Embed Size (px)
Citation preview
16. Computer Applications in Seismic Design 793
Chapter 16
Computer Applications in Seismic Design
Farzad Naeim, Ph.D., S.E.Vice President and Director of Research and Development, John A. Martin & Associates, Inc., Los Angeles, California.
Roy F. Lobo, Ph.D., P.E.Senior Research Engineer, John A. Martin & Associates, Inc., Los Angeles, California.
Hussain Bhatia, Ph.D., P.E.Senior Research Engineer, John A. Martin & Associates, Inc., Los Angeles, California.
Key words: Computer Applications, Earthquake Engineering, Earthquake Ground Motion, Engineering Judgment,General-Purpose Software, Loss Estimation, Instrumented Building Response, Seismic Design, Special-Purpose Software
Abstract: This chapter surveys the state-of-the-art in computer applications in seismic design. The field of computerapplications is rapidly changing. Therefore, a general overview of contemporary applications is providedwith references to the relevant worldwide web site addresses. The ever-increasing reliance on computerapplications requires a re-doubling of emphasis on sound engineering judgment by practicing professionals.Computers can enable us to perform engineering tasks we did not dream to be possible just a few years ago.Blind faith in computers, however, may produce results that are far less reliable than back of the envelopecalculations by a seasoned engineer.
794 Chapter 16
16. Computer Applications in Seismic Design 795
16.1 INTRODUCTION
This chapter provides a sampling ofcomputer applications in seismic design at thetime of this writing. No other field of scienceand technology moves forward faster thancomputer and communication technologies.Therefore, it is vital for the reader to examinethe state of knowledge and practice at the timeof his/her reading because significant advancesmay have occurred in between the time ofwriting this chapter and the time it is beingread. To assist the reader in this task, we willpoint to relevant Internet resources in differentparts of this chapter.
The builder’s need for computationaldevices predates ancient Babylonian, Persian,and Greek empires. Over the ages, as thecomplexity of engineering concepts grew, itinitially created master craftsmen: people whocould design and build magnificent structureswithout an exact understanding of underlyingmathematical principles but a fantastic ability toapply structural proportions found workable innature. For example, it is said that theslenderness of the Pantheon columns werederived from studying the proportions of thehuman female leg-bones. The curves of manymagnificent ancient domes were derived fromthe shape of wild mushroom crowns(16-1). Overmany centuries, remarkable structures werebuilt –without any precise mathematicalformulation– that withstood the test of the time.These designs were based on what we now referto as sound engineering judgment. The design-build practice that is now becoming prevalent inthe United States and other advanced countries,was the only form of construction known formany centuries.
The next stage in engineering evolutionbrought about the multidisciplinary masters.People like Leonardo Davinci who was an
artist, architect and engineer at the same timeexemplify this category. The growth of scienceand engineering knowledge in the 20th centurymade high degrees of specialization necessaryand made multidisciplinary masters extinct.Today, not only we distinguish structuralengineers from civil engineers but we furtherbreak down each field of expertise: structuraldesigners, structural analysts, earthquakeengineers, wind design engineers, claddingspecialists, seismic isolation specialists, designground motion specialists, etc. Therefore, welive in the era of specialists.
Specialization increases the depth of theknowledge but unfortunately reduces thebreadth of it. The grand vision common tomaster builders and multidisciplinary mastersare very difficult to find. At the same time thegrowth of computing hardware and softwareover the past two decades have beenmonumental. It is safe to say that all specialistsnow rely on computing facilities to the extentthat was imaginable just a few years ago. Thecombined effect of reduction in the scope ofknowledge (brought about by specialization)and heavy reliance on computational devices(caused by rapid growth of computingfacilities) can be dangerous. Engineering hasnever been, or can be, a pure game of numbers.Engineering judgment is simply too importantto be lost to blind faith in computing devices.There is a need for balance. We have to findways of maximizing our use of computertechnology without leaving our engineeringjudgment behind. Seismic design students mustbe trained to develop and to value a physicalfeeling for how buildings resist earthquakeforces, why they survive them, and the cause oftheir failure. The best use of computertechnology is only possible if respect forengineering judgment is nurtured andpreserved.
796 Chapter 16
The computer revolution that started in thelast quarter of the 20th century and is stillaccelerating today, has the potential ofimpacting human civilization more than theadvent of printing by Gutenberg(16-2). As will benoticed from reading this chapter, earthquakeengineers are now achieving objectives thatcould not have been even imagined a short fewyears ago. A few examples would beillustrative. The probabilistic seismic hazardmap of the entire United States for default sitesoil conditions is now readily available on theInternet and distributed as a part of the 2000International Building Code (IBC-2000)(16-3) aswell as FEMA Guidelines for SeismicRehabilitation of Existing Buildings(16-4, 16-5). Acompanion CD-ROM to these documentsallows the user to identify design spectralordinates of any site by providing its latitudeand longitude. For more approximateapplications, providing a postal zip code alsosuffices! (Figure 16-1).
Instrumental Intensity maps for significantearthquakes in the southern California regionare automatically produced by the Trinet andCube networks. The Cube maps areinstantaneously sent via e-mail to subscribers.Trinet shake maps may be viewed on theInternet (http://www. Trinet.org) within a fewminutes after earthquakes (Figure 16-2). Aclick-able map for Southern California faultsavailable at a web site(http://www.scecdc.scec.org/faultmap.html)permits users to point to any fault and obtain allrelevant information (Figure 16-3).
In the field of loss estimation, emergencymanagement and post-earthquake response, theGIS based HAZUS-99 software system(16-6)
developed under a grant from the FederalEmergency Management Agency (FEMA) hasprovided a new horizon to various casualty lossscenario and probabilistic analysis (Figures 16-4 and 16-5).
Figure 16-1. Seismic Hazard Map CD-ROM Supplement to IBC-2000
16. Computer Applications in Seismic Design 797
In seismic analysis and design of verycomplex structures, automotive and airplaneproportioning and design software have beenutilized to accommodate the sophisticatedcurvatures in the architectural and structuralsystems (Figure 16-6)(16-7).
Figure16- 2. A TriNet Shake Map for the 1999 HectorMines Earthquake in southern California (www.trinet.org).
Figure 16-3. A click-able Fault Map Available on theInternet (www.scecdc.scec.org).
Detailed nonlinear finite element analysistechniques have been successfully utilized topredict the experimental behavior of proposedstructural connections (Figures 16-7 and 16-8)(16-8).
Figure 16-4. A HAZUS-99 casualty loss estimate for ascenario event in southern California.
Figure 16-5. A HAZUS-99 analysis of liquefactionpotential and dangers posed by hazardous material storage
sites in Alameda county of California.
798 Chapter 16
Figure 16-6. The Disney Concert Hall, under constructionin Los Angeles, California was designed using CATIA, asoftware primarily used in automotive and airplane designapplications.
Figure 16-7. Nonlinear finite element analyses wereinstrumental in shaping a new SMRF connection for the
UCLA Replacement Hospital under constructionin Los Angeles, California.
Last, but not least, up-to-date literaturesearches can be conducted online. Therefore,seismic design engineers rarely need to “re-invent the wheel”. Now, it is not only alwayspossible, but a necessity, to check the relevantinformation on the Internet before one starts toembark on an unfamiliar path. A few web sitesof particular significance in this regard arelisted in Table 16-1.
In short, computer applications havetremendously enhanced our capabilities in allfacets of seismic design and construction. At
the same time, computer applications has to bebalanced with sound engineering judgment anda true physical sense of seismic performance,for it to benefit –and not adversely affect– thesafety and quality of the end product.
Figure 16-8. Full-scale testing confirms the findings of thecompute model shown in Fig. 16-7.
Table 16-1. Important World-Wide-Web sitesOrganization Web Site AddressEarthquake EngineeringResearch Institute
http://www.eeri.org
MultidisciplinaryCenter for EarthquakeEngineering Research
http://mceer.buffalo.edu
Mid-AmericaEarthquake Center
http://mae.ce.uiuc.edu
Pacific EarthquakeEngineering ResearchCenter
http://peer.berkeley.edu
The EarthquakeHazards MitigationInformation Network
http://www.eqnet.org
Applied TechnologyCouncil
http://www.atcouncil.org
Trinet http://www.trinet.orgSouthern CaliforniaEarthquake Center
http://www.scec.org
California StrongMotion InstrumentationProgram (CSMIP)
http://www.consrv.ca.gov
USGS NationalEarthquake InformationSystem
http://wwwneic.cr.usgs.gov
HAZUS User Group http://www.hazus.orgFederal EmergencyManagement Agency
http://www.fema.gov
16. Computer Applications in Seismic Design 799
16.2 EARTHQUAKE RECORDS
A few short years ago, it was very difficultto get hold of a good collection of earthquakerecords for design. That is no longer the case.Naeim and Anderson(16-9) have compiled acomprehensive list of design attributes ofhorizontal and vertical components of availableground motion for North and Central Americaas well as Hawaii. Once the desired designattributes are determined, it takes only a shortvisit to various web sites that contain largedatabases of earthquake records for variousregions of the world. For example, forCalifornia records, the CSMIP web siteprovides time series as well as spectralordinates of a variety of recorded groundmotions (Figures 16-9 to 16-11).
Figure 16-9. Selecting an earthquake record from theCSMIP web site.
Figure 16-10. Time series for the earthquake recordselected in Fig. 16-9 as displayed on the CSMIP web site.
Figure 16-11. Response spectra for the earthquake recordselected in Fig. 16-9 as displayed on the CSMIP web site.
16.3 MONITORING SEISMICACTIVITY
Besides click-able fault maps, seismocams(worldwide web pages connected directly toseismograms or to cameras focused on them)can be found in abundance on the Internet,some very serious work is being conducted inthis area that could not possibly been performedwithout computer assistance. Perhaps the mostsignificant of these experiments is beingconducted by TriNet in Southern California.
TriNet is a multifunctional seismic networkfor earthquake research, monitoring andcomputerized alerts. TriNet is a cooperative
800 Chapter 16
project between US Geological Survey,California Institute of Technology, and theStrong Motion Instrumentation Program of theCalifornia Division of Mines and Geology. Thegoals of TriNet are to provide data for researchin engineering and earth sciences, emergencyresponse applications and development of aseismic computerized alert network. The TriNetnetwork features a dense recording of groundmotions in all frequency bands, dense strongmotion instrumentation with 150 broadband and600 strong motion sensors all connected to acentral processing system. The network canissue automatic post-earthquake intensity mapsvery quickly after an earthquake (see Figure 16-2).
16.4 SEISMIC HAZARD ANALYSIS
There are a variety of software systems withdifferent levels of sophistication available in themarketplace. Arguably, the computer programsdeveloped by the California geologist Dr.Thomas F. Blake (are among the most widelyused at least in the western United States(http://www.thomasfblake.com). We willhighlight Blake’s programs as representativeapplications in this field.
The EQSEARCH program(16-10) contains asearchable catalog of significant earthquakes inwestern United States dating back to 1880.Given a site latitude and longitude, soilconditions and the choice of attenuationrelationship, the program reports historicalevents that have occurred within a given radius(or rectangle) around the site. The program thenuses this information to estimate the peakground accelerations observed at the site as wellas a Gutenberg-Richter recurrence relationshipfor the site (see Figures 16-12 to 16-14).
The EQFAULT program(16-11) can be used toperform a deterministic seismic hazard analysisfor a given site. The input information is similarto that of the previous program. EQFAULT,however, searches a three-dimensional databaseof earthquake faults and reports maximummagnitude associated with each fault and an
estimate of the corresponding maximumaccelerations experienced at the site (Figures16-15 and 16-16).
Figure 16-12. A typical EQSEARCH input screen
SITE
LEGEND
M = 4
M = 5
M = 6
M = 7
M = 8
-100
0
100
200
300
400
500
600
700
800
900
1000
1100
-400 -300 -200 -100 0 100 200 300 400 500 600
E A R T H Q U A K E E P I C E N T E R M A PTest Run
Figure 16-13. A typical epicenter map generated byEQSEARCH
16. Computer Applications in Seismic Design 801
.001
.01
.1
1
10
100
3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
E A R T H Q U A K E R E C U R R E N C E C U R V ETest Run
Cu
mm
ula
tive
Nu
mb
er
of
Eve
nts
(N
)/ Y
ea
r
Magni tude (M)
Figure 16-14. An earthquake recurrence curve generatedby application of EQSEARCH
6.50
6.75
7.00
7.25
7.50
7.75
.1 1 10 100
E A R T H Q U A K E M A G N I T U D E S & D I S T A N C E STest Run
Ma
gn
itu
de
(M
)
D istance (mi)
Figure 16-15. A plot of earthquake magnitudes and theircorresponding distances from a given site generated by theEQFAULT program
.001
.01
.1
1
.1 1 10 100
M A X I M U M E A R T H Q U A K E STest Run
Acc
ele
rati
on
(g
)
D istance (mi)
Figure 16-16. A site acceleration versus distance chartgenerated by EQFAULT
Figure 16-17. A typical input screen for the FRISKSPcomputer program
802 Chapter 16
0.25
0.50
0.75
1.00
1.25
1.50
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
A C C E L E R A T I O N v s . P E R I O D475-Year Re tu rn Pe r iod
Acc
eler
atio
n (g
)
Period (sec)
Figure 16-18. A probabilistic design spectrum generatedby the FRISKSP computer program
0
10
20
30
40
50
60
70
80
90
100
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
P R O B A B I L I T Y O F E X C E E D A N C EC A M P B E L L ( 1 9 9 3 ) H O R I Z . 0 . 0 1 S
Exc
ee
da
nce
Pro
ba
bil
ity
(%)
Accelerat ion (g)
25 yrs 50 yrs
75 yrs 100 yrs
Figure 16-19. A typical probability of exceedance chartfor peak ground acceleration generated by FRISKSP
computer program
FRISKSP program(16-12) is a morecomplicated software than the previous two andperforms a probablisitic seismic hazard analysisfor a given site and is capable of generating aset of probablisitic design spectracorresponding to the desired average returnperiods and dispersions (Figures 16-17 to 16-19). It is also capable of hazard de-aggregation
16.5 LOSS ESTIMATION,SCENARIO ANALYSISAND PLANNING
The loss estimation methodology andapplication was revolutionized by release of theHAZUS-99 software system(16-6),developmentof which was made possible through aconcentrated and prolonged funding by theFederal Emergency Management Agency(FEMA).
HAZUS-99 was intended to provide local,state and regional officials with the toolsnecessary to plan and stimulate efforts to reducerisk from earthquakes and to prepare foremergency response and recovery from anearthquake. The program was also indented toprovide the basis for assessment of nationwiderisks of earthquake loss. HAZUS-99 can beused by a variety of users with needs rangingfrom simplified estimates that require minimalinput to refined calculations of earthquake loss.Since it is totally built around a geographicalinformation system (GIS) technology, itsapplication and enhancement are rather straightforward.
The vision of earthquake loss estimationrequires a methodology that is both flexible,accommodating the needs of a variety ofdifferent users and applications, and able toprovide the uniformity of a standardizedapproach. The framework implemented inHAZUS-99 includes each of the componentsshown in Figure 16-20:– Potential Earth Science Hazard (PESH)– Inventory– Direct Physical Damage– Induced Physical Damage– Direct Economic/Social Loss, and– Indirect Economic Loss.
As indicated by arrows in Figure 16-20,HAZUS-99 modules are interdependent withoutput of some modules acting as input toothers. In general, each of the components willbe required for loss estimation. However, thedegree of sophistication and associated cost willvary greatly by user and application.
16. Computer Applications in Seismic Design 803
Framing the earthquake loss estimationmethodology as a collection of modules permitsadding new modules (or improving models/dataof existing modules) without reworking theentire methodology. Improvements may bemade to adapt modules to local or regionalneeds or to incorporate new models and data.
The modular nature of the HAZUS-99methodology permits a logical evolution of themethodology as research progresses and thestate-of-the-art advances.
HAZUS-99 incorporates state-of-the-artmodels in the earthquake loss estimationmethodology. For example, ground shaking
8. Lifel ines-Uti l i ty
Sys t ems
4 . G r o u n d M o t i o n 4 . Ground Fa i lu re
Direct Physical Damage
6. Essent ia l and High Poten t ia l Loss Faci l i t ies
12 . Debr i s10 . F i re 1 5 . E c o n o m i c14. She l te r9 . Inunda t ion 1 1 . H a z M a t
16. Indi rec tE c o n o m i c
Losses
Potential Earth Science Hazards
Direct Economic/ Social Losses
Induced Physical Damage
7. Lifel ines-Transpor ta t ion
Sys t ems
5 . Genera lBu i ld ing
S tock
13. Casual i t ies
Figure 16-20. Flowchart of HAZUS-99 loss estimation methodology(16-6)
804 Chapter 16
hazard and related damage functions aredescribed in terms of spectral response ratherthan MMI. Modules include damage lossestimators not previously found in most studies,such as induced damage due to fire followingearthquake and indirect economic losses. Anationally applicable scheme is developed forclassifying buildings, structures and facilities.
HAZUS-99 incorporates both deterministic(scenario earthquake) and probabilisticdescriptions of spectral response. Alternatively,it accepts user-supplied maps of earthquakedemand. The software also accepts externallysupplied maps of earthquake ground shaking.The uncertainty in earthquake demand due tospatial variability of ground motion isaddressed implicitly by the variability ofdamage probability matrices or fragility curves.Uncertainty in earthquake demand due totemporal variability (i.e., earthquake recurrencerate) or uncertainty in the magnitude ofearthquake selected for scenario events may bereadily evaluated by the users. Loss estimationusing HAZUS-99 may be conducted on aregional or a national scale.
16.6 EERI/IAEE WORLDWIDEHOUSINGENCYCLOPEDIAPROJECT
Under the joint leadership of the EarthquakeEngineering Research Institute (EERI) and theInternational Association of EarthquakeEngineers (IAEE) and cooperation of engineersfrom over 70 countries an online encyclopediaof earthquake vulnerability of worldwidehousing is under progress. This is amonumental task of immense practicalconsequences. By collecting and comparingvarious types of housing vulnerability acrossthe globe and local techniques currentlydeployed for hazard mitigation, for the firsttime the sharing of experience and expertisemay be exercised in a truly universal scale. Theonline version to be developed and publishedon the Internet can be of immense value to
governmental as well as nongovernmentalagencies. It could be also used by fundingagencies such as the World Bank in rationalprioritization of investments in earthquakehazard reduction projects. The interested readeris referred to the EERI web site(http://www.eeri.org) for more information.
16.7 INSTRUMENTEDBUILDING RESPONSEANALYSIS
Seismic performance of instrumentedbuildings provide a vital link for criticalevaluation of various theories, code provisions,and practices utilized in seismic design.Generally, there are two types of seismicinstrumentation:1. Code Instrumentation whereby according to
mandates of the applicable building code,some significant structures are instrumented.Codes usually mandate a minimal level ofinstrumentation for buildings of certainheight and/or complexity. The requirementsare usually satisfied by installation of a tri-channel accelerometer at the base, mid-height, and roof of the building. Generally,in this type of application the varioussensors are not time-synchronized.
2. Extensive Instrumentation wherebybuildings are instrumented by installation ofa relatively large number of sensors (usuallybetween 10 to 30) throughout the plan andelevation of the structure. The sensorlocations are designed to maximize post-earthquake understanding of buildingresponse. Dozens of buildings have beenextensively instrumented by CSMIP andUSGS agencies in California. The records ofinstrumented response may be downloadedfrom the Internet (see Table 16-1).
To illustrate the lessons that can be learnedfrom studying seismic performance ofinstrumented structures, Naeim developed aninteractive CD-ROM based information system(Figure 16-21)(16-13). This information system
16. Computer Applications in Seismic Design 805
contains detailed information regardingperformance of 20 extensively instrumentedbuildings during the 1994 Northridgeearthquake. However, the database organizationand the overall structure of the informationsystem are readily expandable to include otherbuildings and/or other earthquakes. It providesfacilities for manipulating instrument records ineither frequency or time domain, combiningand contrasting them, identification ofpredominant building frequencies, andgeneration of moving windows fast Fouriertransform (FFT) functions to track possiblestructural damage by identifying significantshifts in predominate building periods.
Figure16- 21. The main folder for one of the 20 buildingscontained in the information system CD-ROM
Figure 16-22. A buckled penthouse brace documented forthe building shown in Figure 16-21.
Figure 16-23. One of the damaged columns for a severelydamaged building documented in the information system
CD-ROM
806 Chapter 16
Figure 16-24. A moving-windows FFT plot generated forthe building shown in Figure 16-23 using the informationsystem utilities indicating a significant softening of thebuilding due to damage. Horizontal axis in the plot showsthe time and the vertical axis shows predominant buildingperiod as a function of time.
16.8 STRUCTURAL ANALYSISAND DESIGN
Structural engineers have always been at theforefront of computer applications. Theadvancement of computer technology in termsof both hardware and software has vastlybroadened the use of computers in seismicanalysis and design. The advent of personalcomputers and availability of very sophisticatedanalysis software on this platform has furtherintegrated computers into routine seismicanalysis and design practices. The large scalefinite element analysis programs that wereavailable only on mainframe computers are nowreadily accessible on ordinary personalcomputers. As a matter of fact, interactive finiteelement analysis software has been eventsuccessfully ported on to some pocketcalculators(16-14).
Two and three-dimensional linear staticanalysis of structures has become so routinethat it is hardly worth extended review in thischapter. It is fair to say, that the generallyavailable competent software systems forperforming these tasks could be primarily
distinguished based on their user interface, easeof use, and the extent to which graphicalmodeling of the structure has been madepossible. The same observation is notnecessarily true for linear dynamic analysiswhere the number of robust software systemsthat can properly model untypical cases withoutill-conditioning and other similar problems isfairly limited.
Nonlinear analysis software systems, on theother hand, are in a revolutionary stage. Theyare undergoing rapid changes to accommodatethe various practical needs that have becomecritical because of the rise in popularity ofperformance based design techniques (seeChapter 15) and application of technologiessuch as seismic isolation and energy dissipationdevices (see Chapter 14).
A structure is said to exhibit nonlinearbehavior when its response is not directlyproportional to the applied load. Generally,three distinct types of nonlinearity may bedistinguished:1. Material nonlinearities account for the
hysteretic behavior of the material. Theircharacteristics are derived from theconstitutive stress-strain properties of thematerial. Commonly utilized materialnonlinearity models include elastic-plastic,hyper-elastic, visco-elastic, or visco-plasticbehaviors. The onset of nonlinear behavior(yielding) is governed by various yieldcriteria and their associated flow andhardening rules such as the Tresca and Von-Mises criteria. Depending on the materialused in the structure, different yield criteriasurfaces and yield to choose from. Exampleand surfaces are selected. Examples includethe Hill’s criterion for anisotropic materialsand the Mohr-Coulomb, Drucker Prager,and Cam-clay criteria for soils and rock.
2. Geometric nonlinearities are the effects oflarge displacements on basic structuralassumptions or on the equilibrium state.They include large deflections, P-∆ effects,and buckling.
3. Boundary nonlinearities model the behaviorof elements in contact, but not connected to
16. Computer Applications in Seismic Design 807
each other. These are specified either by gap(compression only), hook (tension only),sliders (friction at point of contact) or slide-line (friction over line of contact) elements.Generally speaking, computer programs
used in seismic analysis and design can beclassified into two main categories: general-purpose and the special-purpose structuralanalysis software.
General-purpose analysis programs are notspecifically designed for seismic analysis anddesign but can certainly be used for thispurpose. There are many general-purposeanalysis programs that can analyze structureswith any or all of the above mentioned non-linearities. The example(16-15) presented inFigure 16-25 shows the analysis of an anchoredcylindrical storage tank under reversed seismicloading. As the walls of the storage tank aremade of very thin plates of steel, thedeformations of the walls due to hydrodynamicloads are large. Also in many cases, thebuckling of the walls of such tanks are precededby yielding of the steel. Thus, both material andgeometric non-linearities are involved in theanalysis. For unanchored tanks(16-15), theproblem becomes more complex withparticipation of the uplifting of the base platefrom the foundation in dissipating energyduring an earthquake. In such cases, theanalysis program must also include contactnon-linearities as shown in Figure 16-26. Thus,for such seismic analyses, a general-purposeprogram is needed.
Figure 16-25. Buckling of an anchored cylindrical storagetank subject to reversed hyrodynamic pressures during anearthquake. MARC(16-16) was used in the analysis.
Because of the practical utility that specialpurpose software systems provide, their use ismore widespread than the general-purposesoftware. A practicing engineer is usually betterserved by using a special purpose softwaretailored to handle the specific type of project athand rather than using a general purposesoftware to tackle all kinds of projects. Inaddition general-purpose programs by theirnature are more complex and difficult to master.Therefore, training engineering staff on the useof specialty software tends to be lessburdensome. General-purpose programs alsotend to be more costly in terms of initialpurchase and subsequent maintenance.
M
W
W Ws f
kR
(R -r)
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr
R
E leva tion P lan
Θ
−Θ
∗
∗
cre sen tU plifted
808 Chapter 16
Figure 16-26. Buckling and uplifting of an unanchoredcylindrical storage tank under seismic loads. Shown at thebottom is the base plate of the tank uplifting from its rigidfoundation due to the hydrodynamic pressures on thevertical walls of the tank. The resisting moment is due tofluid pressure causes the unique crescent shape of theuplifted portion of the base plate of the tank. MARC wasused in the analysis.
Seismic design of complex projects ofteninvolves application of both general purposeand special purpose software. For example, thedesign of the Staples Center sports arena in LosAngeles(16-17), The Eiffel Tower II in LasVegas(16-18) and seismic correction of the RoyceHall(16-19) all necessitated application of avariety of software from both groups ofcomputer programs (see Figures 16-27, 16-28,and 16-29).
Figure 16-27. The roof truss of the Staples Center wasanalyzed using the RISA-3D computer program. Shearwalls were analyzed using SAP-2000. Special softwarewas developed to pass information from various programsto each other.
Figure 16-28. Royce Hall seismic rehabilitation designutilized SAP-90 computer program in conjunction withBIAX and other nonlinear analysis software. Specialtysoftware was developed for cross-platformscommunications.
16. Computer Applications in Seismic Design 809
Figure 16-29. Eiffel Tower II in Las Vegas was analyzedand designed using SAP-2000. Nonlinear bucklinganalyses for temperature effects and fire scenarios wereconducted using the ROBOT software system
Commonly used general-purpose analysissoftware include SAP-2000(16-20), ADINA(16-21),NASTRAN(16-22), ALGOR(16-23) ABAQUS(16-24),COSMOS/M(16-25), ANSYS(16-26) and MARC(16-16)
The features and capabilities of these systemsare so rapidly changing that a comparativediscussion of their feature in a textbook like thiscould be counterproductive.
In addition to the above-named proprietarysoftware systems, there are a variety ofprograms available in the public domain. Theseare generally programs motivated by academicresearch and are made available by variousuniversities and research institutions.NONSAP(16-27) and ANSR(16-28) are examplesof public domain general purpose computerprograms.
Special-purpose programs, developed foranalysis and design of building structures, are
often used in seismic analysis and design ofbuildings. They are generally faster and provideinformation that could be more readily appliedto design purposes. Perhaps the most popularbuilding seismic analysis software is ETABS(16-
29). Currently a commercial software developedand maintained by Computers and Structures,Inc. of Berkeley, California, ETABS has itsroots in public-domain versions of TABS,TABS-80 and ETABS developed at theUniversity of California at Berkeley, during the1970s. The current commercial version of theprogram, however, is a very powerful and user-friendly program and has little in common withits old university developed predecessors.
A handful of public-domain programs areused extensively in nonlinear seismic analysisof structures. Perhaps the most widely usedamong this class of programs is DRAIN-2DX(16-
30) which is widely used in both professionaland research applications (Figure 16-30).
Figure16- 30. A DRAIN-2D nonlinear beam element
The success of DRAIN-2DX has resulted inthe development of an entire family of DRAINprograms such as DRAIN-3DX(16-31), andDRAIN-BUILDING(16-32). Various hystereticmodels are implemented in the DRAIN familyof programs (Figure 16-31) where the slope ofthe unloading branch is based on the previousmaximum plastic hinge rotation. All plasticdeformation effects including the effects ofdegrading stiffness can now be modeled .
810 Chapter 16
yF
DEFORMATION
X
F
yF−
yX
ey
un kX
Xk
α
max
=ek
maxX
Figure 16-31. Modified Takeda Model(16-33) is one of thehysteretic behavior models implemented in the DRAINand IDARC family of programs
A very promising development recentlyincorporated in the DRAIN family of programsis the incorporation of fiber elements (Figure16-32) that allow modeling of various behaviorstates occurring at the same cross section of abeam or column element.
Figure 16-32. A typical beam modeled by fiber elements
IDARC is another family of very powerfulpublic-domain computer programs developedand maintained at the University of Buffalo.The original IDARC(16-34) was developed fordamage analysis of reinforced concretestructures. The IDARC family of programs(16-35
to 6-37), however, can now be used for nonlinearanalysis of steel structures as well(16-39).
A major difference between IDARC andDRAIN families of programs, is in the
construction of the inelastic element stiffnessmatrices. DRAIN programs use a concentratedplasticity model where the inelastic deformationis concentrated at the locations of plastic joints.The individual member stiffness matrix inIDARC is constructed based on a flexibilityapproach. This permits modeling of plasticitydistributed along the length of the member.Concentrated plasticity models are generallybetter for modeling steel structures whiledistributed plasticity models (Figure 16-33)more accurately represent the response ofreinforced concrete members (Figure 16-34).
A B
L
LAα ( )L1 BA αα −− LBα
AA EI
1f =
BB EI
1f =
00 EI
1f =
Figure 16-33. A distributed plasticity/flexibility model
-1.5
-1
-0.5
0
0.5
1
1.5
2
-1.5 -1 -0.5 0 0.5 1 1.5
DISPLACEMENT (IN)
FO
RC
E (
KIP
S)
Analytical
Experiment
Figure 16-34. Analytical (using the distributed flexibilitymodel) and experimental results compared for cyclic testson a cantilever concrete column
16. Computer Applications in Seismic Design 811
Figure 16-35. Model of a three-story building
Figure 16-36. Comparison of Analytical Responses withExperiment for model shown in Fig. 16-33(16-38).
The correlation among the analyticalpredictions and observed performance has beencontinuously improving. For example, Figure16-35 shows the model of a three story buildingtested at the University at Buffalo. Diagonalbrace dampers were added between floors as aretrofit alternative. As indicated by Figure 16-
36 the analytical and experimental results are ingood agreement. Blind predictions of actualseismic response by analytical means, however,have not generally been as successful.
16.9 CONCLUSION
Computers are inseparable fromcontemporary seismic design. While advancesin computer technology have broadened therange of problems that can be handled byearthquake engineers, they have had theunfortunate side-effect of downplaying theimportance of sound engineering judgment.
Although vital to current seismic designpractice, computer use if not subordinated todesign experience and engineering judgment,is nothing but a recipe for disaster.
REFERENCES
16-1 Pearce, P. (1980), Structure in Nature Is a Strategyfor Design, The MIT Press, Massachusetts Instituteof Technology, Cambridge, MA.
16-2 Naeim, F. (1998), “Internet Frontiers forEarthquake Engineering Design and Education”,Proceedings of the 6th National Conference onEarthquake Engineering, Earthquake EngineeringResearch Institute, Seattle, WA.
16-3 International Code Council (2000), InternationalBuilding Code, March.
16-4 Federal Emergency Management Agency (1997),NEHRP Guidelines for the Seismic Rehabilitationof Buildings, FEMA-273, Washington, D.C.
16-5 Federal Emergency Management Agency (1997),NEHRP Commentary on the Guidelines for theSeismic Rehabilitation of Buildings, FEMA-274,Washington, D.C.
16-6 Federal Emergency Management Agency (1999),HAZUS – FEMA’s Tool for Estimating PotentialLosses from Natural Disasters, CD-ROM set,Washington, DC.
16-7 Naeim, F., Martin, J.A., Jr., Gong, V., Norton, G,Schindler, B.S., Scambelurri, M. and Rahman,M.A. (1999), “ Structural Analysis and Design ofthe Walt Disney Concert Hall, Los Angeles,”Proceedings of the 1999 SEOAC AnnualConvention, Santa Barbara, CA.
16-8 Naeim, F., Patel, K., Tu, K.C., Saiidi, M., Itani, A.,and Anderson, J.C. (2000), “A New RigidConnection for Heavy Beams and Columns in Steel
812 Chapter 16
Moment Resisting Frames,” Proceedings of the 5th
Conference on Tall Buildings in Seismic Regions,Los Angeles, CA.
16-9 Naeim, F. and Anderson, J.C. (1996), DesignClassification of Horizontal and VerticalEarthquake Ground Motion (1933-1994), ReportNo. 7738-96, John A. Martin & Associates, Inc.,Los Angeles, CA.
16-10 Blake, T.F. (2000), EQSEARCH: A ComputerProgram for the Estimation of Peak Accelerationfrom California Historical Earthquake Catalogs,Version 3.0, Thomas F. Blake Computer Servicesand Software, Thousand Oaks, CA.(http://www.thomasfblake.com)
16-11 Blake, T.F. (2000), EQFAULT: A ComputerProgram for the Deterministic Estimation of PeakAcceleration using Three-Dimensional CaliforniaFaults as Earthquake Sources, Version 3.0, ThomasF. Blake Computer Services and Software,Thousand Oaks, CA.
16-12 Blake, T.F. (2000), FRISKSP: A ComputerProgram for the Probabilistic Estimation of PeakAcceleration and Uniform Hazard Spectra usingThree-Dimensional California Faults asEarthquake Sources, Version 4.0, Thomas F. BlakeComputer Services and Software, Thousand Oaks,CA.
16-13 Naeim, F. (1997), Performance of InstrumentedBuildings During the January 17, 1994 NorthridgeEarthquake -- An Interactive Information System --, Report No. 97-7530.68, John A. Martin &Associates, Inc., Los Angeles.
16-14 Naeim, F. (1991), “Interactive Finite ElementAnalysis on a Pocket Calculator," Computers andStructures, Pergamon Press, Vol. 41, No. 2, Oxford.
16-15 Bhatia, H. (1997), Seismic Analysis and Retrofit ofCylindrical Liquid Storage Tanks, Ph.D.Dissertation, University of California, Irvine.
16-16 MARC Analysis Research Corporation (2000), TheMSC/MARC Software System,(http://www.marc.com)
16-17 Martin, J.A., Jr. and Naeim, F. (2000), Structuralanalysis and Design of the Staples Center, ASG-ASCE 6th Annual Symposium, University ofSouthern California, Los Angeles, January.
16-18 Naeim, F. (2000), Eiffel Tower II: Re-engineering aLandmark, Invited Lecture, North American SteelConstruction Conference, Las Vegas, February.
16-19 Naeim, F. (1997), UCLA Royce Hall: Anatomy ofAn Award Winning Seismic Correction, AnnualConvention, American Concrete Institute, April.
16-20 Computers and Structures, Inc. (2000), SAP2000Integrated Finite Element Analysis and Design ofStructures, Berkeley, CA.(http://www.csiberkeley.com)
16-21 ADINA R&D, Inc. (2000), The ADINA System,(http://www.adina.com)
16-22 MSC Software Corporation (2000), TheMSC/NASTRAN Software System,(http://www.mscsoftware.com).
16-23 ALGOR Corporation (2000), ALGOR VirtualPrototyping Solutions for Mechanical Engineers,(http://www.algor.com).
16-24 Hibbitt, Karlsson & Sorensen, Inc. (2000), TheABACUS Software System, (http://www.hks.com).
16-25 Structural Analysis and Research Corporation(2000), The COSMOS/M Software System,(http://www.srac.com).
16-26 ANSYS Inc. (2000), The ANSYS/StructuralSoftware System, (http://www.ansys.com).
16-27 Bathe, K.J., E.L. Wilson and R.H. Iding (1974),NONSAP – A Structural Analysis Program forStatic and Dynamic Response of Nonlinear Systems,Structural Engineering Lab, Report No.UCB/SESM-74/3, Department of CivilEngineering, University of California, Berkeley,California.
16-28 Mondkar, D.P. and Powell, G.H., (1975), ANSR-IGeneral Purpose Program for Analysis ofNonlinear Structural Response, EERC report No.75-37.
16-29 Computers and Structures, Inc. (2000), ETABSLinear & Nonlinear Static & Dynamic Analysis &Design of Building Systems,(http://www.csiberkeley.com)
16-30 Prakash, V., Powell, G.H., and Filippou, F.C.(1992), DRAIN-2DX: Base Program User Guide,Report NO. UCB/SEMM-92/29 Department ofCivil Engineering, University of California,Berkeley.
16-31 Prakash, V., Powell G.H., Campbell S.D., andFilippou F.C. (1992), DRAIN-3DX PreliminaryElement User Guide, Report, Department of CivilEngineering, University of California, Berkeley,CA. December.
16-32 Prakash, V., Powell G.H., and Filippou F.C. (1992)DRAIN-BUILDING Base Program User Guide,Report No. UCB/SEMM-92/31.
16-33 Takeda, T., Sozen, M.A., and Nielsen, N.N.,(1970), “Reinforced Concrete Response toSimulated Earthquakes,” Proceedings, ASCE, Vol.96, No. ST12.
16-34 Park, Y.J., Reinhorn, A.M., and Kunnath, S.K.,(1987), IDARC: Inelastic damage analysis ofreinforced concrete frame -- shear-wall structures,Tech. Report NCEER-87-0008, State Univ. of NewYork at Buffalo, N.Y.
16-35 Valles, R.E., Reinhorn, A.M., Kunnath, S.K., Li, C.,and Madan, A. (1996), IDARC2D: A ComputerProgram for the Inelastic Damage Analysis ofBuildings, Tech. Report NCEER-96-0010, StateUniv. of New York at Buffalo.(http://civil.eng.buffalo.edu/idarc2d50/)
16. Computer Applications in Seismic Design 813
16-36 Lobo, R.F., (1994), Inelastic Dynamic Analysis ofReinforced Concrete Structures in ThreeDimensions, Ph.D. Dissertation. University atBuffalo, NY.
16-37 Lobo, R.F. and Naeim, F. (1998), JAMA-IDARC-3D: A Computer Program for the ThreeDimensional Inelastic Analysis of BuildingStructures, John A. Martin and Associates, Inc., LosAngeles, CA.
16-38 Lobo, R.F., Bracci, J.M., Shen, K.L., Reinhorn,A.M., Soong, T.T., (1993), Inelastic Response ofReinforced Concrete Structures with ViscoelasticBraces, Technical Report NCEER-93-0006,National Center for Earthquake EngineeringResearch, SUNY/Buffalo.
16-39 Naeim, F., Skliros, K., Reinhorn, A.M., andSivaselvan, M.V. (2000), Effects of HystereticDeterioration Characteristics on Seismic Responseof Moment Resisting Steel Structures, A report tothe SAC Joint Venture, Report No. 2000/8428,John A. Martin and Associates, Inc., Los Angeles,CA.
