Engenharia Civil: Estrutura em Aço - SciELO

241REM: R. Esc. Minas, Ouro Preto, 60(2): 241-250, abr. jun. 2007

Roberto T. Leon

AbstractIn late 2005, the American Institute of Steel

Construction issued its most recent Specification forStructural Steel Buildings (ANSI/AISC 360-05). Thisspecification includes updated design provisions in bothallowable strength design (ASD) and load andresistance factor design methods (LRFD), andincorporates the design provisions for hollow structuralsections and single angles. Amongst the major changesare a complete revamping of the methodologies forassessing stability of framed structures, new provisionsfor composite columns and updated materialrequirements. This paper will describe the changes andhighlight those of practical significance.

Keywords: Steel design, specifications, codes, ultimatestrength, allowable stress.

The new 2005 AISC specification(A nova norma americana do AISC 2005)

Roberto T. LeonProfessor - Department of Civil and Environmental EngineeringGeorgia Institute of Technology, Atlanta, GA, 300332-033, USA

E-mail: [email protected].

ResumoNo final de 2005, o American Institute of Steel

Construction editou a versão mais recente da normanorte-americana para estruturas de edifícios em aço,Speci f icat ion for S tructural S tee l Bui ld ings(ANSI/AISC 360-05). Essa norma inclui prescrições paraprojetos com base nos métodos das tensões admissíveis(ASD) e dos estados-limites (LRFD) e incorpora, ainda,prescrições para seções tubulares e cantoneiras. Entre asmais importantes modificações destacamos a completarenovação das metodologias para a verificação daestabilidade de estruturas aporticadas, novas prescriçõespara pilares mistos e atualização das especificações dosmateriais. O presente artigo descreve essas alterações edestaca aquelas com interesse prático.

Palavras-chave: Projeto em aço, normas, resistênciaúltima, tensão admissível.

Engenharia Civil: Estrutura em Aço

REM: R. Esc. Minas, Ouro Preto, 60(2): 241-250, abr. jun. 2007242

The new 2005 AISC specification

1. IntroductionFor the past 20 years, the American

Institute of Steel Construction (AISC)has maintained two differentspecifications: a limited state or ultimatestrength one (AISC LRFD 1999), whosefirst edition dates to 1986, and anallowable or working stress design one(AISC ASD 1989), whose last editiondates to 1989. In the USA, althoughlimited state design of reinforcedconcrete structures has been commonsince the early 1960’s, design for metalstructures has remained for the most partunder the allowable stress approach. Inthis last cycle of its main specification,AISC decided that it could no longerafford to maintain two separatespecifications, and decided to develop aset of unified provisions that mergedboth approaches. With the newSpecifications, AISC intends toconsolidate its design provisions intofour broad documents: basicrequirements (AISC 2005a/ ANSI 360-05),seismic design (AISC Seismic 2005b/ANSI 341-05), nuclear design (AISC 2006/ N690) and contractual provisions (Codeof Standard Practice for Structural SteelBuildings, AISC 2005c).

In order to rationalize the designprocess along an ultimate strengthdesign approach, the 2005 version of thisspecification contains a dual set ofprovisions. One set of provisions willfollow closely the current Load andResistance Factor Design (LRFD) formatand the other set will follow an AllowableStrength Design (henceforth, NEW ASD)approach. Although the NEW ASDbears a similar name to the olderAllowable Stress Design (henceforth,OLD ASD), the two are not to beconfused. In the 2005 AISCSpecification, a single expression will begiven for the nominal resistance of amember or component (Rn), and thatresistance will be reduced by a resistancefactor (f) for LRFD design or divided bya safety factor (W) for the NEW ASD.The general format is:

For LRFD:

Ru ≤ φ Rn

For ASD:

na

RR ≤Ω

Note that the nomenclaturemaintains the use of the terms resistanceand safety factors. While the origin ofthe former is clear and rooted in reliabilitytheory, the latter is based mostly onexperience and cannot be given aconsistent meaning across all forms ofloading. Thus, while for simple loadingconditions such as tension, the NEWASDF and OLD ASD are the same, thisis not a statement that can be generalizedto the rest of the code.

The procedure for determining thenew ASD safety factors (Duncan 2006)is based on the concept of an effectiveload factor. The effective load factor, γ,is determined by setting the ASCE 7(ASCE, 2005) LRFD load combination -for live load (L) and dead load (D) only -equal to the equivalent ASD loadcombination. An effective load factor, γ,may be determined for the ASD side ofthe equation as follows.

1.2D + 1.6L = γ(L + D) (1)

1.2 + 1.6(L/D) = γ ((L/D) + 1) (2)

where

γ = effective load factor.

The AISC LRFD Specification forStructural Steel Buildings (1986) wasoriginally calibrated to the OLD ASD atL/D = 3. For L / D = 3, Equation 2 yields

1.2 + 4.8 = γ (4)

γ = 6/4 = 1.5

Therefore, for calibration at L/D = 3with γ = 1.5 as the target effective loadfactor, and using the following inequalitythat the design strength (φRn) must equalor exceed the required strength (γR), andsolving for the safety factor, Ω, yields:

φRn ≥ γR

1.5nRR

γΩ = ≥ =

φ φ

In the new AISC Manual, color-coded tables are used to highlight thedifference between LRFD and the NEWASD. Figure 1 shows one such table forthe flexural capacity of beams. The bluecolor for LRFD and the highlighted greenbackground for ASD clearly separate thetwo design cases.

In addition to supporting twodifferent approaches, at least thefollowing major editorial changes in thenew AISC Specification should behighlighted:

1. Nomenclature: An attempt has beenmade to coordinate the nomenclatureto a standard one to be adopted byall metal specifications. This wasachieved through the work of a jointcommittee of the American Instituteof Iron and Steel (AISI) and AISC,and gives greater clarity andtransparency to this code.

2. Mandatory language: All non-mandatory language has beeneliminated, and very few locationsremain where designers are given achoice of whether to use or not aparticular clause of the code.

3. Loading: All references to loads havebeen eliminated; all loading is takenfrom ASCE 7-05 (ASCE 7, 2005). Thismeans that AISC has now completelydecoupled the loading andresistance part of the design process,and loads have become materialindependent as they should be.

4. New section types: All requirementspertaining to single angles and pipe/hollow structural (HSS) sectionshave been incorporated into the mainSpecification. These two types ofsections were covered in separatedocuments in the past. While theconsolidation of all the rules into themain specification was deemed asvery desirable, it has led to some verylarge chapters. For example,incorporation of these section typeshas nearly doubled the size ofChapter F - Flexure.


Roberto T. Leon

5. Reorganization: The content hasbeen extensively reorganized, so thateach topic is clearly addressed in achapter. For example, there is now achapter on shear (Chapter G) thatgroups provisions that werescattered through many chapters inprevious codes (mainly section F2).In previous editions, Chapter G wasvery short and only addressed plategirders. In general, one can say thateach chapter now addresses aspecific type of failure mode.

6. Importance: In addition to thereorganization of the material alongfailure modes, the chapters areorganized such that the morecommonly used provisions are at thefront. The intent of the Specificationcommittee was that even though atopic may cover many pages, 90% ofthe everyday design requirementsare covered by the first few pages ofeach chapter.

7. User Notes: The new specificationmakes extensive use of “user notes.”These are the equivalent of a quickcommentary or clarification, and areembedded inside the specificationalthough they are not legally part ofit. The intent of these notes, whichare highlighted inside a gray box inthe text, is to provide designers withguidance where legal requirementsforce an arcane wording of theprovisions. An example of its use isin Chapter D (Tension) where thelimitations on the maximumslenderness have been removedfrom the Specification. A user notehas been inserted to indicate that thislimit should preferably not exceed L/300, but that this recommendationand does not apply to rods orhangers in tension.

8. Commentary: The commentary hasbeen completely rewritten to providemore useful information to designers.Where possible, all historicaldescriptions have been eliminatedand emphasis has been put ondocumenting why changes havebeen made.

In the next sections of this paper,some of the more notable changes arediscussed. The sections referenced aregiven in parenthesis after the title; issuesrelated to stability (Chapter C andAppendices 1 and 7) are left to the latterparts of the paper because they are themore complex. The entire provisions andcommentary are downloadable free fromwww.aisc.org. Designers are directed tothe new AISC Manual and itsaccompanying CD, which containshundreds of design examples and tables,for more details. Finally, it should benoted that this is the first steelspecification issued under the AmericanNational Standards Institute (ANSI)approval. The ANSI process insures amore public development of theprovisions, a careful documentation ofthe changes made, and the developmentof future specifications under aconsistent basis.

2. Highlights of new2005 AISC specification2.1 Scope (Section A1)

A major change in the scope of the2005 Specification is the elimination fromthis section of the traditionalconstruction types (FR and PR for LRFDand Types 1 through 3 for ASD).Differentiation between differentconstruction types is now embedded inappropriate sections, such as in ChapterC for issues related to stability. Inaddition to this major change inapproach, the new Specification nowexplicitly recognizes its applicability fordesign in areas of low to moderateseismicity (Seismic Design Categories Athrough C). This is a significant changebecause as new areas of the USA becomezoned at higher seismic levels, thisclause intends to alert designers thattraditional steel structural systems canstill be used.

Figure1 - Example design table highlighting differences in LRFD and ASD values.



2.2 Materials (Sections A2-A3)The confusing collection of material

and codes referenced in Chapter A ofprevious specifications has beencompletely revamped, with standardsclearly labeled and materialspecifications grouped by type. Giventhe great variety of steel grades availablethroughout the world and the fact that adesigner is unlikely to know the sourceof the steel that will be used in a particularproject, AISC has taken great care to limitthe Specification to the more commonASTM grades, including the new ASTM913 and 992. Perhaps the greatestchanges in the materials area are (1) theallowance of materials with yield pointsup to 100 ksi (about 700 MPa) and (b)the requirements for a minimum impactresistance for thick sections to be weldedusing complete joint penetration welds(20 ft-lbs at 70º F / 27 J 21º C).

In addition, the new specificationtakes into account the large differencesin properties due to the manufacturingprocess. For example, the fact that someof the hollow structural section (HSS)grades (ASTM A500, Grade A) do notmeet the ductility requirements,necessitated that the usable stress belimited to 80% of its yield stress.

2.3 Design Requirements(Chapter B)

Several important changes haveoccurred in this chapter. Amongst themare:

1. Section B2 now contains noinformation on load combinations.For all such information, the readeris referred to ASCE 7. A user notereminds the reader of the ASCE 7section where load combinations canbe found for either LRFD or ASDdesign. This is important becausethe load combination for ASD designhave been considerably expanded inthe latest ASCE 7 document,effectively eliminating the largedisparity between the designmethods inherent in the existingLRFD and OLD ASD approaches.

2. Sections B3 and B4 give concise butcomplete descriptions of the dualapproaches (LRFD and NEW ASD)used in the specification.

3. Section B6 now contains a cleardescription of connection types,which are now divided into simpleand moment connections, with thelatter further subdivided into fully(PR) and partially restrained (PR)ones. It allows the designer to usesimple and rigid conditions in theanalysis of most connections, butrequires the explicit use of moment-rotation curves in the analysis of PRframes.

4. Sections 7 through 11 explicitlyrequire that the designer account forserviceability, ponding, fatigue, fire,and corrosion in the design. Thesesections do not contain specificrequirements for these conditions;those are given in other parts of theSpecification (Chapter L forserviceability and Appendix IV forfire, for example). While a number ofthese have been present in theSpecification for some time, they hadbeen relegated to later Chapter orAppendices, often falling “below theradar screen” of many designers. Byexplicitly calling for theirincorporation into the basic designrequirements, the Specification nowmakes clear that many limit states,besides strength, need to beconsidered.

5. The requirements for local buckling(formerly Section B5 and now SectionB4) have been considerablyexpended and clarified.

6. Section B6 on evaluation of existingstructures has now been reduced toa reference to the new Appendix 5,which contains all the provisions. Inthe past, the move to an Appendixhas implied considerable interest inthe topic and the possible futuredevelopment of a stand-alonespecification. There is now a newAppendix on fire design (Appendix4), which may in the future also be aseparate document or incorporatedinto the main Specification.

2.4 Compression (Chapter E)One of the main logistical decisions

in the development of the 2005Specification hinged on the choice of theformat of the equations for compression,which is the first difficult design topic tobe addressed by the Specification. Whilethe LRFD and OLD ASD curves werevery similar, they were specified indifferent formats and units. Those forLRFD were couched mostly on a strengthbasis (units of load) while those for theOLD ASD where on a stress basis. Thenew Specification opted to adopt thestress format but using the existingLRFD equations with a non-dimensionalslenderness parameter (function of√E/Fy). The specification has removedall limits on slenderness ratios, includesASD provisions for singly and non-symmetrical members, and includes allthe necessary provisions for the designof slender members within the chapter.

2.5 Flexure (Chapter F)The new specification has

collapsed the 5 equations based on theunbraced length of the compressionflange in the OLD ASD into threeequations based on lateral-torsionbuckling conditions similar to those inthe existing LRFD. In most cases, thishas resulted in increases in capacity formembers designed under the ASDapproach and on the removal of a numberof the step functions that characterizedthe design of some wide flange shapesunder the OLD ASD procedures (seeFigure 2). The equations have also beenchanged so that the flexural breakpointbetween inelastic and elastic buckling(Mr) has been substituted by the muchsimpler expression of 0.7SxFy.

The chapter now covers all typesof members, including single angles andsquare and rectangular hollow sections(HSS), as well as unsymmetrical shapes.A especially useful user note, in the formof a table (Figure 3), now clearly givesdesigners guidance on the applicabilityof different failure modes and relevantclauses of the chapter to the flexural


Roberto T. Leon

design of all different types of members. Thus, while the Chapter itself appears asmore dense and complicated in the new version, its usability has been considerablyimproved by careful attention to the governing failure modes.

2.6 Shear (Chapter G)Chapter G, which used to cover only the design of plate girders, now has

condensed all the shear design provisions (including the old section F2 whichgoverned the shear design for most other members) into a single section. Theprovisions have been calibrated to the OLD ASD one, with the result that whileexpressions similar to those in the existing LRFD are used, the limits and resistancefactors have changed in some cases.

Probably the most striking change for those used to the LRFD format, is the useof a φv = 1.00 (and corresponding Ωv = 1.50) for the shear capacity of doubly-symmetric sections and channel sections. In addition, provisions are now given forthe shear strength of HSS sections as follows:

2crg

n

FAV =

where F cr is the larger of:

5 34 2

1.60 0.78cr cr

v

E EF or FL D DD t t

= =⎛ ⎞ ⎛ ⎞⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠

and Ag is the gross area based on the design wall thickness, D is the outside diameter,Lv is the distance from maximum to zero shear force, and t is the design wall thickness(equal to the nominal wall thickness for submerged arc welded (SAW) and 0.93 forelectric resistance welded (EWS) HSS).

Figure 2 - Comparison of OLD ASD and the unified 2005 Specification for the flexuralcapacity of a W36 x 182, Fy = 50 ksi beam (W920x271, 350 MPa)

2.7 Interaction (Chapter H)Chapter H addresses members

subjected to axial force and bendingabout one or both axes, with or withouttorsion, and members subject to torsionalone. The chapter basically preservesthe existing LRFD approach, in whichthe interaction between flexure and axialload is handled by equations involvingthe sum of the ratios of the required tothe available strengths. However, newequations have been developed inconjunction with the stability provisionsof Chapter C. For an unsymmetricalsection, the summation is for the stressesrather than strengths. For double-symmetric sections under uniaxialbending and axial load, the newspecification permits the separation intotwo independent limit states (in-planeinstability and out-of-plane instability/flexural torsion buckling).

2.8 Composite members(Chapter I)

The 2005 version of Chapter Iincludes both extensive technical andformat changes as well as significant newmaterial when compared to previouseditions. The major technical changeconsist of new design provisions forcomposite columns (Section I2), whichnow includes new cross-sectionalstrength models, provisions for tensionand shear design, and a liberalization ofthe slenderness limits for HSS tubes andpipes. Other significant technicalchanges are the new, more rational shearstud strengths values for design (SectionI3.2d), the use of an ultimate strengthmodel for ASD design of compositebeams (I3.2), and new material limitsusable for design (Section I1.2).

The new provisions require that thestrength of composite sections shall becomputed based on first principles ofmechanics and robust constitutivemodels for materials. Two approaches aregiven to satisfy this requirement. Thefirst is the strain compatibility approach,which provides a general method. Thesecond is the plastic stress distribution



Figure 3 - User note from the beginning of Chapter F indicating applicability of differentclauses and failure modes in design.

approach, which is a subset of the straincompatibility one. The plastic stressdistribution model provides a simple andconvenient calculation method for themost common design situations, and isthus treated first.

An example of the differences indesign strength for encased compositecolumns is given in Figure 4, whichcontrasts the capacities given by currentreinforced concrete provisions (ACI 318,2005) and the 1999 LRFD (AISC 1999)ones with two versions proposed in the2005 Specification (AISC 2005a). TheAISC 2005 (1) curve corresponds to thepolygonal approach pioneered in theEurocodes and adopted into the 2005specification, and the AISC 2005 (2)shows a simplified bi-linear approximationuseful for design.

2.9 Stability (Chapter C)Chapter C, which covers stability

provisions, is probably the mostradically changed chapter in theSpecification. While the older methods(the effective length method forindividual members, for example) remainvalid, Chapter C takes an important steptowards recognizing advanced analysisand design techniques rendered possibleby personal computers and advancedcommercial software. The chapter is theresult of several years of work of a jointAISC - SSRC (Structural StabilityResearch Council) committee thatexamined a number of techniques rangingfrom full-advanced analysis includinggeometrical and material non-linearitiesto the nominal load concepts such asthose incorporated in the Australian andEuropean codes. The GeneralRequirements to Chapter C state it asfollows1:

Stability shall be provided for thestructure as a whole and for each ofits elements. Any method thatconsiders the influence of second-order effects (including P-∆ and P-δeffects), flexural, shear and axialdeformations, geometric imperfections,and member stiffness reduction due

to residual stresses on the stabilityof the structure and its elements ispermitted. The methods prescribedin this Chapter and Appendix 7satisfy these requirements. Allcomponent and connectiondeformations that contribute to thedisplacements shall be consideredin the stability analysis.

1 Parts of the Specification and Commentaryare include here because precise language isneeded in order not to confuse the reader;thus rather than paraphrasing these partswith the possibility of distorting the meaning,parts of the introductory material to ChapterC and Appendix 7 are reproduced verbatimin Sections 2.9 and 2.10.


Roberto T. Leon

In structures designed by elasticanalysis, individual memberstability and stability of thestructure as a whole are providedjointly by:

(1) Calculation of the requiredstrengths for members, connectionsand other elements using one of themethods specified in Section C2.2,and

(2) Satisfaction of the member andconnection design requirements inthis specification based upon thoserequired strengths.

In structures designed by inelasticanalysis, the provisions of Appendix1 shall be satisfied.

The commentary says that:

The stability of structures must beconsidered from the standpoint of thestructure as a whole, including notonly compression members, but alsobeams, bracing systems, andconnections. The stability ofindividual components also must beprovided. Considerable attentionhas been given in the technicalliterature to this subject, and variousmethods are available to assurestability (Galambos, 1998). In allapproaches, the method of analysisand the equations for checking ofcomponent resistances areinextricably interlinked. Traditionally,the effects of unavoidable geometricimperfections (within fabricationand erection tolerances) anddistributed yielding at strength limitstates (including residual stresseffects) are addressed solely withinmember resistance equations.Correspondingly, the structuralanalysis is conducted using thenominal structure geometry andelastic stiffness. The Specificationaddresses this traditionalapproach, termed the EffectiveLength Method in this commentary,as well as a new approach which istermed the Direct Analysis Methodand is addressed in Appendix 7. TheDirect Analysis Method includes

nominal geometric imperfection andstiffness reduction effects directlywithin the structural analysis. Ineither the Effective Length or theDirect Analysis approaches, thestructural analysis by itself is notsufficient to ensure the stability ofthe structure as a whole. The overallstability of the structure as well asthe stability of individual elementsis ensured by the combinedcalculation of the required strengthsby structural analysis and thesatisfaction of the member andconnection design provisions of theSpecification.

In general, it is essential that anaccurate second-order analysis ofthe structure be performed. Theanalysis should in general considerthe influence of second-order effects(including P-∆ and P-δ effects asshown in Figure C-C1.1), flexural,shear, and axial deformations. Morerigorous analysis methods allowformulation of simpler limited statemodels. One such example can befound in Appendix 7 where the newDirect Analysis Method is presented

as an alternative method to improveand simplify design for stability. Inthis case, the inclusion of nominalgeometric imperfection and memberstiffness reduction effects directly inthe analysis allows the use ofK = 1.0 in calculating the in-planecolumn strength Pn within the beam-column interaction equations ofChapter H. This simplification comesabout because the Direct AnalysisMethod provides a better estimateof the true internal forces within thestructure. The Effective LengthMethod, in contrast, includes theabove effects indirectly within themember resistance equations.

2.10 Direct Design Method(Appendix 7)

The commentary to Appendix 7says:

The Direct Analysis Method,addresses a new method for thestability analysis and design ofstructural steel systems comprised ofmoment frames, braced frames, shear

Figure 4 - Comparison of the design strength for a 24 in. x 24 in. encased column(f’c = 5 ksi) with 4 # 8 bars and a Grade 50 W10x49 section (610 mm x 610mm concretecolumn (35 MPa) with 4 D25 bars and a W250x73 column (350MPa).



walls or combinations thereof(AISC-SSRC, 2003a). While theprecise formulation of the method isunique to the AISC Specification,some features of it have similaritiesto other major design specificationsaround the world including theEurocodes, the Australian Standard,the Canadian Standard, and ACI318.

The Direct Analysis Method hasbeen developed with the goal tomore accurately determine internalforces in the structure in the analysisstage and to eliminate the need forcalculating the effective bucklinglength (K factor) for columns in thefirst term of the beam-columninteraction equations. This methodis, therefore, a major step forward inthe design of steel moment framesfrom past versions of thespecification. In addition, it can beused for the design of braced framesand combined frame systems. Thus,this one method can be used for thedesign of all types of steel framedstructures used in practice. Themethod can be expanded in thefuture beyond its use as a second-order elastic analysis tool aspresented here. For example, it canbe applied with inelastic or plasticanalysis. Also, it can be used in theanalysis of composite and hybrid

structures, although thisapplication is not explicitlyaddressed in this edition.

Chapter C requires that the DirectAnalysis Method, as describedherein, be used wherever the valueof the sidesway amplification∆2nd order / ∆1st order (or B2 fromEquation C2-3), determined from afirst-order analysis of the structure,exceeds 1.5. It may also be used inlieu of the methods described inChapter C for the analysis anddesign of any lateral load resistingframe in a steel building.

Some of the most importantassumptions embedded in the directdesign method are:

1. Initial member out-of-straightness isL/1000, where L is the length of themember.

2. The initial frame out-of-plumbnessfor a story is assumed as H/500,where H is the story height of thebuilding.

3. The total out-of-plumbness of thestructure is bounded by the limitsgiven in the Code of StandardPractice.

4. The limit states considered includecross section yielding, local

buckling, flexural buckling, andlateral-torsion or torsion-flexurebuckling.

5. For the in-plane buckling check, theeffective length can be assumed tobe equal to 1.

6. Residual stresses are assumed to belinearly distributed and have amaximum value of 0.3 Fy at the flangetips.

The direct design method requiresthat:

1. A rigorous second-order analysis beconducted that accounts for bothP-∆ and P-δ effects.

2. The application of a notional loadNi = 0.002 Yi where Yi is the gravityload from the appropriate loadcombination acting on level i.

3. That the analysis be based on areduced stiffness (EI* = 0.8τbEI andEA* = 0.8EA) in the structure.

Figure 5 shows a comparison of thein-plane beam-column interaction checksusing the two main different methodsallowed in Chapter C. On the left part(Figure 5(a)), the conventional effectivelength approach is shown. In this case,the yield strength (Py) is reduced by theconventional approach (kL effect) andthe usual interaction equations applied.On the right side of the figure (Figure

Figure 5 - Comparison of in-plane interaction checks for (a) the Effective Length Method and (b) the Direct Analysis Method (FigureC-A-7.1 in AISC 2005a).


Roberto T. Leon

5(b)), the direct analysis method is shown. Note that for thiscase, because of the reductions on the axial and flexure stiffnessof the members, the elastic second-order response falls belowthe actual response for much of the load path.

2.11 Inelastic Analysis and Design(Appendix 1)

Inelastic analysis is permitted for LRFD but not ASDdesign. As the design is governed by the ductility of the plastichinge zones, the specified minimum yield stress for membersundergoing plastic hinging shall not exceed 65 ksi (450 MPa)and the compacteness criteria are those of Table B4.1 for λpmodified as follows:

(a) For webs of doubly-symmetric wide flange members andrectangular HSS in combined flexure and compression

(i) for Pu/φb Py ≤ 0.125

h/tw ≤

⎟⎟⎠

⎞⎜⎜⎝

⎛

φ−

yb

u

y PP.

FE.

7521763

(ii) for Pu/φb Py > 0.125

h/tw ≤

yyb

u

y FE.

PP

.FE. 491332121 ≥⎟

⎟⎠

⎞⎜⎜⎝

⎛

φ− (A-1-2)

(b) For flanges of rectangular box and hollow structural sectionsof uniform thickness subject to bending or compression,flange cover plates and diaphragm plates between lines offasteners or welds:

b/t ≤ 0.94 E Fy/

(c) For circular hollow sections in flexure:

D/t ≤ 0.045E/Fy

The laterally unbraced length, Lb, of the compressionflange adjacent to plastic hinge locations shall not exceed Lpd,determined as follows.

(a) For doubly symmetric and singly symmetric I-shapedmembers with the compression flange equal to or larger thanthe tension flange loaded in the plane of the web:

y

ypd r

FE

MM

..L ⎥⎦

⎤⎢⎣

⎡+=

2

10760120

where

M1 = smaller moment at end of unbraced length of beam, kip-in. (N-mm)

M2 = larger moment at end of unbraced length of beam, kip-in.(N-mm)

ry = radius of gyration about minor axis, in. (mm)

(M1 / M2) is positive when moments cause reverse curvatureand negative for single curvature.

(b) For solid rectangular bars and symmetric box beams:

y

yy

y2

1pd r

FE10.0r

FE

MM

10.017.0L ≥⎥⎦

⎤⎢⎣

⎡+= (A-1-6)

There is no limit on Lb for members with circular or squarecross sections or for any beam bent about its minor axis.

The required axial strength of members subjected to plastichinging in combined flexure and axial compression shall notexceed 0.7AgFy. The column slenderness ratio L/r of memberssubjected to combined flexure and axial compression shall not

exceed yF/E.714 .

Connections shall be designed with sufficient strengthand ductility to sustain the forces and deformations imposedunder the required loads.

Stability: Second-order effects may be neglected in frameswhere the ratio of the elastic buckling load to the loadingrequirements specified in Section B3.3

λcr > 10 (A1-7)

For 5 < λcr < 10, second-order effects may be neglectedprovided that the design load effects are amplified by the factor:

cr

11

9.0AF

λ−

=

For λcr < 5, the design shall be based on a second-orderinelastic analysis. Two options are given: direct elastic-plastichinge analysis or distributed plasticity.

Direct analysis: The requirements for direct elastic-plastichinge analysis are governed by the provisions of Appendix 7,with the following modifications:

(1) Inelastic redistribution is permitted in members satisfyingthe provisions of Section 1.2. Moments shall beredistributed based on the assumption of elastic-perfectlyplastic hinge response at the member resistance definedby the provisions of Appendix 7.

(2) The notional load Ni shall be applied as an additive lateralload for all load combinations.

Distributed plasticity: A second-order inelastic analysisthat takes into account the relevant material properties,geometric imperfections, residual stresses, partial yielding of



member cross sections and connectionforce-deformation characteristics may beused to assess the design strength ofstructural frames, provided that theanalysis is shown to model the behaviorand the following requirements aresatisfied:

(1) All contributions to the elasticstiffness of the structure shall bereduced by a factor of 0.9.

(2) The material strength used in theanalysis shall be reduced by a factorof 0.9.

(3) Unless member torsion-flexureinstabilities are captured by theanalysis, doubly-symmetric I-sectionmembers shall satisfy Eq. H1-2 andall other members shall satisfyEq. H1-1 for the limited state of out-of-plane buckling.

Moments shall be redistributedbased on the assumption of elastic-perfectly plastic hinge response at themember resistance defined by theprovisions of Section H1, with thenominal column strengths, Pn,determined using K = 1.0. For memberswhere the required axial strength is lessthan 0.15Py, moments may beredistributed based on the assumptionof elastic-perfectly plastic hingeresponse at a member resistance of φbMp,where φb = 0.9.

3. ConclusionsThe 2005 AISC Specification

represents a significant improvementover previous editions from both theeditorial and technical standpoint. Fromthe editorial standpoint, thenomenclature has been tightened andthe organization of the material followsmore closely how designers would usethe specification in practice. From thetechnical standpoint, the 2005

Specification presents a unified treatment of both the ASD and LRFD approach, andincorporates the latest knowledge in material, member, connection and structuralsystem behavior. In particular, through the new material in Chapters C and Appendices1 and 7, the new Specification brings in advanced analysis methods into the designof steel structures. It is expected that this will lead to more safe and reliable designs,more economical steel structures, more efficient design and use of steel in even moredaring structures (Figure 6).

4. References1. ACI Committee 318. ACI 318-02/ACI 318R-02 Building code requirements for reinforced

concrete and commentary. American Concrete Institute. Farmington Hills, MI. 2002.2. AISC LRFD. Specification for steel buildings - load and resistance factor design (3rd

Edition). American Institute of Steel Construction, Chicago. 1999.3. AISC ASD. Manual of steel construction - allowable stress design (9th Edition). American

Institute of Steel Construction, Chicago. 1989.4. AISC 2005a/ ANSI 360-05. Specification for structural steel buildings. American Institute

of Steel Construction, Chicago.5. AISC 2005b/ ANSI 341-05.Seismic provisions for structural steel buildings. American

Institute of Steel Construction, Chicago.6. AISC, 2005c. Code of standard practice for steel buildings and bridges. American Institute

of Steel Construction, Chicago.7. AISC, 2006/ ANSI N690L-06. Design specification for steel safety-related structures for

nuclear facilities, American Institute of Steel Construction, Chicago.8. DUNCAN, 2006. Private communication (to be published in AISC EJ in 2006).

Artigo recebido em 04/12/2006 e aprovado em 05/12/2006.

Figure 6 - Computer model for the 300m long (200 m clear span) trusses of theretractable roof of the Houston Reliant Energy Stadium (SAP 2000 model by Walter P.Moore and Assoc., Houston, TX)

www.rem.com.br

Engenharia Civil: Estrutura em Aço - SciELO

Documents

Transcript of Engenharia Civil: Estrutura em Aço - SciELO