Breen Precast

8/6/2019 Breen Precast

1/32

0000O000OOOOOOOOO0O0OOOO

D e v e lo p in g S t ru c t u r a lI n t e g r i t y i n B e a r i n gW a ll B u ild in g s

John E . BreenThe J. J. McK etta Professor of EngineeringDep artment of Civ i l EngineeringThe U niversi ty of Texas a t AustinA ustin, Texas

The prestressed concrete industryprovides large volumes of precastfloor and wall units for use in bearingwall buildings. In addition to the wideinterest in precast panel bearing wallbuildings, the industry has a vitalinterest in brick and concrete masonrybearing wall buildings which oftenutilize precast prestressed concretefloor units.For decades, relatively little en-gineering research and developmentattention was paid to bearing wallconstruction. It is heartening to notethat the Prestressed Concrete Institutehas taken a lead in this importantNOTE: T his article is an expanded and updatedversion of a paper p resented at the PC I Seminaron "Advanced D esign Concepts in Precast Pre-stressed Concrete," held in conjunction withPCI 's 25th Anniversary Convention in Dallas,Texas, October 17-18, 1979 .

field. This progress is reflected in thework of PCI Committees and pub-lished reports and papers appearing inthe PCI JOURNAL (see list of refer-ences at end of paper). Other usefulinformation on the subject is con-tained in the current PCI DesignHandbook, the PCI Manual forStructural Design of ArchitecturalPrecast Concrete, and some early is-sues of PCltems.Recently, extensive research pro-grams at the Portland Cement Associ-ation and at the Massachusetts Insti-tute of Technology have helped sys-tematize and extend our knowledge ofbehavior and design requirements forpanel structures. This work has beensummarized by Mark Fintel, andJames Becker and Peter Mueller atthe recent PCI Seminar on "AdvancedDesign Concepts in Precast Pre-

42


2/32

Fig. 1 . G eneva T owers , Lake Geneva, W isconsin . This nine-s tory condom inium andoffice building used 245 6-in. (152 m m ) thick loadb earing wall panels with sizes upto 10 ft high x 42 ft long (3.05 x 12.81 m ). Doub le-tees were used for the floor androof mem bers .PCI JOUR NAU January-February 19803


3/32

Fig. 2a. Atlantis Cond om inium s, Ocean C ity, Maryland. This 22-storybuilding is comp rised of precast floor slabs, wall panels, spandrels, andstair and elevator towers.

stressed Concrete" held in Dallas,Texas, October 17-18, 1979. Undoubt-edly, these and other studies willlead to improved detailing practicesfor bearing wall systems.Much of bearing wall panel de-velopment has taken place in livinglaboratories scattered over NorthAmerica, where innovative design en-gineers and precasters have worked

together on the economic and safesolution of the many design, produc-tion, and erection problems involvedwith this type of construction.The bearing wall structure is simplein concept, but extremely careful at-tention is required in the connectiondetails to ensure safe buildings. Whenproperly designed, the resultingstructure will be attractive, economi-

44


4/32

Fig. 2b. Isom etric view of A t lant is Condominiu ms showing thea rra ngement of the va riou s precas t elements.

cal, functional, and structurally sound.Figs. 1 through 4 show various exam-ples of buildings employing largeloadbearing wall panels. In addition,Figs. 3, 4a and 4b show how hollow-core slabs can be combined very ef-fectively with loadbearing wallpanels.In this presentation, my aim is toupdate you on the general directionthat regulatory bodies are taking withrespect to previous concerns over thepotential for progressive collapse ofbearing wall structures, to outlinegeneral principles for design, detail-ing, and construction which wouldgreatly improve structural perfor-mance under unforeseen loadings,

and to suggest sources for detailedtechnical solutions to carry out theseprinciples. It is not my intent to dwelldeeply on details for individual 'sys-tems, but rather to suggest sources forguidance on such details.

BackgroundIn 1968, a dramatic chain reactioncollapse following a localized gasexplosion on the eighteenth floor to-tally destroyed one quadrant of the

22-story precast concrete panel con-struction Ronan Point apartmentbuilding in England (see Figs. 5a and5b). 1 Such widespread propagation ofPC I JOURNA L/Janu a ry -Februa ry 1980 45


5/32

Fig . 3 . P h i l lips To wer , M inne ap o l is , M innes ota . Th is ten -story bu ilding fo r the e lder lyconsists of 328-8 x 24 to 26-ft x 8-in, solid precast wall panels;316-8 x 42-ft x 12-in, precast hollow-core floor slabs; 54-4 x 42-ft x 8-in, solidraked precast spandrels; and 21-4 x 8-ft x 4-in, precast hollow-core floor panels.(Note: 1 ft = 0.305 m; 1 in. = 25.4 mm.)

46


6/32

Fig . 4a. Th e 20-story Ro ber ts P lazaAp ar tme n t Bu i ld ing in Reg ina ,Saska tchewan , i s com pr ised o f som e2000 p rec as t p res t ressed ho l l ow-co res labs, l oadbear ing wa l l pane ls a nd o the rp recas t e lemen ts .

F ig . 4b . C loseup o f Rober ts P lazaAp ar tme n t Bu ilding show ing tex tu re o fl oadbear ing wa l l pane ls a nd h o l lowco res lab be ing swung in to p os i t ion .

PC I JOURNA L/Janu ary-February 1980 47


7/32

Fig. 5a. R onan P oint Apartm ent Building after collapse, with a second identicalbui lding in b ackground. (Courtesy: London Exp ress News and Feature Services . )

failure following damage to a rela-tively small portion of the structurehas been termed "progressive col-lapse."Today, the term progressive col-lapse has come to signify an incre-mental type of failure, where the totaldamage done is considered out ofproportion to the initial cause. Defini-tion as an incremental type of failureeliminates consideration of the totalcollapse of statically determinate

structures, such as a truss, upon loss ofa single member.While the Ronan Point collapsedrew attention to this failure mode, itwas certainly not a unique occurrence.A large number of progressive col-lapses have been documented in theengineering literature, many occur-ring before 1968 2 There are examplesof vertical collapse propagation as inRonan Point and Bailey's Crossroads(see Figs. 6a and 6b), horizontal pro-

48


8/32

Fig . 5b. C loseup o f dam age a t topp o r tio n o f R o n a n P o i n t Ap a r tm e n tBui ld ing (Reference 1).

F ig . 6a . C as t -in -p lace con cre teapa r tm en t co l laps e (Ba i ley ' sC ross roads, V i rg in ia) . Ver t ica lp ropa ga t ion occu r red f rom a sho r inge r ro r on the upp er s to r ies o f bu i ld ing .

Fig . 6b. Debr is a t base o f tower due to bu i ld ing co l lap se(Bai ley 's C ross roads , V irg in ia) .PC I JOURNA UJ anu ary-February 1980 49


9/32

pagation (see Fig. 7), and combina-tions of the two.The particular type of joint detailused in the Ronan Point apartmentbuilding relied heavily on joint fric-tion between elements. This resultedin a structure which has been termeda "house of cards." This occurrenceindicated that structures with similarjoint characteristics were particularlysusceptible to progressive collapse.

Indeed, occurrences have happenedin all major modem construction ma-terials (reinforced and prestressedconcrete, steel, wood, and masonry).Many types of structures are most sus-ceptible to this type of collapse whileunder construction; however, exam-ples exist of several completed struc-tures which have also undergone pro-gressive collapse.Because the Ronan Point collapsewas initiated by an explosion, whichis a loading condition not generallyconsidered in the design of buildings,

it was termed an "abnormal loading."Several studies have been undertakento predict the frequency and mag-nitude of similar loading conditions.In some studies this category has beenextended to include faulty practice,such as design and construction er-rors.

In the years following Ronan Point,literally hundreds of engineering arti-cles and reports on these subjectshave been published. An extensiveannotated bibliography on abnormalloading and progressive collapse hasbeen published by the NationalBureau of Standards as Building Sci-ence Series No. 67.2Most European countries andCanada adopted some form of regula-tory standard to minimize the risk ofprogressive collapse resulting fromabnormal loading. These standardswere difficult to meet in some bearingwall systems. However, very fewbuilding codes in the United States

Fig. 7. Collapse of cast-in-place post-tensioned parking garag e which was struck atone edge by falling tower crane and debris (Bailey's Crossroads, Virginia).50


10/32

have taken specific action to includeprogressive collapse regulations al-though a number of possible proce-dures have been discussed.Since the Ronan Point collapse, the

Department of Housing and UrbanDevelopment has been a majorstimulus for improvement in resis-tance to progressive collapse in in-dustrialized construction. HUDstimulated research and developedregulation proposals both as a part ofOperation Breakthrough and in con-nection with a major study undertakenat the Portland Cement Association.While these studies have indicatedthat the types of construction suitedfor United States markets differedmarkedly from industrialized con-struction in Europe, only a narrowsegment of the design and construc-tion community showed an interest inundertaking serious studies of re-quirements in this area. Recent pro-grams aimed at improving the resis-tance of bearing wall structures toseismic loads have contributed exten-sively to a parallel resistance to pro-gressive collapse.

A few consulting engineers3 - 5 whostudied the progressive collapseproblem indicated that they felt thejoint details used in Ronan Pointwould be unacceptable in NorthAmerica, and that American practicewould be considerably more conser-vative than that used in the design ofRonan Point. They argued for adop-tion of design principles with respectto jointing and continuity similar tothose used for earthquake design tominimize the danger of progressivecollapse in precast systems. Most im-portantly, they urged that evaluationof the danger of progressive collapseand the method of resistance to thisincremental type of failure should bemade in the United States against thebackground of American practice andbuilding regulations, rather than asimple copying or immediate adoption

of European codes and regulations.Despite this counsel, a study of therecommendations adopted by the

New York City Building Code s indi-cated that this advice was often disre-garded and that the European ap-proach was adopted in some cases al-most verbatim, at least as an interimmeasure.

T he A f term a th of"Ronan Point"The report of the Ronan PointCommission of Inquiry 1 revealed sev-eral deficiencies in existing Britishcodes and standards, particularly asthey applied to multistory construc-tion. The Commission focused on thelack of redundancy or "alternatepaths" in the structure. As an offshoot

of the investigation, the Britishbuilding regulations ? 8 were changed(The Fifth Amendment) to requirethat multistory structures be designedeither to provide an "alternate path"in case of loss of a critical member orto have sufficient local resistance so asto withstand the effects of a gas-typeexplosion.Implementation of these recom-mendations produced a great deal ofboth controversy and uncertainty.Continental authorities9 - i ' were quickto point out that the 1967 CEB Rec-ommendations 1 2 had spoken of panelstructures as a "house of cards" andhad called for mechanically continu-ous networks of reinforcement to pre-vent progressive collapse. Regretswere expressed that the official RonanPoint inquiry report did not specif-ically point out that the collapsedbuilding violated the CEB principles.'It has been claimed that if these prin-ciples had been followed, the progres-sive collapse would not have oc-curred.

The Ronan Point report urged im-proved detailing to toughen bearingPCI JOURNAL/January-February 1980 51


11/32

Fig. 8 . Algerian bearing wa ll concretebuilding which survived a ma jorexplosion (Reference 1).

wall buildings and cited successfulexamples such as an Algerian apart-ment building which effectively con-tained the spread of damage from amajor explosion which destroyedloadbearing panels on the ground andfirst floors (see Fig. 8). Several en-gineers have indicated that the 5 psi(720 psf) (34.5 kPa) pressure requiredin the specific local resistance designprocedures was excessive and that thealternate path method of design wasconfusing and complex.

The reactions of design profession-als in Great Britain to the aftermath ofthe Ronan Point collapse weremixed. 1 3 -' 6 Generally, the design pro-fessionals recognized that there weresubstantial social implications inher-ent in the problem of progressive col-lapse if an occurrence in one dwellingspace of a high-rise structure couldjeopardize the lives and property ofoccupants of far-removed dwellingspaces.

The general concepts of riskanalysis which led to the conclusionthat there was a high probability of anexplosion within a high-rise structurewere not attacked. However, the spe-cific implementations of the "FifthAmendment" to the building regula-tions met a great deal of criticism.Very little factual data were availableas to the economic consequences. Theregulations were basically criticizedbecause of the lack of existing knowl-edge of the effect of gas and othertype explosions on buildings. In par-ticular, substantial questions wereraised as to the dynamic characteris-tics of typical explosions and the re-sponse of various types of structures tosuch internal explosions.The entire concept of designing forredundancy in the construction ofbuildings was questioned because ofthe long history of statically determi-nate structures used in civil en-gineering construction. Much criti-cism had to do with the conflict be-tween building regulations and thepersonal responsibility of the de-signer.Particular criticism was raised con-cerning provisions which stated thattraditional forms of construction suchas steel and concrete framed struc-tures which met existing buildingcodes were "deemed to satisfy" thespecial progressive collapse provi-sions. It was pointed out that nothorough study had been made in thisarea and it was certainly possible todesign statically determinate struc-tures in conformance with existingcodes and regulations.

Reaction to the Ronan Point inci-dent and the amendments to theBritish building regulations whichfollowed quickly spread worldwide.The precast concrete industry becamethe target of jokes (mostly unjustified)such as the cartoon (Fig. 9), which ap-peared in Punch, a British humormagazine. Unfortunately, undue cau-

52


12/32

tion and too many regulations set in-dustrialized construction back.

Similar code provisions wereadopted in many countries and in theUnited States the Department ofHousing and Urban Development"circulated for comment a draft docu-ment to implement such standards inconstruction under its mortgage insur-ance programs. The City of New Yorksamended its building code to requireresistance to progressive collapse byeither the alternate path method orthe specific local resistance to 720 psf(34.5 kPa) method.A National Bureau of Standardsprogram 1 8 - 2 1 documented the frequen-cies of occurrence and risk analysis ofabnormal loadings. A Portland Ce-ment Association program22,23suggested design and constructionconsiderations for large concretepanel buildings and was a movingforce in the development of an overallphilosophy to reduce the risk of pro-gressive collapse by incorporating im-proved overall structural integrity inthe large panel structures. Both pro-grams have emphasized the Americanaspects of the problem presented byU.S. characteristics in loads, spans,building layouts, and constructionpractices.

A bnorm a l Load ingsThe term "abnormal loading" hasbeen used to indicate any loadingcondition not generally considered inthe design of a building. For tra-

ditional construction and typicalAmerican codes and standards, design-ers usually consider dead load, liveload, snow load, wind load, earth-quake load, soil load, hydrostatic load,and effects of temperature and dimen-sional changes.

Abnormal loadings would be loadswhich have generally been consid-ered to have such a low probability of

Fig. 9. Brit ish ca rtoon poking fun a tprefabricated a partment bu i ld ings(Courtesy: Punch).

occurrence as to warrant neglect indesign, such as:1. Violent Change in Air Pressure(a) High explosive detonation

(sabotage, suicide)(b) Service system explosion (gasunit or gas system leaks)

PCI JOURNALJJanuary-February 1980 53


13/32

2. Accidental Impact(a) Automotive (Fig. 10)(b) Crane (Fig. 11)(c) Airplane3. Loads Due to Faulty Practice(a) Construction error (Figs. 12and 13)(b) Unauthorized alteration(c) Lack of maintenance4. Fire5. Flood

6. TornadoA major part of the initial program

in progressive collapse research at theNational Bureau of Standards was todetermine the frequency of occur-rence of various abnormal loadings forresidential and commercial typestructures in the United States.18 -21The results of these studies estab-lished the probability of occurrencefor the various abnormal loadings thatwhile not precise seems to give atleast an order of magnitude assess-Fig. 10. The driver turned the corner of ment.this bu ilding too tightly. In general, the results agree fa-

Fig. 11 . Pa rking ga rage stru ck by fa l l ing crane (Cleveland, O hio).54


14/32

Fig. 12. Egyptian apartm ent bu ilding after loss of a founda tion. N ote that bu ilding isidentical to one in backgrou nd.

vorably with similar studies com-pleted in Canada, England, and TheNetherlands. There appears littlelikelihood of developing significantlybetter information on frequencies andprobabilities for the United Statesuntil a more comprehensive data bankis established.In all the studies, the probability ofoccurrence of a relatively severe gasexplosion is the highest. Next in prob-ability is a high explosive (bomb)explosion. This type of explosionseems to be on the increase and is ex-tremely serious. One major problemin all of the frequency studies of ab-normal loads is that the reporting in-formation concerning the structuralconsequence or the magnitude of theload is so scanty that it is difficult todetermine whether any structural sig-nificance should be attached to thegiven reported incidents.Specific information on the mag-nitude of abnormal loads that mightbe expected can only be characterized

as marginal. The 5-psi (34.5 kPa) pres-sure used in the specific resistancemethod of calculation in the UnitedKingdom's standards following RonanPoint was based on theoretical calcu-lations of probable pressures in a gasexplosion in that size unit and onexamination of the damaged appli-ances and piping. A pressure of 5 psi

Fig. 13. The plum ber cam e late.PCI JOURNAL/January-February 1980 55


15/32

(34.5 kPa) was believed relatively rep-resentative of the pressure that oc-curred in the Ronan Point incident.There is substantial argument con-cerning the magnitude of pressure

that might be expected in a typical in-cident in an American living unit be-cause of size differences, ventingcharacteristics, and other factors. Con-clusions drawn from some studieshave indicated that the combination ofinterior partition venting (if effective)and the general layout of Americanapartments could perhaps combine toreduce explosive pressures to a levelcloser to 1 psi (17.2 kPa) rather than 5psi (34.5 kPa). This conclusion hasbeen vigorously denied.There is an interrelationship be-tween abnormal loads and progressivecollapse. This interrelationship hassignificance in connection with risk.Risks are the products of the prob-abilities of an occurrence and theconsequences of that occurrence. Pro-gressive collapse has higher conse-quences because it is usually accom-panied by an increase in injuries andloss of life or property than would ac-company a local collapse. This meansan increase in the risk to individualsor to society. It is thus necessary toconsider the scale effect.There are other disasters that cancause large damage, such as tornados,hurricanes, floods, and earthquakes.These are termed "Acts of God." Theycannot be prevented but the conse-quences could be reduced and thereis some attempt to do that with ourbuilding codes. However, accidentsthat lead to large damage usually leadto rather immediate reaction from thepublic and from the regulatory agen-cies and they can and usually are pre-vented or steps are taken to preventthem by means of regulatory actions.Because progressive collapse hashigher consequences, loadings oflower probability must be consideredin order to keep the risk constant, as

compared to a failure that does not in-volve progressive collapse.

The entire area of abnormal loadingshould be substantially deempha-sized. There is some need for de-velopment of frequency statistics forthe purpose of risk analysis and regu-latory provision justification. In a fewisolated cases provision of a loadingforce or pressure would be useful indesigning a specific element in anunusual structure. However, in mostcases the level of attention paid to as-sessing the magnitude of possible ab-normal loadings is not justified for themost feasible methods of handling theoverall problem of progressive col-lapse.The British experience indicatedthat the 5-psi (34.5 kPa) loadingadopted in the specific resistancemethod led to costly solutions withoutnecessarily ensuring the desiredsafety. The only area of applicationwhere a pressure loading might behelpful for some engineers was in se-lecting values for proper tie forces orcalculating wall rupture loads. Thisdoes not really justify a large researcheffort in the area of abnormal loadings.More effort should be spent in de-fining the level of damage which isacceptable or is to be contained thanin assessing a magnitude of load. Ac-ceptable damage can consider factorssuch as occupancy, area, volume, cost,and use.Progressive collapses can betriggered by other causes than explo-sive loads. The magnitude of the ab-normal loads becomes less importantif the design approach taken is to pro-vide overall structural integrity whichwill bridge and contain local damage.A positive emphasis on improvedstructural integrity to limit the propa-gation of damage is far more desirablethan using the specific resistancemethod to try to withstand some arbi-trary load. As a result, the importanceof abnormal load magnitude is greatly


16/32

diminished, except as an indicator ofthe amount of initial damage whichmight be expected.Generally speaking, improvedstructural integrity is obtained by pro-vision of integral ties throughout thestructure. The amount of ties can bedetermined from considerations ofdebris loading and the amount ofdamage to be tolerated without usingthe magnitude of the explosion orother abnormal load.

The futility of pursuing the subjectof abnormal loading in depth is bestillustrated by some recent remarks ofan experienced designer. He believedthat a substantial number of structuralengineers would be extremely un-comfortable with a 20-story bearingwall building with no ties even if gassystems were prohibited, the buildingwas erected in the center of a golfcourse miles from any adjacent struc-ture, and armed guards were on dutyat all hours of night and day in thelobby!

P a nel and Bear ingWa l l S tru c tu resThe risk of a progressive collapse in

large panel and bearing wall struc-tures is greater than the risk in tra-ditional cast-in-place structures. Theincreased susceptibility to progressivecollapse is due to a combination of theuse of relatively brittle materials andthe general lack of ductility and con-tinuity in the overall structure be-cause of the details used in assem-bling the pieces. Structures such asthe Ronan Point apartment building,utilizing basically friction connectionsfor resistance to lateral forces, arecompletely unsatisfactory as a struc-tural type.

The addition of suitable ties to de-velop continuity could have probablymade that structure resistant to theconsequences of the abnormal loading

to which it was subjected. A numberof solutions are available to the prob-lem of increasing the structural integ-rity of large panel and bearing wallbuildings. However, a great deal ofdesign ingenuity is required to trans-late these into efficient, economicalconstruction.

The susceptibility towards progres-sive collapse is further increased inAmerican practice because the build-ings tend to contain fewer intermediatewalls and supports than in Europeanpractice. This reduces the chance ofredundancy and development of analternate path. An examination shouldbe made of the general configurationand the joint details of early bearingwall buildings to determine their sus-ceptibility to progressive collapse.Remedial action should be taken todevelop improved continuity in anybuildings adjudged to be particularlysusceptible to progressive collapse.The direct applicability of a largebody of the research findings and de-sign recommendations developed forEuropean practice is reduced becauseof the very large differences whichexist in building layouts betweenAmerican and European residentialstructures. European structures typi-cally have bearing walls surroundingalmost every room, slabs cast to roomsize for a particular job with reinforc-ing steel protruding from the slabs onall peripheries, and thus the de-velopment of continuity is mucheasier with interlacing of the pro-truding reinforcement and a cast-in-place (wet) joint.American architecture calls for avery different structural layout. WhileEuropean walls are often spaced at 15to 20 ft (4.6 to 6.1 m), typical spans inthe United States vary from 22 to 40 ft(6.7 to 12.2 m). Intermediate non-loadbearing partitions are used tosubdivide living units up into rooms.The proportion of walls to, slabs. inAmerican construction is frequently as

P C ,I JOURNAL/January-February 1980 57


17/32

low as one-third the ratio found inEuropean buildings. In addition, hol-low-core slabs produced on longcasting beds and cut to length ac-cording to the needs of the job areprominent in the United States. Withthis form of floor slab construction, noprotruding reinforcement is availableat the ends of the slabs to developcontinuity. Thus, the continuity de-tails must be considerably different.Because of the pattern of labor or-ganization in the United States, largeconcrete panels are often connectedby dry joints utilizing bolting orwelding, so that the iron worker doesnot have to wait for the masonry orconcrete craft to complete the jointbefore proceeding with further erec-tion. Efforts to import the types ofbuilding systems prevalent in Europehave been generally unsuccessful,both because of economic factors andbecause of a general unacceptabilityof the architectural layout prevalent inEurope.

The extensive research program onlarge panel concrete structures at thePortland Cement Association concen-trated on building arrangements,panel connections, and span propor-tions typical of conditions in theUnited States. The generally reducednumber of vertical elements in Ameri-can structures means that the problemof the overall stability of the damagedstructure is probably more severe thanoccurs in European structures. Thus,the design philosophy implementedin the United States should pay care-ful attention to the need of ensuringstability in the overall structure afterthe loss of one or more elements.Des ign Phi losophy to R es is tP rogress ive Col la ps e

Any design and construction re-quirements imposed to reduce theprobability of a progressive collapsemust follow a consistent overallphilosophy to insure effectiveness.

The consideration of the possibility ofa progressive collapse assumes that,due to some overload or weakness ofthe structure, a local failure has al-ready occurred. Supposedly, normalfactors of safety have been set to re-duce the possibility of such a localfailure to a generally acceptable level.Design to resist progressive collapserecognizes that a local failure cannotbe prohibited absolutely. The generalphilosophy is that a structure shouldbe stable under that local damage andbe able to bridge over the damagedarea without complete collapse of thestructure.A fundamental problem in im-plementing this design philosophyis that it is difficult to quantify thevolume of damage which the structuremust be capable of sustaining withoutprogressive collapse. Studies areneeded to define socially and techni-cally acceptable volumes of damage asrelated to use, occupancy, and stabil-ity.The various regulations issued inEurope,'' $ Canada, 2 4 and in draft formby HUD, 1 7 give some indication of theextent of damage being considered.Generally, the structure must be ableto take a reasonable amount of debrisload and resist the loss of a principalbearing member. Practical im-plementation of such regulations inthe design and construction of struc-tures has indicated that specificationof an abnormal load such as a gaspressure is relatively meaningless. Anindication of the volume and type ofdamage which must be contained isfar more effective for designers seek-ing to develop effective resistance inunique situations.There appears to be a general con-sensus that it would be desirable forpanel and bearing wall structures tohave the same degree of resistance toprogressive collapse as traditionalmonolithic construction, such as foundin typical cast-in-place concrete con-

58


18/32

Fig. 14. General structural integrity of a cast-in-place frame structure with 75 percentof the colum ns on o ne story destroyed during a dem olition attempt. Incredibly, thestructure is still standing .

struction. Detailing practices forcast-in-place structures requiring ar-bitrary percentages of reinforcementto be brought to the supports,minimum tension capacity in col-umns, and overall development ofreinforcement impart substantialtoughness in a structure. This capa-bility is demonstrated in Fig. 14which shows a condemned buildingresisting the "attacks" of a demolitioncontractor!This does mean that the design andconstruction procedures for precastpanel and bearing wall structuresshould try to develop overall stability,ductility, and redundancy similar towhat might be expected in cast-in-place construction.In some unusual forms of construc-tion it may be necessary to preventthe propagation of collapse laterallyby provision of frequent expansionjoints which will isolate the structure

into acceptable segments. In general,there are so many different types ofstructures, methods of construction,and possibilities of initial damage thatregulatory groups have found it verydifficult to quantify the functional re-quirements for resistance to progres-sive collapse. It has been far easier torely on the ingenuity of the designerafter assurance that he has developeda sound understanding of the overalldesign philosophy.Design to resist progressive col-lapse is in essence an advanced limitstate, since it already assumes that alocal portion of the structure hasfailed. Because of this, calculationsare usually made using load factorsapproaching unity for foreseeable de-bris load and partial live load (usuallyone-third to one-half of design liveand wind load). The structure must bestable under this load to allow forevacuation and emergency operations

PC I JOURNA L/Jan uary-February 1980 59


19/32

and permit temporary support or re-pair. More specific guidance needs tobe provided as to appropriate safetyfactors under this condition and ap-propriate strength reduction factors forassessing member strength.For satisfactory resistance to pro-gressive collapse, the structure mustmaintain a stable configuration afterthe extensive damage to or loss of amajor member. Design of a structureso that it can develop a substituteload-carrying configuration which willpermit it to remain stable upon theloss of a support makes a substantialcontribution towards eliminating thedanger of progressive collapse. Thevery large tributary areas carried byindividual bearing walls in UnitedStates construction impose substantialproblems in engineering stabilityupon loss of two walls. Preliminarystudies indicate that in many panelbuildings typical of American practicethe structure can be made to bridgeover a missing bearing wall at rela-tively little additional cost. Extremelylarge additional costs would be in-volved if the structure had to bridgetwo or more successive missing walls.It is necessary to ensure stability byprovision of suitable compressionstruts and tension ties to allow thestructure to bridge over the damagedarea, as indicated in Fig. 15. Anystructure suspected to be susceptibleto progressive collapse should becarefully investigated for stabilityunder reasonable local damage. Thischeck may be done indirectly by pro-viding improved overall structuralintegrity by ensuring that proper de-tails are used to develop continuityand redundancy and that an overallstable structure results. In some veryspecial cases, it may be necessary toactually check a structure under sev-eral different configurations corre-sponding to removal of key supports.Much can be done to improve over-all stability in the initial architectural

layout and arrangement of bearingwalls. This can be achieved bybringing the structural engineer in atan early stage of the design. Certainpatterns of wall layout will make itmuch easier to develop bridging.Overall provision of ductility andcontinuity at the joints can assist indeveloping resistance to progressivecollapse equivalent to that of frame-type structures.The concept of limiting damagepropagation means that careful con-sideration must be given to resistingthe debris which might follow thefailure of an adjacent member. InEuropean practice, designers havetried to develop an inherent load-car-rying capacity in a floor, so that it canresist a debris loading equivalent tothe weight of the floor above plus 30percent live load imposed with a sub-stantial impact factor. This debrismust be carried as a superimposedlive load, although it may be carried atexceptionally large deformations andwith substantial damage to themember.With the long spans of Americanpractice, this imposes a substantialload requirement for a member. Onepossibility for reduction is to developtie forces and details which will pre-vent a large part of the debris fromfalling on the member below when afailure occurs.In any case, any debris loading thatis imposed must be carried in shear as.well as in flexure at the limit state.Failures which have occurred in someAmerican large panel structures indi-cate that the weakest point of the sec-tion under debris loading was theshear strength at the supports.

If floor systems are designed to en-sure effective membrane or catenaryaction upon loss of an intermediatesupport, the initially damagedmember must carry its own debrisload. This can generally be done ef-fectively by developing proper tie60


20/32

Tension Reinforcements inFloors Over Walln

Tie RodDiagonalCompression

StoreyHeigh t Compression i TensionCantilever j Capability I Capability orSelf Weight

^ of Structure

WALL REMOVED

n I , .

^ 1 1 A !I

Fig . 15. Br idg ing in a be ar ing wa l l pa ne l s t ruc ture .

forces at the ends and ensuring thatbottom reinforcement in the memberat the point of the missing support isfully effective as tension reinforce-ment by proper anchorage or designedas a splice. In order to provide a suit-able gap to ensure that the membraneor catenary action is effective in pre-venting the debris from bearing on thefloor below, the limiting deflectionunder failure conditions should notexceed approximately one-half theclear story height.

Design of all members in the

structure should include details to en-sure that proper compression and ten-sile resistance is possible, so that thestructure can bridge over a missingouter support, as shown in Fig. 15.Effective vertical ties are required astiebacks and compression struts arerequired for fulcrums.Careful attention must be paid todevelop continuity of vertical walland floor panels, so that shear transferis obtained to allow the bridging wallto act as a deep cantilever beam. Sub-stantial experimental programs and

PCI JOURNAL/January-February 1980 61


21/32

design ingenuity are needed to studyvarious arrangements of bridgingmembers and to determine the mostefficient ways to provide the tensile,compressive, and shear resistance inthe panels and joints.The most important method tominimize the risk of progressive col-lapse in panel and bearing wallstructures is to provide adequate hori-zontal, vertical, and peripheral tiesbetween all structural elements to de-velop improved structural integrity.Emphasis on ductility and continuitysimilar to that used for seismic andwind design is the most usefultechnique for minimizing the risk ofprogressive collapse.Most structures designed and de-tailed to resist seismic loads in UBC'sSeismic Zones 2 and 3 in the UnitedStates would have a low susceptibilityto progressive collapse. However,buildings in low wind and low seis-mic areas might be quite susceptible

to progressive collapse. The PCICommittee report on precast bearingwall buildings 2 5 recommends that allbearing wall structures be designedfor a minimum lateral total designforce equal to 2 percent of the servicedead load.

In structures which are designed forsubstantial lateral forces, the design-ers are used to the concept of de-veloping diaphragm action in the floorand wall elements, in order to provideflow paths horizontally and verticallyfor the lateral forces. However, insome regions of the United States, de-signers are not accustomed to de-veloping diaphragm action. They areprincipally concerned with gravityloads and not as attentive to the needsof tieing all of the elements together(Figs. 16 and 17). The use of panelswith small bearing areas, no protrud-ing reinforcement, and generally in-adequate connections is common inlow seismic and low wind load areas.

Fig. 16. Two-story t i ltup b earing wall building w ith grossly inadequate c onne ctiondetails (Baton Rouge, Louisiana).

62


22/32


23/32

P anel and Bearing-Wal lConstruc t ion Sys tem sThe predominant construction ma-terials used in panel and bearing wallstructures in the United States are

precast prestressed concrete floor andwall panels, reinforced concrete floorslab and wall panels, and masonrywalls with concrete, bar joist, or woodf loor systems.

The Prestressed Concrete Institutehas responded to the problem of pro-gressive collapse very responsibly. Acomprehensive report on design con-siderations for precast concrete bear-ing wall buildings to withstand ab-normal loads was developed by a PCICommittee and published in theMarch-Apr il 197 6 PCI JOU R NA L.25This report recommended de-velopment of a degree of continuityand ductility which can develop rea-sonable resistance to progressive col-lapse without undue economic penal-ties. The report emphasizes that his-torically there has been insufficientattention towards developing overallstructural integrity in panel structuresand recommends that the structure betied together in all directions. Thelevel of tie forces specified is roughlyequivalent to current constructionpractices.*The American Concrete Instituteinvestigated design and constructionrequirements to minimize the risk ofprogressive collapse through parallelactivities in ACI Committee 356 (In-dustrialized Concrete Construction)and through a Task Force on Progres-sive Collapse of Committee 318(Standard Building Code). Since the

*Edge ties on the pe riphery must develop thediaphragm force b ut not less than 16 kips. Hori-zontal t ies should be provided at r ight angles.Those across the floor span must develop at least1500 lb/ft while in the direction of the spa n itshould develop 2' percent of the wall serviceload b ut not less than 1 500 lb/ft . Vertical t iesshould be p rovided in buildings over two storiesto develop any tension b ut not less than 3000lb/ft.

ACI Building Code does not containloads but depends on a generalbuilding code for specific load re-quirements, it appears more appro-priate that it concentrate on factors in-volving details and development ofappropriate continuity and ductilityrather than including performance orfunctional statements regarding theresistance to progressive collapse. Themajor changes from the provisions forprestressed concrete would appear tobe that tie forces could be reducedsince the span lengths are generallysignificantly shorter.A substantial area of concern is thecurrent state-of-the-art concerning theresistance of masonry structures toprogressive collapse. The initial reac-tion of the masonry industry to theHUD draft criteria on increasing re-sistance of buildings to progressivecollapse was to essentially considerthe problem as not applicable tomasonry structures. There seems to bea growing awareness that this is a po-tential problem for masonry structuresas well, which must be met.ACI Committee 531 (MasonryStructures) has recently reported newdesign rules which should improvestructural integrity. 2 6 While substan-tial research programs are underwayin the masonry area to improve overallstructural integrity for resistance ofseismic loads, no formal considerationof provisions to minimize the risk ofprogressive collapse has been gener-ally indicated.

One of the fundamental problems inmasonry design is the lack of a com-prehensive ultimate strength designprocedure. Since calculations for tieforces and bridging assume a limitstate wherein the structure is alreadypartially collapsed and the attempt isto confine the spread of that collapse,the elements are working at very neartheir ultimate strength. It will be dif-ficult to include block, brick and com-posite masonry until a fundamental

64


24/32

understanding of masonry behavior isdeveloped.Review of a number of actualmasonry designs indicated large po-tential susceptibility to progressivecollapse. 2 7 Engineers who have re-viewed several masonry building de-signs expressed disappointment at thelevel of engineering, particularly innon-reinforced masonry design. Whiledesigners utilizing masonry in highseismic zones seem attentive to de-tails to develop diaphragm action,large regions of the country use high-rise masonry structures with essen-tially gravity loads analysis. The re-sulting designs and details may beextremely susceptible to progressivecollapse.

There is a substantial amount of testdata available on the performance ofreinforced and unreinforced masonryelements, such as individual panels.However, there seems to be relativelylittle information on the behavior ofoverall structures and representativewall-floor joints. The large amount ofmasonry construction in the UnitedStates and the increasing use ofmasonry in high rise (greater thanthree stories) construction indicatesthat this is a major area needing re-search. Generally, the manufacturersof masonry have not been attuned tothe need for research on overallstructural behavior.

Much research work has been doneon masonry structure resistance toprogressive collapse in Europe. 2 g 3 0 A sin the previous discussions, thestructural layouts are for relativelyshort spans as compared to Americanpractice. A great deal can be learnedfrom a review of this work but sub-stantial additional input is required todetermine the resistance of typicalAmerican masonry buildings and floorsystems.Properly constructed masonry isvery rigid and the field assemblygives the opportunity for development

of continuity, if proper panel rein-forcement and jointing details are in-cluded. The large variety of mixedconstruction (masonry-concrete-ma-sonry-steel, etc.) indicates that a largervariety of details will have to beexamined. Particularly with non-reinforced masonry, wall capacity nearcollapse may be weaker than the jointcapacity. Masonry structures appear toneed a more extensive testing pro-gram than reinforced concrete andprestressed concrete panel construc-tion, since the wall panels themselvesand not just the joints may be poten-tial failure locations. For this type offailure analysis, it is doubtful that thetension or shear capacity of themasonry should be used in calcula-tions of the shear and tensile forcesneeded to provide adequate ties. Thisindicates that substantial changes inpractice will be required.

Mem brane orCa tenary A c t ionBy the provision of adequate hori-zontal ties it is possible to develop amembrane or catenary action in thefloor slab above the origin of the dis-

aster. This can serve to arrest progres-sive collapse of the overall structureby ensuring that the , damaged floorslabs are held together and do not addto the debris load on floors below. Anextremely large deflection in the slabcan be tolerated under this extremecondition.Typical calculation of the mag-nitude of horizontal tie forces re-quired to ensure membrane or cate-nary action is illustrated in Fig. 19.British practice has been generallybased on spans of approximately 17 ft(5.2 m) and the assumption that two-way membrane action will be avail-able, so that the tie force in either di-rection may be cut in half. Typicalrecommendations in the British Code

PCI JOURNAL/January-February 1980 65


25/32

__I ALTa.-- --P

^ T^_1L Lr--------ost SupportL 1 I L2

Catenary Ac t ion L i = 20 ft.W(LI+LI)Y L2= 24 ft.Tn = 8 W=.2K/Ft.T W (L i + L2 ) 2 0= 4ft.gA T= 12.IK /Ft.

IiGra de60#4l22 Way Action Cut in Half Grade 60#4024"

Fig . 19. De te rm ina t ion o f ti e fo r ces a ssu m ing two-way ca tena ry ac t i on .

of Practice call for ties with a capacityof 4 kips per f t (58.5 kN /m).Tests on typical European struc-tures indicate that floor slabs with tiesactually behave somewhat better thanthe catenary analysis indicates. Theassumption of two-way action withone-half of the load carried in each di-rection seems to be borne out by thesetests and work from the Portland Ce-ment Association. There is a need forfurther testing with longer spans andsubstantial variations in length-to-width ratios to determine the correct-ness of the assumption as to lon-gitudinal and transverse distribution.Typical North American applicationwith longer spans generally results insubstantial increases in tie forces.The catenary analysis is again alimit state analysis for dead load andpartial live loads. While the Britishcalculations have been based on a de-flection of 15 percent of the span, itseems reasonable that a maximumlimit such as one-half of the clear storyheight could be adopted.

There are several recommendationsfor the magnitude of the horizontal tieforces. Many of these are based on acatenary analysis for relatively shortspans. Others are based on simplyproviding tie forces equivalent towhat commercial fabricators are nowproviding in buildings. A substantialamount of research is needed to in-vestigate procedures for proper cal-culation of tie forces, and basic criteriafor loads to be carried by the floorsystem should be codified.Provision of horizontal tie forcesimplies proper anchorage of the ties atall supports. It will be necessary toprevent the tie from pulling out at theedge support and to act effectively asa splice at the central support whichmight be destroyed. This will requiresubstantial improvement of tie details.

C a nt i lever/Bridg ing A ct ionThe flank and corner walls in apanel or bearing wall structure are

66


26/32

both the most vulnerable to manytypes of abnormal loads and are thehardest to replace with an alternateload path. Relatively small edgestiffeners and effective longitudinal,transverse, and peripheral ties greatlyincrease the capacity of a corner slabto cantilever over missing supports.Relatively light partitions can providestrong points to assist the cross wallsin supporting cantilever action. Drywall partitions commonly used in theUnited States may not be effective inthis regard.

Proper design of the wall and tiesystem will allow the structure abovethe damaged zone to cantilever overthe missing support. Comprehensivestudies are required to determinewhat level of ties and shear connec-tors are required to fully develop suchcantilever action in all forms of paneland bearing wall construction.

Tests on large concrete panel con-struction at the Portland Cement As-sociation confirm that shear along thehorizontal joints can be critical.Proper details and analysis for shearcapacity are required to ensure thatbrittle shear failures will be avoided.The transverse ties must work withthe vertical ties to provide effectiveclamping action at the joint.The vertical ties can be extremelyimportant in acting as suspender rodsto carry the floors and walls im-mediately over the damaged area. Ifthis is envisioned in design, appropri-ate anchorage must be provided ateach story level for this function.

Diaphragm A c t ionThe peripheral ties are necessary toestablish an edge beam around thestructure at the level of each floor to

provide proper anchorage for the lon-gitudinal and transverse ties, to im-prove cantilever membrane action atthe corners, and to develop diaphragm

action in the floor slab for the entirestory.While most authorities have rec-ommended that the peripheral tieforces be similar in magnitude to thetie force within the edge strip of theadjacent floor, there is no clearrationale for determination of the ap-propriate level of force in the generalcase.

Joint a nd Deta i lsThere is a very clear consensus thatpanel and bearing wall structures

should be designed and constructed tohave joints with adequate strength,stiffness, and ductility. Specific provi-sions which will ensure the strength,stiffness, and ductility of joints arestill very uncertain and need substan-tial additional research. The recom-mendations of Reference 25 provideinterim guidance.The details which will ensure duc-tility at the joints and continuity withthe wall in various systems must bedeveloped and verified. Designersmust be given insight and analyticaltools to indicate not only the mag-nitude of the tie forces but the mostefficient location and method ofplacement of the tie elements con-cerning both ductility and strength.

Design procedures will of necessitybe based on observed behavior ofvarious tie systems in laboratory orprototype tests. It is essential thatthese tests have realistic anchoragedetails and be conducted with thevery large deformations which are as-sociated with collapse level loadings.In general, it is necessary to deter-mine anchorage and splice behavior atlevels far beyond the point whereprevious testing which has formed thebasis for most current design rules hasbeen discontinued.Anchorage details should carefullysimulate prototype applications. Such

PCI JOURNAUJanuary-February 1980 67


27/32

a factor as loss of ductility due to thenotch effect in threaded connectionsmay prevent formation of the catenaryaction.

The large rotations often associatedwith collapse level loads can causethe point of load applications on sup-porting walls to shift and give sub-stantial eccentricities. Overallstrength of the wall as well as the be-havior of the joint will be affected bythis movement.

Very little is known on the properdetailing to develop the catenary forcesystem in structures typical of practicein the United States. Effects of con-finement and variations in steel over-lap should be studied under extremedeformation conditions. The focus ofthe studies should be the develop-ment of efficient details, and designfor efficiency at large deformationsrather than simply an evaluation oftraditional detail patterns. In theinterim, the designer must provide in-genuity to determine acceptable so-lutions.

As an example of a practical tie ap-plication, the large concrete panel re-search project at the Portland CementAssociation studied the use of un-stressed tendons as a highly efficientway of developing resistance to cata-strophic loads. One of the principal ad-vantages of these tendons is that theycan be embedded in joints and areflexible enough to allow for normaltolerances in abutting the lengths ofthe precast units forming the joint.Further experimental programs areneeded to study the requirements forstrength, stiffness, and rotation of thejoints in typical panel and bearingwall structures after a local collapse.Such studies, will provide a measureof the strength and ductility require-ments for the joints, which will be abasic factor in assessing the adequacyof.any joint. Actual joint tests shouldmodel the complex state of load whichexists on the joint as a three-dimen-

sional body. This indicates that somework will need to be performed at es-sentially full scale in the laboratoryand not on models until complete cor-relations are obtained between pro-totype and model behavior.

Joint C la s s i fica t ionBecause of the wide range of joint-

ing techniques for the large number oftypes of panel and bearing wallstructures, a classification systemshould be established to ensure thatrepresentative joint types areevaluated. Substantial development ofsuch classification systems is under-way in NSF-sponsored studies in theRANN program in earthquake en-gineering.31A large number of systems utilizedin the United States use dry orsemidry joints in contrast to the prev-alent wet joint in Europe. The utili-zation of interlocking hooked bars ismuch less in the United States than inEurope. As a result, relatively little ofEuropean literature on performance ofjoint systems is applicable to Ameri-can practice. However, behavioralideas established in European testscan be of substantial benefit in asses-sing the rotational needs and the fac-tors which contribute substantially tosuch rotational development.The common use of saw-cut ex-truded floor slabs in the United Statesplaces emphasis on the determinationof the importance of grouting in andbetween panels and on the adequacyof grouting ties into the cores. Testswill be required to ensure sufficientbond at the level of deformations ex-pected.Wide use of tack-welded connec-tions raises questions concerning thebrittleness of the connections withimproper quality control. In connec-tion with the notch sensitivity ofthreaded connections, this may pro-

68


28/32

duce zones of weakness under im-pact-type loads.In some cases tie forces may de-pend on erection and manufacturingtolerances. This is particularly true inlateral forces due to out-of-plumbmembers and in floor plank memberswhere relatively inadequate bearingexists under the ends even beforelarge deformations take place. A sur-vey is needed of actual field toler-ances for all types of construction.Throughout the study on details injoints, it is important to get input onthe way that details will affect con-struction practice. Success of any re-search studies in this area will dependon active involvement of designersand constructors as well as competentresearch professionals.

Joints in P recas tP a nel Stru ctu res

In addition to the general questionsregarding joint strength, stiffness, andductility, field experience has indi-cated that the shear strength of hol-low-core slabs immediately adjacentto the joints can be a critical factor incollapse conditions.

Attention should be paid to the pos-sibility of development of additionalshear strength by filling the voidswith mortar or by applying shearreinforcement in the webs. The lon-gitudinal and transverse tie systemscan be used to resist the shear forcesand the section can be checked byshear-friction theory.

Ma sonry Cons t ru c t iostate-of-the-art in design anddetailing of joints in masonry struc-tures for strength, stiffness, and duc-tility is far behind the state-of-the-artin large concrete panel construction.One of the main weaknesses in as-

sessing the behavior of an unrein-forced or reinforced masonry structureat collapse load levels is the funda-mental lack of a behavioral orientedstrength design method for masonry.The absence of an accepted basic re-lation between axial compression andflexure in the presence of strain gra-dient precludes the development ofbasic strength and deformationtheories for masonry, which areneeded before general analyticaltreatment at limit state conditions canbe carried out.

Masonry differs substantially fromreinforced concrete in fundamentalbehavior, in planes of weakness, andin stiffness distributions. The con-struction process makes it easy to de-velop horizontal ties and it is possibleMal me uaSIC patterIl IUr proviu 1 9 allalternate path for loads in masonrystructures might depend more on sus-pending cables from horizontal can-tilever sections than is generally donewith concrete structures.

Specific details for effective jointingin masonry must consider constructionsequence and type of inspection to beprovided. The type of joints selectedmay greatly influence the cost of thestructure if it hampers the productiv-ity of the workmen.The use of unreinforced or partiallyreinforced masonry in high-rise con-struction seems to be contrary to theidea of improved overall structuralintegrity. In addition, the question ofopenings and connections of lightlyreinforced lintels in otherwise heavilyreinforced wall panels provides sub-stantial planes of weakness. Inspec-tion of damage after the San Fernandoearthquake indicated variations in ef-ficiency of grouting the reinforcementinto the masonry cells. Certain typesof masonry bonded well to thegrouted bars while others did not.One of the characteristics ofmasonry construction is that all typesof floor slabs are possible and are

PC I JOURNA L/Jan uary-February 1980 69


29/32

used. Precast and cast-in-place con-crete slabs, bar joists, corrugatedmetal decking, and wood all havebeen utilized for flooring systems.The development of alternate loadpaths and sufficient tie forces in thesewide varieties of structural systemsneeds extensive evaluation. Someconstruction systems may be shown tobe undesirable for this application.

The present status of information onthe behavior at large deformations ofthe wide variety of joints possible isextremely meager. There is a very lowpossibility of being able to assess theactual effectiveness of typical ties andgrouting systems for extremely largerotations in development of catenaryactions from test results currently re-ported in the United States. While anumber of research programs inmasonry have been directed towardsseismic resistance, the joint details aregenerally not typical of the types ofdetails widely used throughout theUnited States. The majority of dataconcerns joints which are muchheavier reinforced and have substan-tially more effective grouting thanfound in practice.

Research work in this area can un-doubtedly benefit from close coordi-nation with research programs under-way to improve the seismic resistanceof masonry structures at several agen-cies and universities on the WestCoast .

Masonry structures have a greatdeal of potential for developing resis-tance to progressive collapse. Place-ment of reinforcement between walland slab units should be easier than inprecast units. Provision of reinforcedbond beams and tie beams shouldsubstantially increase the resistance toprogressive collapse. However, it isnecessary to take a broad look atmasonry structures and consider notonly ideal behavioral characteristics,but effective quality assurance pro-grams.

EconomicsEngineers experienced in design

and construction of panel and bearingwall structures which have tie forcesdeemed sufficient to greatly improvethe overall structural integrity haveindicated that the effort to obtain re-quired toughness and ductility is notvery expensive. Designers report thaton the first project in which they areproviding extra tie forces, there issome extra design cost. However, thisquickly diminishes on subsequentprojects as personnel become familiarwith the concept and typical detailsare repeated.Construction costs could increasefrom 0 to 10 percent when tie forcesand detailing are provided to improveoverall structural integrity. There isgenerally a very small cost if the de-sign starts with the premise of pro-viding suitable tie forces, while thecost is substantially higher if anexisting system which was designedwith very inefficient or insufficientjoints has to be converted to meet thenew requirements.

Fu tu re Direct ionsAfter extensive discussion of variousstrategies for minimizing the risk ofprogressive collapse in panel and

bearing wall structures, the attendeesat a major national workshop on theproblem3 2 felt that the most practicalprocedure was to adopt a positive re-quirement that would encourage thedesigner to develop an integralthree-dimensional structural systemfor carrying gravity and lateral loads.Such an integral system would rely onproper longitudinal, transverse, verti-cal, and peripheral ties, to ensure thatall members interacted and that a highlevel of ductility and continuity wasobtained.

all


30/32

After substantial discussion and inresponse to a specific request for asense of the meeting as to what an ap-.propriate direction for the ACIBuilding Code to take might be, theworkshop passed the following reso-lution with an approximately 85 per-cent affirmative vote:"With regard to large panel struc-tures, it is agreed that:

1. Satisfactory control over pro-gressive collapse can be pro-vided by embodying in ACI318 requirements for hori-zontal and vertical ties.2. These Code requirementscan be of a qualitative nature.3. Commentary provisions canbe quantitative either spe-cifically or by reference.

4. No reference need be madeto `progressive collapse'either in the Code or Com-mentary."

In discussing the motion it was

brought out that this positive approachemphasizing the beneficial effect oftie forces and providing guidance tothe designer was extremely appro-priate for a material code specificationwhen there is no general code re-quirement in an overall building codewhich applies to all materials. It wasthe consensus of the group that an ac-tion of this sort would greatly reducethe danger of a progressive collapse ina large concrete panel structure andprevent the occurrence of a "RonanPoint" in the United States. It wasrecognized that substantial effort wasrequired to provide the quantitativevalues for tie forces for the Commen-tary.This appears to be the way the in-dustry and profession is moving. Thegeneral principles are clear. The de-tails are fuzzy. We must provide over-all structural integrity in our bearingwall structures. It is now up to our in-genuity to come up with details thatdo this dependably and economically.

N O T E: A l is t of references on bearingwa ll bu ildings is p rovided on the nex tcoup le of pag es.

Discussion of this paper is invited.Please forward your comments toPCI Headquarters by July 1, 1980.

P C I JOU RN A L/Janua ry-February 1980 71


31/32

R E F E R E N C E S1. Griffiths, H., Pugsley, A., and S aun -ders, 0., Report of the Inquiry into theCollapse of Flats at Ronan Point,

Canning Town, Her Majesty's Station-ery Office, London, England, 1968 .2. Leyendecker, E. V., Breen, J. E.,Som es, N. F., and Swatta, M., A bnor-mal Loading on Buildings and Pro-gressive Collapse. An Annotated Bib-liography, Na tional Bureau of Stan-dards, Building Science S eries No. 67,May 1975.3. Firnkas, Sepp, "Concrete PanelBuilding System: Progressive Col-lapse Analyzed," Civil Engineering(ASCE) , Novemb er 1969, pp. 60-62.4. Popoff, A., Jr., "Stability of PrecastConcrete Systems Buildings," Pro-ceedings V. III, ASC E-IABSE Inter-national Conference o n Tall Buildings,Lehigh U niversity, 19 72, pp. 571-583.5. Popoff, A., Jr. , "Design C onsiderationsfor a Precast Prestressed ApartmentBuildingDesign Against ProgressiveCollapse," PCI JOU R NA L, V. 20, No.2, March-April 197 5, pp. 44-57.6. Rules and Regulations Relating to Re-

sistance to Progressive Collapse UnderExtreme Local Loads, Housing andDevelopment Administration, De-partme nt of Buildings, The City Rec-ord, New York, August 1973.

7. "Flats Con structed with Precast C on-crete Panels. Appraisal andStrengthening of Existing HighBlocks: Design of New Blocks,"Ministry of Ho using and Local Gov-ernment , Circular 62/68, London En-gland, Novemb er 15, 1968.8. Proposed Amendments to the Build-ings Regulations, 1965, Ministry of

Hou sing and Local Governm ent, Lon-don, England, May 30, 19 69.9. Despeyroux, J., "1'Effondrement deI'Immeuble de Ronan Point et sesCo nsequences en Ma tiere de Codif i-cation" (The Collapse of the RonanPoint Building and I ts Consequences

with Regard to C ode Writing), Annalesde l'Institut Technique du Batiment etdes Travaux Publics No. 263,Novemb er 1969, pp. 1800-1803.10. R obinson, J . R . , "Observat ion sur lesConclusions du Rapport de la Com-mission d'Enquete de Ronan Point"(Observation on the Con clusion of theR eport of the Inquiry Com mission onRonan Point), Annales de l'InstitutTechnique du Batiment et des TravauxPublics No. 263, Novemb er 1969, pp.1797-1799.

11. Saillard, Y., "Le Comportement del 'Imm euble de Ronan P oint en Com -paraison des Principles de Base desRecommandations Internationales`Structures en Panneaux' du ComiteEuropeen du Beton" (Behavior of theR onan Point Building Comp ared to theBasic Principles of the InternationalR ecom men dation "Panel Structures"of the Comite Europeen du Beton),Annales de l'Institut Technique duBatiment et des Travaux Publics No.263, Novemb er 1969, pp. 1804-1806.

12. "Recommandations InternationalesU nifiees Pou r le Calcul et 1'Executiondes Constructions en Panneaux As-sembles de Grand Format" (Interna-t ional Recom mendat ions for the D e-sign and Construction of Large PanelStructures), Comite Europeen duBeton, Bulletin No. 60, Paris, April1967; English Translation by C. V.Am erogen, Cem ent and Concrete As-sociation, No. 137, London, July 1968.

13. "Building Regulations: The FifthAm endment and the Freedom to D e-sign," Editorial Comment, Concrete(London), V. 3, No. 9, Septemb er 1969 ,pp. 343-344.14. "Large Panel Structures," EditorialCom m ent , Concrete (London), V. 3,No. 6, June 19 69, p. 213.15. "Building Regulations and PublicHealth and Safety," Editorial Com-ment , Concrete (London), V. 4, No. 5,May 197 0, p. 165.

72


32/32

16. Structural Stability and the Preven-tion of Progressive Collapse, RP/68/01,The Institution of Structural En-gineers , London, England, Decemb er1968.

17. Structural Design Requirements to In-crease Resistance of Buildings to Pro-gressive Collapse, Proposed HUDHandbook, U .S. Depar tment of H ous-ing and Urban Development,November 1973.

18. Burnett, E. F. P., Abnormal Loadingsand the Safety of Buildings, R eport forthe Structures Section, NationalBureau of Standards, Washington,D.C., August 1973.

19. Burnett, E. F. P., Somes, N. F., andLeyendecker, E. V., ResidentialBuildings and Gas-Related Explosions,NBSIR 73-208, Center for BuildingTechn ology, National Bureau of Stan-dards, Washington, D.C., June 19 73.

20. Fribush, S. L., Bowser, D., and C hap-man, R., Estimates of Vehicular Colli-sions with Multistory ResidentialBuildings, NBSIR 73-175, NationalBureau of Standards, Washington,D.C., April 1973.

21 Somes N. F., Abnormal Loading onBuildings and Progressive Collapse,NBSIR 73-221, Center for BuildingTechno logy, National Bureau 'of Stan-dards, Washington, D.C ., May 19 73.22. Schu ltz, D . M., and Fintel, M., Report1: Loading ConditionsDesign andConstruction of Large Panel ConcreteStructures, Office of Policy Develop-ment and Research, Department ofHousing and Urban Development,Washington, D.C., April 1975.

23. Fintel, M., and Schultz, D. M., "APhilosophy for S tructural Integrity ofLarge Panel Buildings," PCI JOUR-NA L, V. 21, No. 3, May-June 1 97 6, pp.46-69.

24. Taylor, D. A., Progressive Collapse,Technical Paper N o. 450, Division o fBuilding R esearch, National R esearchCouncil of Canada, August 1975.

25. PCI Committee on Precast BearingWall Buildings, "Considerations forthe Design of Precast Bearing WallBuildings to Withstand AbnormalLoads," PCI JOU R NA L, V. 21, No. 2,March-April 197 6, pp. 18-51. (See alsolist of references in this report.)

26. ACI C omm ittee 531, "Specification forConcrete Masonry Construction," ACIJournal, V. 72, No. 11, November 197 5.

27. McGu ire, W., and Leyende cker, E. V.,Analysis of Nonreinforced MasonryBuilding Response to AbnormalLoading and Resistance to ProgressiveCollapse, NB SIR 74-526, Center forBuilding Technology, National Bureauof Standards, Washington, D.C.,November 1974.

28. Morton, J., Dav ies, S. R ., and H endry,A. W ., "The S tability of Load-BearingBrickwork Structures Following A cci-dental Damage to a Major BearingWall or Pier," Proceedings, SecondInternational Brick Masonry Confer-ence, British Ceramic Research As-sociation, 1971, pp. 276-281.

29. Sinha, B. P, and H endry, A. W., "TheStability of a Five-Story BrickworkCross-Wall Structure Following theR emov al of a Section of a Main Load-Bearing Wall," The Structural En-gineer (London), V. 49, No. 1 0, Oc-tober 197 1, pp. 467-474.30. Sinha, B. P., Ma urenb recher, A., andHendry, A. W., "Mod el and Full-ScaleTests on a Five-Story Cross-WallStructure under Lateral Loading,"Proceedings, Second InternationalBrick Masonry Conference, BritishCeramic R esearch A ssociation, 197 1.

31. Zeck, U . I. , Joints in Large Panel Pre-cast Concrete Structures, Seismic Re-sistance of Precast Concrete PanelBuildings Report No. 1 , MIT D epart-me nt of Civil Engineering, PublicationNo. R76-16, January 1976.32. "Workshop on Prog ressive Co llapse ofBuilding Structures," The U niversityof Texas at Austin, Novemb er 197 5.

Documents

Breen Precast