100% found this document useful (1 vote)
310 views288 pages

Tall Building Structures-Analysis & Design

Bryan Stafford Smith and Alex Coull

Uploaded by

elif firuze
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
100% found this document useful (1 vote)
310 views288 pages

Tall Building Structures-Analysis & Design

Bryan Stafford Smith and Alex Coull

Uploaded by

elif firuze
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 288
TALL BUILDING STRUCTURES: ANALYSIS AND DESIGN ANTERSCIENCE PUBLICATION JOHN WILEY & SONS, INC. New York * Chichester * Brisbane + Toronto * Singapore techaoiogy will enable ox wotts of endaring scholly vale Ta pon at long 4 recognition ofthe imponunee of preserving what has been len. isa policy of John Wiey & Sons. Ins fave books, ‘of encurng value published i he United States pried or acid-fee paper. and we exer ou best effort tha ed ‘Sepssigin © 1991 oy sonn Wiley & Sot. nc. All rights reserved, Pubhed sinwlaneousy in Cunads Reproduction or tanslaton of any part this work beyond tat permitted by Secon 107 or 108 ofthe 1976 United States Cepynght Ac wahout tke permission ofthe copyright ouner is unlwil. Requests for emission or fender infomation shouldbe adetesed to ‘the Pemnissons Depanmen. Jobs Wiley & Sonn Ine Library of Congress Ctaoging i Pablcation Daa: Sta’ Stith Bryan, Tal ulding ructure: lis and design nan Saford Smith. ten Cul Bon. “A Wily-Imersceneepublicton“ Includes bisliograpicalreternce. {spy 0471—$1257-0 1. Tall buldings-Design an cor ngineeting. 1. Coull, Alen 4. Tie THIOIL.S59_ 1991 708620 0.13007 cp Printed and bound inthe United Stet of Amica Wysr6sa3 To Beny and Frances PREFACE ‘This book is the indirect outcome of 25 years of research on tll building structures by the two authors. It began with their liaison in the mid-1960s at the University ‘of Southampron, England, and tas since continued in their respective Universities, of Surrey, McGill, Strathclyde, and Glasgow. At the commencement of the period, the evolution of nidically new structural forms gave great stimulus to devising appropriate methods of analysis. In the sue- ceeding quarter-century there have been great advances in the design and construc- tion of tall buildings throughout the world, and in the associsted development of analytical techniques. In the early days, approximate techniques were being devised for specific, largely two-dimensional, structural forms, and the analysis of complex three-di- ‘mensional systems represented a formidable challenge. Since then, there have been significant advances in both computer hardware and software: the power of com- puters has increased dramatically, and a large number of comprehensive general purpose analysis programs have been developed, based on the stiffness method of, ‘analysis. In principle a least, itis now theoretically possible to analyze accurately Virtually any complex elastic structure, the only constraints being the capacity of the available computer, time, and cos. However, the great power of this analytical facility has 1o be handled judi- ciously. Real building structures are s0 complex that even an elaborate computa- tional model will bea considerable simplification, and the resuits from an analysis will always be approximate, being at best only as good asthe cuality ofthe chosen ‘model and method of analysis. It is thus imperative to be able to devise an ana- Iytical mocel of the real structure that will represent an predict with appropriate accuracy, and as efficienlly and economically as possible, the response of the building to the anticipated forces. Models required for the early stages of design will often be of a different, lower level of sophistication than those for checking the final design. The task of siructural modeling is arguably the most difficult one facing the sinictural enalyst, requiring critical judgment and a sound knowledge of the struc- tural behavior of tll building components and assemblies. Also, the resulting data from the analysis mast be interpreted end appraised with discemment for use with the ral stricture, in oner to serve as a reasonable basis for making design deci- ‘The rapid advances in the pest quarter-century have slowed up, and the era is now one of consolidation and utilization of research findings. However, the ma vill prerace: jority of the research findings still exit only in the form of papers in research joumals, which are not generally availae or familiar to the design engineer, There is a need to digest and to bring together in a unified and coherent form the main corpus of knowledge that bas been accumulated and to disseminate itt the struc. ‘ural engineering profession. This task forms the main objective of this volume, tis not possible to deal in a compretensive manner in a single volume with all aspects of tall building design and constuction, and aticnion has been focused on the building structure. Such important related topics as foundation design, con, struction methods. fie resistance, planting, and economics have had to be omit, ted. The intention has been to concentrae on the structural aspects that are partic. ularly affected by the quality of tallness: topics that are of equal relevance to low. rise buildings have generally not been considered in any pends on the ductility of the type and material of the structure, the possibility of fhilure of the vertical load system, the degree of redundancy of the system that \would allow some localized failures without overs failure, and the ability of the secondary system, in the case of dual systems, 10 stabilize the building When the primary system suffers significant damage. ‘The factor W'is normally the total dead load of the building. ‘The value of V from Eq. (3.3) gives the magnitude of the total base shear that rust be distributed over the height of the structure for the equivalent static anal ysis Distribution of Total Base Shear. Having determined a valve for the total base shear itis necessary, in order to proceed with the analysis, to allocate the base shear as effective horizontal loads at the various floor leve's. In deciding on ‘an appropriate distribution for the horizontal load the following factors are consid- cred, 1. The effective toad ata loor level isequal to the product of the mass assigned {0 that floor and the horizontal acceleration at that level ‘The maximum acceleration at any level of the structure in the fundamental mode is proportional to its horizontal displacement in that mode, 3. The fundamental mode for a regular structure, consisting of shear walls and frames, is approximately linear from the base. A reasonable distribution ofthe tou! base shear V throughout the height would be im aocondance witha linear acceleration distribution, as given by wi Zw Rey (3.5) ‘where ; and w, are those portions of W assigned te levels / and x, respectively: that is, the weight at or adjacent to levels /and.x, and assigned to those levels for the purpose of the analysis. For structures whose weight is distibuted uniformly over their height the hor ‘zontal load distribution resulting from Eq. (3.5) forms atriangle, with s maximum, value at the top. Such a distribution has been found to be appropriate for buildings of relatively stocky proportions where only the fundamental mode is significant. In more slender. longer period buildings. however, higher modes become signif icant, casing a greater propnrtinn of the total horizontal inertia forces to act near the top: the intensity ofthis effect is related to the period ofthe building. Conse- ‘quently. ths is reflected in the UBC (3.2), and in other Codes. by applying a part ‘of the total loading as a concentrated horizontal force F, at the top of the building. ‘The remainder of the total base shear is then distributed over the height of the building as an inverted triangle. Torsion in any story of the building is prescribed in the UBC [3.2]. to be taken as the product of the story shear and an eccentricity resulting from the addition of a calculated eccentricity ofthe mass above, from the center of rigidity of the story. and an accidental eccertrcity of 5% ofthe plan dimension of the building perpen- dicular to the direction of the force being considered. If torsional iregularites eist, the accidental eccentricity isto be increased by an amplification factor re- lating the maximum story drift at one end of the structure to the average of the story drifts ofthe two ends of the structure. Explanatory material and related technical information useful to the designer in the application of the cesign procedure for this equivalent static approach is pro- ‘vided in the tentative commentary of the 1988 earthquake code of the Structural Engineers Association of California (3.12). ‘The Applied Technology Council (ATC) produced a report in 1978 with sec- fond printing in 1984, Tentative Provisions jor the Development of Seismic Regu laiions for Buildings, ATC 3-06 Amended {3.13}, for the consideration of building authorities across the United States. Its recommendations indicate the likely de ee velopment in the equivalent lateral force procedures of the major Codes. Details in the approach of the ATC 3-06 have been reviewed by Berg [3.14]. Many of the ATC’ provisions have been used by SEAOC, and consequently the Intemational Conference of Building Officials, ay well as the National Research Council of Canada, as key resource documents to develop their new code editions, Recom= mended Lateral Force Requirements and Tenuative Commentary (3.12), Uniform Building Code [3.2], ané National Building Code of Canada (3.7, 3.8}, respec= tively. 3.3.2. Modal Analysis Procedure “The equivalent static load type of analysis is suitable for the majority of hi siructures. If, however, either the lateral load resisting clements or the vertical cistibution of mass are significantly integular over the height of the building, as in buildings with large foor-to-floor variations of ioternel configuration, or with setbacks, an analysis that takes greater consideration of the dynamic characterst of the building must be made. Usually. in such cases, a modal analysis would be sppropriate ‘A detailed explanation of the theory and procedure of modal analysis is given in Chapter 17 and in other texts [3.15, 3.16]. Reviewing the method briefly, how= ever, in a modal analysis a lumped mass mode! of the building with horizontal degrees of freedom at each floor is analyzed to determine the modal shapes and modal frequencies of vibration. The results are then used in conjunction with an earthquake decign response spectrum, and estimates of the modal damping. 10. determine the probable maximum response of the stmcture from the combined: effect of its various modes of oscillation Buildings in which the mass at the floor levels is highly eccentric from the corresponding centers of resistance will be subjected to torque. causing the pos- ¥ Of significant torsional vibrations and of coupling between the lateral and torsional mode. The modal method can also be applied to the analysis of such a building, by adding to the structural model a third, rotational, degree of freedom. st each floor level ‘The modal method is applicable, in the strictest sense, only 10 linear elastic systems. Consequently, the results for a building structure's response are, at best, only an epproximate estimate, because of its typically being designed to suffer significant inclasti: deformations in only moderate earthquakes. More accurate valves of response may be obtained for some buildings by the modal analysis method, using modified design response specira for inelastic systems [3.16]. 3.4 COMBINATIONS OF LOADING Methods of accounting for load combinations and their effects on the design of members vary according to the Code used and to the design philosophy. The com- bination of dead and live loading with reductions inthe live loading to allow for the improbability of fully louded tributary areas, and considering paitemed tive leading for the worst effects, have already been discussed, ‘The approaches to combinations of loading by two non American Codes, the Uniform Building Code (3.2) and the National Building Code of Canada (3.7). will be referred to as representative of many of the major building Codes. 3.4.1 Working Stress Design ‘The UBC and NBCC both assume that wind and earthguoke louling need not be taken to act simultaneously. The UBC considers the improbability of extreme grav- ity and wind, or earthquake, loadings acting simultaneously by allowing for the combination 2 one-third increase in the permissible working stresses. which is equivalent to a 25% reduction in the sum of the gravity and wind. or earthquake. loading, The NBCC approach to allowing for the improbability ofthe loads acting simul- taneously is to apply a reduction factor to the combined loads rather than (0 allow an increase in the permissible stresses, with greater reductions for the greater num- ber of load types combined, 3.4.2 Limit States Design Im limit states design, the adequacy of the building and its members is checked aaguinst factored loads in order to satisfy the various safety and serviceability limit slates. ‘The UBC requires thatthe strength must be able to resist the actions resulting from the combination of the individually factored dead and live loads, where the Youd factors take into account the variability of the load and load pattems. Ifa wind load or earthquake load is tobe included, a reduction factor is applied tw the combination of the individually factored joads to allow for the improbability ‘ofthe maximum values of the wind or earthquake, and other live loads occurring simukaneously. In the NBC, three factors are required to account for combinations of loading in limit states design: a load factor, which accounts for the variability of the loads as before; a load combination factor, which is applied to loads other than dead loads and accounts for the improbability of their extreme values acting simulta neously; and an importance factor, which allows a reduction where collapse isnot likely to have serious consequences Inboth the UBC and the NBCC the strength requirement is satisfied by ensuring that the factored resistance of the members is not less tha the corresponding ac- tions caused by the factored loads. 3.4.3 Plastic Design In buildings in which plastic design is used for parts or the whole of the stes! framed stricture, available methods of analysis are based on proportional cystems of loading, that is load combinations in which increasing tosds maintain their rel ative magnitudes. Consequently, all the loads within a combination are given the same load factor. SUMMARY Loading on high-rise buildings differs from loading on low-rise buildings mainly in its accumulation over the height to cause very large gravity and lateral load forees within the structure. In buildings that are exceptionaly slender or flexible, the building dynamics can also become imponant in influencing the effective load. ing. Gravity loading consists of dead loading. which can be predicted reasonably sccurately. and live loading, whose magnitudes are estimates based on experience tnd field surveys. and which are predictable with much less accuracy. The prob Ability of net all pars of a floor supported by a beam. and of not all floors sup: poned by a column, being subjected to the full live loading simultaneously, provided for by reductions in the beam loading and in the eclumn loading, re: spectively, in accordance with various formulas. Its sometimes necessary to con: sider also the effects of construction loads, Wind loading becomes significant for buildings over about 10 siories high. and ogressively more so with increasing height. For buildings that are not very tall or slender, the wind loading may be estimated by 2 static method. Modem static ‘methods of determining « design wind loading account for the region ofthe Country where the building is to be located. the exposure of the particular location. the effects of gusting, and the importance of the building in a postwindstorm situation For exceptionally tall. slender. or flexible buildings, itis recommended that & Wind tunel test on a model is made to estimate the wind loading. Boundary layer Wind tunnels, which simulate the variation of wind speed with height, and the sting are used for this purpose. For buildings that do not quite fal into the category that demands a wind tunnel ‘esi, or for those that are in that category but whose budget does not allow such a test, dynamic methods of calculating the wind lead have been developed. Earthquake loading is a result of the dynamic response of the building to the shaking of the ground. Estimates of the loading account for the properties of the tnvcture and the record of earthquakes in the region, For unexceptionally high buildings with unexceptional structural arrangements an equivalent lateral force ‘method is recommended. In this, the loading is estimated on the basis of a simple approximation forthe structure's fundamental period its dead loud, the anticipated stond acceleration oF velecity, and other factors relating to the sol site cond tions, structure type and the importance of its use. The method gives the value of the maximum horizontal base shear, which is then distributed as an equivalent lateral load over the height of the building s0 that a static analysis can, be per formed Ifthe building is exceptionally tall, or iregular in its struciure or its mass dis- {nbution. a modal analysis procedure is recommended for estimating the earth- quake loading. The modal shapes and frequencies of vibration are analyzed: these ‘are used in conjunction with an earthquake design response spectrum and estimates of the modal damping to determine the probable maximum responses. The modal ‘method can also allow for the simultanecus torsional oscillation of the building. Methods of combining types of loading vary according to the design method and the Code of Practice concemed. Although dead load is considered to act in full all the time, live leads do not necessarily do so. The probability of the full gravity live loading acting with either the full wind, earthquike. or temperature loading is low. and of all of them acting together is even lower. This is reflected in the Codes by applying a greater reduction factor to those combinations incor- Porating more different types of loading. Wind and earthquakes are assumed never {0 act simultaneously REFERENCES 3.1. Talt Building Criteria and Loading. Monograph on Planning and Design of Tall Buildings, Vel. CL, ASCE, New York, 1980, 3.2 Uniform Building Code, 1988, Inerrational Conference of Building Officials, Whit tier, California, 3.3 Cook, N. J. The Designer's Guide 1o Wind Loading of Building Structures, Part} Building Research Establishment Repon., Butterworths. London. 1985. 3.4 Simiu, E. and Scanlan, R. H. Wind Effects on Structures, 2nd ed.. Wiley Inersci ‘ence, New York, 1986, 3.5, Davenport A. G. “Gust leading factors," J. Sruet, Div., Proc. AS.C.E. 93, June 1967, 12-34, 3.6 Simis, E. “Equivalent static wind leads for tall building design," J. Struct. Div. Proc. A.S.C.E. W02, April 1976, 719-737, 3.7 Nationot Building Code of Canada, 1990, National Research Council of Canada, Owawa, Canada. 3.8 Supplemen: to the National Building Code of Canada, 1990. Ni ‘Council of Canada, Ortawa, Canada, 39° The BOCA Baste Building Code—1990, Building Officials and Code Adminisrators iemational. Homewood, Hliols, 3.10. The National Building Code—1976, American Insurance Association, New York 3.11 Sandard Building Code, 1988 Eltion, Souhem Building Code Congress Interaa- tional, Birmingham, Alabama. 312 Recommended Lateral Force Requirements end Commentary, Seismology Commit- tee, Structural Engineers Association of California, 1988 3.13. Tentative Provisions for dhe Development of Seismic Regulations for Buildings, ATC 4-06 Amended. Applied Technology Council, National Bureau of Standards, Wash: ington. D.C., 1984 I Research S14 Bmp. G. V. Seismic Design Codes and Procedures, Eanbquake Engineering Re- search Institute, Berkeley, California, 1982, 3S Clough, R. W. and Penzien, 1. Dynamics of Sructures, McGraw: Hill, New York, 1975, 3.16 Newmark, NM. and Hall. W. J. Earthquake Spectra and Design, Eantbquake En- sneering Research Insitute. Berkeley. California, 1982. WME CHAPTER 4 Structural Form From the structural engineer's point of view. the determination of the structural form of a high-rise building would ideally involve only the selection and arrange- meat of the msjor structural elements to resist most efficiently the various com binations of gravity ané horizontal loading. In reality. however. the choice of structural form is usually strongly influenced by other than structural considera tions. The range of factors that has to be taken into account in deciding the struc- ‘ural form includes the intemal planning. the material and method of construction, the external architectural treatment, the planned location and routing of the service systems, the nature and magnitude of the horizontal loading. and the height and proportions of the building. The taller and more slender a building, the more im- portant the structural factors become, and the more necessary it is to choose an appropriate structural form. In high-rise buildings designed for a similar purpose and of the same material ‘ana neight, the eteiency of the structures can be compared roughly by their weight per unit floor area. In these terms, the weight of the floor framing is influenced mainly by the Noor span and is virtually independent of the building height, while the weight of the columns, considering gravity loed only, is approximately pro- ponional ro the height (Fig. 4.1). Bulldings of upto 10 stores designed for gravity loading can usually accommodate wind loading without any increase in member sizes, because of the typically allowed increase in permissible stresses in Design Codes for the combined loading. Fos buildings of more than 10 stories, however, the additional material required for wind resistance increases nonlinearly with height so that for buildings of 50 stories and more the selection of an appropriate structural form may be critical for the economy, and indeed the viability, of the building. A major consideration affecting the structural for is the function of the build- ing. Modem office buildings call for large open floor spaces that can be subdivided with lightweight partitioning to suit the individual tenant's needs. Consequently, the stricture’s main vertical components are generally arranged, as far as possible around the perimeter of the plan and, internally, in groups around the elevator. stair, and service shafts (Fig. 4.2). The floors span the areas between the exterior and interior components. leaving large columa-free areas avsilable for ofice plan ring. The services are distributed horizontally in each story above the par and are usually concealed ina ceiling space. The extra depth required by this space 34 Weight oF steel (Ib/squtt) % “Floor, froning 6 10 2 90 40 50 6 70 @6 90 100 110 Fig. 4.1. Weight ofstet in tall buildings, causes the typical story height in an office building to be 11 ft-6 in. (3.5 m) oF In a residential building or hoiel, accommodation is subdivided permanenily and usually repetitively from floor to floor. Therefore. continucusly vertical col- ums and walls can be distributed over the plan to form, or fit within, the pamti> tioning (Fig. 4.3). The services can then be run vertically, adjacent 10 the walls and columns or in separate shafts, to emerge in esch story either very close 10 where required, or to be distrbuied horizontally from there 10 where required, long ove coridor ceiling spaces. Wi tie exception oF the comdors, hereTore, a cziling space is not required, and the soffit of the slab can serve as the ceiling, This allows the story heights in a typical residential building or hore! to be Kept down to approximately 8f-§ in. (2.7m). A 40-story residential building is, there= fore, generally of significantly less height than a 40-story office building. In addition to satisfying the previously mentioned nonstructural requiremens, the principal objectives in choosing & building's structural form are 0 arrange to suppor the gravity, dead and live, loading, and to wesst at all lvels the external Fig. 4.2. Plan of office block itube-type). ensra) care WME CHAPTER 4 Structural Form From the structural engineer's point of view. the determination of the structural form of a high-rise building would ideally involve only the selection and arrange- ment of the major structural elements to resist most efficiently the various com- binations of gravity and horizontal loading. In realty. however. tke choice of structural form is usually strongly influenced by other than structural considera- tions. The range of factors that has to be taken into account in deciding the struc- tural form includes the internal planning, the material and method of construction. the external architectural treatment, the planned location and routing of the service systems, the nature and magnitude of the horizontal loading, and the height and proportions of the building, The taller and more slender & building. the more im- Portant the siructural factors become, and the more necessary it's to choose an appropriate structural form. In high-rise buildings designed for a similar purpose and of the same material and height, the efficiency of the structures can be compared roughly by their weight per unit floor area. In these terms, the weight of the floor framing is influenced ‘mainly by the floor span and is vinually independent ofthe building height, while the weight of the columns, considering gravity load only, is approximately pro- portional to the height (Fig. 4.1). Buildings of up to 10 stories designed for gravity Toading can usually accommocate wind loading without any increase in member sizes, because of the typically allowed increase in permissible stresses in Design Codes for the combined loading. For buildings of more than 10 stories, however, the additional material required for wind resistance increases nonlinearly with height so that for buildings of 50 stories and more the selection of an appropriate siructural form may be critical for the economy, and indeed the viability, of the building. ‘A major consideration affecting the structural form is the function of the build- ing. Modem office buildings call for large open floor spaces that can be subdivided with lightweight partitioning to suit the individual tenant's needs. Consequently, the structure's main vertical components are generally arranged. as faras possible, around the perimeter of the plan and, internally, in groups arourd the elevator. stair, and service shafts (Fig. 4.2). The floors span the arcas between the exterior and interior components leaving large column-free areas available for office pla: ning. The services are distributed horizontally in each story above the panitioning and are usually concealed in. ceiling space. The extra depth required by this space 34 —" ae ey “ Teale zu ( 2 wt Ze me: aie Floor frano, 0 2 10 2 30 «80 6 70 8 90 100 110 Fig, 4.1. Weight of stet in tall uidings.. causes the typical story height in an office building to be 11 ft-6 in. (3.5 m) or nore Ino residential building or hotel, accommodation is subdivided permanendly and usually repetitively from floor to floor. Therefore, continuously venieal col= Uumss and walls can be cistnbuted over the plan to form, or fit within, the pamti= Woning (Fig, 4.3). The services can then de run vertically, adjacent to the Walls and columns or in separate shafts, 10 emerge in each story either very close to where requited, or to be distrbuted horizontally from there to where required, song dhe comidorcelling spaces. Wit the excepiton of the cormdors, therefore, ceiling space is not required, and the soffit of the slab ean serve asthe ceiling. This allows the story hejghs ina typical residential building or hotel t@ be kept down t approximately 8-8 in. (2.7m). A 40-story residential building is, there= fore, generaly of significantly les height than a 40-story office building. In addition to satisfying the previously mentioned nonstructural requirements, the principal objectives in choosing 2 building's strctural form are to arrange to suppor the gravity, dead and live, loading, and © resist at all levels the external Perimeter colunns 1 clear Floor sence BxEg Fig.4.2 Plan of office bleck (tube-type) Fig. 4.3 Plan of residential block. horizontal load shear, moment, and torque with adequate strength and stifness ‘These requirements should be achieved, of course, as economically as possible With regard to horizontal loading, a high-rise building is essentially « vertical cantilever. This may comprise one or more individually ating vertical cantilevers, such as shear walls orcores, each bendiag about its own axis and acting in unison only through the horizontal in-plane rigidity of the Noor slabs. Alternatively, the cantilever may comprise a number of columns or walls that are mobilized '0 act Compositely. to some degree. as the chords of a single massive cantilever, by vertically shear-esistant connections such as bracing cr beams. The lateral stiff ness.and strength of both of these basi cantilever systems may he farther enhanced ifthe major vertical elements have different free deflection characteristics. in which case they will interact horizontally through the connecting slabs and beams. Within the constraints ofthe selected structural form, advantage may be taken of locating the main vertical members on plan so that the dead load compressive stresses suppress the lateral load tensile stresses. thereby avoiding the possibilty of net tension occurring in the vertical members and uplift on the foundations Particular emphasis is placed in some types of structural form on routing the grav- ity load to the outer vertical members to achieve this purpose. Steel framing has played a pioneering role in the history of tall buildings. Itis appropriate for all heights of structure and, because ofits high strength-o-weight ratio, it has always been the material of consiruction forthe tallest buildings. It allows the possibility of longer floor spans, and of partial prefabrication, leading to reduced site work and more mpd erection. its disadvantages, however, include Aeeding fire and rust protection, being expensive to clad, and requiring costly di- ‘agonal bracing or rigié-frame connections. Ate the earlier use of steel through the fist half of the century. in the form of braced construction, it has evclved in ts structural forms somewhat in parallel with reinforced concrete 0 include rigid-frame, shear wall, wall-frame, tube and braced-tube, and outrigger types of arrangements, as well as in forms more par- ticular to stel such as the suspended siructure and the highly efficient massive space frame — GS CTRTURNET | Se Reinforced concrete tall buildings were introduced approximately two decades after the frst stcel tail buildings. Understandably. the earber concrete building, svructures were influenced in form by the skeletal, column and girder arrangements ff their steel counterparts, but they differed in depending on the inherent rigid- frame action of concrete construction to resist horizontal loading. Subsequently, the Mat plac and flat slab forms were introcuced and these. with the moment. resistant frame. continued as the main reperioire of reinforced concrete high-rise structural form until the late 1940s, ‘A major step forward in reinforced concrete high-rise structural form came with the introduction of shear walls for resisting horizontal loading. This was the first ficant developments in the siructural forms of concrete high-rise buildings, freeing them from the previous 20- to 25-story height limitations of the Figid-frame and flat plate systems. The innovation and refinement of these new forms, together with the development of higher strength concretes, has allowed the height of conerete buildings to reach within sriking distance of 100 stories. Of the following structural forms, some are more appropriate to stec! and others to reinforced concrete: many are suitable for either material, while a few allow or demand a combination of materials in the same structure. They are deseribed in a roughly historical sequence. ‘The structural form of tall buildings, as discussed so far, has concemed mainly the arrangement of the primary vertical components and their interconnections. This topic would not be complete, however. without including consideration of floor systems, because some of them play an integral part with the vertical com: Ponents in resisting the lateral. as well asthe gravity. lading The last part of the ‘chapter is devoted, therefore, to 2 brief review of the floor systems used jn tall buildings. Many of these are commonly used also in low-rise buildings but are included here for completeness. ina senes of si 4.1. STRUCTURAL FORM 4.1.1, Braced-Frame Structures In braced frames the lateral resistance of the structure it provided by diagonal members that. together with the girders, form the "web" of the vertical truss, with the columns acting as the ““chords"” (Fig. 4.4). Because the horizontal shear fon the building is resisted by the horizontal components of the axial tensile or ‘compressive actions in the web members. bracing systems are highly efficient in resisting lateral loads. Bracing is generally regarded as an exclusively steel system because the diag- ‘nals are inevitably subjected to tension for one or the other directions of lateral loading. Concrete bracing of the double diagonal form is sometimes used. how- ver, with cach diagonal designed as a compression member to camry the full ex- temal shear. The efficiency of bracing, in being able to produce a laterally very stiff structure for a minimum of additional material, makes it an economical structural form for Chard mars yt mers Single ¢sagone? Ooubte siagons) — ee + > Story hetght toee ig. 4A Braced frame—showing diferent types of bracing any height of building, up to the very tallest. An additional advantage of fully twiangulated bracing is that the girders usually participate only minimally in the lateral bracing action: consequently the floor framing éesign is independent of its level in the stricture and, therefore. can be repetitive up the height of the building with obvious economy indesign and fabrication. A major isadvantageof diagonal bracing is that it obstructs the internal planning and the location of windows and door. For this wacan heared hents are usvally incorporated internally along wall and panition lines. and especially arourd elevator. stair, and service shafts, An other drawback is thatthe diagonal connections are expensive to fabricate and erect The traditional use of bracing has been in story-heigh!. bay-width modules (Fig. 4.4) that are fully concealed in the finished building. More recently, however, extemal larger scale bracing. extending aver many stories and bays (Fig. 4.5), has been used to produce not only highly efficient structres, but aesthetieally attrc- tive buildings Bracing and its modes of behavior are described in more detail in Chapter 6 4.1.2 Rigid-Frame Structures Rigid-frame structures consist of columns and girders joined by moment-esistant connections. The lateral stifiness of a rigid-frame bent depends on the bending sttiness of the columns, girders, and connections in the plane of the bent (Fig. 4.6). The rigid frame’s principal advantage is its open rectangular arrangement, Which allows freedom of planning and easy fiting of dcors and windows. If used as the only source of lateral resistance in a building, in its typicel 20 ft (6 m)-30 (9 m) bay size, igi framing is ezoromie only for buildings up to about 25 SMories. Above 25 stories the relatively high lateral flexibility of the frame calls for ‘uneconomical large members in order to control the deft Mv os sy TN Fig. 4.5. Large-scale baced fre. Rigid-frame construction is ideally suited for reinforced concrete buildings be- cause of the inherent rigidity of reinforced concrete join. The rigidrame form is also used for steel frame buildings, but moment-esistant connections in steel tend to be costly, The sizes of the columns and girders a any level of arigid frame ate directly influenced by the magnitude of the extemal shear at that level, and they therefore increase toward the base. Consequently. the desi framing cannot be repetitive as itis in some braced frames. A further result is that sometimes it is not possible in the lowest stories to accommodate the required depth of girder within the normal ceiling space. bt Ve dit bed TT Fig. 4.6 Rigid frame. Gravity loading also is resisted by the rigid-frame action. Negative moments are induced in the girders adjacent t0 the columns causing the mid-span positive ‘moments to be significantly less than in a simply supported span. In structures in which gravity loads dictate the design, economies in member sizes that arise from this effect tend to be offset by the higher cost of the rigid joints. While rigid frames of a typical scale that serve alone to resist lateral loading have an economic height limit of about 25 stories, smaller seale rigid frames in the form of a perimeter tube. or typically scaled rigid frames in combination with shezr walls or braced bents, can be economic up to much greater heights. These structural forms are described later in this chapter. The detailed behavior of rigid frames is diseusted in Chapter 7 4.1.3. Infilled-Frame Structures In many countries infilled frames are the most usual form of construction for {all buildings of up co 30 stories inheight. Columa and girder framing of reinforced. concrete, or sometimes steel, is infled by panels of brickwork, blockwork. or castin-place concrete ‘When an infiled frame is subjected to lateral loading, the infill behaves effec tively as a strut along its compression diagonal wo brace the frame (Fig. 4.7). Because the infills serve also as external walls or internal panitions, the system is ‘an economical way of stiffening and strengthening the structure. ‘The complex interactive behavior of the infill in the frame, and the rather ran- dom quality of masonry, has made it difficule1o predict with accuracy the stiffness. and strength of an inflled frame. Indeed, at the tine of miiting, aw wetlod Of analyzing inflled frames for their design has gained general acceptance. For these reasons, and because of the fear of the unwitting removal of bracing infills at some time in the life of the building, the use of the infills for bracing tall buildings has ‘mainly been supplementary to the rigid-frame action of concrete frames. Aa cut- line of a method for designing inflled frames is given in Chapter 8. eC st 7 Te Osagona) strat mt atetonrert inet Fig. 4.7 Infilled frame. AAA Flat-Pl ind Flat-Siab Structures: ‘The flar-plate structure is the simplest and most logical of all structural forms in that it consists of uniform slabs. of $-8 in. (12-20 em) thickness, connected rig- ‘idly to supporting columns (Fig. 4.27). The system. which is essentially of rein- forced concrete, is very economical in having a flat soft requiring the most un- complicated formwork and, because the soffit can be used as the csling, in ereating ‘A minimum possible floor depth, Under latera! loading the behavior of a flat-plate stricture is similar to that of ‘rigid frame, that is, its lateral resistance depends on the flexural stifiness of the ‘components and their connections, wit the slabs corresponding to the girders of the ngid frame. It is particularly appropriate for apartment and hotel construction ‘where ceiling spaces are not required and where the slab may serve directly as the ceiling. The flat-plae structure is economical for spans of up 0 about 25 ft (8 mi), above which drop panels can be added to create a flatslab structure (Fig. 4.28) for spars of up 10 38 fi (12 m) Buildings that depend entirely for ther lateral resistance on fla-plate or flat slab action are economical up to about 25 stores. Previously. however, when Code requirements for wind design were less stringent, many flat-plate buildings were cconsiructed in excess of 40 stones, and are still performing satisfactonly, 4.4.5 Shear Wall Structures Concrete or masonry continuous vertical walls may cere both architecturally ae partitions and structurally to carry gravity and lateral loading. Their very hi plane stiffness and strengih makes them ideally suited for bracing tall buildings. In a shear wall structure, such walls are entirely responsible forthe lateral load resistance of the building. They act as vertical cantilevers in the form of separate planar walls, and 2s onplanar assemblies of connected walls arourd elevator, stair, and service shafts (Fig. 4.8). Because they are much stiffer horizontally than rigid frames, shear wall sructures can be economical up to about 35 stories. In contrast to rigid frames, the shear walls’ solid form tends to restrict planning where open intemal spaces are required. They are well suited, however, to hotels and residential buildings where the floor-by-floor repetitive planning allows the walls to be vertically continuous and where they serve simultaneously as excellent acoustic and fire insulators between rooms and apartments. 1f, in low- to medium-rise buildings, shear walls are combined with frames, it 4s reasonable to assume that the shear walls attract all the lateral loading so that the frame may be designed for only gravity loading. It is especially important in shear wall structures to try to plan the wall layout so that the lateral load tensile stresses are suppressed by the gravity load stresses. This allows them to be de~ signed to have only the minimum reinforcement. Shear wal structures have been shown to perfonn well in earthquakes, for which case ductility becomes an im- portant consideration in their cesign, The behavior and methods of analysis of shear wall structures are Giscussed in detail in Chapter 9. Shear was \/ XR A\ BN \ x LeveL LLL Ly, LELFIF IDI FITTS LEZ AAT AATAA. ‘ee QM Z. Me SA Fig. 48 Shear wall structure Coupled Wall Structures. A coupled wall stricture is a particular, but very common, form of shear wall stricture with its own special problems of analysis and design. It consists of two or more shear walls inthe samme plane, or almost the same plane, connected at the floor levels by beams or stiff slabs (Fig. 4.9). The effect of the shear-resistant connecting members is ta cause the set of walle to behave in their plane partly as a composite cantilever. bending about the commen ‘ceatrcida axis ofthe walls. This esults in a horizontal stiffness very much greater than if the walls scted as a set of separate uncoupled cantilevers ‘Coupled walls oceur often in residential construction where laterl-load resist ‘ant cross walls, which separate the apertments, consist of in-plane coupled pairs, ‘or trios, of shear walls between which there are corridor or window openings. Coupled shear walls are considered in detail in Chapter 10. Although shear walls are obviously more appropriate for concrete construction, they have occasionally been constricted of heavy steel plate. inthe style of mas- sive vertical piate or box girders, as parts of steel frame structures. These have been designed for locations of extremely heavy shear, such as at the base of ele- Vator shafts. 4.1.6 Wall-Frame Structures ‘When shear walls are combined with rigid frames (Fig. 4.10) the walls, which {end to deflect in a flexural configuration, and the frames, which tend to deflect in WARE E SON a Fig. 4.9 Coupled shear wall stricture, Shear volts Rigid frames Fig. 4.10, Wall-rame structure ‘shear mode. are constrained to adopt a common deflected shape by the horizontal Tigidity of the girders and slabs. As a consequence, the walls and frames interact horizontally, especially atthe top. to produce a stifer ard stronger structure, The interacting wall-frame combination is appropriate for buildings in the 40- 10 {60-story range, well beyond that of rigid frames at shear walls alone. An additional, less well known feature of the wall-frame structure is that, in a carefully “tuned” structure, the shear in the frame can be made approximately Uniform over the height, allowing the floor framing to be repetitive Although the wall-frame structure is usually perceived as a concrete structural form. with shear walls and concrete frames, astel counterpart using braced frames and steel rigid frames offers similar benefits of horizontal interaction. ‘The braced frames behave with an overall flexural tendency to interact withthe shear mode of the rigid frames. Deuailed descriptions of the behavior and methods of analysis for wall-frame structures are given in Chapter 11. 4.1.7 Framed-Tube Structures ‘The lateral resistance of framed-tube structures is provided by very stiff moment. ‘sisting frames that form a "tube" around the perimeter of the building. The frames consist of closely spaced columns, 6-12 fi (2-4 m) between centers. joined by deep spandrel girders (Fig. 4.11). Although the tube carries all the lateral load- ing, the gravity loading is shared between the tube and interior columns or walls ‘When lateral loading acts, the perimeter frames sligned in the ditection of loading act as the "webs" of the massive tube cantilever, and those normal tothe direction of the loading act as the ‘‘flanges.”" ‘The close spacing of the columns throughout the height of the structure is usu= ally unacceptable at the entrance level. The columns are therefore merged, or ter- ‘minated on a transfer beam, a few stories above the base so that only a few, larger, ‘more widely spaced columns continue to the base. The tube form was developed ‘riginally for buildings of rectangular plan, and probably its most efficient use is in that shape. It is appropriate, however, for other plan shapes. and has cccasion- ally besn used in circular and triangular configurations. ‘The tube is suitable for both steel and reinforced concrete construction and has been used for buildings ranging from 40 to more then 100 stories. The highly repetitive patter of the frames lends itself 10 prefabrication in stel, and to the use of rapidly movable gang forms in concrete, which make for rapid construction. ‘The framed tube has been one of the most significant modern developments in high-rise structural form. It offers a relatively efficient, easily constricted struc- ture, appropriate for use up to the greatest of heights. Aesthetically, the tube's extemally evident form is regarded with mixed enthusiasm; some praise the logic ofthe cleanly exprested structure while others extiize the grd-like facade as small- windowed and uninterestingly repetitious, The tube structure’s structural efficiency, although high, still leaves scope for improvement because the “flange” frames tend to suffer from ‘shear lag'"; this and entire YateralYousing Bere _SADOD 300000 TED Figs fll Pramedaube, ‘results in the mid-face “flange” columns being less stressed than the comer col- lumns and, therefore, not contributing as fully as they could 10 the flange action. Tube-in-Tube or Hull-Core Structures. ‘This variation of the framed tube consists of an outer framed tube, the “hull.” together with an internal elevaior and service core (Fig. 4.12). The hull and core act jointly in resisting both gravity and lateral loading. In a ste! structure the core may consist of braced frames ‘whereas in @ concrete structure it would consist of an assembly of shear walls, ‘To some extent, the outer framed tube and the inner core interact horizontally 45 the shear and flexure components of a wall-frame structure, with the benefit of increased lateral stifiness. However, the structural (ube usually adopts a highly dominant role because of its much greater structural depth Bundled-Tube Structures. This structural form is notable in its having been ‘sed for the Sears Tower in Chicago—ihe world’s tallest building. The Sears Tower ‘consists of four paraliel rigid steel frames in each onhogonal ditection, intercon= nected to form nine “bandied” tubes (Fig. 4.13a). As in the single-tube structure, the frames in the direction of lateral loading serve as “webs” of the vertical can tilever, with the normal frames acting as “flanges.” {Core (or sonar tuba) Hal} (or exter tube) 3 & i-ff | \ = 3 S sf } seb uy 2% Fig. 4.12 Tube-intube 2 ‘The introduction of the internal webs greatly reduces the shear lg in the flanges: Ba = ‘consequently their columns are more evenly stressed than in the single-tube struc. to) L ture, and their contribution to the lateral stifiness is greater. This allows columns — £ of the fumes to be spiced further apart ard 10 be lest obtrisive. In the Seam PhO} t+-i-e EF Tower, sdvantage was taken of the bundled form to discontinue some ofthe tubes, and So reduce the plan of te building at stages up the height Fig. 413b, c, and ®) Braced-Tube Structures. Another way of improving the efficiency of the framed tube, thereby increasing is potential for use to even greater heights 2s well as allowing greater spacing between the columns, is to add diagonal bracing to the faces of the tube. This arrangement was fist used in a sieel structure in 1969, in (Chicago's John Hancock Building (Fig. 4.14), and ina reinforced concrete struc. ture in 1985, in New York's 780 Third Avenue Building (Fig. 4.15). Inthe sce! tube the bracing traverses the faces of the rigid frames, whereas in the conerete structure the bracing is formed by a diagonal pawtem of concrete window-size Panels, poured integrally with the frame. Because the diagonals of a braced tube are connected 10 the columns at each intersection, they virally eliminate the effects of shear lag in both the fange and + 67-90 Two ads ona cae sto an = Fig. 4.14 steet-braced tube, Lindow spensoas agers) br omittad 9 create _ AA UEVRUCTURAL FORM Se ‘web frames. As a result, the stricture behaves under lateral loading more like a traced frame, with greatly diminished bending inthe members ofthe frames, Cons ‘sequently, the spacing of the columns can be larger and the depth of the spans less, thereby allowing lager size windows than inthe conventional tube structure, In the braved-tube structure the bracing contributes also to the improved per- formance of the tube in carrying gravity loading: differences between gravity 1oad stresses in the columns are evened out by the braces transferring axial loading from, the more highly tothe less highly stressed columns. 4.1.8 Outrigger-Braced Structures “This efficient structural form consists of a central core, comprising either braced frames or shear walls, with horizomal cantilever "outrigger" trusses or girders connecting the core to the outer columns (Fig. 4.16a). When the structure is loaded horizontally, vertical plane rotations of the core are restrained by the cutriggers through tension in the windward columns and compression in the leeward columns (Fig. 4.16b). The effective structural depth of the building is greatly increased, thus augmenting the lateral stiffness of the building and reducing the lateral de- fiections end moments in the core, In effect, the outriggers join the columns to the core to make the structure behave as a partly composite cantilever. Perimeter columns, other than those connected directly to the ends of the outs riggers, can also be made to participate in the outrigger action by joining all the perimeter colurmns with a horizontal russ or girder around the face of the building. butrseger Sraced | @ cs) Fig. 4.16 (a) Outrigger-braced sructure; (b outigger-braced structure ander load, ‘at the outrigger level. ‘The large. often two-story, depths of the outrigger and pe- Fimeter trusses make it desirable to locate them within the plant levels in the build- ing, ‘The degree fo which the perimeter columns of an outrigger siructure behave Fig. 54 (a) Multitay braced frame; (b) equivalent column, column areas about their common cenkroid in te braced or the rigid frame, In this approximation the single curvaure flexure ofthe columns inthe braced and rigid frames. which usually has only a minor influence on the frames’ overall behivior, is neglected in the column model. {fa shear wall has beams connecting (it in-plane. causing i to interact ven= cally, a well as horizontally, with another shear wall or with other parts of the siucture Fig. 5.5a) the wall can be represented by an analogous “wide column ‘Thisis a column placed atthe wall’s centroidal axis and assigned to have the walls inertia and axial area, and having rigid arms that join the column to the connecting beams at each farting level (Fig. 5.5b). In this way the rotations and. vertical dispiacements a the edges of the wall are transfered tothe connecting beams, few, oal1 2 Rigid free Colvmn Gollan ge Fae a hee te We Age A tae Regs Re 7 () © Fig. 55 (@) Shear vatls and frame joined by beams; (b) equivalent wide-column model, I a en eS ne Nonplanar assemblies of shear walls that form elevaior cores (Fig. 5.61 and b) in structures that translate but do not twist under lateral loading. can be simulated by a single column located at the shear center of the section and assigned to have the principal second moments of area of the core section (Fig. 5.72). If the struc ture twists as well as translates. and the core has an effectively closed, box-like section. as in Fig. 5.6b. the single column should be additionally assigned the torsion constant J of the core (Fig. 5.7b) Ifthe structure twists and translates, and the core walls form an 1. U. as in Fi 5.6a. of more complex open-section shape. warping torsional effects may be im- portant. in which case itis possible to use a two-column model (Fig. 5.7¢) to gi an approximate representation of all the beading and torsional properties. Details of such a model are given in Chapter 13, 5.4.2 Approximate Modeling of Slabs In-Plane Effects. in structures that do not depend on the transverse bending, resistance of slabs as part of ther lateral load resisting system. the slabs are taken toserve only as rigié diaphragms that distribute the horizontal loading tothe vertical ®) © Fig. 5.6 (a) Open section nonplanar shear wall assembly: (b) closed section shear wall assembly, @ Fig. $.7 (a) Equvatent flexural column; (b) equivalent fexuralxorsional colume; (@) ‘equivalent two-colums flexural-torsional-warping model elements and that hold the building plan in shape as the structure translates and twists. The slab then serves to constrain the horizontal displacements of the vertical components at each floor to be related to the horizontal two displacements and ‘rotation of the slab. In @ three-dimensional analysis of a structure (Fig. 5.8a) the in-plane rigidity of the slab can be represented at each floor by a horizontal frame of rigid beams joining the vertical elements (Fig. 5.8b) or, ifthe computer program Includes 2 “rigid-tloor” option for simulating a gid in-plane sla. its use is simpler ‘and more accurate Transverse Bending Effects. Flat plate structures. and structures with shear ‘walls coupled by slabs, employ the transverse bending stiffness of slabs as par of the lateral load-resisting system, similar to the girders of a rigid frame, as well as using the in-plane rigidity of the slabs to hold the plan shape of the building. In ‘modeling the structure, the bending action of a slab between in-line columns or Walls can be represented by a connecting beam of equivalent flexural stifiness (Fig. 5.9). This model will result in the correct horizontal deflections, and forces in the vertical members, but it gives only the concentrated moments and shears applied to the slabs. The ineria of equivalent connecting beams to represent te slab bend- ing action is discussed in Chapter 7 and Appendix 1 LL errr. ot Connect sng. tea ‘Shear center fo Soe trier Bis | Fig. $5.9 Equivatent beam repress bering action, Teo coluens sn Soaprration Tete 5.4.3 Modeling for Continuum Analyses So far, all the considered approximations have been for discrete member moxels, that is incorporating individual vertical and horizontal members, for solution by a stiffness. matrix analysis. For certain strictures with relatively uniform properties ver the height, altemative contimum analogy models may te formed that can be analyzed by a closed solution ofthe characteristic differential equation. In a con- tinuur model. the horizontal slabs and beams connecting the vertical elements are assumed to be smeared as 2 continuous connecting medivm—a continuum—baving equivalent distributed stiffness properties. Although continuum methods are tim= ited in their facility to represent variations of a inucture over its height, they can sive very rapid approximate solutions and are valuable in providing @ general uns derstanding of s structure's behavior. Two examples of the types of siructute that can be solved using continuum techniques are « coupled wall and a wall-frame Stnicture (Figs. 5.10a and 5.11a). In the coupled wall. the connecting beams are represented by continuum with equivalent bending and shear properties (Fig. 5.10b), In a wall-frame structure, the connecting links between the wall and the frame are represented by a horizontally incompressible medium, while the beams in the frame are smeared into the general shear propery of the equivalent shear columa (Fig. 5.116). Conesmuum with equivatest feuural Shear rigid Frame representing floor slab olvens etioaet Eousatent Felon reorestcting Teves raced travers Ly 3} Columns representing w 0) Fig. $8 (a) Plan asymmetric structure; (b) epresertaton of slab diaphragm action, 5.10. (a) Coupled shear walls: (6) equivalent continuum model 76 Cooectiog tied lgueaent taunt mis” /tekoe “Than RY fe sl ‘ohm Sines @ © Fig. 5.11 (a) Wall-frame staictare; (6) equivalent continuum mode! 5.8 MODELING FOR ACCURATE ANALYSIS Ivis necessary for the intermediate and fina! stages of designto obtain a reasonably Accurate estimate of the stricture deflections and member forces. With the wide availability of structural analysis programs and powerful compaters it is now pos- sible t solve very large and complex structural madels. Some of the more gross approximations used fora preliminary analysis. such ae repreeneting braced framec and rigid frames by single columns, ate too approximate for a detailed analysis. and they do not yield the detailed forces necessary for sizing and reinforcing the individual members. The structural model foran accurate analysis should represent in a more detailed way all the major active components of the prototype structure The principal ones are the columns, walls, and cores. ard their connecting slabs and beam: ‘The major structural analysis programs typically offer a variety of finite ele- ‘ments for structural modeling. Asan absolute minimum for accurately representing high-rise structures, a three-dimensional program with beam elements and quad- ‘ilateral membrane elements (Fig. 5.12a and b) will sufice. Beam elements are sed to represent beams and colunins and, by making their inertias negligibly small or by releasing their end rotations, they can also be used to represent truss mem- ters. Membrane elements, which are used for shear walls and wall assemblies, should preferably include an incompatible mode option to beter allow forthe char acteristic in-plane bending of shear walls. If muss elements (Fig. 5.122), quadrilateral plate elements, (Fig. 5.124), and ‘combined membrane-plate clements are also available, they can be used to advan- tage in representing, respectively, truss members, slabs in bending, and shear walls subjected 10 out of plane bending. Some typical high-rise structural components and assemblies, and their repre- sentation by finite elements, will now be discussed. Se eee 2 ir ‘© Fig. 5.12 (a) Beim element; (b) quadrilateral membrane element; (e)tass element: () ‘quadrilateral plate bending element 5.5.1 Plane Frames. A plane rigid frame, which is probably the simplest assembly to be modeled, has both its column and beam members represented by beam elements (Fig. 5.13). Shear deformations of the members are normally neglected except for beams with ‘2 span-to-depth ratio of less than about 5. The results of the analysis include the vertical and horizontal displacements, and the vertical plane rotations ofthe nodes, teen elesents ig. 5.13 Rigid frame using beam elements, eo ee eae elements: [= Fig. £14 Braced frame using truss and beam elements, together with the members’ axial fore, shear force, and bending moments, I braced frame (Fig. 5.14) the braces are represented by tuss elements or small- inertia beam elements, the columns by beam elements, and the beams by beam elements with their end rotations released. The results forthe trss elements give anial forces only. 5.5.2. Plane Shear Walls Similar to the modeling of walls for an approximate analysis, a tall slender shear ‘wall that i¢ not connected by beams to osher parts of the structure (Figs. 5.13 and 5.22) can be modeled for an accurate analysis by a sack of beam elements (Figs S.Ib and 5.2b) located on the centroidal axis ofthe wal, and assigned to have the pfincipsl ineria and corresponding shear areas ofthe wall. Skear walls connected by beams to other pans of the structure (Fig. 5.Sa) can he similadly represented by vertical stacks of beam elements located on the centroidal axes of the walls with rigid horizontal beam elements attached at the framing levels to represent the effect of the walls’ width (Fig. 5.56). Inthe case of « beam-connected wall, axial forces will be induced in the wall, so itis necessary to assign to the analogous colurmn an axial area as well as an inertia and a shear area Walls that are not slender. or that have openings. cannot be well represented by simple equivalent columns and are better represented by an assembly of plane stress membrane elements (Fig. 5.152). Because the segments ofa shear wall and the membrane elements that are used to model it are subjected to in-plane bending. incompatitle mode elements that are formed to include this deformation invariably five more accurate results, as well as allowing the wse of reciangulir elements of much greater height-to-width proportions with acceptably accurate results. The resulis fora plane-stress element typically include the horizontal and venical dis- ——--- ee ae asm elements: Lines Ria beans @ ) Fig. 5.15. (2) Shear wall: membrane element model: (b) shear wall: analogous fame model placements of the nodes. and the vertical ane horizontal direct stresses and shear sresscs at either the comers or the mid-sides of the clement If the available structural analysis eomputer program does not include plane stress elements, a shear wall can be modeled altematively using an analogous frame, such as in Fig. 5.15, which can be assembled entirely from beam ele= iments. The stresses resulting from such a model are usually within 1 or 2% of those froma membrane element madel analysis. Details of an analogous frame are ‘given in Chapter 9. ‘Nonrectangular walls can be modeled using quadrilateral elements, and, if more detailed stresses are required in a panicalar region of the wall, a finer mesh can be used in that area, with quadrilateral elements being used to make the transition (Fig, 5.16). For greater accaracy. quadrilateral elements should be proportioned to be as close as possible to equal-sided parallelograms. ‘When modeled by membrane elements, shear walls with in-plane connecting ‘beams require special consideration. Membrane elements do not have a degree of freedom to represent an in-plane rotation of their comers; therefore, a team ele ‘ment connected to a node of a membrane element is effectively connected only by ahinge. A remedy for this deficiency is to add a fictitious, flexurally rigid, aux- iliary beam to the edge wall element, in one of the ways shown in Fig. 5.17. The adjacent ends of the auxiliary beam and the extemal beam are both constrained to rotate withthe wall-edge node. Consequently, the rotation of the wall. as defined by the relative transverse displacements of the ends of the auxiliary beam, and a moment, are transfered to the external beam. ee ee forse 1 oak | | | ar Fig. 5.16 Nonrectangular shear wall with transition, represented by quadrilateral ele- ments. auxstiary beam Connecting beam Fig, 5.17 Comection of beams to mem- brane element shear wall, 5.5.3 Three-Dimensional Frame and Wall Structures ‘The high-rise rigid frame structure has moment-resisting joints. and its columns. ‘and beams are modeled by three- ceance are, therefore. the translations and rotations ef the nodes. the shear forces, bending moments and axial force in the columns, and the shear forces and mo ‘ments in the bees. ‘Three-dimensional shear wall assemblies often form the most important major Iteral load-resisting components in a high-rise building. They occur variously in ‘muliibranch open sectional shapes (Fig. 5.19a), in effectively closed sections (Fig. 5.19b), and in beam-connected sections (Fig. 5.19e). Whether of closed or open section form, the principal actions of the individual walls in an assembly are in- plane shear and flexure, and the principal interaction between the walls of an a8- sembly is vertical shear along the joints. Consequently, plane stress membrane elements are highly suitable for modeling three-dimensional shear wall compo- nents (Fig. 5.20a and bj. Story-height wall-wieth elements give an acceptably accurate representation for most purposes. Plane stress elements alone are not adequate for modeling three-dimensional ‘wall systems because they lack the transverse stiffness necessary at orthogonal wall connections to allow 2 stiffness matrix analysis of the problem. Nor. when used, alone, can plane siress membrane elements provide the out-of-plane rigidity r= {aired to maintain the sectional shape of the core. as itis held in reality by the in- (a) Membrane clement and auxilcry beam model (b) model for besm-connecte shear aS ee aes oe plane rigidity of the floor slabs. ‘The remedy for these deficiencies is 10 add at each odal level a horizontal frame of fictitious. rigid auxiliary beams (Fig. $.20a). If any of the walls are connected in-plane to each other, oF to other parts of the structure, by beams. the auxiliary beams adjacent to the wall edges ean be made vertically rigid also, to cause the transfer of moment (Fig. 5.20b) as described in Section 5.5.2. ‘Another action, which would automatically be accounted for if shell elements were used for the model, but not in the ease of plane stress elements, is the tor- sional stiffpess corresponding to twisting of the walls. Although this is usvally relatively insignificant, in open-section wall assemblies it can be important and should be incorporated. Itis introduced by adding to the model a fictitious column located on any one of the vertical sets of nodes (Fig. 5.20a) and assigning it a torsion constant with a value equal to the sum of the individual walls” torsion ‘constants, as discussed in Chapter 13. The arial area and inertia of the column are assigned 10 be zero, ‘An alternative way of representing beams connecting shear walls in the same plane is to represent them by story-height membrane elements with a vertical shearing stiffness equal to the yertical-displacement stiffness of the represented ‘beam, as shown in Fig. 5.21. In such 2 model. auxiliary beams are stil required to form a horizontally rigid frame around each level of the wall assembly. but the beams adjacen: to the openings do not have to be vertically rigid. septate eneting Meadrane etenars representing walls Fig. 5.21. Beam-connected wall assemblies: membrane elements representing beams. 5.5.4 P-Delta Effects ‘Second order P-Delta effects of gravity loading can be included in a single frst ‘order computer analysis of the structure by adding to the first-order model a fiti~ tious column with a negative stiffness. The translational P-Delta effects in a nontwisting structure can be incorporated in the two-dimensional model by adding a shear column, connected to the model by rigid links a the framing levels (Fig. 5.22a), The column isassigned a negative Actus anoniy nigie Tinks gat Tnere Cola () Fig. 5.22 (2) P-Dela negative shear colurmn mocet; (b) P-Deta negative inertia column model shoar area to simulate the laicral sofiening of the structure due to gravity loading. ‘The column is assigned to be rigid in flexure. Altemaiively, the Sime result ca be achieved by using a flexural column with its rotation restrained atthe framing levels (Fig. $.22b) and its inemia assigned a negative value. The cofumn is spec- iffed tobe rigid in shear. The resulting deflections aned member forces inthe model then include the P Delta effocts of gravity loading. Details of the telugu ale given Chapter 16, ‘When making a full three-dimensional analysis of an asymmetric stricure, the P.Delta effects of twisting, as well as of translating parallel tothe building's major axes. can also be represented in the mode by aBcitious negative stiffness column, ‘The column is located in each story at the centeoid ofthe resultant gravity lon acting through the story, and is assigned to have either negative shear areas, oF negative inertias. as described before, corresponding to the directions of the bulld- ing’s two major axes. The columns adeitionally assigned a negative torsion con- stant to allow for the twisting P-Delta effects. This technique also is discussed in detail in Chapter 16, 5.5.5 The Assembled Model By combining the previously described techniques, a complete three-dimensional ‘model can be formed for any high-rise stricture consisting of a combination of frames, walls, and cores with beam and slab connections: If the bending resistance of the slabs contributes to the lateral load resistance of the structure, it is usval to model the slabs by beams of equivalent flexural stiffness connecting the venical components. Although an even more accurate ‘model could be formed by representing each slab as an assembly of plate elements, such a detailed representation would vastly increase the size of the problem. 8B MODELING FOR ANALYSIS. In the complete detailed model. therefore, beam elements are used to represent beams and columes. and story-height plane-stress membrane elements are used to represent shear walls and cores. At all floor levels an auxiliary beam is added to the tp of each membrane clement. The auxiliary beams, and the real beams, are assigned extremely high axial areas and horizontal bending inerias in order 10 simulate the rigid diaphragm effect of the slab. Auxiliary beams are also used at tach floor level to interconnect frames. walls and cores, as well as any isolated columns. Where a real beam connects in plane with a wal. the auxiliary beam on the connected wall clement is assigned to be rigid in the vertical, as well as the horizontal. plane so 2s to transfer moment between the wall and the external Beam, For each open section shear wall assembly, « vertical column essigned to have the: ‘walls torsion constant is added to the assembly. ‘The requirement for providing auxiliary beams. joining the columns, walls. and cores to form a rigid horizontal diapiragm at each floor. and to connect shear walls to beams in their planes, has been avoided in at least one tall building structure analysis program (5.1}- 5.6 REDUCTION TECHNIQUES When the detailed model of 2 high-rise structure is s0 large and complex that its analysis presents a formidable task of bookkeeping and computation, mt may be preferable to try to simplify the model, provided the accuracy of the results is not seriously compromised, The following techniques are among those used to sim plify the model. Some of the techniques ¢o ot diminish at all the accuracy of the analysis, while others, although losing a litle in accuracy, are sill good enough for a final design analysis. The reductions are therefore applicable to both deuailed and to simplified models for anlaysis. 5.6.1 Symmetry and Antisymmetry A structre that is symmetric in plan about the axis of horizontal loading (Fig. 3.23a) can be analyzed as a half-siucture, to one side of the line of symmetry, subjected t0 half the loads (Fig. 5.236). The ends of the members cut by the line of symmetry must te constrained (© represent the omitted half of the staictare: ‘That i, they must be constrained agains! rotation and horizontal displacement in the plane perpendicular wo the direction of loading, and against rotation about @ ventical axis, while simaltancously being free to displace vertically and o translate in the direction of the loading. The results for the deflections and forces for the analyzed half-structare will apply symmetrically to the corresponding nodes and members in the omitted half structure ‘A structure that is symmetric in plan about a horizontal axis perpendicular to the axis of horizontal loading (Fig. $.24a) behaves antisymmetrcally about the ‘xis of symmetry. In this case only half ofthe structure. to one side of the axis of symmetry, and subjected to loads of half vale, needs to be analyzed (Fig. 5.24b). ts of symetry stiuctore ond loading translet ier rotation oe L J- @ 5.23. (4) Plan aymmetric structure with synmetric A deve aneisymetrtcal nea against z z 5 z © Fig, 5.24 (a) Antisymmettic y behaving structure: (b 86 REQUCTION TEGHMIQUES 91 ‘The ends of the cut members are constrained or the line of symmetry to represent their connection to the omitted antisymmetrically behaving other half of the siruc- ture. That is, they are constrained against vertical displacement, but are free 10 rotate in the vertical plane parallel to the direction of loading. The values of the results for the analyzed half-structure apply antsymmetrcally to the omitted half- structure, ‘Thus, if a structure is doubly symmetric on plan and subjected to horizontal loading along one of its axes of symmetry it can be analyzed by considering just ‘one-quarter of the structure, with appropriate constraints applied to the ends of members cut on the lines of symmetry, to represent the symmetrical and antisym- metrical aspects of behavior, 5.6.2 Two-Dimensional Models of Nontwisting Structures ‘The assumption that the floor slabs are rigid in plane. which permits the horizontal displacements of all vertical elements at a loor level to be defined in terms of the slab’s horizontal translation and rotation, allows the possibility of representing a three-dimensional structure by @ two-dimensional model. An explanation of this can be developed by first considering techniques for the planar representation of nontwisting structures, ané then extending them to twisting structures. ‘Symmetrical Structure Consisting of Paralte! Bents. A structure that is symmetric on plan and symmetrically loaded does not twist. Adding to this the assumption of the slab’s in-plane rigidity means that the horizontal displacements of all the vertical components ai a floor level are identical. Now considering the symmetrical structure in Fig. 5.25a, and allowing for symmetry by analyzing only ‘one-half of the structure. the identity of displacements at the floor levels can be established in a planar model by assembling the bents in the same plane. in any order and at an arbitrary spacing. as in Fig. 5.25b, and providing a horizontal ‘constraint between the bents at each level. The constraint can be formed in two alternative ways. If the analysis program has a dependent node option. seis of nodes. one in each bent. at the same level, can be assigned 10 have the same horizontal displaccmenis. If a dependent node option is not available, pairs of nodes at the same level in adjacent bents may be joined by axially rigid pin-ended links, asin Fig. 5.25b. The half-structure model is then subjected ut the floor levels to loads of half the valve of the total load per level AS far as the validity of the assumptions allow, the resulting moments. shears, and vertical axial forces in the model will correctly represent those in the structure ‘The shear in the slabs between bents must be found by considering the differences between the shears in successive stories of each best, and the relative plan location Of the bents. The axial forces in the beams and links of the model are not mean- ‘ngful because both the application of the leading and the horizontal connections in the planar model do not properly represent their on-plan locations in the real structure, 92 MoDELING FoR ANALYEIS ate of eyematry : Pa a ' tyyddddeaed talistracture waded co} Fig, $.25 (@) Symmetric stricture with parallel bents: (b) equivalent two-dimensional model Symmetrical Structure with Connected Orthogonal Bents. Structures that consist of an orthogonal system of connected tents, which are symmetrically Jocated about the axis of horizonial loading, as in Fig. 5.26a, can be modeled for analysis by an extension ofthe planar modeling technique described above. Considering haf the structure, and assuming that, perpendicula: 01 the bens have negligible stiffness, the strcture’s shear resistance in the of loading is provided by bents AB and CD. as they displace horizontally in their planes paraliel to the direction of loading (Fig. 5.26a). Bents AE and BF. perpen- dicular to the loading, do not displace horizontally in their planes, but interact vertically with bents AB and CD along their vertical lines of connection A. B, C, and D. This vertical interaction causes the perpendicular dents o act as “flanges” to the parallel bent ““webs,”* as par ofthe structure's overall flexural action. In the equivalent hal-structure planar model, half the parallel tents and the perpendicular half-bents are assembled in plane, with the parallel bents in one ao 35 20° lec h 4 Dents perpendicu ® aon wo z : i 93 94 MODELNG FoR ANALYSIS: {g10up and the perpendicular half-bents in another (Fig, $.26b). A column at the intersection of orthogonal bents appears twice in the planar model, once in a par- allel bent and once in a perpendicular bent. In each bent the columr is assigned an inertia appropriate to its bending in the plane of that bent. So thatthe axial area of an intersection line column is not represented twice, it is arbitrarily assigned cmtirely to the column in the parallel bent with a zero area assigned to that in the perperdicular frame. The nodes in the model are numbered so that those on the vertical lines of intersection, which are represented twice, are assigned two difer- ent numbers. as in Fig, 5.266, ‘The identical horizontal displacements of the parallel bents are established it the model as before. either by using the dependent node facility oF by including fictitious axially rigié links, as between B and C in Fig. 5.26b, ‘The compatibility of vertical displacements between the parallel ard perpendic: lar bents may also be achieved in altemative ways. Ifa dependent node option is available, vertical compatibility can be established by constraining the connection odes that afe duplicated in the parallel and horizontal bents to have the same vertical displacements. The zer0 horizontal in-plane displacement of the perpen: dicular bents is arranged by constraining horizontally atleast one vertical line of nodes in each of those bents. If a dependent node option is not available, there are two alternative ways of using fictitious members to establish the connection in the planar model. ‘The frst is t0 dimension the model horizontally so that the vertical intersection lines om cach perpendicular frame, as for example lines A and lines C in Fig. 9.200, are located immediately adjacent. say a5 close as 1/200 of the adjacent span, to the duplicate incrsection lines of the connected “*paralle!" frames. Each pair of dus plicated connection nodes is then joined by a very stiff horizonial beam with horizontal and rotational relesse at one end, as. for example, nodes 32 and 38 Fig. $.26b, The alternative is to cimension the model horizontally so that the vertical connection lines on the perpencicular frames are. in effect, coincident with those on the paraite! frames, as, for example, lines B and lines D in Fig. 5.26b, and to dimension them yerically so that the connection nodes on the perpendicular frame are displaced upward slightly, sey 1/100 story height, from the correspond {ng nodes on the parallel frame, Each pair of duplicated connection nodes. a8, for example, nodes 46 and 35 in Fig. 5.26b, is then joined by a vertical axially rigid link. In either of these ways, the fictitious links establish vertical compaiibilty, while avoiding horizonal interaction and vertical plane rotational interaction be- tween the orthogonal bents. “The technique can be used for structures whose bents consist of walls, or frames, or combinations of both. 5.6.3 Two-Dimensional Models of Structures That Translate and Twist ‘The common assumption for analysis, that the floor slabs are ngid in their planes, implies that for an arbitrary origin and a pair of axes parallel 10 the onhogonally ‘oneented bents of a laterally loaded structure (Pig, 5.27) the resulting displaced @dvsplaces se * direction te final posstvon “5. @ fotated atout Inver} posttion F c Fig. 5.27 Displacements of bending and twisting structure location of any floor slab can be defined in terms of the rotation of the slab about the origin, and two displacements parallel to the axes. Further, for the horizontal ‘equilibrium of any slab, the extemal X- and Y-direction forces on the slab and their combined moment abou the vertical axis through the origin must be in equilibrium With, respectively, the X- and ¥-drection resultants of the reactions from the bents and their resultant moment about the origin. ‘Assuming that the stricture consists ofa plan-asymmetric system of orthogonal bents that are stiff in their planes but have zero transverse and torsional sifinesses (Fig. 5.28a), a two-dimensional model can be formed to satisfy the above condi- tions of displacement and equilibrium, as follows, Fit, select an arbitrary omigin 0 (Fig, 5.282) that is located 10 the left of and below the lower lefi-hand comer of the structural plan, Bents AB and CD are Parallel 10, and at distances x and x from the ¥ axis, while the orthogonal benis AAC and BD are paraliel to, and at y; and ys from the X axis. ‘Next, form the two-dimensional model by assembling all the bents in the same plane wih the X-direction bens in one group and the ¥-

You might also like