8/13/2019 03(2)0051

1/9

51

EARTHQUAKES AND TAL L BUIL DINGS +P a u l C . J e n n i n g s

Editor 8s Note: Professor Jenn ings 1 contribution is, he says, relativelynon-technical, and is aimed primarily at laymen and engineers not directlyengaged in structural engineering . Papers of this sort are so rare thatwe are delighted to offer this one from so respected an authority asProfessor Jennings, not only for those of our members that the Professorintended should benefit, but also for structural engineers. All of us ,deeply involved with technica l deta il, tend to lose something of ourappreciation of the overall problem.

In troduct ionThere seems little doubt that the responseof tall buildings to strong earthquake motionoccupies a special place in the public mind.

In April of last year in Los Angele s, when pop ular attention was focused again on the earthquake problem by the dire predictions of avariety of soothsayers, the two questions thatinvariably arose in conversa tion were Willthere really be an earthquake in April? andWhat about the tall buildings in downtownLos Angeles ? April passed without a largeshock, so the first question has been resolved.A big earthquake will almost certainly occur

San Francisco and the Pacific Nort hwest.Buildings over about 6 stories did not appearin Tokyo until the construction of the 36-storey Kasumigaseki Building in 1966 . TheTower Latino Americano introduced tall buildings to Mexico City in 1957 and a 40-storeybuilding is under construction in Lima, Peru.The few buildings in Caracas in the 30-storeyrange also are recent additions. There are nobuildings of this size yet in chile or inNew Zealand, and the tallest buildinqs inAnchorage at the time of the Alaska earthquake of 196 4, and now, are fourteen stories.The trend towards tall buildings is so strong,however, that it seems likely that every major

8/13/2019 03(2)0051

2/9

52frame structure are shown in Figure 1.

Tall buildings share many Similar propertiesand it is quickly observed that they differ fromshort buildings in more way s than just inheight. The tall buildings have more regulargeometry and are nearly all symmetrical,usually being rectangular. This results fromeconomic and architectural considerations. Theconstant cross-sectional area of the buildingsleads to a weight distribution that is approximately uniform with height. In addition, coderequirements usually dictate that all the tallbuildings have a moment resisting frame ratherthan, or in addition to, the shear walls thatare often used in shorter build ings. As wouldbe expected, the weight distribution and thetype of structural system are reflected in thedynamic properties of the building.

Both steel frame and reinforced concreteframe construction have been used for high-risebuildings and most engineers agree that earthquake-resistant structures can be made fromboth of these construction materi als. Statistically speaking, however, nearly all of thebuildings in the 30-storey and up class aresteel-frame structures. The dynamic propertiesof both types are similar, but for structuresof the same height, the reinforced concretebuildings are usually stiffer than buildingsframed of steel.

Mexican earthquake of July 1 957, with insignificant damage even though several shorterbuildings collapsed. The earthquake occurredafter the structural frame was complete, butbefore the glass was completely installed.None of the installed panes of glass wasbroken. In the Caraca s, Venezuela earthquakeof July 1967 , the buildings in the 30-storeyrange all came through the earthquake with nosignificant damage, whereas five buildings ofintermediate height collapsed with major lossof life.

Longer ag o, the Call Building in SanFrancisco withstood the Magnitude 8.2 ea rthquake of 1906 with almost no dama ge, as didother tall buildings and building frames-Figure 3 shows the Call Building and MarketStreet after the earthquake and during thefire. The building appears to be completelyundamaged by the vibration. The damage inthe right foreground of Figure 3 is thoughtto be from dynamiting, a technique used totry to control the spread of the fires.Effects of Height on Earthquake Response

Because really strong earthquake motionsare so rare at any given site, it does notmake sense to design all structures to with stand the strongest earthquake without da mage.The currently accepted design philosophy forearthquake-resistant structures is that the

8/13/2019 03(2)0051

3/9

ing, - these are the properties needed topredict the earthquake response of tall buildings.Many more tests are needed, but when the dataavailable are gathered together and examinedsome patterns do emerge as wil l be seen fromexamination of Table I . ^ It is found, forexample,that the ratios of the higher frequencies to the fundamental tend to followvery nearly the odd integers; Table I showsthe ratios for the second through the sixthNorth-South frequencies of the Union BankBuilding to be 2.9, 5.09, 7.18, 9.25 and 11. 5.In this respect the building has properties muchlike a simple beam that deforms only in shear.It is found also that the mode shapes of thetall building s, particularly those associatedwith the higher frequencies, are quite similar.As an example, Figure 4 shows the second modeshapes of a number of tall b ui l di n gs . ^ Forcomparison. Figure 5 shows the first three modesof a 15-storey building in one direction atdifferent stages of construction, following thecompletion of the structural fr a m e . ^ Althoughthere are some other factors involved, comparison of Figures 4 and 5 does suggest thatthe differences in mode shapes of differentbuildings is not much larger than the errorsinvolved in trying to find the actual modeshapes of a single building.

The similarities in frequencies and modeshapes appear to hold fairly true for all tall

53arly with height and is determine dessentially by the fundamental mode.

(4) The maxim um value of the accel erati on onthe top of tall buildings ten ds, on theaverage, to decrease approximately asthe inverse square root of the height.For a very tall building, the motion onthe top is expected to be less severethan at the base.Computer calculations of the response of

tall buildings to recorded earthquake motiontend to support these conclusions and it isinteresting to note in this regard also thatProfessor Charles Derleth of the University ofCalifornia at Berkeley reported in his studyof the 1906 earthquake that M.. . the base ofsuch a structure (relatively tall buildi ngs)oscillated and moved with the earth while thetop tried to remain quiet. This statement issubstantiated by the general evidence that lessshock was felt in the upper stories of highbuildings than in floor levels near the street.It is the view of the writer that this effectwas somewhat fortuitous for buildi ngs so short bypresent standards, but it is a good descriptionof the type of phenomena that is expected tobe more common as taller buildings are considered. This effect is illustrated by Figure 6which shows, for several building s, the ratioof the maximum acceleration at the top of the

8/13/2019 03(2)0051

4/9

54structed buildings , including tall build ings,can withstand strong earthquake shaking without excessively endangering the occupan ts.On the other hand, earthquakes have demonstrated repeatedly that under-designedbuildings, or poorly built buildings, of allheights are a definite hazard during strongearthquake motions. The earthquake successof tall buildings is reinforced by studiesusing average properties of earthquakes andtypical properties of tall buildi ngs. Thesestudies show that the properties of tallbuildings are such that there are no specialearthquake hazards that arise simply as aconsequence of height. In fact, for the verytall buildings the lateral forces required bythe codes in Los Angeles are such that thewind forces exceed the code earthquake loadings by a substantial amount. This increasesthe margin of safety against earthquake motionsto the extent that the amount of yielding instrong shaking is quite limited; much less thanfor most shorter structures. These reasonshelp the writer to conclude that carefullydesigned and constructed tall buil ding s canbe built safely in seismic regions.

In a general sense, the earthquakeresistance of most of the cities of the worldis improving. The increase in understandingof the earthquake,motions, a better knowledgeof the dynamic properties of buildin gs, andthe increasing capacity to perform complex

requirements of wind resist ance.Designf Tali Building s

Because tall buildings are relatively newin seismic areas , the engineering professionhas limited experience with their earthquakeperformance. Furthermore the existing building codes , based in substantial part on earthquake experience with much smaller buildings,are not completely applicable to the design oftall structures. Responsible engineers areaware ofthis,of course, and give the designof tall buildings the special care that theirimportance deserves. The expanding knowledgeabout earthqua kes, structures and calculationtechniques, and the growing capabilities of thedigital computer place the engineers in anincreasingly better position to design tallbuildings and other structures to withstandstrong earthquakes successfully.In the design of tall bui ldings, the

structural engineers nearly alway s make moreextensive analyses than required by the code.One technique of analysis that seems certainto become more common is the use of thedigital computer to calculate the response ofthe proposed design to recorded earthquakemotions and artificially generated motionsthat model the shaking expected in the event ofa great earthquake. (The motion in theepicentral area of a magnitude 8 or greaterearthquake has not yet beenrecorded).

8/13/2019 03(2)0051

5/9

is improving, it would be a mistake to have toobright a picture of the effects of strong earthquakes. Strong earthqua kes are a frighteningand hazardous experience and are likely toremain so for many years to com e. We shouldtry, however, to put the various sources ofhazard into proper perspective so we can focusour efforts on those that are most serious.For a variety of reasons, the very tall buildings appear to be in a relatively good po sitionin this regard, especially when compared tothe tens of thousands of old buildings builtbefore earthquake-resistant design was firstrequired . A sizeable fraction of these buil dings are substandard and represent, in thewriter*s opinion, the greatest hazard duringthe next strong earthquake in the metropolit anareas of California or New Zealand. The nextlargest hazard probably comes from speculativelybuilt buildings of intermediate height.

55REFERENCES1. Jennings, P.C., Spectrum Techniques for

Tall Buildin gs , Proceedings of the FourthWorld Conference on Earthquake Engineering,Santiago, Chile, January, 1969.2. Rea, D., Bouwkamp, J.G., and Clough, R.W. ,The Dynamic Behaviour of Steel Frame andTruss Buildings , Structural Engineering

Laboratory Report No.66-24, College ofEngineering, University of California,Berkeley, Sept., 1966.3. Housner, G.W., Behaviour of StructuresDuring Earthquak es , Journal of theEngineering Mechanics Division, AmericanSociety of Civil Engineers, Vol . 85, No .EM4,Oct., 1959.

8/13/2019 03(2)0051

6/9

56

8/13/2019 03(2)0051

7/9

57

8/13/2019 03(2)0051

8/9

58

L

8/13/2019 03(2)0051

9/9

TABLE 1MEASURED PERIODS OF TALL FRAMED STRUCTURES

Period RatiosBuildin g Stori es Type Plan Di recti on T_ sec.) T,/N T _ A T./T. T. /T T / T c T /T. 1 1 1 2 1 3 1 4 1 5 1 6Dimensionsft.)

Jet Propulsion Steel 220 X 40 Long. 0 .99 0 .110 2 ,97 5 .02 7. 2Laboratory, Pasadena 9 Frame Tr an s. 1 .02 0 .113 3 .26 6 .37Union Bank 40 Steel 196 X 98 Lon g. 3 .10 0 .078 2 .90 5 .06 7. 18 9 .25 11 .5Los Angeles Frame Trans. 3 .54 0 .088 3 .03 5 .58 7. 96 10 .5 13 .1Canadian Imperial 44 Steel 140 X 100 Long. 4 .65 0 .106 2 .90 5 .06 7. 71 9 .61 11 .9Bank of Comm erce,Montreal Frame Trans. 4 .65 0 .106 2 .72 4 .78 7. 06 8 .93 11 .1CIL House, 34 Steel 168 X 112 Long. 4 .46 0 .131 3 .05 5 .58 8. 5 11 .1 13 .6Montreal Frame T r a n s . 3 .94 0 .116 2 .84 5 .04 7. 36 9 .85Univ. of Cal, 15 Steel 107 X 107 E-W 1 .18 0 .078 2 .65 4 .60 6. 54 8 .47Berkel ey, Frame N-S 1 .18 0 .078 2 .65 4 .60 6. 54 8 .47Medical Center,East Bldg. Summer1964 Test s, frameand slabs only.)Alexander Bldg, 15 Steel 68 X 60 Long. 1 .27 0 .085 3 .10 5 .29 7. 7San Francisco Frame Trans. 1 .37 0 .091 3 .05 5 .27 7. 21Built in 1920*s)Ottawa Post Office 10 Reinforced Long. 0 .695 0 .070 2 .87 4 .45 6. 30 7 .83Building Concrete Trans. 0 .592 0 .059 2 .77 4 .16 5. 8 7 .30Frame 266 X 74Canadian Dept. of 18 Steel and Long. 0 .99 0 .058 3 .75 5 .17Health and Welfare Concrete 140 X 88 Trans. 1 .28 0 .075 4 .26 6 .55Building, Ottawa Core, SteelColumns,

ReinforcedConcrete FloorSlabsSan Diego Gas and 21 Steel 70 X 180 Long. 2 .54 0 .121 2 .8 5 .1 7. 6 10. 8 13 .0Electric Co. Bldg, Frame T r a n s . 2 .63 0 .125 3 .2 6 .0 9. 0 11 .3 13 .0San Diego.

eno