REPORT NO.

UCB/URC-89109

SEPTEMBER 1989

?B92-14:J064

EARTHQUAKE ENGINEERING RESEARCH CENTER

FEASIBILITY AND PERFORMANCE STUDIES ONIMPROVING THE EARTHQUAKE RESISTANCEOF NEW AND EXISTING BUILDINGSUSING THE FRICTION PENDULUM SYSTEM

by

VICTOR ZAYAS

STANLEY LOW

LUIS BOZZO

STEPHEN MAHIN

,----..................--+---1--------/',

COLLEGE OF ENGINEERING

UNIVERSITY OF CALIFORNIA AT BERKELEYREPRODUCED BYU.S. DEPARTMENT OF COMMERCE

NATIONAl'~NICALINFORMATION ~ERVICESPRINGFIELD. VA 22161

For sale by the National Technical InformationS"rvice, u.s. Department of Commerce,Springf;"ld, Virginia 22161

S"e back of report for up to date listing ofEERC reports.

DISCLAIMERAny opinions, findings, and conclusions orrecomme.,dations expressed in this publica-

tion are those of the authors and do not nec-

essarily reflect th" views of the National Sci.ence Foundation or the Earthquake Engineer-ing Research Center, University of Californiaat Berkeley.

~02n 101

OP'1'IOHAL FO 272 (.....77\F..' ....,ly I'fI'lS-~51Department 0' Comm.fC.

5.. '",'rucrlon. 01'1 11..",,"_I

(So. ANSI-Z39_111

REPORT DOCUMENTATION 11 RPORT NO. IZ-3.

P892- 143064PAGE NSF!ENG-890264. Titl. and Subtltl. 5. R_port Oat.Feasibility and Performance Studies on Imp~oving the Earthqua.ke September 1989Resistance of New and Existing Buildings Using the Friction LPendulum System

7. A~lhorr.1 L Porlorm,,,, O'1o"lnll... Ropt. No.V. Zayas, S. Low, L. Bozzo, s. Mahin UCB!EERC-89!099. p,tform'"c O....nlz.tion Name lind Add' 10. "",,_/T.ok/Work Un,l No.Earthquake Engineering Research CenterUniversity of California. Berkeley 11. Can,t.cUe) or C,."ICC) No.1301 S 46th St.

leIRichmond, CA 94804

ttl) 151-8860953I%. S""~onn. O,...."i.J:.tIOft N.me and Add~.. 13. Type af R.port .. "ori"" Cor.dNational Science Foundation1800 G. St. NWWashington, DC 20550 14.

1S. Suopl.m."tary ~ot.a

1!50. Abwacl (Umit: 200 _nil' The Friction Pl:ndulum System (FPS) is an innovative technique forimproving the earthquake resistance of buildings, which uses steel connections to isolateseismically a building by means of small amplitude pendulum motions. The anticipated seismicperformance of building structures using the FPS steel connections was investigated-analytically and experimentally. Buildings designed to have approximately equivalentconstruction costs as conventional building designs were studied. The earthquake responsesof the FPS supported buildings were compared with those of conventional code design.

The FPS was assessed to be feasible and cost effective for improving the seismicresistance of new buildings. The flexibility to select any isolator ~eriod makes theapproach suitable for a wide range of applications. The compact size and high strength of-the FPS isolators permit a versatility in installation details that helps to achieveconstruction which is cost equivalent to non-isolated buildings, yet provides substantiallyimproved seismic resistance. A cost eqUivalent example building, designed with the FPS anda reduced seismic .design load of 50%, demonstrated 86% less building damage during severeearthquakes as compared with the full strength design without the FPS.

The FPS was also assessed to be a feasible and attractive technique to improve theI _

seismic resistance of existing hazardous buildings. The flexibility to achieve relativelylong isolator periods of 3 to 4 seconds offered improved performance for the cases of weakbuildings studied. The compact size and high strength of the FPS isolators also offeredadvanta~es in retrofit details

17. Docum."t .....al~. Do.:"pto..

t1. Id."lln.rI/Opo",Eftd.d T......

c. COSATI ro.ld/Groull

lL Avai'~b,ltty Star.,".,,~ I', s.curity Cia.. (Tho. F.ogertl 21... _ .1 ... I. I

Release unlimited unclassified Iza. Security' Cia... (1I','. ") z. ,..... Iunclassified

~ ...

FEASIBILITY AND PERFORMANCE STUDIES ON

IMPROVING THE EARTHQUAKE RESISTANCE OF NEW AND EXISTING B11ILDINGS

USING THE FRICTION PENDULUM SYSTEM

by

Victor Zayas

Stanley Low

Luis Bozzo

and

Stephen Mahin

A Report to Sponsor:National Science Foundation

Report No. UCB/EERC-89/09Earthquake Engineering Research Center

College of EngineeringUniversity of California, Berkeley

California

September 1989

I) ,

ABSTRACT

The feasibility of using an innovative earthq 'Jake resi.stantconstruction technique to improve the earthquake resistance ofbuildings .... as in vestigated. Th e tech niq ue, called the FrictionPendulum System (FPS), uses steel connections to seismicallyisolate the buildings using small amplitude pendulum motions. Theanticipated seismic performance of building st:.ructures using theF!'S steel connections Io'as investigated using analytical andexperimental studies. Buildings designed to have approximatelyeq~ivalent construction costs as conventional building desi~ns Io'erestudl.ed. The earthquake response of the FPS supported buildingsIo'ere compared to those of conventional code design.

The FPS .... as assessed to be a feasible and cost effectiveconstruction techrdque fllr improving the seismic resistance of nelo'buildings. The fJe;.i.hility to select any isolator period makes theapproach suitable to a Io'ide range of applications. The compactsize and high strength of the FPS isolators, permits a versatilityof installation details which helps to achieve construction whichis cost equivalent to non-isolated buildings, yet providessubstantially improved seismic resistance. A cost equivalentexample building. designed with the FPS and a reduced seismicdesign load of 50%, demonstrated 86% less building damage duringsevere earthquakes as compared to the full strength design withoutthe FPS.

The FPS was also assessed to be a feasible and attractivetechnique to improve the seismic resistance of existing hazardousbu~ldings. The flexibility to achieve relatively long isolatorperiods C'f 3 to 4 seconds uttered improved performance for thecases 01: weak buildings studied. The compact size and highstrength of the FPS isolators also offered advantages in retrofitdetails.

Model size FPS isolators were tested at velocities up to 20inches/second, and at varied pressure loads. They consistentlyachieved ideal linear stiffnesses and dynamic friction coefficientsof less than 5%. Analytical studies of example building casesshowed that the shear loads, story drifts, ductility demands. andstructure inelastic energy dissipation in FPS supported structureswere substantially reduced as compared to similar structureswithout the FPS. The studies showed that the torsion motionsoccurring in asymmetrical structures Io'ith large mass eccentricitiescan be substantially reduced using the FPS. The FPS offers thepotential to improve the seismic pel"formance for a wide variety ofstructures.

ACKNO WL E DGM ENTS

This research was funded by the National Science Foundation,Small Business Innovation Research Program, under SBIR Phase Ig I'an t No. ISI-8860953. Th e NS F P rog ram official for this g rant isA. J. Eggenberger. The SBIR Program manager is Ritchie B. Coryell.

The support of the National Science Foundation in funding thisresearch, and the advice of Dr. Eggenberger and Mr. Coryell, aregreatly appreciated. The opinions, findings, conclusions andrecommendations expressed in this report, however, are those of theauthors and do not necessarily reflect the views of the NationalScience Foundation.

The research work reported was a cooperative research effort byEarthquake Protection Systems, San Francisco, and the Department ofCivil Engineering, University of California, Berkeley. Thecollaborative research effort was made possible by the NSF SmallBusiness Innovation Research Program, the University of CaliforniaEarthquake Engineering Research Center, and Earthquake ProtectionSystems, which encourage joint industry and academic research.

Copyrights: Permission to reprint part or all of this document isgranted on the condition that full credits are given to the authorsand source. All other rights reserved.

ii

TABLE OF CONTENTS

Acknowledgments

Chapter 1. Summary of The Phase I Research

Chapter 2. Feasibility of the FPS for New BuildingConstruction

Chapter 3. Feasibility of the FPS for Seismic Upgradingof Existing Buildings

Chapter 4. Component Tests of Low Friction FPS Isolators

Chapter 5. Torsion Response uf FPS Supported Structures

Chapter 6. Effects of Displacement Restraints onBuilding Response

Chapter 7. Response of Inelastic Multistory StructuresSupported on FPS Connections

Chapter 8. Response of Elastic Single-Degree-of-FreedomSystems Supported on FPS Connections

Appendix A. Design and Analysis Calculations, andAnalytical Models for the Example Building

1

Page

ii

3

26

58

86

107

135

154

189

255

SECTIONS

CHAPTER 1

SUM~ARY OF THE PHASE I RESEARCH

1.1 Introduction

1.2 Identification of the Need

1.3 Background

1.4 Phase I Research objectives

1.5 summary of the Research and Findings

1.6 Technical Feasibility Conclusions

1.7 Potential Applications

1.B Future Research Needs

1.9 References

Preceding page blank3

1.1 Introduction

The Friction Pendulum System (FPS) is an innovative approach forimproving the earthquake resistance of structures. structuressupported on FPS connections respond to earthquake ground motionsas ...,ould a simple pendulum. These simple pendulum motions are easyto control and predict. The feasibility of using the FPS approachto improve the earthquake performance of buildings is examined inthis report.

The feasibility of using the FPS in new building construction ina cost effective manner is examined in Chapter 2. The feasibilityof using the F PS approach to improve the seismic resistance ofexisting hazardous buildings is Axamined in Chapter 3. Thefeasibility assessments are based on the technical performanceevaluations presented in Chapters 4 through 8. In each of thesechapters a specific technical issue is investigated. Theconclusions, references, and figures pertinent to the issue arepresented at the end of each chapter.

This report summarizes the results of the Phase I researchsponsored by the National Science Foundation, Small BusinessInnovation Research Program. Consistent with the feasibilityassessment objectives of the Phase I program, simple examples andlaboratory tests, and simplified analytical models are used toassess the overall performance characteristics of the FPS. Thesimple building examples, tests and analytical models have thebenefit of being both instructional and illustrative of theanticipated overall behavior of buildings supported on FPSisolators. More detailed investigations of different buildingtypes, localized structural behavior, and the response to variedearthquake loadings are planned for the Phase II program.

1.2 Identification Of The Need

It has been estimated that a major earthquake occurring in alarge metropolitan area in the U.S. could result in upwards of 70billion dollars in damages and in tens of thousands of fatalities.Even more moderate seismic events can cause collapse of existinghazardous buildings, damage to new buildings, businessinterruptions, and disruptions of vital public services, all of...,hich have been proven to have profound long-term impacts on theeconomic and social ...,ell-being of the affected communities.Because of this, considerable research has been performed todevelop reliable techniques to design and analyze earthquakeresistant structures.

4

studies by the U.s. Geological Survey (USGS) have indicatedthat earthquake shaking can produce forces in structures 10 to 20times higher than those currently used in building design.Economic considerations have lead to a design approach forconventional buildings which relies on inelastic energydissipation, rather than strength, as the means of resisting theeffects of such large earthquakes. As a result, properly designedand constructed buildings would remain standing followil"g a majorearthquake, but substantial damage to the structure and to contentswould be anticipated. Research on these types of structures hasshown the need for costly details and construction practices inorder to achieve the desired behavior.

During the past decade, greater focus has been placed ondeveloping design strategies which reduce the need for costlystructural details while providing greater protection againstdamage to the structure and contents. With this in mind, a varietyof seismic isolation systems has been proposed in which thestructure is supported on components or devices whichpreferentially modify the dynamic characteristics of the supportedstructure and/or 1 imi t the ampl i tude of forces which can betransmitted from the ground into the structure. By reducing thedynamic response of the structure, lower design forces may beconsidered, seismically induced deformations and damages can bereduced, and critical contents (such as those involving toxicchemical or biological operation1=., data processing andtelecommunications equipment, and high technology manUfacturingfacilities) may be economically protected from earthquake damage.

An important application of the seismic isolation approachrelates to the mitigation of hazards pc::;ed by structures builtprior to the development of modern design and constructionpractices. Reduction of seismic defi ciencies in such structuresusing conventional means, is not onl.. technically difficult andcostly but is highly disruptive to the building occupants. Theappl icabil i ty of seismic isolation techniques to existingseismically hazardous buildings appears promising due to thepossible reduction in design forces and the localization of thestructural modifications to one floor level. However, improvedmethods for earthquake resistant construction of new or existingbuildings which increase the cost of construction are not widelyused in the building industry.

1 . 3 Background

The Friction Pendulul4 system (FPS) is an innovative seismicisolation system which appears to offer improvements in strength,longevity, versatility, ease of installat~on, and cost as comparedto previous systems. Moreover, the approach adds several inherentperformance benefits not available before. The FPS is based on

5

wall known anqtnaarinq principlas and is constructad ofconventional materials with demonstrated longevity and resistanceto environmental deterioration. The desirable isolationcharacteristics e>ehibited by FPS components hold the promise of aneconomical Clnd effective system for significantly increasing theseismic resistance of new structures and for substantially reducingthe earthquake hazards posed by e>eistinq structures. In order toachieve the potential benefits of this and other innovativesystems, careful attention must be placed on economic,architectural and construction aspects as well as on the moretraditional technical issues.

The Friction Pendulum system (FPS) offers a simple approach forincreasing a structure's earthquake resistance. A cross sectionview of an FPS steel connection is shown in Figure 1.1. The FPSconcept is based on an innovative way of achieving a pendulummotion. Fig. 1.2 schematically illustrates how the FPS achieves apendulum response for a supported building. The building respondsto earthquake moticns with small amplitude pendulum motions.Friction damping effectively absorbs the earthquake's energy. Theresul t is a simple, predictable, and stable earthquake response.Examples of the FPS hysteretic loops are shown in Fig. 1.3.

The connections can be installed at the bottoms or tops of lowElrstory colul'lns, or between the building and its foundation. Theoperation of the connection is the same whether the concave surfaceis facing up or down. Fig. 1.4 illustrates the operation of theconnection when installed at the top of a column, with the concavesurface facing downward.

Previous research by the investigators has addressed how the FPSconnections behave under simulated earthquake conditions (Ref.1.1). The FPS connections serve as shear links which absorb thedamaging earthquake motions and energies. When the earthquakeforces are below the threshold level, the building responds like aconventional structure. Once the threshold is exceeded thebuildings ductility response and energy absorption are controlledby the F PS connections.

The lateral restoring stiffness of the activated FPS connectionis:

k=(~1

where W is the supported weight and r is the length of the radiusof curvature of the concave surface. This is the same as thestiffness of a simple pendulum.

6

The weight proportional stiffness is a unique property of theF PS which greatly reduces the torsion response of a structure. Thecenter of lateral stiffness of the FPS connections coincides withthe center of mass. since the friction force is also proportionalto the supported weight, the center of rigidity of the connectionsacting as a group always coincides with the center of mass of thebuilding. This property makes the FPS connections particularlyeffective at 4ninimizing adverse torsional motions which wouldotherwise occur in asymmetrical buildings.

The FPS connections also serve as seismic isolators. Seismicisolation is achieved by shifting the response of the structure toa range where the lateral loads are reduced. The extent of theperiod shift is controlled by the radius of curvature of theconcave surface.

The natural period of vibration of a rigid mass supported on FPSconnections is determined from the pendulum equations and is:

where g is the acceleration of gravity. This is the sliding periodof the isolators. It is also the sliding or activated period for ashort and relatively stiff building.

The fact that the period is independent of the structure mass isanother unique property of the FPS which can have advantages incontrolling the response of a building. The desired structureperiod can be selected by simply choosing the radius of curvatureof the concave surface. The period does not change if thestructure weight changes or ~ different than assumed.

Another unique property of the FPS connections, as compared toother sliding supports, is the design of the articulated slider.The semi-spherical design of the slider results in uniform contactpressures between the slider and the concave surface for anycombination of lateral and vertical loads. This avoids edgegouging, and reduces high frequency stick-slip motions which occurwith other sliding support systems.

Seismic isolation is an emerging technology which has beenreceiving increasing attention in recent years by researchers anddesign professionals. The research presented herein builds onprevious research on the FPS concept by Zayas, Low and Mahin (Ref.1.1). During this previous research effort simple building modelswith FPS connections were tested on the University of Califo~-niaBerkeley shake table. The research presented also builds onextensive worldwide research 0", seismic isolation by other

7

investigators. These include work on sliding isolation systems byKelly (Ref. 1.4), Mostaghel (Ref. 1.5), and Constantinou (Ref.1. 6): and isolation systems with displacement restraints by Kelly(Ref. 1.7): and related research efforts by other investigators inthe U.S. (Refs. 1.8 and 1.9), Japan (Ref. 1.10), France (Ref.1.11), China (Ref. 1.12) and New Zealand.

One of the potential advantages offered by the FPS approach,compared to other available techniques, is the cost ofinsta llation. The connection size, strength, materials andversatility shO\.lld make the approach easier and less expensive toinstall. Potential reducti.ons in other construction costs couldcompletely compensate for the cost of installing the FPS, therebyachieving a cost-equivalent but more earthquake resistant building.

The need for cost effective ways to mitigate seismic hazards ofeXisting buildings has recently been recognized by many municipalgovernments in California. California Senate Bill 547 (Ref. 1.2),which was passed in 1986, requires that California cities adopt aprogram by Jan. 1, 1990, to reduce the hazards from unrein forcedmasonry buildings. Ci ties have begun to identify seismicallyhazardous buildings and to notify the owners. Many building ownershave resisted city efforts to require seismic upgrades because ofoccupant disturbance and cost issues.

Cost is also a primary factor affecting the implementation ofimproved seismic resisting methods in new buildings. Technologiesthat increase the earthquake resistance of new buildings but alsoincrease the construction costs have not been widely used. Thestructural Engineers Association of Northern California (SEAONC)has developed Tentative Seismic Isolation Design Requirements (Ref.1.3) which would permit rp.ductions in the cost of seismicallyisolated buildings. These or other equivalent criteria are likelyto be adopted into the Uniform Building Code by 1991, which shouldstimulate commercial applications of the seismic isolation approachfor buildings.

1.4 Phase I Research Objectives

The primary objectives of the Phase I research are to assess thetechnical feasibil i ty of using the FPS to improve the seismicperformance of new and existing buildings. The following technicaland engineering issues were investigated:

1. The feasibility of achieving significant increases in theearthquake resistance of new buildings without increa"'ing theconstruction costs.

2. The feasibility of applying the FPS to existing hazardousbuildings, thereby helping to protect these buildings from damageand collapse.

8

3. The interrelationships between the FPS desiqn parameters(friction coefficient. radius of curvature and displacement travel)and the building design parameters (period and strength) forearthquakes of different strengths.

4. The ability of FPS components to achieve dynamic coefficients offriction which are less than 5%, for relative slidinq velocitieswhich are representative of earthquake loadinqs.

5. Invest iqat ions of building parameters which control thestructure's torsional response, and to estimate the potential forreductions in damaging torsion motions tor aSYmmetrical buildingsusing the FPS approach.

The research work and principal findings are summarized below.

1.5 Summary of the Research and Findings

The feasibility of applying the FPS to new building constructionis examined in Chapter 2. Details which could be used for theinstallation of the FPS isolators are proposed. An examplebuilding was selected and redesigned using the FPS. ThE:! examplebuilding was a steel moment frame building, a common type ofconstruction used for commercial and institutional buildings. TheFPS offered the versatility of installation details which helpedachieve an isolated b'lilding which was cost equivalent to thenon-isolated building, and yet had significantly improvedearthquake performance. Analyses results indicated that seismicdamage during a severe earthquake would be reduced by 89%.

The feasibility of using the FPS to retrofit existing hazardousbuildings is examined in Chapter J. Considerations for applyingthe techn ique to the different types of existing hazardousbuildings are examined. The seismic performance of the FPS isassessed for a case study of a structurally weak building.

Preliminary design considerations were proposed on the selectionof FPS parameters suitable to weak buildings. When properlydesigned, the FPS was found to reduce ductility demand by 86%.Analyses results indicated that the expected performance would bebetter than conventional strengthening techniques used alone.However, it was unclear if the anticipated reductions in earthquakeforces and ductility demands alone, would be sufficient to preventcollapse of weak hazardous buildings. strengthening of portions ofthe building may also be required.

The feasibility assessments put forward in Chapters 2 and 3 arebased on studies of the technical properties and performance of theFPS which are presented in Chapters 4 to 8. Results from tests ofindividual FPS isolators art! reported in Chapter 4. Low frictionFPS assemblies were tested at high sliding velocities of up to 20inches/second. The dynamic coefficients of friction of the low

9

friction assemblies were found to be less than 5\. These dynamicfriction values were considered suitable for seismic retrofits orcost equivalent new building construction. The hysteretic loopswere observed to have an ideal bi-linear response with linearlateral stiffness throughout the displacement travel. The lateralstiffness was directly proportional to the vertical load, andexactly matched the theoretically predicted p':!ndulum stiffness.The loops were observed to be stable and non-degrading over manycycles of loading at varied vertical loads.

Analytical studies on the torsion response of f' f'S supportedstructures are presented in Chapter 5. Analytical moclels aredeveloped and found to compare well with experimental results. Theweight proportional stiffness properties r-f the FPS reduced thetorsion motions of asymmetrical structures. The weightproportional stiffness directly compensated for the effects of masseccentricities. The reductions in torsional motions of up to 80%were observed, as compared to linear elastic structures withequivalent eccentricities. Parameter studies indicated that thereduction of torsion motions would be substantial for a widevariety of building configurations. The unique ability of the FPSto reduce torsion motions is one of the Most attractive aspects ofthe approach.

In ve stigations on the effects of en ga g i ng the lateraldisplacement restraint of the F PS are presented in Chapter 6.Engagement of the lateral displacement restraints occurs when theseismic loading demand exceeds the lateral displacement capacity ofthe FPS. Analytical models were developed which could simulate theeffects of engaging the displacement restraints, and were found toagree well with experimental results. Time history analyses ofnonlinear models of the example building, inclUding thedisplacement restraint model, were used to assess the effects onthis type a f building. The displacement restraint was found tobreak up the modes of vibration, prevenUng resonant dynamicresponses from building up in anyone set of modes. Whendisplacement loading demand exceeded displacement capacity hy 50%,the F PS ret:lined 85% of its effectiveness in reducing ~ nelasticductility demand, 95% of its effectiveness in reducing structuralinelastic ener' y dissipation, and 80% of its effectiveness inreducing first story drift.

The inelastic responses of multistory structures supported onthe FPS isolators are presented in Chapter 7. Bi-linear elasticplastic models of the building structures are used to investigatethe relationship between the primary building and FPS properties,and to assess the relationship between yielding in the structureand response of the FPS. The anal}tical model of the examplebuilding presented in Chapter 2 is used as a basic case. Thestructure strengths and structure periods are varied to simulateother building cases. Nine different cases were considered, and

10

the responses with and without the FPS were compared. Theresponses for total displacements, structure ductility demand,structure story drift, and structure energy dissipation wereassessed and compared. The total lateral displacement at the rooflevel of the FPS isolated structures was found to be less than, orapproximately equal to, the roof displacements occurring in thesame structures without the FPS. The struct:.lrc ductility demand,structure story drift, and structure energy dissipation was foundto be always less than, and usually substantially less than, thoseoccurring in the same structure without the FPS. For structureswith periods less tha'1 0.86 sec., ductility and drift values in FPSsupported structures were found to be of similar magnitudes tothose occurring in equivalent non-isolated structures with fourtimes the strengths. Weak structures representative of existinghazardous buildings were given particular attention. Lengtheningthe period and lowering the friction coefficient of the FPS wasfound to reduce the yielding and ductility demands in thesestructures.

Whereas example building cases and responses to particularearthquake motions were studied in Chapters 2 through 7, a broaderand more theoretical approach was taken in Chapter 8. Equivalentsingle degree of freedom models representing elastic structuressupported on FPS isolators were developed. The equations of motionfor the equivalent system were formulated and used to investigatethe interrelationships of the FPS parameters of period and frictioncoefficient, to the period and strength of the structure, and thestrength of the ground motion. A systematic study of the responsesof the equivalent system to ten different ground motions wasundertaken, and the results for each of the individual motions, aswell as the means and coefficients of variation of the responsesare reported. The strength required to maintain an elasticresponse in the building was identified, as well as the base shearand displacement responses.

When the structural period was greater than 0.5 sec., the totaldisplacements of the equivalent systems relative to the ground werefound to be approximately equivalent to the total displacement inelastic structures without the FPS. The total drift displacementfor these cases could be estimated from displacement responsespectra. Once the total drift was known, the sliding displacementin th~ FPS and drift displacement in the structure could becalculated from the basic relationship between the FPS period,friction coefficient, and building period. The total drift wasrelatively independent of the selection of FPS period and frictioncoefficient. The selection of longer FPS periods and lowerfriction coefficients was found to reduce the drift displacement inthe structure, while increasing the drift displacement in the FPSby a comparable amount. For structures with periods less than 0.5sec., total displacements in the FPS supported structures wereqreater than those of the elastic structures without the FPS, andthe total displacement was affected by the selection of the FPSperiod and friction coefficient. The structural drifts and forcesin the FPS supported structures were always less than those in theelastic structures without the FPS.

11

1.6 Technical Feasibility Conclusions

The FPS was assessed to be a feasible and cost effectiveconstruction technique for improving the seismic resistance of newbuildings. The flexibility to select any isolator period makes theapproach suitable to a wide range of applications. The compactsize, high strength, and articulated joint of the FPS, permits aversatility of installation details which helps to achieveconstruction which is cost equivalent to non-isolated buildings,yet provides substantially improved seismic resistance. The costequivaleu\: example building investigated, designed with the FPS anda reduced ::;..;;

1.7 Potential Applications

Because of the inherent simplicity, versatility, stability anddurability of the FPS concept, it should become a major tool forthe seismically resistant design of buildings once the engineeringdesign and detailing issues are resolved, and simple yet reliabledesign procedures are daveloped. Reduction of the seismic hazardsin new and existing buildings is a problem of national importanceand one which is expected to become increasingly critical in theyears to come. The development of a reliable, economical andpractical FPS seismic isolation system is expect~~ to significantlyincrease seismic safety. In addition, the FPS will provide anef feet i ve method for mit igat ing damages to non-structuralcomponents and building contents, and for overcoming difficultstructural problems such as those as associated with inelastictorsional response. Illustrations of possible applications tobuildings are shown in Figs. 1.5 to 1.7.

In addition, the versatility of this system has yet to beexplored. Innovative i::'pplications of this system with highpotential include very irrc'J'l1!.ar or unusual structures (Fig. 1.8),or industrial: manufact\lring or chemical processing facilities,critical structu~es W.Lth stringent performance requirements,bridges, and suppor~ed equipment (Figs. 1.9 to 1.13). Developmentof a feasible and economiGal FPS solution for conventionalbuildings will naturally lead to these other applications.

1.8 Future Research Needs

The Phase I research established the ability of the FPS toimprove the overall seismic performance of new or existingbuildings. Consistent with the Phase I objectives and limitations,the investigations used simple analytical models to assess theresponse of limited building types (ie: moment resisting frames),to a f,w representative earthquake loadings.

Additional research and engineering evaluations are needed toanswer a substantial number of engineering questions on the effectsand performance of the FPS within buildings. Investigations areneeded to assess the performance of the FPS when applied todifferent building types, including braced frames, shear walls, andbearing wall structures. Since the lateral forces in the FPS areproportional to the supported weight, this changes thedistributions of forces and stresses within structures. Thus, thelocal effects on individual structural members and standardstructural details and design practice need to be addressed.

13

Experimental investigations of the responses of realisticmUltistory building models sUbjected to shake table tests areneeded. The dynamic responses of low friction FPS isolators undershake table tests need to be investigated. The effects of highermode responses need to be assessed, examining the differencesbetween FPS supported structures and conventional structures. Theinfluences of friction coefficients and FPS periods on the highermode responses need to be investigated, as well as design methodsand installation approaches which may influence and control highermode responses. Also, for buildings with sensitive equipment theoccurrence of high frequency vibrations and in-structure responseneeds to be investigated.

Shake table tests are also needed on the responses of weak andnon-ductile buildings supported on FPS isolators, with attention tothe interactions between the responses of the buildings and theFPS.

There is also the need to investigate the effects of varied anddiverse ground motions on all the response characteristics.Moreover, before the FPS technique can be accepted and appliedwithin the industry, there is the need to develop simple yetreliable design guidelines and procldures. A thoroughinvestigation of these important engineering consider",tions isrequired to insure the safety of the pUblic in the application ofthe proposed new construction methods. These investigations areplanned for the Phase II research program.

1.9 References - Chapter 1

1.1. Zayas V. A., Low S. 5., and Mahin S.A.,"The FPS EarthquakeResisting System: Experimental Report," Report No. UCB/EERC-87/01,Earthquake Engineering Research Center, University of California,Berkeley, June 1987.

1.2. state of California seismic Safety Commis:;ion, "Guidebook:To Identify and Mitigate Seismic Hazards in Buildings," Report No.SSC 87-03, December 1987.

1.3. structural Engineers Association of Northern California,"Tentative Seismic Isolation Design Requirements," September 1986.

1.4. Kelly J. M. and Chalhoub M. S., "Sliders and TensionControlled Reinforced Elastomeric Bearings Combined for EarthquakeIsolation," Earthquake Engineering Research Center, University ofcalifornia, Berkeley, 1988.

14

1. 5. Mostaghel N. and Tanbakuchi J., "Response of SlidingStructures to Earthquake Support Motion, II Earthquake Engineeringand Structural Dynamics, Vol. 11, 729-748, 1983.

1.6. Constantinou M.C. and Tadjbakhsh 1. G., "The OptimumDesign of a Base Isolation System with Frictior.al Elements,"Earthquake Engineering and Structural Dynamics, Vol. 12, 203-214,1984.

1.7. Kelly J. M., Griffith M., and Aiken I., "A TensionRestraint for Upl i ft Control, II Report No. UCB/EERC-B7 /03,Earthquake Engineering Research Center, University of California,Berkeley, 1987.

1.8. Tarics A. G., Kelly J. M., Way D., "The SeismicRehabil itation of Existing Buildings Using Base Isolation, II JointReport Reid and Tarics Associates and Earthquake EngineeringResearch center, University of California, Berkeley,

1.9. Mayes R. L., Jones L. R., Kelly T. E., and Button M. R.,"Design Guidelines for Base-Isolated Buildings with EnergyDissipators," Earthquake Spectra, Vol. 1, No.1, November 1984.

1.10. Kitagawa Y., "Base Isolated Building Structures in Japan,"International Organization for the Development of Concrete,Prestressing and Related Materials and Techniques, New Zealand,August 1987.

1.11. Jolivet F. and Richli M., "Aseismic Foundation System forNuclear Power Stations," Departements Genie Civil, Spie-Batignoles,France, 1977.

1.12. Li L., "Advances in Base Isolation in China," Presented atthe 3rd International Conference on Soil Dynamics and EarthquakeEngineering., Princeton University, USA, June 1987.

15

....a-

BEARING MATERIAL ARTICULATED FRICTION SLIDER

SPHERICAL CONCAVE SURFACE

Fig. 1.1 FPS Isolator Section

M

PENDULUM MOTION

M

~NG PENDULUM MDTIQM

EQUATIONS. PERIOD T=2IrJr/9

STIFFNESS k = \J/r

W WEIGHT,. =RADIUS OF'CURVATURE

Fig. 1.2 Basic Principles

17

-0.14

o I I:::> .< I :::>' .< I-0,12

In0,15. C)

0,16

o

oI I I I I

0.30 i T = 3 SEC. u = 0.10i I

0,30 i T = 3 SEC. u = 0.05I !

I II I

-0,3~ io -3.9 0 5.2 10105.0o-0.18

-0.20

o I ( I / ,

0.30 I T = 2 SEC. u = 0.10i

O.Z3 I

0.30 i T = 2 SEC. u = 0.050.24 i ~

,l-----:!-:,;--:~'~'1-030! -3.3 0. -10

"Vl....L:1.......

Q

BUILDING

FRAME

FPS CONNECTIDN

BASe: PLATE

---- LO\/ER COLUMN

CENTERED POSITION

. BUILDINGFRAME

DISPLACED POSITION

Fig. 1.4 FPS Operation

19

:"""1r=::I. --11-H

.....

,..,I

,III I ,..,

,..II1.1

r-nV)

Cl.~

II 1.1I I '-In I

~

~\ I

20

~~

,'>,~

- ~

1--11--1

~.

-~

......~

-- P

1-= ~-~

~'-- b

l

21

=CO8

E5o

00

tl()c:::.-

"'0-.-

-n.-

Il-

-~ ~ 1 - ~

r

"Ih

~ 1

~

-

r- ~

~ I ~ ~1 n

L-

rr = ~ ~

I ,_ I

=~

I

~

I I I I I I I I I.

I-Fl :::J

boOc::.a-'SCO

11I

22

il1

rl~t1=ti-~~=~~~tJLl ~=11_ ~rmttH=UlJli" i=I~=__..I ....:lIb...IIIL.J,LJIIL_JL,JLJ_IJlJLl==- ,::=. ~I I I II I I ;-....J~~Ii-' -4.-

!::

23

(1)be

""0.-....CQ

z ~[J- ~....4: ..=> bew .-...J ::t:w

Q:l

..-4

bil~

F"PS CONNECTION

Fig. 1.10 Water Tank

~J-L,I ,

\ I\ I: L-- $cutot....l I\ f\ ,I ,I I /P...tDI

,... l __ J_L__ -,I I _"'S IIDIDI ..,

r-~ ~., Plollor...

);,;;;;;;;; ;~

Fig. 1.11 Museum Sculpture

24

TAPE DRIVE

DISK DRIVE

~ FPS CDNNECTION

Fig. 1.12 Computer Equipment

PRINTER

SUPPORTING PLATFORM

FLOOR

CIRCU)T BREAKERS

SUPPORTING PLATFORM

FPS CONNECTIDN

FOOTING

Fig. 1.13 Electrical Circuit Breakers

25

CHAPTER 2

FEASIBILITY OF THE FPS FOR NEW BUILDING CONSTRUCTION

by

victor Zayas

and

stanley Low

SECTIONS

2.1 Installation Details

2.2 Example Building

2.3 FPS Isolation Approach

2.4 Cost Equivalent Design

2.5 Comparison of Building Performance With and Without the FPS

2.6 Conclusions

2.7 References

26

2.1 Installation Details

The application of the FPS technique to new buildings requiresthe development of practical installation details which arecompatible with existing construction practices. Details whichcould be used for the installation of FPS isolators within buildingframes are shown in Figs. 2.1 through 2.6.

Example details for installing the FPS connections within abuilding frame are ill ustrated in Fig. 2.1. Installation of theisolators between the foundation and first floor is the approachwhich has been used in completed installations of rubber bearingseismic isolators. This approach offers the advantage ofsimpliL ed architectural and structural details to accommodate theseismic movements. The steel girders directly above the isolatorscan resist the P-Delta moments which occur in the rubber isolatorsbecause of the eccentricity of the gravity loads during earthquakemovements. However, for buildings which do not otherwise require acrawl space, the addition of this partial story can be expensive.

For many buildings, it would appear that installation of seismicisolators at the tops or bottoms of the first story columns couldbe a more cost effective way of incorporating seismic isolation.The design and size of the FPS isolators appears to be well suitedfor this installation approach. The steel components of the FPSfacilitates the fabrication of high strength, relatively compactisolators, which could readily fit at the top or bottoms ofcolumns. The steel cylinder which encloses the articulated slider,provides an inherently redundant mechanism for carrying lateral andvertical loads. When located at the top of a column, the FPSconnection can be installed with the concave surface facingdownward. In this orientation the articulated slider remains atthe centerline of the column during earthquake movements. Thisavoids P-Delta moments in the column:i caused by eccentric gravityloads from seismic movements of the building. When installed atthe bottom of a column with the concave surface facing upward, thegravity loads rem,;tin concentric on the column. A disadvantage ofinstalling the isolators at the tops or bottoms of columns, ascompared to the crawl space approach, is that additional seismicgap details for architectural components which cross the seismicgap are required.

A possible combination of the bottoms of columns approach, andthe between the foundation and first floor approach. is illustratedin Fig. 2.2. Placing the FPS isolators at the bottoms of the firststory columns, and using a concrete flat slab for the first floordirectly above the isolators, can reduce the structural costs ofthe crawl space approach, and reduce the architectural detail costsassociated with the bottoms of columns approach.

In some installation applications it may be desirable to includean uplift restraint detail as part of the FPS connection. Anillustration of a possible uplift restraint detail is shown in Fig.2.3. The tension' rods would be able to carry tension loads and

27

limit uplift displacements, while permitting lateral seismicmovements within the isolators.

Details which could be used to accommodate seismic movements forelevators and stairs which cross the seismic gap are shown in Fig.2.4.

The versatility of the FPS isolators to accommodate differentinstallation details, and any desired isolator period, canfacilitate the application of seismic isolation to buildings.Illustrations of possible building applications are shown in Figs.2.5 and 2.6. Because the FPS can absorb torsional masseccentricities as well as the seismic deformations, theasymmetrical building shown in Fig. 2.6 could achieve seismicresista nee capabilities which exceed those of a symmetricalbuilding which relies only on ductility detailing.

2.2 Example Building

In order to investigate the technical feasibility of applyingthe F PS technique to new buildings, an example building wasselected and redesigned using the FPS approach. The examplebuilding was used to illustrate the possible design approaches andthe potential benefits. The comparative earthquake performancesand construction costs are evaluated.

Earthquake Protection Systems asked Cygna Consulting Engineersof San Francisco to provide an example case of an actual buildingwhich had been designed according to the conventional building codeapproach. cygna offered an example case of a three story steelmoment frame building which the owner considered critical forbusiness operations. Cygna engineers had recently designed thisimportant "Operations Building" according to the 1985 UniformBuilding Code (UBC) earthquake criteria. Since the ability tocontinue operations of the building after an earthquake wasimportant, at the request of the owner, the building had beendesigned as an "Essential Facility" according to UBC procedures.An importance factor I of 1.5 was used, resulting in a design baseshear coefficient of 0.14. The layout and details of thebuilding's structural frame as designed by Cygna are shown in Figs.2.7 to 2.10. Additional details on the building's design areprovided in Appendix A- Sections 1 and 5.

As is common in commercial buildings of this type, thearchitectural design included a first story which was taller thanthe upper stories. The building also had relatively long girderspans of 28 ft. to 32 ft. As usually occurs in such buildings, thetall first story was more flexible than the upper stories. The UBCsetsmic design was controlled by drift, not strength, as isgenerally the case for steel moment frames. In order to controlthe first story drift, the design engineer was forced to usespecial square tubular steel columns, and imbed these columns intothe foundation grade beams to create a fixed-base connection. Thespecial tubular columns and foundation connections were high cost

28

structural items in the non-isolated design. Tt'.e resulting seismicdrifts at a base shear loading of O.14W, were the maximum permittedby code.

The elastic and plastic strengths of the building werecalculated based on the actual member sizes, and includedstructural redundancy effects (Appendix A - Section 5). Theelastic strength was computed to be O.31W, and the ultimate plasticstrength was O.45W. These calculated strengths are consistent withtest results of the building frames, which typically show ultimatelateral strengths of 3 to 5 times the design loads.

The building site was located approximately 8 miles from the SanAndreas Fault. Fig. 2.11 shows the U.S.G.S. average responsespectra for magnitude 5, G, and 7 earthquakes for a stiff soil siteat a distance of 8 miles from the fault (Ref. 2.1). The UBCrequired design strength for a moment frame, designed with animportance factor of 1.5 is also shown. The average magnitude 7earthquake was noted to be 5 to 10 times stronger than the requiredUBC design strength.

A magnitude 7 or stronger earthquake was considered to be arealistic event to expect during the life of the building. Thesteel frame was considered to have sufficient strength andductility capacity to withstand the seismic de formations of amagnitude 7 earthquake with light to moderate structural damage.However, large drift deformations, SUbstantially exceeding codeallowables, and high accelerations would be expected to causesignificant damage to the non-structural components and operationalequipment items within the building. Such damage could render thebuilding incapable of functioning after a severe earthquake. Forthese reasons, the Cygna engineers believed the application of theF PS technique would be of particular value for this type ofbuilding. The objective was to determine the extent to which theF PS isolators could reduce the estimated damage caused by largedrifts occurring during severe seismic events. Cygna engineersworked together with Earthquake Protection System engineers todevise the isolation approach, and review the comparativeperformance and costs.

2.3 FPS Isolation Approach

Different schemes for installing the FPS isolators 1n theexample building were considered. The advantages and disadvantagesof the different schemes were considered to be as follows:

1.. Tops of Lower Level Columns. Advantages: Least impact onexisting design concepti and most cost effective. Disadvantage:Seismic gap details required for stairs and architecturalcomponents Which cross the FPS level.

2. Bottoms of Lower Level Columns. Advantages: Small impact onexisting design concepti and the design and details are similar toa standard base plate connection. Disadvantages: Special details

29

would be required for stairs and and architectural components toaccommodate the seismic gap at FPS level; and the pinned columnbase would require strong and stiff tubular columns with equalstiffness in both the x and y axes to control the deflections ofthe lower story.

3. Between Foundation and First Floor. Advantages: Simple detailsfor stairs and and architectural components~ and the accelerationsof the first story floor would also be reduced. Disadvantages:Configuration and expensive construction changes would be requiredto add the isolation story~ including deeper excavation for thefoundation and the added structural floor framing.

After comparing the alternatives, the design with the FPS ontop of the lower level columns was considered the preferredisolation scheme for this: building. This approach was the mostcost effective, and also minimized the need for design andconfiguration changes. The structural frame and isolator detailsusing the FPS at tops of the lower level columns are shown in Figs.2.12 to 2.15. Architectural details to accommodate the seismicmovements are shown in Figs. 2.16 to 2.19.

Based on dynamic analyses results (Appendix A, sections 5 to 8),a sliding period of 2.25 sees., a dynamic friction coefficient of10%, and a lateral sliding capacity of 6.5 inches were selected forthe design of the FPS isolators. The dynamic analyses predicted anisolator sliding displacement of 4.7 inches for the magnitude 7earthquake loading. The required lateral displacement capacityaccording to section B of the Tentative Seismic Isolation DesignRequirements (Ref. 2.2) was computed to be 4 inches (Appendix A,Section 3). The isolator design exceeded the minimum requirementsof the SEAOC guidelines because the magnitude 7 earthquake loadingexceeded the seismic loading used to develop the minimum criteriaof the SEAOC guidelines.

Placement of the F PS isolators at the tops of the lower levelcolumns resulted in a hinge connection at this joint. The columnsbelow the FPS were changed from special steel columns to reinforcedconcrete columns which were integral with the foundation. Thehigher section modulus of the concrete columns reduces the drift atthe first level, and results in equal column stiffness in the twoloading directions. Since the F PS isolators would provide thenecessary deformation and ductility capacity, the lower ductilitycapacity of the concrete columns versus the E'+:eel columns was notconsidered a disadvantage for the isolated design.

Two different structural design criteria for the isolatedbuilding were examined. In the first, the strength and stiffness ofthe structural frame were assumed to be the same as the originalnon-isolated design ("Ful: Strength Isolated Designll). In thesecond, the sizes of the :..cructural members were reduced to thedegree that the structural savings could offset the cost of addingthe F PS isolators (II Cost E quivalent Isolated Design"). Theredesigned structural frame of the cost equiValent design is shownin Figs. 2.12 to 2.14, and is discussed below. The full strength

30

isolated design was assumed to have the same structuralconfiguration as the cost equivalent design, but with the strenqthand stiffness of the structural members equal to those of theoriginal non-isolated design.

2.4 cost EquiValent Design

The "Cost Equivalent." isolated building design was used todetermine the feasibility of achieving increased earthquakeresistance capacity without increasing construction costs. Thedesign base shear coefficient was 0.067, as compared to 0.14 forthe original non-isolated design. This SO\ reduction in the designbase shear permitted enough savings in the structural frame tooffset the added structural and architectural costs associated withthe isolated design. The base shear coefficient of 0.067 is thecalculated design base shear required by the 1985 UBC if theisolator sliding period of 2.25 seCF. is used as the buildingperiod in the UBC formulas. The cost equivalent: isolated designwas also checked against the "Tentative Seismic Isolation DesignRequirements" of SEAONC [Ref. 2.2], and also complied with theseguidelines. Additional details on the checks with the UBC andSEAOC design criteria are provided in Appendix A.

As was the case for the non-isolated design, the seismic designof the cost equivalent building was controlled by interstorydrift. The resulting seismic drifts at a base shear loading of0.067 were the maximum permitted by code. The elastic and plasticstrengths of the cost equivalent design were calculated based onthe reduced member sizes, inclUding structural redundancy effects.The elastic strength was computed to be O.14W, and the ultimateplastic strength O.31W (Appendix A - Section 5).

The savings in the structural steel and foundation detailspermitted by use of the isolation approach was estimated to offsetthe cost of the 32 FPS isolators and the architectural andstructural details required for the seismic gap. A summary of theestimated structural savings and isolation costs is given in Table2.1. Thus, the estimated construction cost of the isolated designwas approximately equal to the cost of the non-isolated design.The construction cost of the full strength isolated design wasestimated to be approximately 1.8% greater th':!n the non-isolateddesign. Additional details on the redesign detllils and costestimates are provided in Appendix A, Section 4.

2.5 Comparison of Building Performance With and Without the FPS

Time history dynamic analyses were used to compare theearthquake performance of the isolated designs to the originaldesign. To do this comparison, scaled earthquake loadingsrepresenting magnitUde 5, 6 and 7 earthquakes were used in thenon-linear analysis program Dynin (Ref. 2.3).

The Dynin models of the non-isolated and isolated structures

31

are shown in Appendix A, Section 5. A stick model was used withnodal masses representing the second floor, third floor, and roof.The structures were analyzed as elastic upper structures onnon-linear FPS isolators. The stiffness of each building level wasderived from the section properties of the correspondingstructure. The nonlinear properties of the FPS isolators wereincluded in the modeling of the first level. Analyses results forseismic drifts occurrinq within the structural frame were used toestimate damage to the building.

Additional analyses of this building, including inelasticmodeling of the upper structure, are presented in Chapters 6 and7. As discussed in Chapters 7 and 8, for a building in this periodranqe, the total seismic drifts which occur are approximately thesame whether the building is isolated or not, and whether the upperstructure remains elastic or is permitted to yield. The primaryeffect of the isolators is that a reduced percentage of the totaldrift oc..:curs within the structural frame. Similarly with aninelastic model of the upper structure, it is primarily thedistribution of drifts among stories which is affected by theinelastic response, not the total drift. The elastic upperstructure model was, therefore, considered sufficient for designdevelopment, and for non-structural damage estimates Which arebased on total drift. The non-isolated building would not becapable of achieving the story shears predicted by the elasticupper structure model. However, the predicted elastic shears area measure of the earthquake loading demand on the structuralframe. Analyses results for the ductility demand, energyabsorption demand, and inelastic structural drifts using inelasticstructural models of the upper structure are presented in Chapters6 and 7. The inelastic analyses results presented in Chapters 6and 7 confirm the overall results and conclusions drawn from theanalyses of the elastic models presented in this Chapter~

The input earthquake loadings were scaled to approximatelyrepresent the average sPeCtra for magnitude 5, 6, and 7 earthquakesat a distance of 8 miles from the site. Values obtained for baseshear, story shears and story drifts were used to compare theresponse of the isolated and non-isolated buildings. Comparisonsof the base shear responses for the simulations of the threedifferent magnitude events are shown in Fig. 2.20.

The earthquake loading used to represent a magnitude 5earthquake was the El Centro earthquake scaled to a PGA of 0.159and the time scale compressed to one-half. The response spectraplots and the tabulated analytical results are given in Appendix A,Section 6. This low intensity earthquake shaking is of comparablestrength to the URC design strenqth. The resulting base shear was0.08, with or without the FPS. Since the FPS isolators had athreshold friction force of 0.10, the FPS isolators did not slide,and the building- response was the same with or without the FPS .Analysis results indicated that story drifts for the isolated andnon-isolated designs would be less than the UBC code allowabledrifts, and no building damage was anticipated for any of thedesigns at this strenqth event.

32

The earthquake loading used to represent a magnitude 6 eventwas the E1 Centro earthquake scaled to a PeA of 0.34g and run atfull time scale. The response spectra plots and the tabulatedanalytical results are shown in Appendix A, Section 7. The resultsshow that in the isolated design structures, the base shear wasreduced from 0.54g to 0.14g, and the 1st story drift was reducedfrom 2.28" to 0.54". This is a 74\ reduction in base shear forceand story drift. The responses of the cost equivalent and fullstrength isolated designs were approximately the same. The FPStravel displacement was calculated to be about 1.7 inches.

The earthquake loading used to represent a magnitude 7 event wasthe El Centro earthquake scaled to a PGA of 0.70g and run at fulltime scale. The response spectra plots and the tabulatedanalytical results are shown in Appendix A, Section 8. The resultsshow that in the isolated structures, the base shear was reducedfrom l.11g to 0.21g and the 1st story drift was reduced from 4.64"to 0.74". This is a au reduction in base shear force and storydrift. The FPS displacement travel was calculated to be 4.70inC'hes.

A summary of the analytical results for both the isolated andthe non-isolated structures is shown in Figs. 2.20 ~~ ~.~2. Fig2.20 shows the large reductions in base shear demand that werepredicted for the isolated structures for the magnitude 6 and 7earthquakes. Figs. 2.21 and 2.22 show the reductions in storyshears and dr~fts within the building for a magnitude 7earthquake. These predicted reductions in story shears and driftswould be expected to reduce building damage during the magnitude 7event.

Fig. 2.23 shows damage estimates for the magnitude 6 and 7earthquakes. Damage estimates for the non-isolated building weremade using the ATC-13 "Earthquake Damage Evaluation Data ForCalifornia", as estimated for low rise steel moment framebuildings. Damage estimates for the isolated designs werecalculated by using the damage estimates for the non-isolateddesign, but for reduced earthquake strengths which woUld result inequivalent structural drifts as those of the isolated designs. Thecost of construction of the operations Building was approximately$6.5 million. Based on the comparative drifts for the reducedstrength isolated design and the full strength non-isolated design,it was estimated that the FPS would reduce building damage for themagnitude 7 event from $1,300,000 to $138,450, a reduction ofapproxir.lately 89t. Additional details and calculations of thebuilding damage estimates are provided in Appendix A, Section 9.

The predicted reductions in story shears and accelerations(Figs. 2.21 and 2.22) would be expected to also reduce damage tothe equipment and other building contents. Hl.wever, within thescope of this study it was considered that there were no reliableways of making preliminary damage estimates for the buildingcontents.

33

2.6 Conclusions

The versatility of installation details offered by the FPSisolators could help achieve isolated buildings which are costequivalent to non-isolated buildings, and yet have significantlyimproved earthquake performance.

The analysis results for the example building indicated thatthe cost equivalent isolated design would provide better seismicperformance than the full strength non-isolated design. Buildingdamage was reduced by an estimated 89%.

In summary, it was feasible and practical to incorporate theFPS into the design of a new building. A cost equivalent designcould be achieved, and the seismic performance was SUbstantiallyimproved as compared to the non-isolated design. However, furtherevaluations are required to evaluate the application of thetechnique to different types of structures, and in buildingsdesigned for normal occupancy as opposed to essential facilities.

2.7 References -- Chapter 2

2.1 Joyner, W. B., and Fumal, T. E., "Predictive Mapping ofEarthquake Ground Motion," u.s. Geological Survey ProfessionalPaper 1360, U.S. Geological Survey, Menlo Park, California 94025,1985.

2.2 "Tentative Seismic I~olation Design Requirements," BaseIsolation Subcommittee of the Seismology Committee, structuralEngineers Association of Northern California, September 1986.

2.3 Khatib, 1. and Mahin, S., "DYNIN Interactive computerProgram for the Nonlinear Dynamic Analysis of Buildings," SEMMComputer Program, University of California, Berkeley, 1987.

34

Table 2.1 Summary of Savings and Costs

Structural Savings

1. Reductions in structural framedue to reduced lateral loads.

2. Reductions in foundation costs.

Added Structural Costs

1. 32 FPS Seismic Isolators

2. Bracing for exterior panelat 1st. level.

Net Structural Savings

Added Architectural Costs

Contingencies & Miscellaneous

Net Savings and Costs

35

$82,000

$3.,000

$116,000

$6.,000

$15,000

$79,000

$37,000

$11,000

$20,000

$0

L.)

'"

UiC.... I .. F1JD11NIi

~~-

ca.&IlN

snn CiIRIEIt

F'PS ClNlttTDl

~CIIUMI

-1--1-'

BElWEEN ~UNDAll0N AND FIRST FLOOR AT TOP OF COLUMN

Fig. 2.1 FPS Installation Details

AT BOTTOM OF COLUMN

==ll=~=~~~~F:!:PS CONNECTION

COLUMN

CONCRETE SLABGAP 1SLOPE-

SLAB

CONCRETE FOOTING

Fi.g. 2.2 Installation With Flat Slab Floor

CDLUMN

rps CONNECTION

1 - TENSION ROD

Fig. 2.3 FPS Connection With uplift Rest'aint

37

,.... ~

$ECOID qvEL

7i}'y'~

(:;~tuYtR

V RIGID flWtE

I

h

ELEVATIItCAl

]

~r~J~ ~~:

1,.0)

00

~",..." ~

r:,-... ,. .. " ..... '. .. ~ . 1, ....>3;

ELEVATOR DETAIL STAIR DETAIL

Fig. 2.4 Elevator and Stair Details

r---.~

~

l-

I'-

..,r-

~

V)Q..I..

/-- I J ....

-

~\

39

,,~......--

~~

~

w.....-

I...-

,~

~

f

-....

~~

:=: ~

--~ ~

--,

!=:II...."~

40

.-...--

i)l

0- -- - - -

0- -

bOc:.-

:9.-~

0:enc:oez~

~c..o....oc:o.~

~~

=='o

::....

"BoW~-oen-tc:Z

II IIII IIII II" II" IIII IIII II

I II II

========W======$========I II III II III II III II III II III II II I

========W======$========I II II II II II JI 1/ II II II II II II ~DII II II ~,. II II II II " II II II I

========~======~========I II II II II " II II II II 1/ " II II II J

========~======~========I II II II /I /I II II II II II II II II 1/ II II II II II II II II II I

========~======~========I II I I II II II II /I II II II II II /I II II II II I

========W=======$========I II " II II II II II II II II II IJ II II IJ II II II II II I

II II

l------~---------142

-----1 1-----

II

1----1 1----1

i iI cgjO------M .M....-----...0

.llllOal ,ILIIlIl1l

01..-- - .....11IIllDl

Ec:l...~

"i"'@'0tfJ-II::oZ

-----1----1-----

i-----1----I-----I

.1......n....a..------tH~---- ...H;...-----....a

43

""/

U.S.G.S. Average Response Spectra ForMagnitude 5, 6, and 7 Earthquake atDistance of 8 Miles From the Fault

1.50 ,.-------------------------.)

3.00

usc0= 1.5, K=0.67, S= 1.5)

1.00 2.00PERIOD (sec.)

o.ooj . =:=:=;-1--'-:: - . I0.00 I I I i I

M=5

zo~a:::w-.J

~ 0.50

~

_1.00!'J01'-''""VI'

Fig.2.11 Site Spectra

,......' .' .4:'

~.--.. 4, -.,1""1

III )111.1'1 ~1-7.

'4

t.1 r:- ~

*~

~ 'lc

~ ~ .'I-J, :J. ~ -.Q

-. P3'- ..... - Q .o(J". . .~

.... .....IW'I~'1 --.

~

,i INT t~ .a

~).. It

~ ~

26- SQ. REINFORCED EX'TERIORCONCRETE COlUMN

.po....,

II II " II " "II II II II " IIII II " II II IIII II II II I I III I I I I I I I [l[VATIIt I I I I I

I II II II II 0 II II II II II " II II II I

I --------~a"------~DII--------~all-----~a"--------~a" ~all I--------~------~--------~-----~r--------~------tnr--------

I \I II II II II II II II II II II II II I

I \I \I \I ~I JO- SQ. '~FORCED INlrR~ II I I I I I I I J CONCRE~ ICOLUMN I I I

I --------~all------8a

ll--------.8.

0

11II ~DiI ------~D'I I

--------~------~--------~----- --------~------~--------I II I I I I II II " II II II II II " II II II II 11 II II II II II II " II " II II II II II II \I 1\ II II II II II " II II II II \I II II II I

II II II \I \I II

Fig. 2.13 FPS Isolated Foundation Plan

..Lza "aM ~

II I It It

iI .~It p

I IH' ...... ...

.....

.... H H ...J "aM ..LIII tlPLIIiI II I I

,!'I I I I~

IHI ..... IH' ..."Lza "Lza ~

48

TOP PLATE

GASKET

I 0 0 rJ

I VI 0 0

ENCLOSINGCYLINDER

BOTTOM PLATE

FPS ISOLATOR ELEVATION

......------ 22.0 -----'-1

r 20.0' DIA.~II = \.00' I16.0'

J_ARTICULATEDSLIDER

r---GASKET

L6.0' DIA.FPS ISOLATOR SECTION

Fig. 2.15 FPS Isolator Details

"49

VIa

CIDUMl

f1ISCClMNEC1IlIN

ClClNCX'f[~

Slm'IIEAM

COl..lllINCONCIlE1E !US OVERMETAL DEaetlG

FPSCQtiNEClION

lASE PLAlE

CQNalElE 00l..LWl

MOIltC'MW..-.oow

NIQf1EC1UIW.PANO.

CONNECllON AT EX1[RIOR COlUMN CONNECTION AT INlCRlOR COLUMN

Fig. 2.16 FPS Connection Details

GRAll[ BEAM

CONNECTION AT EXlERIOR PANEL

V'...

STEEL BEAM

HUNG CEILING

STEEL CHANNELFOR BRACINGPANEL

ARCHITECTURALGLASS

~EXPANSION GAP// MATERIAL

~'" GAP

ARCHITECTURALPANEL

Fig. 2.17 Architectuml Facial Connection

d~~ .J

52

VAv

~ __ - HYDRAULIC CYLINDER

Fig. 2.19 Elevator Detail

53

,.....~-()IL:1;':;mc0~

0GL(J)

VIG~00m

Operations Building1.5 .....,---------------

1.4

1.3

1.2

1.1

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1o y //1'0;2>J7t?'fiII 1///1 ,,'j >'v///fl 1/ / / I ), ? ':,VUUL!

Magnitude 5

IZZJ Non-Isolated

Magnitude 6 Magnitude 7

[SSJ Full Strength Isolated ~ Cost Equivalent Isolated

Fig. 2.20 Comparisons of Base Shears

2nd. Story .3rd. Story

rs:sJ FulJ Strength Isolated IZZJ Cost Equivalent Isolated

IMagnitude 7 Earthquake

2.6

2.4

2.2

2

1.8

- 1.6~'""L.. 1.4-0.t:.

(,I) 1.2~

V' 0 1V1 ~(/l

0.8

0.6

0.4

0.2

0

1st. Story

IZZJ Non-Isolated

Fig. 2.21 Comparisons of Story Shears

Magnitude 7 Earthquake5 ' .--------------------

4

1

o L-.',- I '\. ). '\.VO'l/A 1/ / / I ... f' '\V////4 V / / I '\ f' '>1/////4

,...L. .30c

V......L0

VI ~ 2Q\ 0..en

1st. Story 2nd. Story 3rd. Story

IZZ1 Non-Isolated lSSJ Full Strength Isolated ~ Cost Equivalent Isolated

Fig. 2.22 Comparisons of Structural Frame Drifts

Operations Building....----------~===-==~-----------I'2 i1.9

1.8

1.7

1.6

1.5 "1i

1.4

1.3

1.2

1.1

1

0.90.80.70.6

0.50.4

0.3

0.2

O.~ j'l/L//] , '.' , W/##4 r////l'\."0-"V//PL/A

OtoEocOtc:0'5m'0

!oE~w

-c~~..'oj

UI-..l

Magnitude 6 Magnitude 7

[ZZJ Non-Isolated lSSJ Full Strength Isolated faZJ Cost Equivalent Isolated

Fig. 2.23 Comparisons of Building Damage

CHAPTER 3

FEASIBILITY OF THE FPS FOR SEISMIC UPGRADING

OF EXISTING BUILDINGS

by

Stephen Mahin

and

Victor Zayas

SECTIONS

3.1 Introduction

3.2 Some Basic Considerations for Seismic upgrading

3.3 Application of FPS Seismic Isolation Techniques to Existing

Structures

3.4 Installation Details

3.5 A Simple Case Study

3.6 Summary and Conclusions

3.7 References

58

3.1 Introduction

One of the most serious and challenging problems to be facedby earthquake engineers in the coming decades is that posed byexisting seismically hazardous buildings. Based on research andlessons learned from recent damaging earthquakes, building codeprov islons applicable to new construction have been substantiallyimproved over the past thirty years. The designer of new buildingsalso has at his/her disposal vastly improved analytical tools, morereliable construction practices and materials, and new seismicisolation techniques, such as the FPS connection, to help mitigateseismic hazards. As a result, it would be expected that, While thedesign of new structures still involves complex technical issuesand professional challenges, in general these structures willperform significantly better during severe seismic excitations thanthose constructed in earlier eras.

However, many existing buildings will also perform well, asdemonstrated by past earthquakes, due to the skill or conservatismof the designer, selection of structural systems with provenseismic resistance, and the use of inherently ductile details andmaterials. ThUS, identification of seismically deficientstructures must go beyond mere evaluation of compliance with basiccode requirements, and consider their likely seismic performance. Avariety of professional reports detailing methods for screeningexisting buildings and for evaluating their seismic vulnerabilityhave been published (e.g., see Refs. 3.1 and 3.2). Issues such asinadequate strength, limited ductility, excessive drifts, impropertying together of structural as well as nonstructural components,excessive seismic demands due to irregular structural features andresonance with local soil conditions have been identified as someof the possible deficiencies of existing buildings.

Once a building has been identified as being seismicallydeficient, additional considerations to those used for theconstruction of new buildings must be taken into account toachieve economical and effective upgrading procedures. Inparticular, issues related to cost, retention of functionalcapabilities (or avoidance of disruption of services),preservation of special architectural or historic features, and soon have had an increasingly important role in determining the typeof upgrade strategy to implemented. Augmenting the seismicresistance of old structures also raises new and special designproblems.

The feasibility of using FPS connections in upgrading theseismic performance of existing buildings was studied. Technicalconsiderations were explored along with some practical issuesrelated to installation of FPS isolators in various types ofexisting building:,;. Finally, a simple example structure wasanalyzed to assess the effectiveness of FPS isolators in improvingseismic resistance. Additional information related to this topicis presented in Chapters 7 and 8.

59

3.2 So.. Basic Considerations for Seismic Upqradinq

As indicated in Ref. 3.3, the desig'ner may choose between twobasic approaches when considerinq options for upqradinq theseismic performance of existing structures. The first, and morecommon approach, is to increase the resistance of the structure tothe demands imposed by the seismic excitation. Reference 3.3contains a detailed description of these strategies. They consistof qeneral items such as increasing' the strength or ductility ofthe existinq structure by the addition of new elements orstructural systems, or by modification of existing' elements. Theuncertainty with which the intensity of future earthquakes can bepredicted raises questions about methods that rely solely onstrengthening' as a means of improvinq seismic performance.

The second basic approach is to reduce the seismic demands onthe structure. Techniques for accomplishing this include reducingthe mass of the structure (by removal of heavy cladding orequipment, and by reducing the overall heiqht of the structure),and by seismic iSOlation ot the structure. In general, bothstrength increases and load demand reductions can be used incombination. For example, a seismicallY isolated structure mayalso be strengthened.

The selection of an appropriate upgrade strategy depends onthe type of structural system and materials used, the nature ofthe seismic deficiencies detected, the dynamic characteristics ofthe structure, the expected ground motion (including the effectsof local soil conditions), the availability of local materials andqualified labor, in addition to economic, architectural andfunctional constraints. Many of the requirements for upgradingexisting buildings differ in fundamental ways from those consideredfor new construction.

While economic considerations are always important in design,existing older buildings are often only marginally profitable sothat the economic ramifications of the seismic upgrading approachto be selected must be carefully considered. If the re~uirem could substantially interfere with themovement of people already working within the structure. Forexample, addition of nUmerous shear walls may be a technically andeconomically feasible solution in some cases, but it would violatethe existing functional utility of most structures. Similarly,most owners would prefer to keep the building occupied during theconstruction process. Because of this, it is desirable to selectan upgrading strategy that can be implemented successively in small

60

localized zones in the structure with minimum disruption ofservices to the existing occupants.

Also, many existing buildings are considered historic andmajor changes cannot be made to their exterior or interiorappearance without destroying important aspects of a community'sheritage. Thus, the retrofitting technique selected for thesestructures should be as unobtrusive as possible.

3.3 Application of FPS Seismic Isolation Techniques to ExistingStructures

Seismic isolation, especially with FPS isolators, appears tobe a particularly appealing means of improving the seismic safetyand performance of existing structures for a variety of technicaland functional reasons. Some of these are discussed below.

1. The strength of existing buildings is often sUbstantially lessthan that of new buildings. The ability of FPS connections to actas structural fuses which limit the force that can be transferredfrom the foundation to the supported structure makes them suitedfor these weaker structures. In addition, the energy dissipationand period shifting characteristics of the FPS connections alsohelp reduce the seismic demands on a structure.

2. Existing structl'res often possess little or no ductility.However, code design procedures rely on a significant amount of theenergy input by a major earthquake to be dissipated through ductileyielding of members and connections. To add this requiredductility at the local level in all the existing structural membersrequires massive structural intervention which is functionallydisruptive and economically p ....ohibitive. As a result, it isgenerally more common to strengthen these types of non-ductilestructures by the addition of new shear walls, augmentation ofexisting stru"t~":"al walls, addition of new steel braces, and soon. Be("auze of economic considerations, many retrofit strategies(e.oJ. Ref. 3.1) suggest design forces and ductility requirementsfor existing buildings which are lower than those considered fornew buildings. As a result, one would generally expect the levelof damage in structures retrofitted in this manner to be higherthan that found in comparable new structures.

Moreover, future earthquake ground motions can not bepredicted with great certainty. Consequently, the design loadsconsidered in a retrofitting strategy may be SUbstantially smallerthan those that could develop in a future seismic event, therebyreSUlting in even higher strength and ductility demands thanstipulated by design recommendations. Methods to make theperformance of strengthened structures relatively insensitive tothe likely changes in code strength requirements have not beenadequately investigated.

61

FPS isolators can provide the structure with not only a meansfor limiting and controlling earthquake loading demands, but canalso provide an effective and dependable means for inelastic energydissipation. As demonstrated in the experiments (Ref. 3.4 andChapter 4) as well as in recent analyses (e.g. Chapters 6, 7 and8), it is possible to design FPS supported structures so that mostor all of the energy dissipation occurs in the connections. Insuch cases, the structure remains elastic or nearly so. Thisability to achieve overall ductile structural behavior Whilecontrolling the degree of inelastic damage in the struc..ture is aimportant attribute of the FPS. This ability is examined in moredetail below and il~ sUbsequer.t chapters.

3. Past earthquakes have dei!lonstrated that a major problem causingdamage to structures is irregularities in plan. The reSUltingtorsion has caused serious damage in nearly all major earthquakes.Reducing stiffness and mass eccentricities by conventionalconstruction procedures usually requires major structuralmodifications. An important aspect of the response of the FPS isits unique ability to reduce torsional responses in asymmetricstructures (see Chapter 5). Since the lateral load resistance ofan individual FPS isolator is proportional to the axial load (ormass) supported, the center of lateral load resistance of theisolator group as a whole will be at the center of mass of thesupported structure. While eccentricities between the centers ofstiffness and mass will continue to exist before the connectionsare activated, once sliding commences, eccentricities between thecenters of mass and lateral load resistance disappear. ThUS, anFPS supported structure would initially exhibit the torsionalresponse characteristics of Lhe original building, but thetorsional response will be sUbstantially reduced once sliding isinitiated. This desirable response phenomenon was conclusivelydemonstrated in the previous experimental studies (Ref. 3.4) andin the analytical investigations reported in Chapter 5.

4. Many of the correction techniques used in upgrading conventionalbuildings are performed in the field, and require specializedequipment and skilled workers. ThUS, to be assured that thedesigners intentions are fully realized, stringent qualityassurance programs become a vital aspect of the retrofittingprocess. The need for high quality is, of course, not reduced byuse of the FPS. However, overall quality of the project is mucheasier to achieve, since the PPS isolators and much of theirhardware are: (a) manufactured under stringent quality controlstandards possihle only in modern industrial manUfacturing plants,(b) SUbjected to substantial research and development programs, and(c) SUbjected to pre-installation testing programs. ThUS, FPSisolated structures would be expected to perform reliably anddependably in practice once issues related to the design,performance, and integration into existing buildings are resolved.

5. As discussed in Chapter 2, FPS isolators can be installed in astructure in a variety of ways. The specific location will dependon a range of fd~~ors including the type of structural system to besupported, the strength and ductility of the existing structure,

62

restrictions on location imposed by architectural and functionalrequirements, the proximity of other adjacent structures,availability of an existing basement, and so on. Theseconsiderations will be discussed in more detail later.

It is clear that FPS isolators can be installed in a basementlevel as with most other isolation systems. However, a specialcharacteristic of FPS isolators is that they freely permitrotation. They thereby lend themselves to installation in columnswithout the need for supplemental beams to restrain rotations ofthe isolators. Installation in columns avoids the need for costlymodifications to the existing foundations or the addition of newbeams attached to the isolator. Moreover, by placing the FPSisolators at the top of the first story columns (as done in theprevious experimental program (Ref. 3.4 or in the columns ofupper levels, problems associated with egress conditions or withproximity to adjacent structures can be sUbstantially mitigated.Versatility is further enhanced by the ability to locate the FPSconnections along the midspan of a column. This flexibility of theinstallation and the ability of the FPS connections to rotatefreely are expected to be important attributes to be considered inselecting seismic upgrading strategies for existing buildings.

6. Typically, FPS isolators are added to a single level of astructure. Localization of the seismic upgrading work to a singlelevel will help minimize the disruption of occupants that hasproven to be such a problem with more intrusive and globalrehabilitation methods. Furthermore, where FPS isolators are to beadded to columns, construction work can be logically sequencedbet .... een successive zones of a building, further reducingdisruption. Of course, it may be necessary to make changes toother portions of the structure. But the magnitUde of thesechanges should be reduced in comparison to those required by othertechniques.

7. Another feature of FPS isolators is that they are physicallysmall. While the width of the isolator is governed by the amountof displacement to be accommodated, the height is generally quitesmall. Unlike other types of isolators, the height is nearlyindependent of the amount of lateral displacement expected. Thismakes the isolator very stable against overturning modes of failurethat might bp. associated with large lateral displacements. Thecompact size also makes FPS isolators easier to install andarchitecturally more attractive in an existing building.

8. Becau~e of the above features and the inherent simplicity ofthe FPS, it could provide a partiCUlarly cost effective method forupgrading the seismic safety and performance of existingstructures.

The ability of the FPS isolators to limit seismically inducedforces in a structure, to dissipate SUbstantial amounts of energyand to mitigate the adverse effects of torsional response providethe primary technical motivation for their application to existing

63

buildings. Their dependable operation, versatility and the mannerwith which they could be installed provide the practicality andeconomy necessary for serious consideration as a means forretrofitting seismically deficient structures.

To use FPS isolators, the building should be situated on asite that would accommodate the sliding displacements. Detailedstudies of frame structures responding in the elastic and inelasticranges are included in Chapter 7. Analytical studies on the designof FPS isolators to support structures that should remainessentially elastic are also presented in Chapter 7.

3.4 Installation Details

An important consideration in selecting a particular seismicrehabilitation strategy is the feesibility and economy of theconstruction details. Even simple conventional retrofittingtechniques, such as the addition of shear walls or braces, canbecome extremely complex to implement, depending on the detailsneeded to add collectors, secure new elements to the existingstructure, provide adequate foundation support and accommodatepre-existing conditions. Thus, when considering the basicsuitability of FPS isolators tor the rehabilitation of existingbuildings, it is desirable to identify potential problems andsolutions related to their installation. Due to the preliminarynature of this investigation only basic issues will be addressedfor a number of standard types of structures.

The versatility of installation of FPS isolators has beenhighlighted in Chapter 2 as it relat.es to new construction. Thisversatility is an especially important attribute when consideringexisting buildings. Each building will have a different geometry;structural systelt.; strength, stiffness and ductilitycharacteristics; proximity to adjacent buildings; performanceexpectations~ seismic exposure; and so on. Because of the manyfactors to consider, the designer will not have as much freedom tomodify the structure as would be the case in new construction. Asa result, ingenuity and engineering jUdgment are especially neededto find simple, yet effective retrofitting strategies that arepractical and economical.

A number of different categories of potentially hazardousbuildings have been identified by various authors (e.g., Ref. 3.1).In this study, two basic groups of buildings are addressed: bearingwall systems and frame systems. Information on application of theFPS to these systems constructed of different materials ispresented, as is additional information on special problemsassociated with retrofitting which are not associated with buildingtype.

Bearin9 Wall Systems. The first type of structureconsidered was the bearing wall or box system. These systems havebeen 10n9 rec09nized as havin9 inherently less desirable seismiccharacteristics in comparison too other types of buildings.

64

Accordingly, new bi.:.~ldings which employ bearing walls for lateralresistance are designcfi for higher seismic force levels thanconsidered for other building types.

within this catagory of building systems, greatest concern hasbeen expressed regarding unreinforced masonry structures. Suchstructures are usually quite stiff, but exhibit brittle responseswhen loaded beyond the elastic range. They have repeatedlydemonstrated poor performance during past earthquakes throughoutthe world. Some research has been performed (e.g. Ref. 3.2) onmethods for evaluating and retrofitting these types of structures,and design recommendations related to conventional constructiontechniques have been published (Ref. 3.5).

Some masonry and other bearing wall structures can be quitestrong in the direction of the walls due to the large size, closespacing and large number of walls used. Nonetheless, seriousdeficiencies may exist which are related to out-of-plane wallcapacities, wall attachment to diaphragms, and diaphragm integrityand flexibility. These types of problems would also have to beaddressed and solved prior to using FPS isolators. Theintroduction of the FPS may change the nature of the internalforces and deform