonic method. The MCE is between MMI and peak ground velocity is almost independent of the property of local soil. The
derived on the basis of maximum earthquake potential inside the seismotectonic provjnce of the relationship between peak velocity and peak ground acceleration is dependent sHg?tly on local
site or adjoining sejsmotectonic provinces associated with or not associated with specific tectonic soil condition. But the MMI is strongly correlated with damage and the damage IS dependent
structures. and combined probabilistic and seismotectonic approach may also be used based on on 10cal soil condition. Hence. local soil eondition is implicitly taken into account in arriving
available data on earthquake occurrence. However, design earthquake has to be prescribed so at peak ground velocity. When faults exist around the site, the earthquake of maximum
that duration and frequency content of ground motion is included for orOllnd motion considered magnitude associated with each fault is determined from past seismic data. However,
specification. There can be more than one design earthquake for a particular ;ite. care should be taken to ascertain the type of magnitude (e.g. body wave magnitude. surface wave
magnitude, local or Richter magnitude, moment magnitude~ etc.) in the reported data. From the
known length of faults from tectonic map, it is ascertained that the identified faults. associated
5.4,5 GeGlngiml, Geophysical and Geotechnical Data
with highest magnitude, are capable to release that amount of energy. It is customary to increase
TIle tdcntifieation of Sc1SJllic wurces is of prime importance for evaluation of ground motion these magnitudes by lI2 for Me. The epicentres of these earthquakes are assumed to be the
~eveL Various geological and geophysical parameters are studied to identify seismic sources. closest point on the faults from the site. The depth of focus of an earthquake gIves the depth
rhe~e studies include earthquake hi~tory, geological record of past seismic activity, tectonic at which the strains build up in the earth's crust andlor Upper Mantle resulung in fracture
maps, t'cr:nt (,..cloniC movemenl, surflJce landforms jndicators., lineaments map from remote generating seismic waves. The rocks in the immediate vidnity of a site are often not strong
sens~ng, elc. .It also inc1udes other parameters such as abrupt change in ground water level, :'jteep enough to store the energy for earthquake of magnitude of engineering sig.nificance (magnitude
gravJt:y gradJent, magnetic gradient, difference in seismtc wave velocities in the region, etc. greater or equaJ to five) in a few km of tbe upper depth due to weatherIng and other natura!
Moreover, locu! topogrl.lphy, properties of soil and its strength, area of subsidence andlor proce"cs. TIli, depth plus half of the idealized downdip rupture width, W is taken for the depth
settlement, N.c. shall be used. 11 is necessary to af)certain faults by ground verifkution of of focus of eacb earthquake in a causative fault jf no specific information of fault d1p IS
remotc:iy sensed lineaments map. The tectonIc (structural geoJogy) map of an area gives
Chapu:r ti E",a.luoti01'l of Seismic Design Po.rometef'.s ) MU
avaHable. The effect of these distant earthquakes on the faults to the site, expressed as peak where,O <: PI <: 1, P, = 1  P, and <P(.) is tile probabilily distribution function of N(O, I). The
ground motion (e.g. acceleration, velocity and displacement). is obtained vla attenuation law. estimation of parameters (PI' O'it Oi, v) arc formulated as mjnimum Chisquare problem. The
Various attenuation Jaws are available in the literature (e,g. [2]. [9] etc.), One of them is due magnitUde, ML of earthquake is independent of rate of occurrences of earthquake. The
to Esteva and Villaverde [lOl and is given by magnitude distribution is estimated in two ways in accordance wjth availabJe data. The
a = 5600 exp(O.8Mt )/(R + 40)2 (5.1 7) probability density function corresponding to the bilinear frequency magnitude relation with an
upper and lower threshold is given by
v = 32 exp(M,)/(R + 2S)' (S.l8)
d (l + 2()()~') v'la (5.19)
f.",(m) ~ K,p exp[(mo P+ (m,  m)AU(m  m,), mE (mo. m,l
m) (5.21)
where, 11K, ; [A + P A exp I(mo m) P}  [3 exp ((mo  m,) P+ (m, m,) A}]/(A + fJ), and
where, a is PGA in gal, v is PGV (emls), d is PGD (em) and R is focal (hypocenlral) distance
mo < ml < mz, The U() is Heaviside function. It is assumed that mo;;;: 5 is the lower threshold
(km). ML is the Richter magnitude and 40 km and 25 km are empirical constants to account for
magnitude. The estimation of parameters (A, ~. mj. m:z.) are obtained using minimum Chisquare
the volume of lithospheric rock that partjcipates in releasing the stored energy. The above laws
estimation method, Phy~ica] interpretation of equation (5.2 t) is that there ex.ist two different
indicate more T'dpid reduclion in value of high frequency component of ground motion,
processes, one leading (0 release of energy below magnitude mj; and other above that. A linear
relation can also be used with upper threshold magnitude m2 that takes care of the ultimate
srrengrh against ruprure of underlain strata, The pl:lrameler. m2 is inferred from historicaJ and!
5.6 PROBABILISTIC APPROACH
or geological data. Normahsatjon of linear frequencymagnitude relation with upper and lower
Vanous worker< (Cornell [8J, Esccva 191, Aigermissen and Perkins [1]. McGuire [17] and Basu threshold leads 10 the probability density function as
l3L etc.) developed methodology and techniques tor the probabilistic estimations of ground fM,(m) = K, Pexp [(mo  m) Pl, mE lmo, m,J (5.22)
motion. The combined statistical and seismotectonic approach for the evaluation of ground
motion parameter (acceleration, velocity and displacement) at a site involves identification of where, llK2 = 1  exp l(mo  mo) Pl. The attenuation law correlating peak ground acceleration
seismotectonjc province of the site and seismic sources in whjch future significant earthquake wilh earthquake parameters for a source including the effect of scatter in the past data is assumed
can originate, determine the rate at whjch earthquake can occur in different sources, obtain the lObe
frequency distribution of depth of focus and magnitude in varIous sources" and establish a y a exp[bm  cln (r + d) + 6] (5.23)
ground motion atrenuatlon to account for the effect of focal distance of earthquake on the site.
where, m is the magnitude. r is the focal distance (km), y is the peak ground acceleration (gal)
The analysis is carried out with the assumption that the available data is not exhaustive and
contain error in locations (say 0.1 degree). depth estimates, magnitude etc. The statistical tool
e
and is a nonna] random variable. Esteva and VilJaverde [1 O} have suggested values of a, b,
available for analysis is Bayesian analysis, and can be carrjed out on the Hnes of Basu [3J for c and d as 5600, 0.8, 2.0 and 40 Ian. respeclively and nonnal random variable 6 is of mean 0.04
and variance OA096. The occurrence of an earthquake is assumed to be in accordance with
evaluation of ground acceleration in following steps:
Poisson process with intensity, J.1i~ for magnitude greater than five. The posterior intensity of
(I) The data are s011ed out for different seismjc sources. earthquake arrival is estimated through Bayesian statistics and by using past regional (Newmark
(ii) A modular source of arc length 150 km al the surface ofthe earthquake wilh the project and Rosenblueth [24]) seismic data, Under the assumption of mutUal statistical independence of
site as it~ center and of 150 km depth is taken and seismically active faults lying withill various sources at project site, it can be sho\vn that, the probability distribution function of
modular 'i(\un::e are con.l>idered as area 50urces. maximum acceleration. YmftP can formally be
Lrn;3.liOlls of Hoating earthquakes (not associated with faults) jn the modular source are
considered t('mporarily stationary and spatial1y homogeneous: and occuncnce of earthquake is (5.24)
eqaally hkely in the latitude and longitude direction. The foeal depth data with assigned value
of 33 km (ZJverage depth or Moho) arc assumed to be distributed unifonnly wjthin 16 to 51 km
rOf (;stimalioJ of fo;.'.a! dislribulicn. /\ mhed truncated lognormal distribution is fitted in in which !1i is the intensity of earthquake arrival in lis SOUfCr:, n i~ the ntlITlher of faultll in the
"h? moiui;rr ~()llTC/' fmd !\rPfl ~C'llrCf~:; The proh<lbil1ty density function of focal depth for h madular SQuree at site, and pry, > y] is the probHbility of eXI.X".e,ding peak acceleration, y, at the
to, hn1 i~,., site due to the i lh source, Prom l".qllation (5.24) I~yr.tlr return period of PGA, )" i~ obtained as,
,
!H(!;) ~ L ,~, " ~
. cxp[(ln hv)'1(2ot)]
"J .:Jr(i1nc,bj(ln ho v)laJ
;rl
(520) (5.25)
Substituting t, numerically the Iyear acceleration is obtained at the site solving equation (5.25)
~umencany for y. In general peak acceleration (velocitYI displacement) for 100 years service
lIfe of the structure for various exceedence probability are evaluated from equation (5.24). The
level of probab1l1t~ .1S ~hosen with the consideration of the consequence of faHure. A low
T where, r is the hypocentraI distance (km) and m is the magnitude. AU empirical coefficients.
reported by McGuire [18) for projections of hori7..ontal PCA, PGVand PCD also. are reproduced
in Table 5.2. The probabilistic approach for the same site gives PGA of 0.23 g with exceedence
probability of 0.25 for lOOyear service life of the structure. The PGA of 0.25 g corresponding
exceeden~e probabIlity IS taken for atomic power plant stnlctures (say 0.05). For dams Jess to MCE can be recommended for the site by combjning both detenninistic and probabilistic
conservattve. exceede~ce probability (say 0.25) may be taken. The choice of the ground motion approach for evaluation of acceleration response spectra.
para~et.er WIth prescnbed exceedence probability is made on an engineering judgment based on
pemusslble damage to the structures and prescribing the levels of design ground motion. TABLE 5,2 Parameters for attenuation laws
Observation CoejjicienlS
5.6.1 Example
b1 I b, L b,
Him~iayan be~t. induding parts of IndoGangetic planes, can be divided into three seif'mic I !
~ro:mces. Major earthqUake...;; up to magnitude 6.5 have OCcurred in western Himalayas. The
Acceleration tg)
Veloci,y (m/,)
0.482
O.0564 ' 0.278
0.401 ~'
1.202
L?Wl
limned data on depth of focus of earthquakes, for which fault plane solutions are available do
not sho~' cl~arcut relationship with probable extension of the known thrusts and faults. Both
Dj;;placemeni (mm)
~'
3.93
L __ __
.
0.434
. .
0.885
.
the longitudmal ~nd transverse features are capable of future seismic activity in the region, In
t~e absenc~ of I~strumental evidence and precise depth evaluation of the seismological 5.7 RESPONSE SPECTRA
lmeaments m rehmon ~,o the project site for evaluation of design earthquake parameters can be
selected on the foHOWIng criteria: Earthquake engineers prefer to report interd.etion between ground aceeleration and structural
I. As fault plane solution indicate~ probabilities of seismic slip both along Ille longitudinal ::;ystems through response spectrum as popularised by Housner (12]. It reflects frequency content,
as weI.l as transverse tectOniC lIneaments in Himalayas, deep seated extensions of the amplitude of ground motion and effect of subsequent filtering by the structure, Acceleration
tecto.mc ~lanes :ejated ~ith the major thrusts and faults can act as capable faults. spectrum is a plot of natural period of vibration of a Single degree of freedom (SDOF) oscillator
2. The Idealil.ed WIdth of sbpped fault is about 30 km for magnitude 6.5. The focal depth with a specific value of damping versus peak absolute acceleration of o$;cillawr mass when
IS estunated to he 20 km based on the assumption that the rock in first 5 km depth in subjected to a base acceleration equal to the earthquake accelerogram (i,e .. ground acceleration).
the ImmedIate VIcinity of site may not be strong enough to store the strain energy. The value of the speetral acceleration at zero periods, known as zero period acceleration (ZPA),
TIlough the depth of focus of damaging earthquakes in the region are noted to vary is the PGA because oscillator is composed of infinitely stiff bnear spring. The relative
from about 25 km to about 100 km. a conservative estimate of depth of focus of displacement response spectrum asymptotically approaches maximum ground displacement for
20 km could be adopted for the MCE. Based on the seismotectonic setup of the region, highly flexible structure. This implies that the mass remains stationary for all practical purposes
the follOWing set of earthquake parameters reported in Table 5.1 are considered for and only the ground moves as the linear elastic SDOF system is composed of spring with negligjbJe
evaluatl?n of PGlt due to the MCE along the major faults around site based on the stiffness. Inbetween the two extremes period, the value of spectral acceleration at a particular
attenuatIOn law proposed by McGuire [18 J period is a constant multiplier. known as amplification factor. of peak ground acceleration. The
amplification facror at shortperiod increases with increase of period and reaches a maximum
at the subsoil period and then jt decreases with increase of period in general The amplification
(5.26) factor for rocky sjte condition is higher than that of alluvium site condition at short periods and
v ice versa at Jongperiods. TIle amplification factor reduces with increase of hypocentral
di::;tance from the site and peak amplification oCCUrS at longer reriod.
I
Three straight lines bound the general shape of the smooth spectra on a logarithmic tripartite horizontal component of earthquake. Newmatk et a1. [22] proposed transition from amplified
graph as shown in Figure .5.2. At low frequency range the spectral displacement Sa = maximum ground acceleration to ground acceJeration begin at 6 Hz.. for an
damping values and end at 40.
ground displacement d; and in the high frequency range, the spectral acceleration Sa ':': maximum 30. 17.0 and 9.0 Hz, for critical damping ratio 0.5, 2.0. 5.0 and 10.0 per cent, respectively.
ground accelerat10n a. As we proceed from low to high frequency, there exist five different Corresponding to I g ZPA, the peak ground velocity is 122 cml. and displacement is 91 ,m f()t
regions, These are: alluvial soil, and 58 cmls and 30 em for the rock. The measure of width of the spectrum is
adlv'l = 6 for both type of spectra. Figure .5,2 shows the spectrum for alluvial soil recommended
(i) a transition from maximum ground displacement to amplified spectral displacement, by Newmark et aI., for 1 g ZPA. Figure 5.3 shows the 84,} percentile (i.e, mean + one standard
(ii) amplified displacemen~ deviation), 5% critical damping spectra for the horizontal component of earthquake motion by
(iii) amplified velocity, Seed et al, [27], Mohraz [19J studied three components of ground motion. The mean value of
(iv) amplified acceleration and the ratio of smaller and larger horizontal component (RS) is 0.83 and that of vertical and larger
(v) a transition from amplified spectra) acceleration to ground acceleration, horizontal component (RV) is 0.48. The 84.1 pen:entile values are RS; 0,98 and RV; 0.65
The design spectrum Can be obtained from maximum ground velocity, displacement and which indicates that borh horizontal components are almost equal and the venical component
acceleration jf the amplifications are known. Table 5.4 gives the ampJiiication factors for larger is approximately 213 of the larger honzonta) component. Figure 5.4 shows average spectra
normalized to I g ZPA for 2% critical damping, Tables 5.35.5 give the ground motion para
meier. ampHfication factor for anuvium and proposed site design spectra coefficients, Given the
PGA is a, using Tables 5.35.5. 50 or 84.1 percentile horiwntal design response spectrum can
be obtained, The spectral values are Sd::; factor x d, Sv = factor x v and Sa :::; factor x a where
d, v and a are PGD, PGV and PGA respectively from Table 5.3. From Table 5.5, design spec
..r~~, ~ .
Total number of records analysed 104 Spcetra for 5% damping
4
Frequency. cps to 15 20 25 3D
FIGliRE 5,2 r)esig,~ s{.I(>ttra l"l!'wmmended by 'Newmark ("l at. rZZJ for 1 g PGA at 84.1 I FIGURE 5.3
Period, sec
Design spedra recommended by Seed U7J for 5% damping at 84.1 pen:endJe.
l
perc:cnti!~.
Mlii. ( E(J.rt~.ke Re,mtant DeJrign of Strud~".r::.",,_ _ _ _~_ _ _ _ _ _ _ _ _
) Chapter 5 Evaluation of Se';mic Design P~mmeters ) MlihW
5r~r_r_._~~ TABLE 5.5 Site design coefficient (after Mohraz (1976))
Alluvium CoejficiertfS
,  Less than 30 fl alluviwn on rock Velocity Acceleration
Site category Displacement
 ..  30200 ft, alluvium on rock
  Rock Rock 0.50 0.50 1.05
Alluvium underlain rock 0_75 0.75 L20
\'\
\\ trUm value can be obtained as site coefficients x design spectrum value for alluvium site.
\\\ \..

I'.
\. ''t \. .
According to the geolOgical condition of sire, taking averaged spectral acceleration as. a guide,
the spectral acceleration of each faults are drawn, An envelope, of all these spectra) acceleration
~~:::~~:::::2~:?:;::,:.~::::o, =:.._____ _
of various causative faults for a particular site, lS called acceleration spectrum of MCE. The
acceJer3tion spectrum corresponding to DBE is obtained by multiplying a fmerion less than
, equal to ha)f to the spectral acceleration of MeE. Figure 5.5 corresponds to the s.pectral accel
, ! . ' ! eration for MCE of rocky site with earthquake parameters given in Table 5.1. The spectral
O.l I.l 2 2.5 3 acceleration for DBE is used for working stress design and that of MCE is used for ultimate
Period, SeC. design.
FIGURE 5.4 Average d($i~n spei':tra recommended by Mohraz fur 2% damping.
Sile condition
TABLE 5.3 Ground motion parameters (after Mohr"" (1976
(m/~)/g:
Larger horizontal
iilVA:g':I'adIV',
!
v
mfs
I
=d
rom
via
(mJs)/g:
Vertical
I~dl'v'=rlvrd
,mfs  rnm
  Recommended spectra
  From Srinagar thrust
b "' From MBF (NB)
c ~ FromMCT
d    From MBF (8B)
SUe (:ontiiJion
SUMMARY
l
84.1
The seismic design parameters and ground motion characteristics are discussed in thIs chapter.
Rock Both deterministic and probabilistic approaches to determjne ground motion level are presented.
Various factors influencing the design ground motion parameters are discussed. Finally. the
Alluvium underlain by COnstruction of design spectrum needed for earthquake resistant design calculations is elaborated
rock < 9 m deep 2.53 3.30 1 33 209 260 3.38
at the end.
Alluvium underlain by
rock between 961 m deep 1.85 273 1.47 219 229 2.94
~_lu_vi~", ~ ___ .__ L.._._.c2c.07._ _ _27~ __ ~~ 20~.~ 2 02., ~ ..~2_5~
.,ItM lEarthquake
.. .
Resistant Design oj StructU'I"CS c Chapter 5 Evah.Lation of SeiMnic De81.gn Panl1netC'l"S ) _ ..4
REFERENCES [17] McGuire, R.K.~ "Seismic Structural Response Risk Analysis, Incorporating Peak
Response Regressions on Eanhquake Magnitude and Distance", Technical Report
II] Algermisen, ST and Perkins D,M., "A Technique for Seismic ZoningGeneral 755 I, Department of Civil Engineering, MIT Press, Cambridge, Massachusetts, 1975,
Consideration and Parameter". In Proceedings of the intemGJional Conference of [18] McGuire. R.K., '''Seismic Design Spectra and Mapping Procedures Using Hazard
Microzonation for SaferConstruction, Research and Application, VoL II, Pl', 865878, Analysis Based Directly on Oscillator Response", Earthquake Engineering and
Seattle. Washington, 1972. Structural Dynamics, 5: 211234, 1977,
[2] Arnbraseys, N,N .. "The Correlation of Intensity witl1 Ground ?I.1otions'. In Advances [19) Mohraz, B., "A Study of Earthquake Response Spectra for Different Geologie
in Engineering Seismology in Europe, Trieste. 1974. Condition", Bulletin of lhe Seismolagical Society of America, 66: 915932, 1976,
13] Bmu. S., "Statistical Analysis of Seismic Data and Seismic Risk Analysis of Indian [20) Murphy, J,R. and O'Brien, L.J" "The Correlation of Peak Ground Acceleration
Peninsula", Ph.D thesis, Department of Civil Engineering, lIT Kanpur, India, 1977. Amplirude with Seismic Intensity and Other Physical Parameters", Bulletin oj the
[4] Bolt, B,A. and Abrahamson, N,A., "New Atlenuation Relations for Peak and Expected Seismological Society of America, 67(3): 877915, 1977.
AcceleratiDns of Ground Motion", Bulletin of fhe Seismological Society oj America, [21] NEHRP, "Recommended Provisions for Seismic Regulation for New Buildings and
72(6): 23072321, 1982, Olher Structures", Technical Report, Building Safety Council for Federal Emergency
[5] Boore, D,M" Joyner, W.B" and Fumal, TE., "Equations for Estimating Horizontal Management, Washington D.e., 1997,
Response Spectra and Peak Acceleration for Western North American Earthquakes: A [22] Newmark, N,M., Blume, I.A .. and Kapur, K.K. "Seismic De. . ign Spectra for Nuclear
Summary of Recent Work", Seismological Research Letlers, 68(1): 128140. 1997. Power Plants", Journal of Power Division. ASCE, 99(02): 873889, 1973,
[61 Burridge, R, and Knopoff, L. "Body Force Equivalents for Seismic Dislocation", [23] Newmark, N.M. and Hall, W,J" "Earthquake Spectra and Design", Technical Report,
BulleTin o.fthe Seismological Society of America, 54; 18751888,1964. Earthquake Engineering Research Institute, Berkeley, California, 1982,
[7] Campbell, K.W" "Empirical Nearsource Attenuation Relationships for Horizontal and [24J Newmark, !\I.M. and Rosenblueth, E., "Fundamentais of Earthquake Engineering",
Vertical Components of Peak Ground Acceleration, Peak Ground Velocity, and Pseudo Prentice Hall, Inc" New Jersey, 1971,
absolute Acceleration Response Spectra", Seismological Research LeUers, 68(1): 154 [25] Richter. C.E. Elementary Seismology, W.H. Freeman and Co., San Francisco,
i79. 1997, California, 1958,
[81 Cornell, C.A .. "'Engineering Seismic Risk Analysis", Bulletin of lhe Seismological [26) Seed, H.B, and Idriss, I.M" "Ground Motions and Soil Liquefaction during
Societv ~rAmerica, 58(5): 15831606, 1968. Earthquakes", Technical Report, Earthquake Engineering Research Institute, Berkeley,
[9] Esteva, L. "<Bases Para 13 Formulacion de Decisiones de Diseno Sismico", Technical Californill, 1982.
Report, Institute de Ingenieria, UNAM, Mexico, 1968. (21) Seed, H.B" Ugas, C .. and Lysmer, J.. "Site Dependent Spectra for Earthquakeresistant
110] Esteva, L and Villaverde, R, "Seismk Risk Design Spectra and Structural Design", Bulletin of the Seismological Society of America, 66: 221243, 1976.
Re1iahiiity'. In Proceedmgs of Fifth World Conferenc(; on Earthquake Engineering. [28] Slemmons, D.B., "Determination of Design Earthquake Magmtudes for MlCrozo~
Rome, pp. 25862596, 1974. nation", in Proceedings of lrd international Earthquake MicrozolUltion Conference.
Ill] Hanks, Te. and Kanamori, H., "A Moment ?I.1agnitude Scale", Journal of Geophysical pp. 119130, 1982, ,
Research, 84(B5): 23482350. 1979. [29) Tocher. D" "Earthquake Energy and Ground Breakage", Bulletin oflhe SeismologIcal
[12] Housner, G,W .. "Calculating the Response of an Oscill.tor 10 Arbitrary Ground Society of America, 48(2): 147153, 1958.
Motion''. Bulletin afthe Seisnwlngical Snciety 0/ America, 31:143149, 1941. 130] Trifunac, M.D, and Brady, A.G" "On Ihe Correlation of Seismic Intensity with Peaks
113] Housner, G,W" "Measures of Severity of Earthquake Ground Shaking", In Proceedings of Recorded Strong Motion''', Bulletin of the Sei$moiogicai Society of America. 65:
Of 1/1(" US Natirmai Conference on 1:.arthquake Engineering, Etlflhquake Engineering 139162, J975.
Research Institute. Ann Arbor, Michigan, pp. 2533, 1975, [31] Trifunac, M.D. and Brady, A.G" "A Study on the Duration of Strong Earthquake
'14) IS1R93, Indi<ln Standard Criteria for Earthquake Resislant Design of Structures~ Ground ?I.1otion", Bullelin of the Seismological Society ofAmerico, 65: 581626, J975,
Parr J: General Provi,>;irms and Buildings, Bureau of Indian Standards, 2002. [32] Wells, D.L. and Coppersmith, KJ" "New Empirical RelationShips among Magnitude,
nS] }oyner', W,B. and Boore. D.M" "Peak Horizontal Acceleration and Velociry from Rupture Length, Rupture Width, Rupture Area and Surface Displacement", Bulletin of
Strongmotion Records lncluding Records from the 1979 lmperial VaHey. California lhe Seismological Society of America, 84(4), 9741(){)2, 1994,
Earthquake", Bulletill n,f the Seismni.rJgica/ Societ), of America, 71: 20112038. 1981.
I i ~1 tv'laruyama, T. "On the Force EqulvaJents of Dynamic Elastic Dislocations with
Reference to the Earthquake Mechanism", Bullr?tir1 of Earfhquake Rl'search Institute,
Tokyo t:l1ivcrsity, 41: 467~4R6, 1963,
Structural Dynamics
Chapter 6
6.1 INTRODUCTION
Any study of vibrations and related topics requires an under~1anding of the basic questionwhat
is vibration? Basically, vibration is an oscillatory motion of a particle or a body about a reference
position. Such motion may be simple harmonic (sinusoidal) or complex (non~sinusoidal), The
most strtking feature of any vibrating body is the effect of inenia which comes into play by
virtue of Newton's second law of motion, which states that the rate of change of momentum
of any body in motion is equal to the external forces acting on it, that is,
d (6.1)
F(t) = (mv)
dt
where, F(t) is the external (time varying) force applied on the body, m denotes the mass
(a measute of inenia) of the body, and v represents the instantaneous velocity of the body.
Equation (6.1) describes the instantaneous equilibrium that exists between various forces acting
on the system, if we assume the time rate of change of momentumwhich has the units of
forceto be a fictitious inertia force. Energy considerations often play an important role in
vibration problems, From the consideration of the principle of
conservation of energy, any vibrating system will, in general, have
three constituents: (i) a mechanism to store the Kinetic energy; (ii) a
mechanism for energy dissipation/Joss; and (iii) a mechanism to store
,
f '
J:
,, ,,
,,1" 8" \
,
r
the potential energy, responsible for the elastic restoring force in the
vibrating system. An example of the most elementary form of a ! \Img'ine
dynamical system, which every student of science is familiar with, is
a simple pendulum, shown in Figure 6.1. This simple system is a ~ >1,mgoose
classic demonstration of the energy conversions that take place in any
mg
I mg
dynamical system. once it is set into motion. Let us recapitulate some
of the basic aspects of the dynamics of this simple system. Tbe FIGURE 6.1 A simple
pendulum is set into motion by taking the bob to the extreme po~ition pendulum.
111
,.." ~7C;::r.llpter 6 Initiation into Structural Dynamic.:
and then releasing it. When the bob is at the extreme position the potential energy of the system For example, let us consider the develop~ Negligible axial
is at its maximum while the kinetic energy is minimum. As the bob approaches the mean position deformation in beam
men! of a mathematical model for the lateral "' Heavy mass concennted
during its downward swing. the potential energy is gradually converted into the kinetic energy load analysis of a simple portal
frame shown in ~ r;:==~=:;;:r_~.!!at!.!s~la'b level
of the bob so much so that at the mean position the kinetic energy is at its maximum with Figure 6.2. Since the mass of columns is very
, .
~, . _"'..['
. . . . ~jgid jomts ....
potential energy being zero. This energy conversion goes on in every half cycle as long as these small in comparison with that of the slab, It 1S (00 relative
oscillations persist. The oscillations eventually subside due to frictional losses caused by the reasonable to assume that the entire mass of the rouHlon) NegligibJe mass of
resistance offered by the air to the motion of bob. l
portal is concentrated at the slab level. Fur : the columns
hl::iT""":::':":
Having observed the physical phenomenon, it is then natural to enquireis it possible to
develop a mathematical model for describing it? The answer is a resounding yes, In fact,
mathematical modelling is an integraJ part of the study of structural dynamics. The solution of
ther, we note that the axial rigidity of the beam
and slab is very large in comparison wit the
stiffness of columns in the lateral deforma~
h jmal
i / Negligible,
stretch X
in columns
the differential equation governing the mathematical model correspond to the observed physical tions. Thus it is a good approximation to as .>;<;7:7
phenomenon. The governing differential equation, also known as the..equation of motion, is a sume that the beam/slab is infinitely rigid and FIGURE 6.2 A simpl~ portal frame.
second order differemial equation in time. The most important (and many a time quite difficult entire lateral deformation is due to the flexural ,
one) aspect of structural dynamics is the formulation of equation of motion, In general, the deformation!' in columns. Since) the change in length of the columns due to lateral deformatIOns
guveruing equatiun of a vibrating system can be developed by adopting any of thc foHowlng (assumed to be small) is nor very significant, it is a good firs[ order approximation to assume
fi ve approaches: that the axial stretch in the columns is negligible, Moreover, as the beams are ~sua1]y c.ast
L Newton's second jaw of motion, monolithically with the columns, the joint can be assumed to ,be rigid a~ the, re:atlVe, rotatl~n
2 Application of d' Alemben's principle. between beam and column at the joint win be negligible, With these slmphfymg kmematlc
3. Principle of vil1ual work, constraints. the lateral displacement of the rigid beamlslab is the only possible ~ode. of defor
4. Hamilton's principle, and mation in the system. Smce the entire mass is concentrated at the slab level: the mernal effects
5. Lagrange's equation in the model can be completely determined from the knowledge of the mollon of the slab. The
model resulting from aU the above mentioned simplifying assumptions is known as the shear
Of these, the fust two approaches are based on the principies of vector mechanics, whereas building model. The origin of this nomenclaUlre is that the shear force is constant .acr~ss the
the latter two approaches are based On variational principles, The approaches based on vector height of the column. It cannot be overemphasized here that the ~ode of deformatlon 10 c~l
mechanics are physically intuitive but invariably become intractable in ca.o;;e of complex umns is purely flexural" This also brings us to the concept. of dynam~c degrees affreedom. wh~ch
configurations. On the other hand, the variational approaches depend on scalar work~like is defined as the total number of displacements (and rotatlons) reqUlred to completely detenmne
quantities and can accommodate very complex systems without any djffieulty, However, the the inertial effects in a dynamical system. Accordingly, the
valiationai approaches are more abstract and lack the physical intuitive appeal afforded by the lateral deformation of the portal frame of Figure 6.2 under
vector mechanics approaches. The principle of virtual work is an extension of the equilibrium the influence of a lateral load F(t) can be represented as the
methods in the sense that it is a statement of no work being done by a system of forces, in response of a singledegreeof~freedom system shown in
equilibrium, in movIng through 3. set of vmual displacements consistent with the geometric Figure 6,3, This is a typical discrete springmassdashpot
constraints. For the purpose of introductory exposition to structural dynamics the methods based mechanical analog for the response of portal frame to lateral
on vector mechanics will suf1ke for estabHshjng the equation(s) of motion, FIGURE 6.3 Equivalent single
loads. The parameters of discrete model are related to the
dtgrte of fret'!dom system.
physical system (portal frame) as:
602 MATHEMATICAL MODELliNG mass m is: the total mass of the beam and slab of the frame and serves as the storage
for kinetic energy,
The study of HructuraJ dynamics involves deveJoping an insight into the dynamic behaviour of spring of stiffness k represents the combined stiffness of two columns ~or lateral
lhe structural systems by investigating the behaviour of thelr models under the influence of defonnattons and stores the internal strain energy due to co1umn dcformatlOns, and
dynarnlr loads, such as blast. winds, eanhquakes, heavy rotating machinery, etc. The models
used in these jnves.tigalion~ can be either smallscale laboratory models for experimental studies,
or tan be, mathematical models for analytical studies, The development of an appropriate lA:'ually it is also po1>,;,ible to aCCount for the massJinertla propertie& af the columns by adding on:lhi.'c of the
mathema.tical model for a specific study requires an understanding of the basic phenomenon and tot: ' f the columns and infill panels to the mass of the beam/slab, Thi:. "onethird" rule whIch .~ wldely
ill mas!'; 0 . ' 1 k' , 'valence
a clear idea of the basic mechanics, adopted in aU codes of practice for earthquake resistant design ha!l: jli, bailis in the lota tneH~ energy t m
1
criterion.
_ " _ L~~.~~hquak Resistant Design oj StNdti/res
dashpot with damping coefficient c represents the energy dissipation due to various
sources.,
thc excitalion F(t) is the lateral force F(t) applied on the portal frame.
Chapter 7
Since the essential properties of the dynamital system have been segreBated into
independent, discrete elements, such a model is also known as a lumped parameter model as
against a distributed parameter model or continuous system wherein all the properties are
Dynamics of Single Degree
distributed continuously throughout. Though all physical systems are essentially distributed
parameter system, it is nevertheless possible to get a fairly good estimate of the response of a
continuous system by investigating lhe behaviour of a suitable lumped parameter model. We
of Freedom Systems
shall see, jn the next chapter, how the dynamic behaviour of such a system can be described by
means of a second oroer linear differential equation with constant coefficients.
SUMMARY
This chapter provides a general introduction to the study of vibration problems, The process 7.1 INTRODUCTION
of rransfonning a physical phenomenon into a mathematical modeJ suited for numerical
experimentation is described. This should enable a reader to put in proper perspective, the In the preceding chapter we have noted that from the considerations of conser/ation of energy
detailed mathematical formulation of vibration problems in the following chapters. any vibrating system will. in general. have three constituents: (1) a mechanism to store the
kinetic energy, which is also responsible for the generation of inenia force~ (ii) a mechanism
for energy dissipationlloss; and (iii) a mechanism to store the potentiaJ energy. responsible for
REFERENCES the elastic restoring force in the vibrating system. In the simplest possible idealisation of a
vibrating system, these three mechanisms may be considered to be lumped into discrete elements
[IJ Clough, R.W. and Penzien. J . DYlUlmics of SlruclareS, 2nd ed., McGrawHili, as shown In Figure 6.3. The massJinertia element(m) stores lhe kinetic energy. the spring
New York, 1993. element (k) stores the potential energy, the dashpot (e) represents the viscous damper for
[21 Craig, R.R, Jr., Siructural Dynamics, John Wiley & Sons, New York, 1981. dis,ipationlloss of energy, and F(I) is an external time varying force. The system shown in
[31 Humar. J.L, Dynamics of Siruc/ures, Prentice Hall, Inc., 1990. Figure 7.1 is a Single Degree of Freedom (SDOF) system because it is only required to monitor
[4J Hurty, w.c. and Rubinstein, M.P., Dynamics of StruCfUI'r'S, PrenticeHall of India. a single quantity. viz., the movement of the mass (m). to completely describe the vibration of
New Delhi, J967. the system. In generaJ the number of independent displacements required to define the displaced
[5] Warburton, G,B .. The Dynamical Behaviour of Slructures, 2nd ed., Pergamon Press, positions of all the masses relative to their originaJ positlon is called the number of degrees of
1976. freedom for vibration analysis. A single degree of freedom (SDOF) system is the simplest
possible mathematical model in structural dynamics.
The motion of the mass m is governed by Newton's second law of motion. In order to
derive the governing equation of motion, let us consider the various forces acting on the mass
m as shown in lhe accompanying Free Body Diagram (FBD). The forccsf, andfv represent the
115
_ , , _ { Earthquake Resistant Design of Structures ) Chapt;;.r 7 Dynamics of Single Degree of Freedom Systems _'M
elastic restoring force and the force developed in the damper, respectively, These forces, along the mass is then termed as the free vibration and is given by the solution of differential
with the externa1 force F(t) act on the mass (m), which then cause the change in momentum of equation
the mass in accordance with Newton's second law as,
mx +Cx +kx=O
~(mX) ~ F(t)  Is  Ie (7,1)
or,
'i + 2~"' .i: + "'~X (l (7.5)
dr
where. x(t) denotes the displacement of mass m from its original position of rest For the specjal . . d fi ed bv ",'  kim and r is defined by t;; clecr with c,,; 2m",. = 2k1m ~ 2Jkffl.
where,l.t\,!s em , ., ct' dJ' risa
case of civil engineering structures, the mass of the system does not change with time, and there~ . alled the undamped circular natural frequency, measure In ra s, ~
The constant run IS C . 11 d th ,. I d ping
fore, the rate of change of momentum can be considered to have the same effect as that of dimensionless uantity caned the viscous damping factor; and CO IS ea e e cnUca am
applying afictitious inertia force fj directed opposite to the direction of motion for considering coeffi~ient. Th~ parameters I.t\, and (; play important roles in determining the response of SDOF
the instantaneous equilibrium of forces acting on the mass. This simplified interpretation of
systems l' . h f
Newton's second law of motion (valid only when the mass of the system is time invariant) is I~ 'order to detelmine the solution of Equation (7 .5), l~l liS assume a ,so unon m.t e ?r~
popularly known as the d' AI.mbert's principle. xU) = xi'. Substituting this assumed solution in the govemmg equation YIelds a charactcnsttc
(7.2) equation
(7.6)
Thus the inertia forcefj = mx
(mass x acceleration), spring force Is kx (spring constant
X spring deformation) and damping force ID :: eX (coefficient of viscous damping X relative
velocity between the two ends of the dashpot).' By substituting these relations in Equation (7.2), which has the roots at
we get. (7,7)
mx + eX + kx = F(t) (7.3)
The magnitude of the damping factor (0 can be used to distinguish ~hree cas~s: under
Equation (7.3) is a linear second order differential equation with constant parameters, The da d (0 < r < 1) criticallv damped (~= 1), and overdamped (~> l). F,gure 7.2 Illustrates
general solution for this equation is given by,
the;;':onse f~r thc:e three c~ses. For the underdamped case the rnotio~ is ~sclllatory m ~tu!t"!
(7.4) with ~Pdeca ing amplitude. For the overdarnped ca:e there ,is n.o oscdlatJon, an? amph:ude
decays slowiy. For the critically damped system there 1S no oscllIatlOn, and the amphtude de...ays
where, Xh(t) is the solution of the corresponding homogeneous equation obtained by making the
right hand side zero; and xp(t) is the particular solution which depends on the specific form of more rapidly than in either the underdamped or overdarnped cases.
applied force F(t). The homogeneous solution of this second order differential equation contains 1.2 ~~~ ... Undcrdampe<l  
two arbitrary constants which can be evaluated by using the initial conditions x(0) and (0), x }.O Critically damped 
i.c.~ the displacement and velocity of the mass at the onset of vibration. _.2'{~d~f~d.   ::_'
~S .
J
7.2 FREE VIBRATION OF' VISCOUSDAMPED
SDOF SYSTEMS
~~ ~~t..
02
1
7i
For a particuJar case when there is no external force acting on the system. it is still possible Q
00
..()2
f
to make the mass vibrate by giving some arbitrary initial conditions. The ensuing motion of
IThb. is only a convenient maihcmatiul model for the energy dissipation mechani~m in II vibrating system and i~
actually an e'Xpro:>~~.~i('ln of the viscous drag in a laminat Dow. In real strllctural systems, energy dissipation actually
0.6L
0.4,' .
..{l.S
0.0
'
0.5 ].0 1.5
~~ )'I\~ce \!\ OlHerenl ways and the mo~t important of them is the energy loss due to inter granular friction
Tunc (s)
between the p.3rticles of a vibrating system. However. incorporating a dry friction damping behaviour would
result in a nonlinear equation of motion. Since the magnilude of damping force iJ; generally very small in FIGURE 7.2 Response (lof single degree of freedom sysf"ms.
comparison to the other force$ acting on the system, a viscous damping model (leading to a linear equationD
serves well to get a iiluonal estimate of system response for engineering design.
( Chllpter "! Dynamicl'l ,of Single Degree of Preedom Systems) .'p_
7.2.1 Underdamped Case (' < 1) where, (j}'" = (t)nJS2 1 and the coefficients Al andA 2are determined from the initial conditions
For S< 1, it is convenient to write Equation (7,7) in the form Xo and xo'
The trend is similar to that for the critically damped systems except that the system
returns to the zeroposition more slowly.
A1.2 (OJ" i()).~12  (
= '14. i{f). (1.8) Practical ntility of free vibration analysis
where ffidis known as the damped circular naturaljrequency. Using E\ller's formula, the general Although the preceding discussion has been fairly mathematical in nature. the basic concepts
solution, x(J). can be written in the form, presented there in have immense practical utility. A point in the ease is the experimental
determination of dynamic characteristics (i.e" natural frequency and damping) of real
(7.9) structures. The dynamic parameters, so estimated~ can serve as a valuable check for validating
The coefficients AI and A., are determined from the initial conditions x(t = 0) = Xo and i (t = 0) the mathematical model of the structure and verify some of the modelling assumptions. Figure
r r: and the ~ojution is then glven by 7.3 shows a comparison between the free vibration response recorded during an actual test and
that for a mathematical model using viscous damping model for the same set of parameters, Note
the Similarity in the general trend of response even though the two curves differ in detail. The
xU) (
'{{I1~;l'  ,<; "(io + ~ m"x(l \.
, "'0 cO'(Ud l 'i ,"'W ~)smmdtJ
:
(7.10) response from the viscously damped model is a simplified representation of actual behaviour
d which is a combined effect of several different energy loss mechanisms simultaneously at work.
Although the value of , has an effect on the frequency rod. the most pronounced effect of 1.0r~~~"'" 1.0
damping h on If.Jc rate at which tht: motion dies out, that is. on the e(wnt tenn. 0.8 j=2.58 Hz 0.8 j=2.S8 Hz
~ 0.6 '=2.85%
Q
~I
7.2.2 Critic.allydamped Case (' = 1) "S. 0.4 15. 0.4
0,2 E 0.2
'*
Q
=
When' I, the Equation (7.7) gives only one solution as A = ()). The response for the case 0.0
."
Q 0.0
of repeated rOots is then given by " 0.2 ] 0.2
~ ~
E 0.4' E 0.4
~
xU) = ew'(A, + A,I) (7.11) i. 0.6 :i 0.6
0.8 0.8
where, Al and A2 are determined from the initial conditions.to and io' The solution in terms 1.0 "'.~~~~ .. 1.0 , 1
of the initial conditions can be derived as, 0.0 2.0 4.0 6.0 8.0 to.O 0 2 4 6 8 to
Time (5) Time (s)
x(1) = e~' [x,,(1 + ()).tl + i,t] (7.12) (aJ (b)
FIGURE 7.3 Free vibration response (a) from field test data, and (b} theo~tical curve for
The free~vibration response of a criticaHy damped sys.tem does nO[ involve oscillations about
visoously damped system.
the zero~displacem~nt position, instead the system returns to zerodisplacement position
as,ymplotlcally by vIrtue of the decayed exponential term of Equation (7.12). However, there The determination of (damped) natural frequency foHows from its definition and is
:,:iI, be prec~selY one crossing of the zero~dispJacement position jf the initial displacement and estimated by determining the number of cycles completed in the vibration record per Untt time.
mllJ(l:1 ve]ocllY were of opposite sign!:,
The determination of damping from free vibration records, unfortunately, is not so stra;ght
forward. However, the procedure does simplify a lot, if we decide Lo ascribe energy loss to only
7.2.3 aucrdamped Case ((; > j) one damping mechanism. For example, assuming viscous damping behaviour the equivalent
f:';,r ,,,> ; thr Equl:Ition C!.7} give:; twu distincl, negative real roolS, The solution j~ then given viscous damping parameler may be el'llmated by using the method of logarithmic decrement Let
b).
us consider any two positive peaks An and x/H! which occur at times n( 2") and (n + 1) 21t
(J) d (fJd
re;.;pectively. The ratio of these two amplitudes is given by using Equation (7.10)
(7.13)
(7.14)
( Chapkr '1 Dynamics of Single Degree of fuedom Systems) M@_
equation will be harmonic of the same frequency as that of the excitation albeit with a phase
By taking tbe natural logarithm of botb sides and substituting for (J)", we get,
lag. Thus the particular solution (or, steadystate response) is given by,
cycles (say, m) apart. It can be shown that in this case the equation for logarithmic decrement (J)C (7.19)
changes to tan';= 1
k (J) m
(7,20)
which is c?Jllmonly used in practice,
Real
(7.21)
7.3 FORCED VIBRATIONS OF SDOF SYSTEMS FIGURE 7.4 Force polygon.
The dynamical systems may be set into motion by several types of excltations. These forcing where, X,,; Folk is the static deflection if the force Fo was applied statically, 1); roIm,. is known
functions may either be harmonic or nonharmonic, periodic or aperiodic, etc. Further. the as the tuning ratio and S is the damping ratio defined earlier. The steadystate magmficatlOn
response may also differ on account of the duration of exposure to the applied e.xcitation. The factor D, and the phase angle'; are plotted in Figure 7.5. From Equations (7.17) to (7.21) and
nature of the response of a SDOP system to harmonic excitation is significantly different from Figure 7.5 the following significant features of steadystate response can be observed:
that for a finite duration excitation, We shall begin with the response of SnOF systems excited (i) the motion described by Equation (7.17) is harmonic and is of the same frequency as
by harmonic excitations, which has great practical significance since any periodic function can
the excitation.
be decomposed into a sum of harmonic functions by using Fourier series2 The results of this
section have an important bearing on the design of vibration recording instruments, industrial
vibration isolators. and shock absorbers.
Let us assume the initiaJ conditions given as XfI and xo. Moreover. the particular solution of thjs
0.94 '_L_.l~~..... ~~~,~
(I 0.2 0.4 0.6 0,8
lCnde,' a fairly gene;"J sel of condilinn~. :my periodic ftmction fU) can he expre",,~rl :1.'>: Frequency ratio (11) Frequency ratio (11)
,=0.00 l;= 0,50 
~, 0.00 s=0.60 0.65
/(1) =. (10 + LlUj c().~ ("=0.10 s= 0,75  " 0.70 . S=1.00 .
j""l
FlGURE 7.5 Magnificatiun and phase of response of SDOF system to different adtation
where T is the period 'Jf tV), N)( more details consult any lexl on Applied Mathematic);, suen as, E, Kreyszig,
Advanced Entinl:'f'rinx Mothematir~, Wiley lnlerscience,
frequencies.
(ij)
( Earthquake ~e$utant Design oj StM.lctures
The ampli~ud~ of the steadystate response is a function of the amplitude and frequency
of the eXcitatIOn as well as that of the natural frequency and damping factor of the
)
( Chapf.~r If Dynamics oj Single Degnre oj medom Syste"m8)
The above equation resembles the equation of motion for forced vibration for 0 :$ t T and
u
system. The sleadystate magnificatjon factor can be considerably greater than unity or describes the free vibrations of SDOF oscillator for I > T. Accordingly, tbe response of the
Jess than unity. SDOF system can be considered in two phases:
(,',,',') The steady,slate response xp(l) X sin (WI  ) and the excitation F(t) = F sin!JJt 0) response during 0 $ t $ T, the forcedvibration era, which comprises a homogeneous
'h
are not In p ase, th
at 'IS, they do not attain their maximum values at the Same0 instant. solution and a particular solution:
The response lags the excitation by a phase angle ' This corresponds to a time lag of
lw, X(I) ~ xh(l) + xp(t); 0 ,. I ,; T
(iv) At resonance, l) = 1. the amplitude is limited only by Ihe damping force, and CD,)".,
== 112<:'. Also, at resonance the response lags ex.citation by 901>. ::; e{Wnf [Al cos CtJdt + A2 sin {Od1] + ~o
~e ~otal response .of the SD?F system can be given by superimposing the homogeneous 'w 1 [( Fo) Xo + sw.(xo
S~hutl~n (JJ] the partIcular soJutlOn (xU) :::: Xh(t) + xp(t)). The homogeneous solution is the free =e'" ft lXo I;: COS{Odt+~ .... Wd
VIbratIon ~o]utjon ~s deri:ved earlier. The unknown parameters in the homogeneous solution can
be determmed by Imposmg the initial conditions on the total response solution of the SDOF where Xo and Xo are the initial displacement and initial velocity of the oscillator mass at t = 0,
system. S,jn~e Lhe homogeneous part of the solution gradually decays with timein a damped and
l>ystemJt IS referred to as starting transient. (li) free vibration response subsequent to the removal of the applied external force:
7,3.2 Excitation by Base Motion x(11) c,,'w.,,n c4, cos aMI 1) + A, sin {J)d(1 1)J; I > T
leI us now co~sld.er the ~ituation when the extemal force acting on the mass is F(f) = O. Instead
the support p~mt IS ~OVl~g as xii) ,= Xg sin (i){, Let xm{t) denote the absolute displacement of
the mas~ reJatlve to as ongmal posItion of rest and x(t) be the motion of the mass relative to
where, XT and XT denote the displacement and velocity at the end of forced vibration era, at
th~ mOVing base, t'.e. x,,lt) = x(t) ... x;.;(l). The equation of motion for this system can now be
wntten as, . I T
The eff""t of tbe duration of application of excitation pulse on the response of a SDOF
m i + c( X m X It) + k(xm  xA') 0 (7,22) oscillator lS shown in Figure 7.7. wherein the dynamic response has been normalized by the
Thil> equation ..:an be rewritten as,
static response Folk, Four pulses of duration T = [O.2ST" O,5T 2T., 4T.l (where T. = natural
period of the oscillatorassumed to be 0.1 5 for numerical study) have been considered. It is
worthwhile studying these plots in some detail, The following points emerge (with respect to
mx + eX + kx mx, (7,23)
the finite duration rectangular pulse):
This equation is now completely analogous to the standard equation of the motion of a SDOF
system and can be solved for the relative displacement response l;ly the standard procedures. (i) The response of oscillator to a pulse of very shon duration (with respect to the natural
period) closely resembles the free vibration response of oscillator due to nonzero
initial velocity.
7.3.3 Response of SOOF Systems to a Finire Duration
(li) The time of occurrence of the maximum oscillator responsefor the pulse width
Excitation
T'$ O.STtl' the maximum response occurs in the free~vibration phase after the excitation
Let us now cOTIsJder the rel>ponse of a SDOF has ceased to act on the oscillator, It is. therefore, prudent while studying the dynamic
!'lystem subjected to a rectangular pU]i>e of 1'(1)
response of a structunU system to finite duration excitations to compute the response
finite duration. T, as !>hown in Figure 7,6. The beyond the duration of excitation for a few cycles. This is particularly important for
governjng equation of motion for thlS sy<'tem lightly damped systems,
j<: given hy , (iii) When the pulse duration is relatively large (T> O.5TJ, the response during the forced
vibration era oscillates about the static response (unit normalized response), and the
1(5) free vibration response oscillates aboul the zerolevel. This shift of mean position
(,).24 ) FlGURE 7.6 SDOF system excited by occurs instantaneously after the excitation ceases to ex.ist.
finite duration rel'tanguiar pulse.
These observations can be used to derive analytical procedure~ for computing dynamic response
of osclllators for arbitrary loadings as discus~ed in the foJJowing.
.U. (Earthquake Resjstant Design 01 Stf'UctuJ"es
! ~:~'.1~~mI
"g tlO ~_,   __ _
I ~
:a
o. 5
0, 0 1\ /\ f\ VOvA
, A general method for eValuating the response of SDOF systems to an arbitrary form of the
excitation can be derived on the basis of unit impulse response function developed earlier.
The method is based on the principle of superposition and hence is strictly valid for linear
~ 05 ,MM' I"
.
'll
.~ 0. 5
E1. 01 ...
V
V
t_
systems only. Let us consider a SDOF system subjected to some arbitrary forcing function F{t).
This arbitrary excitation can be considered to be a
2~:O;;;'.,;::~f:;:'::
0.2 OA 0.6 O.S 1.0
~ I
:r
2~.0
 .
)
{ 0.5
~r4Of
1fEQ
, ___ I
denoted by dx(t) and is given by.
(7.26)
Time
A /\
~ 0.0 I ~!I\7""'"
rtJ,\J IV  j ... _ j 
. . .1. , . !i
~i
~<l.5CI
;;
I
I
E'LO I :++
..
++.
,I
.......ii ~ 1.5 :  '  prior to time 1. Therefore.
0.2 2.0.';;',;;,;,}:}:+:::
0.4 0.6 0.8 1.0 0.0 0.2 OA 0.6 O.g 1.0 xlt) = [' F(T)et;"nU,,') sin ill (t  T)dr (7.27)
Time (5) Time (s) mCO
d
Jo 'd
FIGURE 7.7 Effect of finite duration of excitation on respo~ of SDDF system (Til = 0.1 s, and This integral is known as Duhamel integraJ and can be used to compute the response of SDOF
~ ~ 0.05).
system to any arbitrary excitation. This equation can also be written as
..~
""
.~
0.0
~.5
r
. _.. j _.+'~I~C+_
_ F,

[1 + (ZS'1l2]liZ
0 [01)2)' + (2(I)/J'12
(7.30)
i
1.0 .; where 11 and ,respectively denote the tuning ratio and the damping ratio of the mounting
Z
1.5 system,
Let us define transmissibility ratio (TR) as the ratio of the amplitude of the force
2.0
00 0.2 04 0.6
transmitted to the support to rhe amplitude of the applied hannonic force,
0.& 1.0
TIme (s) R~
TR=
FIGURE 7.1} Effed of step size on numerical evaluation of Dubamel i.ntegral. F"
of Jag is proportional to the choice of AI. It can, therefore, he concluded that 61 :> 0.1 T, for
reasonable accuracy in the computed response. (7.31)
The formulaljon for numerical evaluation of Duhamel integral is discussed in a later
chapter on the numerieal evaluation of dynamic response. whict, is shown in Figure 7.11 for a tllllge of damping and tuning parameters. It may he noted
from this figure that the TR curves for all damping ratios inte"""'t at I) = ..fi for TR = 1.0.
Further~ following inferences can be drawn from these plots:
7.4 VIBRATION ISOLATION
(i) For T] < ..fi. the amplitude of the transmitted foreds always greater than the amplitude
Several industrial units have machines with reciprocating of imposed harmonic force and adding more damping to the mounting system
parts installed at various places in the structure. These contributes to the reduction in amplitude of the transmitted force.
machine installations should be carefully pJanned and proper
(ii) Fon..fi. the amplitude of the transmitted force is always smaller than the amplitude
mountings should be designed lest these machine vibrations
of imposed harmonic foree. Further. adding more damping to the isolator system is not
shoutd transfer to other parts of the structure and interfere
beneficial for vibration isolation.
with the daily operations. Let us consider the schematic
representation of a SDOF system shown in Figure 7.10, R(t) ~ kt(I)  cx{l)
where the mass now represents the vibrating machine" the ,=0.02  
force F(t) := Po sln mt represents the harmonie force generated FIGURE 7.10 TraIL,mission of ,'0.05 ....
vibrations fTom reciprocating ,,0.tO ..........
during its operation. and spring and dashpol denote the
machines. (= 0.20_
properties of the mouming system to be designed such that
'=0.40   
the force fransmitted to the base (reaction) is kept under some specified limit. {,~ 0,80 
The concepts developed in the study of response of SooF system to harmonic excitations
are uyed for designing the vibration isolators. Since the machines typically operate for long
durations ar a certain frequency, the starting transients can be neglected. The disp~acement, xU),
ef machine c::m be given by Equ<ltions 7.}7 to 7_21. The total force transmitted to the support
i<; f;iven hy (assuming that the sUpporting system if; rigid enough not to deform significanrJy due
to these transmineJ forces):
 ....~~~~
2 3 4 '5
Tuning ratiQ (11)
R(t) = kx(t) + cx(t)
FIGURE 7.n Vibration transmissibility ratio.
~ kX 'in(J)[  I/>l + cruX cos(llJI 1/ (7.29)
It might appear that one should always aim for a flexible isolator/mounting pads (11 > J'i)
with very small damping so that the amplitude of the force transmitted to the support can have
minimum isolation problem~ However, such an isolator would Jead to excessive displacement Chapter 8
of the ~chine frame itself. Therefore it is often preferred to have stiff isolators/mounting pads
j
(11 ~ ~2) ,so that the machine frame itself does not vibrate so as to hinder it." operation. For
deslgnmg lsolator to operate in this range~ adding more damping to the system improves the
performance of vibration isolators.
Theory of Seismic Pickups
SUMMARY
The behaviour of the most elementary form of a dynamical system is described. The basic
concepts of vibration analysis are presented with reference to this elementary system. Different
aspects of the vihration prohlem are introduced such as, free vibration, forced vibration
harmonic and transient excitation, vibration isolation, etc. This background paves the way r; 8.1 INTRODUCTION
assimilation of concepts from dynamics of more complex systems. 
Often, 'Nmall insignificant vibrations can excite the resonant frequencies of some structural parts
and be amplified into major vibration/noise sources. Vibrations may also have adverse effects
REFERENCES on human beings. The primary effects are task~performance interference. motion sickness.
breathing and speech disturbanee, and a handtool disease known as whitefinger. in which
[1] Clough, R.W, and Penzien, J" Dynamics of Structures, 2nd ed" McGrawHill. nerves in the fingers are permanemJy damaged, reSUlting in the Joss of touch sensitiVIty. It is
New York, 1993, always desirable to minimise the hannful effects of vibration through a suitable engineering
[2] Craig, R.R., Jr" Structural Dynamics, John Wiley & Sons. New York, 1981. design of products and systems. A fundamental requirement in all such vibration related
[3J Humar, J.L, Dynamics of Structures, Prentice Hall, Ine" 1990, problems is the aVailability of the charaeteristics of the expected vibrations. These vibration
[4] Thomson, WT" Theory of Vibration, 3"d ed" CBS Publishers, New Delhi, 1988, characteristics are derived from the analysis of the records obtained from various vibration
measuring devices, known as seismic pickups or transducers.
4.0
is given as.
3.5
mE + e(i  x) + k(z  x) = 0 (S.l) 3.0
where, z is the absolute displacement of mass m, x is the displacement of housing~ c is the 2S
damping coefficient of dashpot, k is the stiffness of tile spring and a (") represents differentiation ~
>; 2.0
with respect to time. Defining the displacement of seismic mass relative to the base as y(t) ;;; 1.5
z(t}  xU). and rearranging the terms in the above equation.
1.0
(8.2) 05
0.0
where, , and ill" represent the critical damping ratio and natural frequency of the osciHator. 0 2 3 4
Transforming the lime domain equation of motion as given by Equation (8.2) to the frequency Frequency ratio (11)
domain, 1;" 0.00 1;,0.25  S" 050
1;" 0.70 .. , .. ,"1.00
FIGURE 8,2 Amplitude and pbase of the transfer function of transducer.
y (w) = X(w),~.:::.a._ (8.3)
and (8,5)
Velocity (v)
Thc variation of the~e fUllctiom with respect to the ratio of the frequency of the base " ',
Displacement (d)
mOfion io l11e narurR] f,eqnency of [he oscillator, also known as the tunjng ratio 1] r= ~ \
",
ill'

fl\r v;l;'iolJ~ "a:ucs of lhe osciHulOr damping as shown in Figure 8.2. The parameters of the Frequent:y
osc!lator a;'c ::cjeC'!ed depending On the quantity to be recorded and the uRable range of
FIGURE 8..3 'fYpicaI vibration ~ctra.
frequencic!;.
'""""".'~_____ """___ ~"_"".. oC"h~,p:t::"::;8'T;;;h~e:;:.::ry;:;o;;tS;,;:e;;i;sm;;;
....~ic;p;Pi~kv.ps)  _
eM ( Earthquake Re.tii$tani Design nl Sif"tl.ctu1"f!$ ) \.. ~~~ .. .
. motive force e.mJ, is proportional to the rate of intersection of the ~agnetic
spectrum. in the frequencyrange of interest, in order to best utilise the dynamic range the mdured el<Ctro I . propcrtional to the velocity of Lbe vibrating body. FIgure 8.4
(the difference between the smallest and the largest values that can be measured) by the Ii ld the generated vo tage is . I' f a
e , . f typical velocity pickUp. The seismic mass conSIst'> mam) ~
~I'
I;J x,~x.
I;~O.OO
0.70 S'"'O,7S_
FIGURE 8.5 Amplitude spectrnm of the transfer function of an accelerometer.
Thus the natura} frequency of an accelerometer should be at least 34 times the hlghest
frequency of the vibration to be recorded and all the smaller frequencies will be accurately
K 53 M 1 '11 VanabJe capacltance
DvolWm
measured by the accelerometer. Further, there may be phase shift between the ground motion
and the relative displacements measured by the transducer. If the phase shift is same for waves
of all frequencies, then the reSulting signal would simply be sbifted a little in time, which would
x,
rEt';~,A'l
be of no consequence for structural response calculations, If the phase shift is different for
different frequencies, however. component waves will add up to give a distorted signa) which
wiH be quite different from the input signal However. if the phase shift can be made a linear
I IG i ~[d
L
\ l E"
11
k t ~~
outpu, vottag'
<>
function ofrhe frequency, then the resulting output signal will have the same shape as the input
signal. with;) small. constant shift of phiiSC. Fortunately, it so happens thal a transducer element
Force
generator
G(N/Amp)
L I
__
aT) ~
Ro
1_0
having a dumping of ahout 70% of critical possesses a phaseshift~frequency curve that is a good _~_
approximation to a straight line. Thus the same value of damping that gives an optimum FIGURE 8.6 FOl1:e Balante :"u.~('''elerometer block diagram. Xc is the absolute grou,nd
amplitude re.:;ponse curve is also the be:<>l value fmm the stand point of phase shift. displacement, XQ denotes absolute displacement of the mas;;: M and Xy represents the relative
displacement of the mass.
m (Eorthqua" Resistant Design of Structures
SUMMARY
The alms and ohjective~ of vibration recording and monitoring are discussed. This is followed
by a detailed description of the principle of operation of vibration pickups. The important . h d 'c response of SDOF systems(iJ by
Two disti~Cl approache~ ~lst to e~aluatea t :x=~~n of the derivatives in the differential
criteria for choosing an appropriate type of vibration pickup depending on the application are
interpolation of the eXCitatiOn. or (n) ~Y b pp d D hamel integral and involves interpolation
discussed.
equations of motion. The nrsht apPl~~:h lSI aansed ~:re~ore is strictly applicable only for linear
f th . !egrand of the Du ame m~gra, ' .
~yste~~~ The second approach is appl1cable for both linear and nonlInear systems.
REFERENCES
[1] Crede, C.M. and Piersol, A.G., Harris' Shock and Vibration Handbook, 5th ed .. NUMERICAL SOLUTION BASED ON
McGraw,Hill Professional. 2001. 9.1
[2J De Silva, C.w., Vibration: Fundamenlals and Practice, CRC Press, 1999. INTERPOLATION OF EXCITATION
[3] Rao, S.S., Mechanical Vibrations, 4th ed., Prentice Hall, Inc., 2003. .' . . 'fied as a set of discrete valuesf, = f(t i ) for i = 0 to N spaced
[4] Thomson, W.T., Theory of Vibration with Applications, 3rd ed" CBS Publishers, If the eXCitatiOn funcuonJ{t) IS S~l. . ti f the excitation function within an
New Delhi, 1988. at uniform interval of A L AssumIng a ~lOear vana on 0.. .
interval. the variation of I('!) within an mterval may be gIVen as,
r f" Vi +t;WnVi ]
e~Q),,t lV;COSWdt+ smaJdl
. rod
+ __1_
mm d
J'' ' f(~)e(W"'H) sin(Od(t  ~)d~
I,
dVi + 1
and Vi+l =: dt
Substituting forf(r) and evaluating the integral, the recurrence relations for di!>placement and
137
) ~_ _ _ _ _ _ _ _ _,C"h::.r<.:':::er:....:9,,~~merieal Et>aluation 0/ Dynamic Response) _ . _
velocity may be obtained as.
of these deviations in amplitude and phase properties of the computed response from those of
the exact solution depends on the ratiodtIT. Generally, the choice of dt:5 O.lT. has been found
Vi+1 = At + Bf;"'l + Cv; + DVf
to gi ve satisfactory results for all practical pwposes.
and Vi+l =Ah+ B'li+l + C'Vj+ D'vj'
AJtematively, the recurrence relations may be arranged in a matrix form as,
9.2 NUMERICAL SOLUTION BASED ON
APPROXIMATION OF DERIVATIVES
If the variation of acce1eration response within a time step is assumed to be approximated by
a simple functional form. then it is possible to derive the corresponding variations in the velocity
and displacement response which .is consistent with the assumed variation of the acceleration
response. For example, if the acceleration within a time step can be assumed to be constant
given by the average value of acceleration at the beginning and end of the time step. Thus the
~cceJeration during the interval Ii to li+1 is given by,
v( 'C) = _V,i+,=;;:=i .c
1
2
where, r = (t  ri)/AI is the dimensionless time and varies from 0 to 1 during the interval I; to
t",. The associated velocity and displacement during this interval are given by,
~
vIr) = Vi + Jo ii(6)d6
,
v('!) = Vi + J v(lJ)dl!
o
, 1"2 Vi + i;l+l
= Vi +"l)Ii+T 2
Considering the dynamic equilibrium at the end of the interval; i.e., at ti+lth instant,
For re.asonable a(;~'urac:y in the computed response, it is necessary that the :;ampJine:
!IIlcrvaJ 111 < 0 1T T b' h "I" '"
<" ;. l' ,.: fI
" ClOg r e. naturat penod of the SDOF os.elllator. Lsc of too large
. ",
"atIl~m.g lmerval1ead~ to the Joss 10 temporal resolution and then it is possible that the correct
maxmmm re,;pOD:e mIght not be captured. Further. due to the finitc size of pulse duration or, dt dtl
J..:v. ( c+~k
f 2 4
J"'}
v
I
used for .sl1rnmatWl) approximation of the COJlVOlution integralthe numerically computed
rcsp~nse IS only an approximation to the exact (analyticaUy derived) response. 1n particular the
~?lphtude of the computed response may differ from that derived analytically. Also the fi'nite wherc, m= m + ~/ c + All k has been I'ubstituted for the sake of brevity, This equation can be
SIZe of the pulse has th~ effect of introducing a phase shift in the compuled response. The extent
'
so1ved fo~ the acce,lerafion at the end of the interval, and wbich, in tum can be used to compute
the velocIty and dIsplacement at the end of the time step as; scheme may be written as,
where A is known as the amplification matrix~ and L is known as the load operator matrix.
1), _ ill ",,"J , 81 J ..., I [~t 1
,+;  T m j 1+ ,m h+1  6tm kVj + 1 _61 1 (2c+ tJk)J v
~  2 I
9.3 STABILITY AND ACCURACY CONSIDERATIONS
In these relations the accelerat'o .. _ if.r . . .
. . _. I n v  CVi  kVi) has been ellmtnated by making use
m Vi 
of t~e dynarruc eqUIlIbrium equation at the instant t" Thus the solution marches in time from As with any numerical procedure. the time marching schemes fQr numerical integration of
one mstant to the next Again these recurrence relations can be arranged in matrix fonn as, equation of motion provide an approximation to the actual soJutton of the equations. The quality
of the computed solution depends on the choice of timemarching algorithm and aiso on the I.ime
For a single degree of freedom system, the matrix A is of si:re 2 x 2 and hence will have 2
D = 1It[1~ ~t mJ(2c+iltk)] eigenvalues and corresponding eigenvectors. Let us assume that these eigenpairs are given as Alij,
{~(I)} and AA2, ('1'(2), Thus, invoking the linear independence property of eigenvectors, the
lnitial state vector can be written as,
A' ::; 6t ;;;.J
2
jB';:,j,;;nJ (9,2)
lD' = [1~ml(2C+l11k)] Ji
~~"~
whe.re.~ in (93)
Where, AA is. a diagonal matrix of eigenvalues AAj of matrix A. It is, therefore. clear that the
' 5ilHlila: recurrence relationR can be derived in the case of other forms of assumed variation
() t acee eratlOn 'th' . contributions of original coefficients c are reinforced by the powers of the eigenvalues, A'A" In
_ , WI In a tlmestep leading to various timemarching schemes for numerical
tntegra.tlOn of equatlOn of m t; S h ' such a situation, bounded response can only be expected if and only if the modulus of [he
o...on. ,c ematlcally. the recurrence relation for any timemarching maximum eigenvalue does not exceed 1, Mathematically, it translates into stating that for
. , . (Earthquake Re~tant Design of StractUT'e8
([) A unit injtial displacement with zero initial velocity, i.e., to'o = 1.0. Vo = 0.0
(F) ,7.ero initial displacement with unit initial velocity, i.e., Vo = 0.0, Vo = to.
r Chapter 10 Re8ponse Spectra) Mli4
(10.1)
Assuming ground acceleration as nonzero in t E. (0, tal, the Equation (10.1) can be written as.
Chapter 10
X( w) = f: x(t) cos(wt)dr  i foux(r) sin (wt)dt (10.2)
Response Spectra Fourier amplitude and phase spectra of strong motion are defined using Equation (10.2) as
rr
x(t) cos(wr) dr'
IJI(OJ) =  tan1,_O'_.............................. J. (lOA)
10.1 INTRODUCTION If: f(r) sin (wr) dt
Earthqualce causes ground to vibrate and structures supported on ground are subjected to Ibis
Although phase spectrum (Equation (10.4) is considered to be reiatively of less
motion. Thus the dynamic loading on the structure during an earthquake is not extemalloading
importance than amplitude spectrum (Equation (10.3, both amplitude and phase spectra are
but due to motion of supports" In general, the ground motions have three translational and three
required for unique defInition of ground accelerauon. Foun~r ampht~de spectra of velo~lty and
rotational component~, Not much infonnation is available regarding the properties of rotational
dispiacement can be obtained by dividing acceleration Founer amphtude spectrum ordmate by
components of ground motion due to difficulties in recording those, On the other hand, the
frequency and square of frequency value respectively. . .
characteristics of translational components: are relatively better known as they are routinely Figures 10.110.3 show nonnalized Fourier amplitude spectra for acceleranon, velocity and
recorded and processed during strong earthquakes by an instrument known as accelerograph.
displacement of N15W component of accelerogram recorded at Uttarkashi during Uttarkashi
Moreover. it can be deduced from the analysis of vibration records obtained from different
earthquake ofOcwber 20, 199I.It is to be noted that the bandwidth of predominant frequency
elevations in a structure during an earthquake that the rotational components are quite small in
for displacement is narrowest and that of acceleration is the ,broadest. It also demoll~trates that
magnitude in comparison with translational components. Therefore. the effects of rotational
acceleration, velocity and displacement are controlled by different frequency band.
components of ground motion are usually neglected in seismic analysis of structures. Further,
the response of a slTucture is often obtained by subjecting structure to one component of ground E
~ 1.0
translation (acceleration). The total response of structural system is obtained by combining U Accn.
u 0.9
response due to individual component, as method of superposition is valid for linear elastic ~
0.8
systems. Since the natural frequencies of the structure are not known a priori. complications G
~
arise in the design of structure necessitating several iterations, In such a situation, a design . 0.7
<i 0.6
engineer requires a prescription of seismic loading that reflects frequency conlent, amplitude of ~
a 0.5
ground motion and effect of subsequent filtering by the structure. This infonnation is provided
c~ 0.4
by Response Spectra.
In this chapter frequency content of ground motion is discussed first by means of Fourier ...g 0.3
Spectrum and later the concept of response spectrum is introduced as a Lool for quick dynamic
analysis, ..
'0
.~
0
144
.,. Earthquake Re~istant Design of Struetures
(~~~ Chapter 10 Response Speetm) .,.4
I"c.
l.0
0.9 Vet
10.3.1 FonnulLLtion
O.R The equation of motion of a SDOF system subjected to support (ground) motion. as shown in
"
~ (n Figure lOA. may be written as
C. 0.6
E ~
~
0.5
mz .j. c(z  i).j. k\z  x) = 0 00.5)
.""
.0
."
0.4
0.3
0.2
Reference
frame
~
[j 0.1
y
0 0.0 m
'" 6 8 10
Frequency (Hz)
12 14 [~~~~~~~~~) I:
FIGURE 16.~ !'\onnaJized Fourier amplitude velocity tru f Nl S"W
tlttarlcashi.  spec mOM component at
I II
_ 09
l.0
. DiSP.j ,,,
,
,I" ~:~ I, hW}/,// ,0J//./ ,%///
f 0.0 fxj
t; 0,5 ~
";:
~
i5
""
] 0; f
'L0.4
0.3.
0'1'
t'
FIGURE 10.4 Single degret'! of freedom system excited by support motion
FIGURE 10.3 Normalized Fourier amplitude dispIactment spectrum of NlS"W component at Where (f}n is undamped natural frequency of the system and "is the damping ratio. The
Uttarkashi.
Equations (10.5) and (10.6) are ordinary second order differential equations with constant
coefficients. Either of these can be used to find response of the SDOF system. The Equation
(10.5) needs ground displacement x(t) and velocity i (I) as input suppon motion. While the
10.3 RESPONSE SPECTRA x
required support excitation is ground acceleration (1) for the Equation (10.6)_ The solution
of these Equations (l0.5) and (l0.6) requires initial condition of the reRpon,e and ground
Earthquake engineers prefer to report interae[ion between ground acceleration and structural motion at the start of ground vibration. Thus the computation of response spectrum is an initial
,;"emR through re'pon,e 'pectrum first proposed by Biot 11, 2,J and later popularised by value problem. The systems represented by Equations (10.5 and 10.6) are incrementally linear
~:lU~~e; I~I. Respllnse spet:trum is a :)et of ordinates that describes maximum response of a set systems. These systems violate zero~infzeroout property of linear SYSTem U5] if the systems are
'" Slug,c ucgree Freedom Sysiemg {SDOF) subjected tu a prescribed ground mot jon, Often a not initially at rest. The complexity of this solution was realised by Pecknold and Riddell I J6J,
res~onse, <;pcctrum j\ presenled as a plot of maximum re!>ponse of a set of SOOF systems Malhotra [7) and Mylonakis and Syngros [9J among others. This initialvalue problem is not
sub,ecteo " , , , (g"ourrd)
, ) " io a liUPPOT! '" mOll. on as. ord'mate an d correspondmg
. natural trequencjes
.' (or, well posed because support metlon and/or initial condition are not properly defined. TIle ill
pi~.Ji)d:i) of the SDOI~ ;.,y~!em a~ abscissa.
POSed problem creates complications in the solution if:
(i) the system and ground motion are not initially at rest, and
.II#W (Earthquake Resistant De8ign of StructUf"l!:'~
(ii) the prescribed ground motions (acceleration, velocity and displacement) are not Thus for lightly damped system (i.e. O>d = 0>,)
compatible.
Sd(, "',) '" 1 S,...(I;''''n) = _I_S,. (,m,,) (10.10)
The accelerograph generally starts recording motion after a fixed threshold of acceleration level (J) d (1)"
set up in the instrument is exceeded by the base motion. Thus. at the s.tart of recording, ground
displacement and velOCity are not zero. Moreover. reported processed histories of ground The relative displacement response spectra asymptotically approaches maximum ground
acceleration, velocity and displacement are often incompatible (i.e. reported histories of velocity di.splacement for highly flexible structure. Formally, the limiting value of Sdls, ill,,) is,
and displacement ean not be obtained by Integrating reported aece1erauon) due to the use ~f (10.11 )
extra correction procedure employed for VelOelly and displacement correction than that of
acceleration history correction of standard ground motion processing. Note that solution for well
Thi~ implies that the mass remains stationary for all practical purposes and only the ground
posed Ploblem will give same solution using either of the Equation (10.5) or (10.6). These
moves as the linear elastic SDOF system is composed of spring with negligible stiffnes.'L
complications are ignored by engineers by assuming that the ground and structure are initially
Differentiation of Equation (10,7) with respect to time t gives,
at rest and using only ground acCeleration 8..<; input.
; lI{l'
(i)d \
x(T)efm,(H) sino>,(t 
\I.
rldr}11
m&
(10.8)
From the Equation (lO.9) the undamped relative pseudo velocity response spectrum can be
obtained as,
~here> natuml period of vibration Tn = 2rrJro". For a prescribed aecelerogram the Equation
nO.7) is numerically integnited and the resulting maximum relative displacement value gives
?ne value of Sd for a specified set of ~" and t;. The numerical scheme for evaluating Duhamel's
S",,(O, )n) = If' o
X(!}Sinm"U!)drl
Ill<>);
(10.15)
mtegral, originally proposed by Nigam and Jennings [13], has been discussed in the previous
chapter. Typically, this integration is carried out at uniform frequency interval in a prescribed It is easy to see from the Equation (10.14) that as ~ + 0 the relative velocity spectrum S" >
range of frequencjes for different values of ,. It is worth noting that the quantity within the IxU)lm3J> and from the Equation (10.15) the relative psuedo response Spv 1 0 becau!'e the mass
curly brackets of Equation (10.8) has the unit of velocity. The absolute maximum of this does nol move. Hudson [4, 5] has shown that numerically S. (0, (LIn) and S,.,)O, ro,,) are almost
quantity is termed as pseudo relative velocity response spectrum (psv) Sp,(t;, 0>,) and is fonnally equa' except for very long period structure. However, variation is considerable in case of highly
glven a~, damped structure. Figure 10.5 shows relative velocity spectra ,.)11 and pseudo relative velocity
spectra Sfl~ for damping ratios , ; ; ;: 0.02 and 0.20 of N lS"W compo~ent of UHarkashi record.
"~,,.(I;, 0>,;) S".(I;, T,,) = if x(r}eL:""U~"
)
sinO>d(t !ldrl
11'1:'(
(10.9)
By re<rrnmging the terms in Equation (10.6), the absolute acceleration response of the lmear
elastic SDOF system can be obtained as,
Miii. \ Ear~ Reei$~!.1c.t..;D:...e'".:'ign="".fc..::;S;;;.t""'=t:::.::;""::'~_ _ _ _ _ _ _ _ _ _ _ =:J Ch4p~er 10 Re6pQnse Speclm) MUM
(10.19)
The maximum spring force developed in the oscillator is kSd (', w,,) ~ mS",,(" w,,), while mS.(',
ron) is the maximum of total elastic and damping forces. The maximum strain energy input is,
(10.20)
OJ 10
Period (s)
and the maximum strain energy per unit mass is,
FlGt:RE 10.5 Comparis.n of spectral and pseudospectral velocity of N1S"W component
"~corded at Uttarkashi fof' different damping ratios.
(10.21)
tV) x(t) + y(l)
The total energy of the system is,
"'~ y(1)  2'w"j(tj (1.16)
The absolute acceleration spectra is simIlarly defined as, Er(t) ;ly(t)]2 + ~1y(t)J2 (10.22)
SO<" w,,) '" Si', Tn} If(I)I_ (10.17) For an undamped linear elastic SDOF system, substitution of y(l) by the Equation (10.7) and
It may be seen that for damping ratio 'E (0.0, 0.20) YU) by the Equation (10.12), the Equation (10.22) reduces to,
S.<,. "'n) ~ "'nS",(', w")ES,,,(' w,,) (l0.18)
where. Spa(~, Wn) is calle.d absolute pseudo acceleration spectral response and the relation (10.23)
(10.18) reduces t? equalIty for ,~ O. Absolute pseudo acceleration spectra S",,(" w,,) $
Sat', "'n) This dIfference mIght be ImpOI1ant for rigid systems. Figure 10.6 shows absolute which a[ the end of accelerogram t ; ; ; td is iden
tical to Fourier amplitude spectrum lX(w)1 of the 0.14 .~r.~'~
FS .
j ::: f~I~ ~I
ground acceleration evaluated at frequency COrt. 0.12
The maximum of the Equation (10.23) is pseudo ';;' Sp,<' = 0.0)
.
$ 0.10
relative velocity spectrum Sp,(O, w,,). If the rela hl
~)
5
1 0.60,
~
In general, IX( w,,)1 $ Sp,(O, "'.). Figure 10.7 ~ 0.06
B {lAO
shows relative velocity response spectrum SpY for
an undamped system and Fourier spectrum lX(m~ 1 0.(;4
~ 0.20 ~
0.02
of N IS'W component recorded at Uttarkashi.
0.00 .....
n.oo ~~::,:~~:..."", ___ ~..J
0 5 to 15 20 25
0.01 0.1 10
10.3.3 Solntion: General Frequency (Hz)
I()Q
P~od {s)
Conditions FIGURE 10.7 Comparison of pseudospec.
lral velocity spectrum for 0% damping and
FIGURE 10.6 Comparison of spectral and pseudo~speetral ac(~leratiQft of N15W component The concepts of stiff and flexjble system are
1 ecorded
a! lTttarkashi for 5% damping. the Fourier spa."trum of N1S C W oompooent
necessary to obtain response spectrum When the
recorded 01 t:ttarkashi.
ground motion are incompatible Or when the
. . ( Earthquake Resista.nt Design of Stru.dures
systemlground is not initially at rest. A SDOF system excited by support excitation is considered  x(a)
to be stiff (flexible) if the natural frequency is larger (smaller) than the support excitation and yeO) = _.... . (10.28)
frequency, Response spectrum by defmition encompasses response of both stiff and flexible
SDOP systems as it prescribes maximum response for a set of SDOF systems, Moreover,
1.0+(::)
absolute dispJacement response for a very flexible structure is zero and relative displacement The Equation (10.6) is solved with the derived frequency dependent initial condition given by
response of a very stiff structure is zero. Thus, for a specified nonzero initial support motion Equation (l0.28). However, specification of parameter Cf. and representative ground frequency
both absolute and relative displacement cannot be simultaneously zero. This requires a (J) need to be defined. Using simulation teChnique, the parameter a= 21S found to be satIsfactory
reformulation of the problem. Let w, be a predominant (representative central) frequency of the fotr response spectrum computation, There are several definition of predominant frequency !Or
excitation and f(t:Un, lOt) be a function such that f( (J)fl' we) approaches unity as Orr ~ 0 and of ground motion records in tbe literarure [6, S, 17]. The predominant ground frequency may be
vanishes as l.Orr t """. One such generalised function is, defmed as the frequency at which relative velocity respo~se of 5% damped SDOF sy~tem IS
maximum in the entire range of frequencies of ground molion. Thls frequency IS msensitlve to
1.0
j(m", mJ (10.24) the initial condition of ground m~tion, Representative ground frequency can a180 be defined as,
_ IPGA
m,  VPGD
where. a> 0 and the function satisfies the specified limit conditions for both flexible and stiff
SDOF systems. Define a dii;placement u by mixing absolute and reJative displacement as, where, PGA is Peak Ground Acceleration and PGD is the Peak. Ground Displacement. Useof
Equation (10.29) is criticised on the ground that PGA and PGD occur at djfferent frequencIes
of ground motion and thus it does nOl represent a single frequency. Use of forme: defimtJ.?n
(10.25) of predominant ground frequency needs. additional comp~tation of 5% damped relatIve velocu:y
spectrum of ground motion with initially at rest condltlOn,
Addition of Equation (10.6) and Equation (10.5) multiplied by (fbc1m.)"lm results in, The SE component of 1940 EI Centro accelerogram [14] is used to compute 2% damped
response spectrum for initialJy at rest and at non zero condition. The accelerogram has ~ak
ground valUes as PGA ~ 0.3484 g, PGV = 0.334 mls and PGD 108.7 mm. The predomI~ant
(10.26)
ground period from 5% relative damped velOCIty spectrum IS 0.85 s. The EquatJon (l 0.29) gIves
period as L 1 s. These two values are considered to be comparable for all practIcal purposes.
The Equation (10.26) is solved with the generalised initial condition u(O) ';(0) = O. Note that The initial ground velocity and displacement are 0.04664 mls and 21.59 mm. Use of two
the solution requires acceleration, velocity and displacement of ground motion as excitation. definition of predominant ground period gives almost ident~c~l, result. ~~gure 1?8 shows .1%
Replacing absolute displacement z by y + x from Equation (10.25) the relative displacement is damped displacement spectrom. The solution using nonzero lTIlual condloons satIsfies phYSICal
obtained as,
60 .. 1 I I ,nAt rest ~itial1y n~M_1 ,
. (10.27) : ~ Nonzero initially  
lot~'I~F'r ".'r~l~~~rLLI_.
mixed displacement u. The required spectral displacement is the maximum of y at any time" In
most of the cases, a simpler approximate method of solution than that of the general method
presented above can be used satisfying limiting behaviour of SDOP system. In this method a
loro~~tc
frequency dependent initial condition is obtained from Equation (10.27) by using generalised
initial condition "(0) = u(O) = \) as, 10 ' / ,["'''''[1''1'+ i=""
 x(O) o . i ' . .,j~:::':;'~;":;;::;;'
y(O) = ' 0.00 5.00 10.00 15.00 W.oo 25.00 30.00 35.00 40.0045.0050.00
1.0 + (j), I Period (s)
mi' ) FIGURE 10.8 Displacement spectrum (damping 2%).
(
Chapter 1 (j Response Spectra) 'p.
constraint for both stiff and f1exibJe systems. The initially at rest solution satisfies physical
condition of stiff systems.
J~ .,I
~.
_J
... ,I
.,"
..!:
.'S
u
0
!i
. l
tI'
.~ .. ~~::~':i~'
1
;
'<
:.\,
1
",~
l
"
~
u 0.01 At rest ~~~Iy ..~
~ 1.00
Q.
en
!
Non~wo 1ll1tlally 
Smooth  ...
Peak ground value ........
i
~
11
'!' 0.50
O.OQI
0.0
~.~.
0.1 1.0
Period (s)
t
10.0
........."
J
_ ~~~~~~~~~~~~~~,
0.04 0.06 0.100.120.14.
Spectral displacement (m)
FIGURE 10.9 Pseudo ,eloclty spectrum (dalnping 2%).
FIGURE 10.11 Demand diagranl for vertical component (damping 5%).
[2] Bior, M.A" "A Mechanical Analyser for the Prediction of Eanhquake Stresses",
Bulletin of the Seismological Society of America, 31: 151171, 1941.
[3) Housner, G.w., "Calculating the Response of an Oscillator to Arbitrary Ground Motion".
Bulletin of the Seismological Society of America, 31: 143149, 1941.
Chapter 11
[4] Hudson. D,E.~ "Response Spectrum Techniques in Engineering Seismology". ]n
Proceedings of the First World Conference on Earthquake Engineering, 4: 112.
Eanhquake Engineering Research Institute, Los Angeles, California, 1956,
[51 Hudson, D.E" "Some Problems in Ibe Application of Spectrum Technique to Strong
Dynamics of MultiDegree
Motion Eanhquake Analysis", Bulletin of the Seismological Society of America, 52(2):
417430, 1962.
ofFreedom Systems
[6) Kramar, S.L., Georechnical Earthquake Engineering, Prentice HaU, Inc., New Jersey,
1996.
[7] Malhotra, P.K., "Response Spectrum of Incompatible Acceleration, Velocity and
Displacement Histories", Earthquake Engineering and Structural Dynamics, 30: 279
286, 2001.
[8) Miranda, E. and Bertero, V. V., "Evaluation of Strength Reduction Factors for Earth
quakeresistant Design", Earthquake Spectra, 10(2): 357379, 1994. 11.1 INTRODUCTION
r9J Mylonakis, G. and Syngros, C., "Discussion of Response Spectrum of Incompatible
Acceleration, Ve]ocity and Displacement Histories t " Eanhquake Engineering and A MultiDegreeofFreedom (MDOF) system, as the name suggests, is one that requires two or
Structural Dynamics, 3 J: 10251031, 2002. more independent coordinates to describe its motion. The coordinates normally used to describe
[10] Nau, l.M. and Hall, W,J., "Scaling Methods for Earthquake Response Spectra", the motion of a structural system, may be related to each other via some constraints, which could
Joumal of Structural Engineering Division, ASCE, 110(7): 15331548, 1984, either be simple kinematic relations between various coordinates. or they could arise from the
[ll] NEHRP, "Recommended Provisions for Seismic Regulation for New Buildings and consideration of equilibrium of forces. The number of generalised (independent) coordinates is
Olber Structures", Technical Report, Building Safety Council for Federal Emergency given by the difference between the total coordinates describing the motion of a system and the
Management Agency, Washington D.C., 1997. number of constraint relations. For example, consider
112] Newmarlc, N.M.. "A Method of Computation for Structural Dynamics", Journal of y
the case of a doubJe pendulum, which is constrained to
Engineering Mechanics Division, ASCE, 85: 6794, 1959. move in XY plane as shown in Figure 1 L L 1n Cartesian
(13] Nigam. N.C. and Jennings, P.C., "Calculation of Response Spectra From Strong coordinate system the positions of two masses ml and mt:
Motion Earthquake Records". Bulletin of the Seismological Society of America, 59(2): are described by twO pairs of Cartesian coordina.tes
909922. 1969. (x" Y:) and (x" Y2) respectively. These four coordinates
/14] NOAA, Earthquake Strong Motion: CDROM, National Geophysical Data Center,
Boulder Colorado, March 1996.
viz. Xb y,. X2. Y2. however. are related to each other
through two constraint relations:
[15J Oppenheim, A.V., Willisky, A.S .. and Young, I.T., Signals and Systems, Prentice Hall, x~
Inc .. New Jersey, 1983. x1 + Yl ; 11, (11.1)
F1Gt;RE II.I A double pendulum,
[l6] Pecknold, D.A. and Riddell, R., "Effect of Initial Base Motion on Response Spectra". and (x, x,)' + (y, y,)2; 1', (11.2)
Journal of Engineering Mechanics Division, ASCE, 104(2): 485491, 1978. Thus the number of degrees of freedom (or, generalised coordinates) of the structural system
[17] Rathje, E., Abraltamson, N., and Bray, J., "Simplified Frequency Content Estimates is 4 ~ 2 ;;; 2. The angles 6 j and 8.t can be taken as the two independent generalised coordjnates
of Earthquake Ground Motions", Journal of Geotechnical and Geoenvironmental to describe the motion of the masses. In certain systems it is possible to eliminate dependent
Engineering, ASCE, 124(2): 150,,159, 1998. variables by using constraint relations and derive a set of generalised coordinates. are said
to possess hoionomic constraints. On the other hand, there may exist some constraints, caHed
nonholonomic for which it is not possible 10 derive a set of independent coordinates.
Nonholonomic constraints are rarely encountered in prdlcrice, so it wiH be a$sumed in the
following that the equations of the dynamic equilibrium of the system are specified in the
unconstrained c00rdinarc system.
157
C~(l.r~quake Residant D~st"gn::..:QC.<.I,S,,t:.:",::,::t"""=.,,_ _ _ _ _ _ _ _ _ _ _ __
11.2 SYSTEM PROPERTY MATRICES
As mentioned earlier. every dynamical system comprises
(1) a mechanism for storing strain energy due to deformations,
(ii) some means of sloring kinetic energy of the system in motion, and
(iii) an energy dissipation mechanism.
In MDOF system, described by a set of N generalised coordinates (say, VN)' these v" v" ....
energy functionals depend on the motion of the system described by the generalised coordinates.
iJ'U
'2 L
. 1 N
Potential energy:::::; U(l'j, 1'2, VN, t);;: avavo l',V f
(J:o} I J The nature of damping forces is assumed to be of the viscous type primarily as an.appr~xim.ate
representation of the combined action of all energy dissipation mechanisms present m a. vlbrating
system, Since the extent of damping in structural syste~s is usually very sm~ll, preCIse nature
of the damping force is not very important for dynamlc response computatlons.
. ... . ..
RayleIgh dISSIpatIOn function = R(vl v, . ."
.
V N, t) = 
1 IN ~
iJ'R ..
v,v; ,Fosiu6lt
. 2 avav' j
11.3 DYNAMICS OF TWO DEGREE
~
t
1,)=1 ( )
'"2
OF FREEDOM SYSTEMS
where Rayleigh dissipation function represents the energy loss through velocity proportional
viscous damping force.
Let us consider the response of a hannonically excited 2DOF " '" . .
FlGURE 11.3 Hannonic ud
The element Cu of damping matrix (C) is given by the coefficient aiJ',R and represents (undamped) system as shown in Figure 11.3. The governing tali"" of 2DOF system.
. ~~ equation of motion for thIs system can be given as;
the damping force at i lh DOF corresponding to the unit velocity at i h DOF with the velocities
(11.4)
at all other DOFs remaining zero. The coefficient a~!!
'c, is the element kij of stiffness matrix
avpvj
(K) and represents restoring force at the ,.Jll DOF corresponding to the unit displacement at /J This system is the characteristic of an industrial building with a reciprocating machine installed
DOF with displacements at all other DOFs being constrained to zero. Similarly, the coefficient at one of the floors. Since such machines typically operate at a fixed speed. the force exerted
by these machines on the building floor will be hannon~c and the steadystate response of the
"'I ~iJ. is tbe element mtj of inertia matrix (M) and represents the inertia force at the ,1h DOF
system to this harmonic excitation wiH also be harmomc ~f the same frequ*ncy. ,
O"iOVj
corresponding to the unit acceleration at the lb DOF with accelerations at aU other DOFs being Thus assuming the hannonk response as v = [v" vz] = sin ro'IV,. V,l , Equation (11.4)
!;onstrained to zero. The governing differential equation of motion for an MDOF system can can be solved for the response amplitudes V, and Vz of the two masses as:
be derived from the same principles as used in the case of SDOF systems. As an illustration.
let us consider the 2DOF system as shown in Fjgure 11.2. By invoking the d' Aiembert's
principle. the equations of motion for free vibration of this system can be written as,
k 2 ](Vl)'
" V2,
sm(J)t = (FoSin<illi
0
k2 ar"m2 j
=~ Fo , (! 1.6)
V2(t)
Jim,' [w'  w' (w,.. + w,(l
' .. *
+ 11) + (w, w,l']
sin fiJt
l! may be noted from Equation, 11,6 that it is possible to force the amplitude of response of
the first mass (mJ) to vanish by a suitable choice of parameters (also refered to as tuning)
kz and mz (or, )1). This. concept can be exploited in designing vibratjon absorbers for industrial
structures and can be achieved by attaching an auxiliarylsecondary mass to the primary struCture
which is subjected to a hannonic excitation. This can be quite effective, when the operating
frequency of the reciprocating machine and a natural frequency of the supporting structure are
nearly equal, causing large amplitude vibrations of the supporting structure due to resonance,
Let us assume for simpHcity. that the second mass mz in Figure 11.3 is an auxiliary mass
attached to the primary structure, which is excited by a reciprocating machine installed on the
floor. The amplitude of displacement response (normalized with respect to the static defl~ction
Folk,l of the primary mass is shown in Figure 11.4(a~for a range of operating frequencies. The
unbounded amplitude for ldw7 = l.0, corresponds to the condition of resonance and it should
be avoided (w~ :=. JkJ/11lt represents the natural frequency of primary structure alone), Let liS
now consider the use of a tuned mass damper (or, vibration absorber) in altering the dynamic
response of primary structure~ Figures lJ.4(b'i) show the response characteristics of primary
structure y"ith an auxiliary/secondary structure for different parametric variations. The objective
is. to find a suitable set of parameters for secondary structure, so as to limit the normalized
respon~ of primary structure to be less than unity in the neighbourhood of alW;::;: L The effect
of addmg a secondary structure is to split one resonant peak into two resonant peaks.
corresponding to two natural frequencies of lbe combined 2DOF system. A light (J.t < I)
secondary system attached to the primary structure by means of a relatively flexible attachment,
does take away a significant part of the vibration energy from the primary structure at }/w ~
= 1. as can be seen in Figure 11.4(b). Howev~.r, such an arrangement is only effective (in
reducing. the vibration amplitude of primary mass) for a very small range of operating
frequencIes, by means of increasing the deformations in the secondary structure. Attaching a 5,0
heavy secondary mass with a flexibJe spring is not at all effective~ as can be infered from Figure
11.4(c), Figure 11.4(d) shows that the use of a light mass and relatively stiff attachment does FIGURE 11.4 Performance of vibration absorber/tuned mass damper.
?ot Jead [0 any significant change in the resonant frequency of the original structure and hence
lS nut an effective solution for the vibration problem. Figures 11,4{e) and (f) show the perfo.rm~
ance for secondary structures comprising heavy mass with stiff attachments_ By comparing these Thus a secondary structure with heavy absorber mass can be very cffectlve in controlling
two fj~re&, it may be noted that the two resonant peuks are wen separated and are sufficiently machine induced vibrations in a building/structure. However, the large size associated with a
away 110m the original location of resonant peak (for the primary structure alone) at aim; = heavy mas.s can sometimes impose a practical limitation on the usable range of operating
1, Further, the amplitude of displacement of primary structure vanishes at catw ~ J and is le......, frequencies. The reqUired stiffness (k::) of the secondary attachement can be calculated from
than the static deformatjon (Folk,) for a wide range of operating frequencies. This effective the maximum allowable displacement of the secondary system (at co CO;) VZmax
f'dnge of operating frequencies increases with increase in the mass rdttO (Ji;; m2/mj)' Folk"
. , . ( EorthquoJre Resistant Design of Structures
L.
11.4 FREE VIBRATION ANALYSIS OF MDOF (I LI2), which is a set of N simultaneous linear homogeneous equations in unknowns ~. The
problem of determining constant (w) for which the Equation (1l.12) has a non,trivial solution
SYSTEMS is known as the characteristic value or eigenvalue problem. The eigenvalue problem may be
By considering the fact that the damping levels are usually very small in structural systems, let rewritten. in matrix notation as)
us consider the respense of an undamped MDOF system. The effect of damping will be dealt (l Ll3)
with at a later stage. The equation of free vibration then reduces to,
A nontrivial solution for Equation (1 Ll3) is feasible only when the determinant of the
Mv .. K,,=O (11.7) coefficient matrix vanishes~ i.e.,
We look for a solution in the form Vi = q(l)~" ; = 1,2, ... , N, where the dependence on time IK  w2Ml = 0 (I L14)
and that on space variables can be separated. This implies that the ratio of amplitudes of any
The expansion of the determinant in Equation (lLl4) yields an algebraic equation of 1'1' order
two coordinates is independent of time. PhysjcaHy~ it implies that aU degrees of freedom
in {f)2, which is known as the characteristic equation. The roots of characteristic equation are
perform synchronous motion and the system configuration does not change its shape during
known as the eigenvalues and the positive square root of these eigenvalues are known as the
mouon but only Its amplitude changes.
natural frequencies (Wi) of the MDOF system. It is only .t these N frequencies that the system
Substituting for v, the equation of motion may be written as.
admits synchronous motion at all coordinates. For stable structural systems with symmetric and
M{ifJ}ii(t) .. K{q,)q(l) = 0 (11.8) positive definite stiffness and mass matrices the eigenvalues will always be real and positive. For
each eigenvalue the resulting synchronous motion has a distinct shape and is known as naturaV
which is a set of N simultaneous equations of the type no17fU21 mode shape or eigenvector. The normal modes are as much a characteristic of the system
N N as the eigenvalues are, They depend on the inertia and stiffness. as reflected by the coefficients
I. m'il/Jii(t) .. I. kijtP jq(l) = 0; i = 1, 2, ... , N (11.9) mij and kij' These shapes correspond to those structural configurations. in which the inertia forces
imposed on the strUcture due to synchronous harmonjc vibrations are exactly balanced by the
j=l
elastic restoring forces within the structural system. These eigenvectors are determined as the
where the separation of variables leads to nontrivial solution of Equation (11.13). Since the determinant of the coefficient matrix
evaluated at one of the natural frequencies js singular, a unique solution for eigenvectors can
noOt be found. It is, however, possible to compute the amplitudes of the synchronous motion at
_ ij(t)
1.2, .. ,. N (11.10) N  J coordinates relative to the amplitude of motion at the remaining coordinate, which may
q(l) be selected arbitrarily, Thus an additional constraintknown as normalisation conditionmust
be supplied in addition to Equation (l L13) to completely determine an eigenvector. Two of the
Since the tenns on either side of the equality sign are independent oOf each other, this equality most commonly used nonnalisation procedures are:
can hold only when each of these terms are equal to a positive constant (say. ar).l Thus we have. (i) assume the amplitude of synchronous motion at the first degree of freedom as unity.
ij(t) .. wq(1) =0 (1 J.lI) (ii) constrain a length measure of the eigenvector to be unity. Fot example, for any eigen
vector (4)(')) it is pessible to determine elements of {tP(,)} such that {4>'}TM{;!>W} = I.
N Such a normalisation. using mass/inertia matrix (M) is known as mass renormalisation
I. (k" w'm,)tP j 0; i:;: 1, 2. , .. , N. (lJ.J2) and the resulting mode shape is known as mass orthonormal mode shape.
j=f
It can be shown that the N eigenvectors of an NDOF system completely span the Ndimensional
The solution of Equation (I L11J is q(1) = sin(al'  <X), a harmonic offrequency w. Thus we vector space, and therefQre~ can be used as basis vectors foOr representing any Nh order vector.
may conclude that the motion of all coordinates is harmonic with same frequency (f) and same Since the condilion of orthogonality is a necessary condition for any set of base vectors, it will
phase djfference a. However, it !aiJ) needs to be established if sl1ch a synchronous, harmonjc now be shown that the eigenvectors also satisfy this condition.
1)luliuf) is possible for aU frequencies. To investigate this issue let us consider the Equation
{(j)}TK{~(i) = ro;{qIJ,}TM{#'l} (ILl8) For a particular case when the vector u represents the arnpJitudes of the harmonic oscilIations
of the NDOF system or the Rayleigh quotient, p, corresponds to square of the frequency of
(#'1jTK{qIJ) W] (~i))TM{(i)} (J L19)
harmonic oscillations. This result follows from the principle of conservation of energy by
Subtracting Equation (ILl9) from Equation (lLl8) and noting {.pWJ'K{~'l} = ({"lj'KT{qlJl))' equating the maximum potential energy stored in the system to the maximum kinetic energy,
and the fact that K and M are symmetric matrices we have, Further, the Rayleigh quotient has the property of being stationary in the neighbourhood of the
natura] modes of the system. It is a global minimum for the fundamental mode and global
(1 L20) maximum for the highest mode of vibrationalso known as the minimax property of RayJeigh
For all modes i ;t j with distinct eigenvalues (w,;t "'i), Equation (11.20) can be specified only quotient.
if the matr]x inner product {qlJl)TM{ ~"} vanishes. This proves the first half of the proposition
stated in Equation (1 LIS). The other half follows by substituting this result in either of 11.5,2 Stodola Method
Equation (1 Ll8) or (I L19).
By o'ansfonning the generalised eigenvalue problem to the standard ejgenvalue problem.
Since the computed mode shapes of al'DOF system form a set of orthogonal vectors, they
span the Ndimensional space completely. In other words, these mode shapes can be used a;; a (11.24)
set of basis vectors in the Ndimensional space and any vector in this space can be represented
as a linear combination of these modeshapes. where D == K1M is known as the dynamical matrix of the system and t. ~,Stodola method
W
The orthogonality property of mode shapes leads to a very powerful theorem, modal
expansion theorem, which states that any vector x in Ndimensional vector space can be start, with the choice of a trial vector, say, (i>(0)). Premultiplying (i>(O,) by the dynamical
represented as a linear combination of modeshape vectors, matrix, D yields another vector {{l)}, which is an improved estimate of the eigenvector. An
.v estimate of the eigenvalue is obtained by taking the ralio of any element of new vector {~(l)}
:< = I. q( {(d) (ILlI) to the corresponding element of the trial vector, i.e.,
;""1
(1)
We/Jere JA,U),
l V'. j represents e,. mode shape and
th"h .
q; denotes the correspondmg modal'
coordinate. ,(1) = !J._ (1 L25)
A "'(O)
For a given vector x, the modal coordinates qi may be computed by using the property of lJ
\ll\hogunality of modeshapeR as,
if {~(l)} were lI: true eigenvector. this ratio would be constant for any choice of the elemen~ of
these vectors, In general, however, this ratio will be different for different choice of elements
(11.22) of these vectors, In the special case of symmetric coefficient matrices, the minimum and the
maximum values of this ratio provide the upper and lower bounds on the eigenvalue. The
iteration resumes with the new trial vector of {(1)} ';:; J{~(1l}.
,1.(1)
Thus the equation for i lh where, use has been made of t.'1e modal expansion theorem (Equation (l1.27, The coefficient
q) so determined, gives the extent of representation of the mode~shape {tiP}} in t"'le trial vector
iteration is given as~
{~}, Let us define a new trial vector {or} ~ (~) ql(ID'n} by sweeping out the traces of known
(11.26) eigen vector (lilt,}, Using tillS purified trial vector in the iteration procedure we would converge
to the next lowest mode i.e. 211d mode, {<2)}. This process can be repeated to compute any
The Stodola method can be viewed as an iterative solution of a system of simultaneous equations desired eigenvector by sweeping out the traces of all the previous lower mode eigenvectors from
to arrive at that configuration of generalised displacements for which the inertia forces are the trial vector. A geometrical interpretation of the process of sweeping is to detertrtine a tria1
exactly balanced by the elastic forces in the structural members, vector which is orthogonal to aU the previously determined eigenvectors and this approach is
known as vector purification/deflation.
Why should iterative procedure converge to the first mode always? In theory, though it is possible to sweep out completely the traces of a known eigenvector
To answer this natural query, let u<:o take recourse to the modal expansion theorem and expand from an assu;ned trial vector, in practice, however. it is necessary to sweep oui the known
eigenvectors from trial function before !.he beginning of each iteration. This precaution is
an arbitrary trial vector (~) as, necessary because the round off errors due to finite precision arithmeric, on a computer always
(11.27) leave some small traces of swept out eigenveclor(s} in the trial vector at the end of the iteration,
It is possible to automate the process of sweeping in each iteralion by sweeping out the [races
where (<,l'''), (1/>(2)), "" [ID'N denote the eigenvectors of the dynamieal system, The first of known modes from the coefficient matrix. Let us consider that first n(< N) modes are known
iteration results in, and it i:: required to converge 10 the 11+ J I/: mode via iteration. Since the need for sweeping the
traces of known modes from tria} veClor at each iteration may be computationally expensive.
D{) ~ 4{I/>(I)}+ q~ {~(2+",+ q~ {~(N)} (11,28) it is worthwhile to look for the possibillty of a more elegant formulation for this procedure. Let
WI U)2 ruN us consider that rtf} be the trial vector from which the traces of first n( < N) modes are to be
Thus each iteration results in amplification of the lib teoo in the modal expansion by a factor removed. We have, by modaJ expansion theorem,
1101. So that after p successive iteratlons. N
For first n modes, which are known, the coefficients q) can bc computed by using the orthogo
Assuming that the natural frequencies (Wi) are all distinct and are numbered in the ascend~
jng order i.e. oJ < ~ < ... < roN. it foHows that after sufficient number of cycles nality propeny, The purified trial vector (or) can then be given ""
~~ ~ ... ~J.._. Therefore the first teoo in the modal expansion becomes progres~ N
... ~'lp
..... 1
ro'lp
2
(j)'lP
N (VrJ = (iii)  I. (I/!(}'}q i
sively m.ore dominant with each iteration and eventually converges to the first mode {4'(0). i"'l
chosen triai vector exactly resembles a higher natural mode. Therefore to determine the higher ~ S(i,iI)
modes using iteration procedure, it is necessary to sweep out all thp lower modes. For example.
let us as.sume that the fIrst mode shape has already been determined and has been mass where S is known a~ the sweeping matrix and {he entire process of ~\I;"eeping out [he known
m()dc~ from a trial vector ht.lS been reduced to a simple matrix tnUlliplicalion. In practice, the
onilOnorm.lized (such tbat (Ol) "M (ID" ') = 1,0), Considering any arbitrary trial vector (~) and
Coefficient matrix of the eigenvalue problem is postmultiplied by thc sweeping matrix and the
premUltiplying it by (qi")TM and invoking the orthogonality of mode shapes,
resuIting updated coelTkicnt matrix is u.\:Ied in the iteration procedure to converge fo rhe n + I In
1(l)l'M[~) = 1'i))lM(Q)'I)}q, + (<fI!)7M(<f21)q, +" + {(J'M{I'"VI)qN mode. The sweeping matrix: is then updated to sweep OUl [he fir'll n + 1 modes by extendjng
I the l>ummation in Equation (J L 3 J) to include the n + JIh mouo;::,
(11.30)
1
i
. , . ( Earthquake Resi6tant Design of Structures c Chapter J 1 Dynamics oj Multi~DegreeofFr~~;;'~ ~fJ$ie;;!) .100.
For the demonstration of the procedure, let us consider Table I 1.2 shows iteration for second mode and the coefficient matrix for the iteration for
a three~storey shear building shown in Figure 1 L5. The second mode IS given by,
=
system parameters are given as m 3500 kg, k, = k 1500
kNlm, k, = 15k. and k, = 2<Ok< The mass and stiffness D, = DS, = 7
On3 0.189
OJ 89 0.246
0.1 37
0<067
1
matrices can be written as,
r
,0.137 0.067 0.296 J
I
TABLE 11.2 Iterations (or second mode
25
tS D,
m
NO! ) (,p(i)) { lji"} [,p(2)) {if 2l J (.;I") {ljI"i l (.pI'I) {1f14 I } (it"')
.. ~~ .. ~ ~.<
For inverse iteration, the system coefficient matrix for the 0.173 D J89 0.137 1.00 0302 LOO 0.49 J,OO 0.496 LO 0.498 1.00 0.499
standard form of eigenvalue problem is given as, FIGURE 11.5 A 3slorey shear O.~ 89 0.246 0.067..0.50 ,0.329 1.09 ,0.51 :.04 D.506 1.02 0.505 ,LOt 0.503
building.
.(U3? 0<067 0<2% 0.25 {l.245 OS} 0.45 0.92 {l.479 0.%6 0.49 ~O_986 0.497
<.~~<
2.167 Ll67
0.5'1' The approximation to eigenvector after SID iteration lS {Vpi}'t::= rl.000, 1.008, 0,9961
7[ U67 1.167 05
Thus, as the iterations proceed, the iteration procedure converges to the second eigenvalue
05 0.5 0.5 i
t., = O.5mlk and the correspooding eigenvector is (q)[2)} = [1.00. l.00,  Loo)' Accordingly,
and then deriving the coefficient matrix for iteration for the third mode (D z) by premultiplying
Sweeping
the new :;weeping matrix by D, I.e.,
The sweeping matrix for removing the fir:;;t mode is given by, D;, = DS,
The remaining eigenpair may then be computed via iterations,
If we relax the assumption regarding simultaneous occurrence of peak response in all modes,
and assumjng that the natural frequencies are not very closely spaced then the maximum
response in physical coordinate system can be estimated as, = r~:~:~lm
,0.159
It must be empbasized bere that these modal combination rules are approximate procedures for
combining the maximum modal responses to get a probable estimate of the maximum response
of physical system. These modal combination rules may be used to estimate the probable which, in case of current problem leads to,
maximum value for any response quantity of interest such as, shear force, bending moment,
drifts, etc. However, care should be taken to ensure that the maximum of each desired response r r }' + 1(1
r + 15221 x ", 0.342 x '::T
1 57 l'
parameter is first calculated for each mode and then these modal maxima are combjned 1{(1  0.548) x !.303 x 2.940 \
according to a modal combination rule. An example which illustrates this procedure is given
L 3.88' 9.10 J
') ~O.5
below. 393 }'
+ {(l+6.260)xO,037X',
15.31
J I
Example I Consider a 3storey shear building shown in Figure 11.5 with the following 2
[t
properties: '
(0548 0.198) x L303 x ~:~:~}"{ 1"71
+ (1,522 + 0.872) x 0.342 x 9.;'SQ
30.0 0.0 LOOO }2 : . 05
J
! 0.0 15.31}
'{(O.l98) x 1.303 x 2.9 40 1' + {(O.872) x", 0.342 x 1057, }'
Compute the floor displacement.:;, interstorey dnftR. slorey shears and overturning moment'i of
this building when excit~ by an earthquake. The pseudospectral acceleration ordinates of the l 3,88' J .
21.5
9.15'
earthquake ground ace,<le"'tion for the three modes are given as 5" = 2.94, 1.57, and 3.93 m/sZ. 3.93 } ..
+ {(12.100)xO.037X~
Assume the storey heights to be 3.0 m and use SRSS rule for combining modal responses. 15.31' !
\
Solu~n The modal mass in the r: mode of vibration can be computed as m ~
h
'0.116\
{qt'j)'M{'''). For this problem, the modal masses are = 40.185, m;
= 122.306, and m; m;
\ = [ O.089 1m
;:: 5597.92K The mode participation fach"" l( rr I , "'jT M .
.,
m,
IlP' r i can be computed as I!
)
\0.051,
1.303,12 D.342, and rj = 0,037. The maximum storey shears are given by,
The maximum floor displacements arc given by,
', J05
v'"""
l.mM':
I~ r {I)
l I
, .2 1
r,.(SuJw,}}
Chapter 11 Dynamics of MultiDegreeof~Freedom SYBte.m8 ) . ,
or. in vector form
trans1ationa1 components of ground motions simultaneously at a support point. Computations
([(30,0 x LO x 1.303 X 2,94)' .;. (30,Q x 1.0 x  0.342 x 1.57)' for any response quantity of interest for simultaneous excitation by mUltiple components would,
in general, yield different estimates than for any single ground motion component acting alone .
.;. (30,0 x 1.0 x 0.037 " 3.93)' J0.5
It is not adequate, for design purpose, to consider the maximum response out of the three
{{30.0 x (I.;. 0.548)" 1.303 x 2.94)' estimates obtained for different ground motion components independently, To find the response
Vm. ~ + {3D,Q x (I  1.522) x  0342 x 1.57} , parameters for use in design, the response estimates for excitation by individual components
.;. {30.0 x (I 6.260) x 0.037 x 3.93)'JO.5 may be combined together by SRSS rule. For any generic response quantity of interest, say~ R,
({30D x (l .;. 0.548.;. 0.198) x 1.303 x 2.94f the value to be adopted for design calculations Rdes can be obtained as,
.;. {30,Q x1.522  0,872) x 0.342 x 1.57)'
(J
= IR' + R'y + R'z
RdestJx
T {30.0 x (1 6.260.;. 12.100) x 0,037 x 3.93)'f5
where, R,_, Ry. and Rz represent the estimate of response R due to excitation by ground motion
1116.13' in x, )', and z directions, respectively.
~ !
17 0"'_,'</11,"1>,'
'.n.J"
I {30.0 x (I x (9.0  3.0) + 0.548 x (6.0  3.0).;. 0.19& x 0,0) X 1.303 X 2,94)'
+ 130() x (l x (9.0  3.0)  1522 x (6.0" 3.0)  0.872 x 0,0) x 0342 x 1.57)' where. it <:: <:: N, The decision about the number of modes to be included in the response
computations may be based on the foHowing two criteria:
pO.!, x (i x (9.0 3.0)  6.260 X (6.0  3.0) .;. [2.1 00 x 0.0) x 0.037 x 3.93)' ]05 J
([) All modes having natural frequencies less than or equal to the highest frequency in the
"q '1
',','' ~
excitation should be included in the modal summation,
~ 348391 tN."., (il) At least 90% of the total mass of the structural system should be included in the dynamic
~~~\r.5S ) response computation, This criterion in assessed by considering the cumulative effective
modal rnass (= m ~ r;') for all modes included in the summation. which should be
'f"](' ~:,:!::.:~!; .aioil;. Jor nlJxirnum overturning momeni <Ii [be base Gl!1 also be more than 90% of the toml mass of the system.
In ,he 1.I1)()VC~lllentioncd .. ~. I .
. , ." c .... amp1e, on y one component of nround...acceJeralion was: consjd inference can be drawn by consi.dering tbe mll.ure of response of SPOF systems to harmonic e:xdtations. The
. It;) 'Ui ':~'('H'llj()fl Tn oen"ral thO' .[ 1d be . 0
 "'. ~ .. e  !; 'i ructure wou subjected to three mutually orthogonal dynamic amplification factor for the oscillator response approacbes unity as the ratio of frequency of e