ACI 352.1 R-89

(Reapproved 1997)

Recommendations for Design of SlabColumn Connections inMonolithic Reinforced Concrete Structures

Reported by ACI-ASCE Committee 352

James K. WightChairman

James R. Cagley*Marvin E. Criswell*Ahmad J. DurraniMohammad R. EhsaniLuis E. GarciaNeil M. Hawkins*

Norman W. HansonSecretary

Milind R. JoglekarCary S. Kopczynski’ 1)Michael E. Kreger*Roberto T. Leon*Donald F. Meinheit

Recommendations are given for determining proportions and details Chapter 4-Methods of analysis forof monolithic, reinforced concrete slab-column connections. determination of connection strength, p. 6

Included are recommendations regarding appropriate uses of slab-

column connections in structures resisting gravity and lateral forces,procedures for determinat ion of connection design forces, proce-dures for determination of connection strength, and reinforcementdetails to insure adequate strength, ductility, and structural integrity.The recommendations are based on a review of currently available

information. A commentary is provided to amplify the recommen-dations and identify available reference material. Design examples il-lustrate application of the recommendation% (Design recommenda-tions are set in standard type. Commentary is set in italics.)

4. I-General principles and recommendations4.2-Connect ions without beams4.3-Connections with transverse beams4.4-Effect of openings4.5-Strength of the joint

Keywords: anchorage (structural); beams (supports); collapse; columns (sup-ports) ; concrete slmbs: connect ions; earthquake-resistant structures; joints( junctions); lateral pressure: loads (forces); reiaforctd codcnte; reinforcingsteels; shear strength; stresses; structural design: structures.

Chapter G-Reinforcement recommendations, p. 105. l-Slab reinforcement for moment transfer5.2-Recommendations for the joint5.3-Structural integrity reinforcement5.4-Anchorage of reinforcement

Chapter 6-References, p. 166. I-Recommended references6.2-Cited references

CONTENTSChapter l-Scope, p. 1

Chapter 2-Definitions and classifications, p. 22. l-Definitions2.2-Classifications

Chapter 3-Design considerations, p. 53.1-Connection performance3.2-Types of actions on the connec t ion3.3-Determination of connection forces

AC1 Committee Reports, Guides, Standard Practices, and Commentariesrue intended for guidance in planning, designing, executing, and inspectingconstruct ion. This document is intended for the use of individuals whoare competent to evaluate the signif icance and l imitations of i ts con-tent and recommendations and who wil l accept responsibil i ty for theapplication of the material i t contains. Tbe American Concrete Institutedisclaims any and al l responsibil i ty for the stated principles. The Instituteshall not be liable for any loss or damage arising therefrom.

Reference to this document shall not be made in contract documents. Ifi tems found in this document arc desired by the Architect/Engineer to bea part of the contract documents, they shall be restated in mandatory ian-guage for incorporation by the Architect/Engineer.

Jack P. Moehle, Sub-Committee Chairman for Preparationof the Slab-Column Recommendations

Robert Park*Clarkson W. PinkhamMehdi Saiidi*Charles F. ScribnerMustafa Seckin

Gene R. Stevens*Donald R. StrandS. M. UzumeriSudhakar P. VermaLoring A. Wyllie, Jr.Liande Zhang

Examples, p. 17

Notation, p. 22

CHAPTER l-SCOPEThese recommendations are for the determination of

connection proportions and details that are intended toprovide for adequate performance of the connection ofcast-in-place reinforced concrete slab-column connec-tions. The recommendations are written to satisfy ser-viceability, strength, and ductility requirements relatedto the intended functions of the connection.

*Members of the slab-column subcommittee.Copyright 0 1997, American Concrete Institute.All rights reserved including rights of reproduction and use in any form or by any

means, including the making of copies by any photo process, or by electronic ormechanical device, printed, written, or oral, or recording for sound or visual reproduc-tion or for use in any knowledge or retrieval system or device, unless permission inwriting is obtained from the copyright proprietors.

352.1%1

352.1 Ii.2 MANUAL OF CONCRETE PRACTICE

Design of the connection between a slab and its sup-porting member requires consideration of both the joint(the volume common to the slab and the supportingelement) and the portion of the slab or slab and beamsimmediately adjacent to the joint. No reported cases ofjoint distress have been identified by the Committee.However, several connection failures associated withinadequate performance of the slab adjacent to thejoint have been reported.“’ Many of these have oc-curred during construction when young concrete re-ceived loads from more than one floor as a conse-quence of shoring and reshoring.d-‘O The disastrousconsequences of some failures, including total collapseof the structure, emphasize the importance of the de-sign of the connection. It is the objective of these rec-ommendations to alert the designer to those aspects ofbehavior that should be considered in design of theconnection and to suggest design procedures that willlead to adequate connection performance.

Previous reportsJ*” and codes (AC1 318) have sum-marized available information and presented some de-sign recommendations. The present recommendationsare based on data presented in those earlier reports andmore recent data.

The recommendations are intended to serve as aguide to practice.

These recommendations apply only to slab-columnconnections in monolithic concrete structures, with orwithout drop panels or column capitals, without slabshear reinforcement, without prestressed reinforce-ment, and using normal weight or lightweight concretehaving design compression strength assumed not to ex-ceed 6000 psi. Construction that combines slab-columnand beam-column framing in orthogonal directions atindividual connections is included, but these recom-mendations are limited to problems reldted to thetransfer of loads in the direction perpendicular to thebeam axis. The provisions are limited to connectionsfor which severe inelastic load reversals are not antici-pated. The recommendations do not apply to multi-story slab-column construction in regions of high seis-mic risk in which the slab connection is a part of theprimary lateral load resisting system. Slab-columnframing is inappropriate for such applications.

These recommendations are limited to slab-columnconnections of cast-in-place reinforced concrete floorconstruction, including ribbed floor slab construction’2and slab-column connections with transverse beams.Recommendations are made elsewhere (AU 352X) forconnections in which framing is predominantly by ac-tion between beams and columns.

The recommendations do not consider connectionswith slab shear reinforcement, slab-wall connections,precast or prestressed connections, or slabs on grade.The Committee is continuing study of these aspects ofconnection design. Relevant information on these sub-jects can be found in the literature. (See References 5,II, and 13 through 18 for slab shear reinforcement,References 19 and 20 for slab-wall connections, andACI 423.3R. and References 21 through 26 for pre-

stressed connections.) Although structures having con-crete compressive strength exceeding 6000 psi are withinthe realm of this document, the recommendations limitthe assumed maximum value of compressive strength to6000 psi.

Slab-column framing is generally inadequate as theprimary lateral load resisting system of multistorybuildings located in regions of high seismic risk (such asZones 3 and 4 as defined in ANSI A.58.1 and UBC)because of problems associated with excessive lateraldrift and inadequate shear and moment transfer capac-ity at the connection. In regions of high seismic risk, ifdesigned according to provisions of these recommen-dations, slab-column framing may be acceptable in low-rise construction and multistory construction in whichlateral loads are carried by a stiffer lateral load resist-ing system. In regions of low and moderate seismic risk(such as Zones I and 2 as defined in ANSI A.58.1 andUBC), slab-column frames may be adequate as the pri-mary lateral load resisting system, provided the con-nection design recommendations in this document arefollowed.

CHAPTER 2-DEFINITIONS ANDCLASSlFlCATlONS

2.1 -DefinitionsJoint-The part of the column within the depth of

the slab including drop panel and having plan dimen-sions equal to those of the column at the intersectionbetween the column and the bottom surface of the slabor drop panel.

Connection-The joint plus the region of the slaband beams adjacent to the joint.

Column-A cast-in-place vertical supporting ele-ment, including column capital if provided, with orwithout construction joints, designed to resist forcesfrom the slab at the connection, and having a ratio oflong to short cross-sectional dimensions not exceedingfour.

Column capital-A flared portion of the column be-low the slab, cast at the same time as the slab, and hav-ing effecfive plan dimensions assumed equal to thesmaller of the actual dimensions and the part of thecapital lying within the largest right circular cone orpyramid with a 90-deg vertex that can be includedwithin the outlines of the supporting column.

Drop panel-A thickened portion of the slab aroundthe column having thickness not less than one-quarterof the surrounding slab thickness and extending fromthe column centerline in each principal direction a dis-tance not less than one-sixth of the center-to-centerspan between columns.

Shear capital-A thickened portion of the slabaround the column not satisfying plan dimension re-quirements for drop panels.

Slab critical section-A cross section of the slab nearthe column, having depth d perpendicular to the slaband extending around the column (including capital). Acritical section should be considered around the col-umn so that its perimeter 6, is a minimum, but it need

\4T%

3521 R-4 MANUAL OF CONCRETE PRACTICE

oheor capital,

droppandr - - e--m- alabI criticd

r - -I I,* recthlr

IFI7i!c

L--J !1L - - - - - - J’ *

(d)

c .

/

Discontinuous

alob edgel - - - - - - - - - 1

is,- ----- 1-----ta than f+ d

Note: For exterior connections, the dab critical sectionshould extend to the slab edge 05 shown in (e)

if such extension will reduce the critical section perimeter.Otherwise, the slab crlticol section Is OS shown in (f)

Fig. 2.2~Examples of slab critical sections.

Note : The rocommmdatl~* apPlY

only I f c, / c2< 4

Fig. 2.3-Limitation on column aspect ratio

cent support dimension. A connection not defined as anexterior connection is considered to be an interior con-nection.

Openings or slab edges located close to the supportinterrupt the shear flow in the slab, induce momenttransfer to supports, reduce anchorage lengths, and re-duce the effective joint confinement. The distance of

four times the slab thickness is based on considerationsrelated to strength of the slab near the support.” Sev-eral examples of exterior connections are in Fig. 2.5.

Where openings are located closer than four slabthicknesses, the connection may behave as an exteriorconnection, depending on the size and proximity of theopening. To gage approximately the effect of the open-ing, radial lines are drawn from the centroid of thesupport area to the boundaries of the opening [Fig.2.5(e)]. If the length of the slab critical section enclosedwithin the radial lines exceeds the adjacent support di-mension, the connection is classified as an exteriorconnection. In the preceding, if there are no shear cap-itals, a support should be interpreted as being the col-umn plus column capital if present. If there are shearcapitals, the effect of the opening should first bechecked considering the column to act as the support,and secondly, considering the shear capital to act as the

fone-way

slab

direction ofmain slab

PlUr\,

Fig. 2.4~Moment direction for one-way slab

support. For the purpose of classifying u connection asinterior or exterior, the effect of openings on the criti-cal section around a drop panel need not be consid-ered.

Where distances to openings and free edges exceedthe aforementioned requirements, the connection maybe defined as being interior. In such cases, the diameterof the longitudinal bars should be limited so that ade-quate development is available between the column andthe opening or edge. Recommendations givenelsewhere” suggest that bars should be selected so thatthe development length is less than half the distancefrom the column face to the edge or opening.

2.2.2 A connection is classified as either Type 1 orType 2 depending on the loading conditions of the con-nection as follows:

(a) Type 1: A connection between elements that aredesigned to satisfy AC1 318 strength and serviceabilityrequirements and that are not expected to undergo de-formations into the inelastic range during the servicelife.

(b) Type 2: A connection between elements that aredesigned to satisfy AC1 318 strength and serviceabilityrequirements and that are required to possess sustainedstrength under moderate deformations into the inelas-tic range, including but not limited to connections sub-jected to load reversals.

The design recommendations for connections are de-pendent on the deformations implied for the designloading conditions. A Type 1 connection is any con-nection in a structure designed to resist gravity andnormal wind loads without deformations into the in-elastic range for expected loads. Some local yielding ofslab reinforcement may be acceptable for Type 1 con-nections. Slabs designed by conventional yield-linemethods may be included in this category, except if re-quired to resist loads as described for Type 2 connec-

(a) Edgo Ccmrtkvn (b) Cornw Connoctbn

(c) Edge Connection withTmnmern (Spandral) Beam

(d) Edga Connaetkn withshalt slab owrhang

rodw lho to boundary oi .

0 - iongth of crltlcelrlthln r&la! Ilna

‘b - door dbtmm bc.-w..uppert and openIn

e - column dimension

Not.: Connretlan consldwrcl axtrforIf a > c

and b <4h

(0) Connection with Signifkont Opening

Fig.

3-DESIGN CONSlDERAtlONS3.1 -Connection performance

The connection should be proportioned for service-ability, strength, and ductility to resist the actions andforces specified in this chapter.

3.2-Types at actions on the connection3.2.1 The design should account for simultaneous ef-

fects of axial forces, shears, bending moments, andtorsion applied to the connection as a consequence of

2.5~Examples of exterior connections

tions. A Type 2 connection is a connection betweenmembers that may be required to absorb or dissipatemoderate amounts of energy by deformations into theinelastic range. Typical examples of Type 2 connec-tions are those in structures designed to resist earth-quakes or very high winds. In structures subjected tovery high winds or seismic loads, a slab-column con-nection that is rigidly connected to the primary lateralload resisting system should be classified as a Type 2connection even though it may not be considered dur-ing design as a part of that primary lateral load resist-ing system. As noted in Chapter I, these recommenda-tions do not apply to multistory frames in regions ofhigh seismic risk in which slab-column framing is con-sidered as part of the primary lateral load resisting sys-tem.

CHAPTER

DESIGN OF SLAB-COLUMN CONNECTIONS 352.1 R-5

952.1 R-6 MANUAL OF CONCRETE PRACTICE

external loads, creep, shrinkage, temperature, andfoundation movements. Loads occurring during con-struction and during the service life should be consid-ered.

The connection should be designed for the forces dueto applied external loads and due to time-dependentand temperature effects where they are significant. Ef-fects of construction loads and early concrete strengthsare of particular importance for slabs without beams,as demonstrated by several catastrophic failures duringconstruction.‘-‘ Effects of heavy construction equip-ment and of shoring and reshorin~z7~za should be con-sidered. Effects of simultaneous bidirectional momenttransfer should be considered in design of the connec-tion, except wind or seismic lateral loads generally arenot considered to act simultaneously along both axes ofthe structure in design.

3.2.2 Moment transfer about any principal axissh,Juld be included in evaluating connection resistanceif the ratio between the factored transfer moment andfactored slab shear at the slab critical section exceeds0.2d, where d is the slab effective depth. The momentshould be taken at the geometric centroid of the slabcritical section defined in Section 2.1. Where biaxialmoments are transferred to the support, the 0.26 limi-tation can be applied independently about both princi-pal axes of the connection.

Moment transfer at a connection can reduce theshear strength of a slab-column connection. However,the strength reduction for eccentricity less than 0.2d iswithin the experimental scatter for nominally identicalconnections transferring shear only.”

3.3-Determination of connection for&s3.3.1 Forces on the connection may be determined by

any method satisfying requirements of equilibrium andgeometric compatibility for the structure. Time-depen-dent effects should be evaluated,

3.3.2 For normal gravity loads, the recommenda-tions of Section 3.3.1 may be satisfied using the DirectDesign Method or the Equivalent Frame Method ofAC1 318. For uniformly loaded slabs, slab shears at theconnection may be determined for loads within a trib-utary area bounded by panel centerlines; slab shears atfirst interior supports should not be taken less than 1.2times the tributary area values unless a compatibilityanalysis shows lower values are appropriate.

The design should account for the worst combina-tions of actions at the connection. Analysis for connec-tion forces should consider at least (a) loads producingthe maximum slab shear on the slab critical section, and(6) loads producing the maximum moment transfer atthe slab critical section.

Factored slab shear at the connection can be deter-mined by several procedures, including yield line andstrip design methods13.z9 and the equivalent framemethod. However, in typical designs, simpler proce-dures such as the use of tributary areas are acceptable.The designer is cautioned that the shear at first interior

supports is likely to be higher (by as much as 20 per-cent) than the tributary area shear9e3’ because of con-tinuity effects.

3.3.3 For lateral loads, effects of cracking, compati-bility, and vertical loads acting through lateral dis-placements (P-delta effects) should be considered.

Cracking in the connection has been shownlJ-34 to re-duce connection lateral-load stiffness to a value wellbelow the stiffness calculated by the elastic theory.3zJJThe reduction in stiffness can result in lateral drift ex-ceeding that anticipated by a conventional elastic anal-ysis. Effects of gravity loads acting through lateral dis-placements (P-delta effects) are consequently amplifiedand may play an important role in behavior and stabil-ity of slab-column frames. Methods of estimating re-duced lateral-load stiffness are discussed in References32,33, and ACI 318R.

CHAPTER 4-METHODS OF ANALYSIS FORDETERMINATION OF CONNECTION STRENGTH4.1 -General principles and recommendations

Connection strength may be determined by anymethod that satisfies the requirements of equilibriumand geometric compatibility and that considers the lim-iting strengths of the slab, the column, and the joint. Inlieu of a general analysis, strength of the slab includedin the connection may be determined according to theprocedures given in Sections 4.2, 4.3, and 4.4, andstrength of the joint may be determined according toSection 4.5.

Methods of computing strength of the slab in shearand moment transfer have received considerable atten-tion in literature in recent years. Available methods in-clude applications of yield line theory, elastic piate the-ory, beam analogies, truss models, and others.5*3M1 Theexplicit procedures given in Sections 4.2, 4.3, and 4.4provide acceptable estimates of connection strengthwith a reasonable computational effort. It is noted thatmoment transfer strength of a connection may be lim-ited by the sum of the strengths of columns above andbelow the joint; hence, connection strength should notbe assumed to exceed this limiting value.

4.2-Connections without beamsThe connection should be proportioned to satisfy

Sections 4.2.1 and 4.2.2.4.2.1 Shear

L2,l.I Connections transferring shear-Shearstrength V, in the absence of moment transfer is givenby

v, = 4 V,, where V, = C,V, (4-l)

in which @ = 0.85, V,, = the nominal shear strength,v, = basic shear strength carried by concrete, and C, isthe product of all appropriate modification factorsgiven in Table 4.1 and is taken equal to 1 .O if none ofthe modification factors of Table 4.1 are applicable

v, = (2 + 4/&) fl A,Y < 4 A, fl , (4-2)

DESION OF SLAB-COLUMN CONNECTIONS 3521 R-7

in which & = ratio of long to short cross-sectional di-mensions of the supporting column, A, = cross-sec-tional area of the slab critical section = b,,d, and f: =concrete compressive strength in units of psi and not toexceed 6000 psi.

Eq. (4-l) defines shear strength in the absence ofmoment transfer. The presence of moment may resultin decreased shear strength. Therefore, the designer iscautioned when computing the required connectionmoment strength to consider effecrs of pattern loads,lateral loads, construction loads, and possible acciden-tal loads.

Eq. (4-l) is based on a similar equation for two-wayshear strength as presented in the ACI 318. However,modification factors not included in ACI 318 are in-cluded in these recommendations. The basic shearstrength should be multiplied by each of the applicablemodificction factors in Table 4.1 to arrive at the nom-inal shear strength V,. The modification factors reflecthow each variable individually affects shear strength.There is little experimental information to show thatthe effects are cumulative. The Committee recommen-dation is intended to be conservative.

The maximum value of 4fi A, for the basic shearstrength given in Eq. (4-2) exceeds the nominal strengthof 2fibdti used for beams largely because of thegeometric confinement afforded to the slab shear faii-ure surface. As the supporting column cross section be-comes elongated, the confinement due to lateralcompression along the long face is diminished. Theterm flc in Eq. (4-2) reflects the reduction in strengthdue to reduction in lateral confinement. A similar phe-nomenon arises if the critical section perimeter b,greatly exceeds the depth d of the slab,” as occurs forthe critical section around drop panels and shear capi-tals. The values of the modification. factors as a func-tion of b,/d are based subjectively on trends observedin References 42 and 43. Research on interior connec-tions with shearhead reinforcemenP shows that thenominal strength decreases as the distance between thecritica! section and the column face increases. An evai-uation of the data by the Committee indicates that thereduction may also have been attributable to the in-crease in the ratio of the critical section dimension toslab depth.

Lightweight aggregate concretes have been observepto exhibit lower shear strengths relative to normalweight concretes having the same compressive strength.

Connections subjected to widespread flexural yieid-ing have been observed” to exhibit shear strengthslower than those observed for connections failing inshear prior to jlexurai yielding. Nominal shear strengthfor this case is reduced by a factor of 0.75. This provi-sion should be applied for ail Type 2 connections andfor some Type 1 connections. Included in the lattercategory are slabs designed by yield-line methods. Thepossibility of yield should be considered in flat-slab andflat-plate floor systems for which column layouts areirregular.

The basic shear strength given by Eq. (4-2) is written

- Modification factors for basic shear

as a function of the square root of the concrete com-pressive strength. Some researcff*“ suggests that the re-lation should be in terms of the cube root of concretestrength rather than the square root. Thus, it is possi-ble that shear strength given by Eq. (4-2) is unconser-vative for concrete strengths exceeding 6000 psi, theupper bound of strengths reported in tests of siab-coi-umn connections.

During construction, young and relatively weak con-crete may need to carry heavy loads. Low concretestrength has a greater effect on shear strength thanfiexurai strength. Thus, there is G tendency towardconnection shear failures. In checking resistance toconstruction loads that occur before the full designconcrete strength develops, it is important to use theconcrete strength corresponding to the age at which theload occurs rather than the design strength.

4.2.1.2 Connections transferring shear and mo-ment-Any connection may be designed in accordancewith the recommendations of Section 4.2.1.2(a). Con-nections satisfying the limitations of Sections 4.2.1.2(b)or 4.2.1.2(c) may be designed by the procedures listedin those sections in lieu of the procedure in Section4.2.1.2(i). All Type 2 connections should satisfy therecommendation of Section 4.2.1.2(d) in addition to theother recommendations of this section. All connectionsshould meet the recommendations of Section 4.2.2.

(a) The fraction of the transfer moment given by

l-1

YY = l+MJKj (4-3)

should be considered resisted by shear stresses acting onthe slab critical section. In Eq. (43), 8, is the ratio ofthe lengths of the sides of the slab critical section mea-sured parallel and transverse to the direction of mo-ment transfer, respectively. The shear stresses due tomoment transfer should be assumed to vary linearlyabout the centroid of the slab critical section. The al-gebraic sum of shear stresses due to direct shear andmoment transfer should not exceed the value of VJA,.

(b) Corner connections, and edge connections trans-ferring moments only perpendicular to the slab edge,may be assumed to have adequate shear strength if thefactored direct shear transferred to the column does notexceed 0.75V,, with V, defined by Eq. (4-l).

(c) Connections supported on columns having a ratioof long to short cross-sectional dimensions less than or

952.1 R-8 MANUAL OF CONCRETE PRACTICE

quai to two may be assumed to have adequate shearstrength to transfer the factored connection shear andmoment if

in which b, = perimeter of the slab critical section, I’,= factored direct shear on the slab critical section, andMvb, and iUvbl are the factored moments transferred si-multaneously to the support in the two principal direc-tions at the geometric centroid of the slab critical sec-tion. For exterior connections, moments perpendicularto the slab edge may be taken equal to zero in Eq. (4-4)if V, does not exceed 0.75 I’,, with V, defined by Eq. (4-1). The value of (Y should be taken equal to 5 for inte-rior connections and 3.5 for edge connections.

(d) For all Type 2 connections, the maximum shearacting on the connection in conjunction with inelasticmoment transfer should not exceed 0.4V,.

Shear strength may be reduced when moments aretransferred simultaneously to the connection. In Sec-tion 4.2.1.2, several alternate procedures for consider-ing the effects of moment transfer are recommended.The most general of the recommended procedures,which can be applied to connections of any geometryand loading, is described in Section 4.2.1.2(a). How-ever, connections can be designed with less computa-tional effort if they satisfy the loading and geometricrequirements of Section 4.2. I .2(b) or 4.2.1.2(c).

The design method described in Section 4.2.1.2(a) isidentical to the eccentric shear stress mode1 embodied inACI 318. It is assumed that shear stresses due to directshear on the connection are uniformly distributed onthe slab critical section. In addition, a portion of theunbalanced moment given by Eq. (4-3) is resisted by alinear variation of shear stresses on the slab”critical sec-tion. The algebraic sum of shear stresses due to directshear and moment transfer should not exceed the valueof V,/&. The portion of moment not carried by ec-centric shear stresses is to be carried by slab flexural re-inforcement according to Section 4.2.2. The method isdescribed in detail in several references (e.g., AC1318R, and Reference 13).

For corner connections, and for edge connectionstransferring moment only perpendicular to the slabedge, a simple computational design procedure is givenin Section 4.2.1.2(b). The procedure is based onresearch46 on slab-column edge connections for whichthe outside face of the column is flush with the slabedge. For such connections, moment transfer strengthperpendicular to the slab edge is governed by slab flex-Ural reinforcement within an effective transfer width,and apparently is not influenced significantly by shearon the connection. Failure apparently occurs when theconnection moment reaches theflexural strength of slabreinforcement, or the connection shear reaches theshear strength of the slab critical section. In cases wheremoments induce yield in slab flexural reinforcement,shear failure can apparently occur for shear less thanthat given by Eq. (4-I) because of loss of in-plane re-

straint when the flexural reinforcement yields. For thatreason, an upper limit equal to three-quarters of thevalue given by Eq. (4-l) is recommended. Recommen-dations for moment tmnsfer reinforcement are given inSection 4.2.2.

For interior or edge connections having a ratio be-tween long and short column dimensions less than orequal to two, effects of moment transfer on shearstrength can be accounted for by proportioning theconnection to satisfy the recommendations of Section4.2.1.2(c). Eq. (4-4) of that section essentially emu-lates, in algebraic form, the eccentric shear stress modeldescribed in Section 4.2.1.2(a). The form of Eq. (4-4)was originally presented by ACI-ASCE Committee426,” which recommended the equation for interiorconnections with a value of a equal to 5.2. The valueof a has been modified to 5.0 for interior connections.For edge connections transferring moment only paral-lel to the slab edge, a value of a equal to 3.5 is appro-priate. For edge connections also transferring momentperpendicular to the slab edge, the shear V, is usuallyless than 0.75V, in which case moments perpendicularto the slab edge can be ignored in Eq. (4-4). This equa-tion may be unconservative for connections not satis-fying the requirement for column cross section aspectratio.

The recommendation in Section 4.2.1.2(d) should beapplied to all connections without beams for which in-elastic moment transfer is anticipated. The recommen-dation is based on a review’? of data reported in Refer-ences 33, 34, and 48 through 52, and some previouslyunpublished tests, which reveal that lateral displace-ment ductility of interior connections without shear re-inforcement is inversely related to the level of shear onthe connection. Connections having shear exceeding therecommended value exhibited virtually no lateral dis-placement ductility under lateral loading. The recom-mendation of Section 4.2.1.2(d) may be waived if cal-culations demonstrate that lateral interstory drifts willnot induce yield in the slab system. For multistory con-struction, stiff lateral load resisting structural systemscomprising several structural walls may be adequate.

4.2.2 Fiexure-Slab flexural reinforcement should beprovided to carry the moment transferred to the con-nection in accordance with Section 5.1.1.

4.3~Connections with transverse beamsIf a connection has beams transverse to the span of

the slab, shear and moment transfer strength of theconnection may be determined as follows:

4.3.1 Shear strength is the smaller of the following:(a) Design shear strength-limited by beam action with

a critical section extending across the entire slab widthin a plane parallel to the beam and located a distance dfrom the face of the beam, where d is the slab effectivedepth. Design shear strength for this condition is cal-culated according to AC1 318 for beams,

(b) Design shear strength limited by the sum of de-sign strengths in shear of only the transverse beams.Design shear strength of the transverse beams at a dis-

DESIGN OF SLAB-COLUMN CONNECTIONS 353.1 R-9

tana d- from the support face should be computedconsidering interaction between shear and torsion,where d- is the beam effective depth.

4.3.2 Moment transfer strength is the smaller of thefollowing:

(a) Design flexural strength of the slab at the face ofthe support over a width equal to that of the columnstrip.

(b) Sum of the design flexural strength of the slaband the design torsional strengths of the transversebeams. Slab design flexural strength is computed overa width equal to. that of the support face.

The procedure described is based on concepts of thebeam analogy as presented in Reference 38. The pro-cedure assumes the shear strength is limited by eitherbeam action in the slab or by development of shearstrengths of the beams at the side faces of the connec-

.

tronsvsr5e

column

tion. For connections having substantial transversebeams, it is unlikely that the beams and slab will de-velop design shear strengths simultaneously, so shearstrength should be limited to the contribution of thebeams only.

Flexural strength is limited by development of a flex-Ural yield line across the slab column-strip width, inwhich case the transverse beams do not reach their de-sign strengths [Fig. 4.1(a)], or by development of ayield surface around the connection that involves flex-Ural yield of the slab and torsional yield of the traits-verse beams [Fig. 4.1(b)]. Beam torsional strength iscalculated considering interaction between shear andtorsion. The beam shear may be determined by theprocedure given in Reference 16, or more simply, allshear may be assumed distributed to beams in propor-tion to their tributary areas if the beams have equal

r-Jab

U e*d/

Mu - Slab flexural strength

for width of the

column strip

boom-M, = Momant transfer

w St1tength = M,,

(a) Strength Limited by Slab Column-Strip Capacity

f, - Boom torsional strength

Mu - Slab flexural strength

for width c2

M5 t Moment transfer

strength = M, f 21,

(b) Strength Llmitsd by Combined Flaxural/ k~~iotw~ Capacities

Fig. 4.1--Unbalanced moment strength of connections with transverse beams

3!R.lR40 MANUAL OF CONCRETE PRACTICE

Fig. &I-Illustration of cases where balanced and un-balanced connection moments predominate

stiffness. Combined shear and torsion strength may berepresented as in ACI 318 or can be based on othermethods such as those described in References 53 and16.

4.4-Effect of openingsWhen openings perpendicular to the plane of the slab

are located closer to a slab critical section than fourtimes the slab thickness, the effect of such openingsshould be taken into account. This may be done usinga general analysis that satisfies requirements of equilib-rium and compatibility. In lieu of a general analysis,Section 4.2 or 4.3 should be followed as appropriate,except that portions of the slab critical section enclosedwithin lines from the centroid of the support area to theextreme edges of the opening should be considered in-effective. The eccentricity of the applied shear causedby the opening should also be taken into account, ex-cept where the ineffective length of the slab critical sec-tion is less than either d or half the length of the adja-cent support face. The support should be consideredthe column including column capital if the critical sec-tion under consideration is adjacent to the column, andshould be considered the shear capital or drop panel ifthe critical section under consideration is adjacent tothe shear capital or drop panel.

Slab perforations and embedded service ducts dis-rupt the flow of flexural and shear stresses in the vicin-ity of the connection and generally result in decreasedstrength. The influence is a function of proximity andsize of the disruption. Effects of slab perforations andof embedded service ducts are described in Reference54:

4.5-Strength of the joint4.5.1 Axial compression--If the design compressive

strength of concrete in the column is less than or equalto 1.4 times that of the floor system, strength of thejoint in axial compression can be assumed equal tostrength of the column below the joint. Otherwise, ax-ial strength should be determined according to Section10.13 of AC1 318. The column longitudinal reinforce-ment should be continuous through the joint, with or

without splices, and the joint should be confined asspecified in Section 5.2.2 of these recommendations.

4.5.2 Shear-Calculations for joint shear strength inslab-column connections are not required.

The committee is aware of no cases of joint shearfailure in flat slab or flat plate connections. The ab-sence of joint shear failures is likely to be attributableto two phenomena: (I) For slabs of usual proportions,the magnitudes of moment transfer that can be devel-oped, and hence of the joint shear, are not excessive;and (2) confinement afforded by the slab concrete en-hances joint shear strength.

CHAPTER 5~REINF0RCEMENTREQUIREMENTS

5.1 -Slsb reinforcement for moment transfer5.1.1(a) Interior connections-Reinforcement required in

each direction to resist the moment +fMb, where 7, =1 - yy, should he placed within lines 1.5h either side ofa column (including capital), where Mub = the momenttransferred to the column in each principal direction, h= the slab thickness including drop panel, and 7, =fraction of moment transferred by flexure. The rein-forcement should be anchored to develop the tensileforces at the face of the support. Reinforcement placedto resist slab flexural moments or placed as structuralintegrity reinforcement (as recommended in Section5.3) may be assumed effective for moment transfer.

The optimum placement of reinforcement for mo-ment transfer has not been clearly established by avail-able experimental data. Current practice (AC1 318)considers reinforcement placed within I.5 slab thick-nesses both sides of the column to be effective in trans-ferring the flexural moment y,&, and observed per-formance of connections designed by this procedure hasgenerally been acceptable. Whether the reinforcementrequired for moment transfer is placed totally as topreinforcement, or whether some bottom reinforcementshould be used, is less clear and requires judgment onthe part of the engineer. As guidance, consider the twoextreme cases illustrated in Fig. 5.1.

In Case A of Fig. 5. I, the connection loading is pre-dominated by a large balanced moment. If a small ec-centric loading is introduced, the slab moment in-creases on one side of the connection and decreasesslightly (but still remains negative) on the other side ofthe connection. In this case, the designer would be pru-dent to pluce all the moment transfer reinforcement astop steel.

In the other extreme (Case B of Fig. 5.1). the con-nection is loaded by a small balanced moment and alarge moment transfer due to lateral loads. In this case,the loading results in nearly equal slab moments of op-posite sign on opposite sides of the column. Conse-quently, the total area of reinforcement required bySection 5.1. I(a) for moment transfer should be dividedequally between the top and bottom of the slab. Be-cause the loading condition shown in Case B of Fig. 5. I

DESIGN OF SLAB-COLUMN CONNECTIONS 352.1 R-l 1

normally involves moment reversals, both the top andthe bottom reinforcement should be effectively contin-uous over the column.

(b) Exterior connections-For resistance to momenttransfer parallel to the edge of edge connections, therecommendations of Section 5,1,1(a) for interior con-nections should be followed.

For resistance to moment transfer perpendicular tothe edge, including corner connections, sufficient rein-forcement should be placed within a width 2c, + c,,centered on the column, to resist the total moment tobe transferred to the column at the centroid of the slabcritical section, unless the edge is designed to transferthe torsion due to required slab reinforcement outsidethis width. The quantity c, is the distance from the in-ner face of the column to the slab edge measured per-pendicular to the edge, but not to exceed c,. In caseswhere the edge is designed for torsion, recommenda-tions of Section 5.1.1 (a) for interior connections shouldbe followed.

Experimental results’6~sJ~J6 indicate that slab rein-forcement for moment transfer perpendicular to theedge is fully effective in resisting the edge moment onlyif it is anchored within torsional yield lines projectingfrom the interior column face to the slab edge (Fig.5.2). Because of the large twist that occurs in the edgemember after torsional yield, reinforcement beyond theprojection of the yield line cannot be fully developeduntil large connection rotations occur. For the typicaltorsional yield line having a projection of approxi-mately 45 deg, only that reinforcement within the width2c, + c, is considered effective, as shown in Fig. 5.2.

If the edge has been designed for torsion, the edgemember is likely to possess greater torsional stiffness sothat reinforcement beyond the torsional yield line mightbe effective. In this case, the column strip should becapable of resisting the total moment, and sufficientreinforcement should be placed within the effectivewidth as defined in Section 5.1.1(a). There is some ex-perimental evidence to verify the performance of thistype of connection.‘6

5.1.2 At least two of the main top slab bars in eachdirection and all the structural integrity reinforcementrequired by Section 5.3 should pass within the columncage. Maximum spacing of slab flexural reinforcementplaced in both directions in the connection should notexceed twice the slab thickness.

5.1.3 Continuous bottom slab reinforcement shouldbe provided at the connection in accordance with thefollowing:

(a) Where analysis indicates that positive slab mo-ments develop at the connection, sufficient bottom re-inforcement should be provided within the column stripto resist the computed moment.

(b) Where moment transfer alone develops positiveslab moments, and the maximum shear stress on theslab critical section due to moment transfer computedin accordance with Section 4.2.1.2(a) exceeds 0.4VJA,or when the quantity 5(M,,, + Mub2)/boVu computedaccording to Section 4.2.1.2(c) exceeds 0.6, bottom re-

(a) E* Cannedon (b) Camr Connection

Fig. 5.2-Plan views showing yield lines at edge andcorner connections

inforcement should be provided in both directions. Thevalue of p’f, for that reinforcement within lines 2heither side of the column in each direction should benot less than 100 psi, where p’ is the reinforcement ra-tio of bottom slab reinforcement.

(c) Structural integrity reinforcement should be pro-vided according to provisions of Section 5.3.

Slab reinforcement is required through the columncage to insure that there is continuity between the slaband column. Minimum reinforcement in the slab sur-rounding the supporting column is necessaty to controlcracking. Concentration of reinforcement at the con-nection delaysjlexural yield of reinforcement and, thus,enhances shear strength.*9 For exterior slab-columnconnections in which the slab extends beyond the outerface of the column, the slab overhang should be pro-vided with temperature and shrinkage reinforcement asa minimum.

In designs where lateral loads are of sufficient mag-nitude that positive slab moments are computed at thecolumn face, reinforcement should be provided in thecolumn strip to resist the computed moments (Case Bin Fig. 5.1). This can occur even in buildings withstructural wall systems designed to resist the lateralload.

In designs where moment transfer is of lesser magni-tude, the total slab moment at the column face may becomputed to be negative (Case A in Fig. 5.1). How-ever, it is stil! possible that positive slab moments willdevelop near the column,s and reinforcement [Section5.1.3(b)] should be provided to resist this moment.” Atedge connections where the column is flush with theslab edge and the connection is loaded by an unbal-anced moment that produces tension at the top of theslab, the provision of Section 5.1.3(b) does not apply.

The recommendations for continuity and anchorageof bottom reinforcement presented in this and othersections of this document differ from minimum re-quirements of many codes (e.g., ACI 318). Minimumrequirements of these codes are considered to be inad-equate for many common design situations.

5.1.4 Where bottom reinforcement is placed to sat-isfy the recommendations of Section 5.1.3(a) or

252.1 R-12 MANUAL OF CONCRETE PRACTICE

5.1.3(b), the sum of the top and bottom reinforcementwithin the width c, + 3h should not exceed threequar-ters of the balanced reinforcement computed for thearea having total width c, + 3h and depth d, unlessboth the bottom and top flexural reinforcement can bedeveloped within the column.

The upper limit on the sum of continuous top andbottom reinforcement applies for cases where the coi-umn dimension is not sufficient to develop the rein-forcement, according to Section 5.4.5. In the presenceof significant moment transfer at such connections, abar in tension due toflexurai stresses on one face of thecolumn may, because of inadequate anchorage, be intension also at the opposite face of the column. Thus,both the top and bottom reinforcement may be stressedin tension on a single face. To insure that the extra ten-sile forces will not result in local crushing of slab con-crete, the sum of fop and bottom reinforcement ratiosshould not exceed three-guarters of the balanced ratio.

5.1.5 At discontinuous edges of exterior connections,all top slab reinforcement perpendicular to the edgeshould be anchored to develop the yield stress at theface of the column, and the edge should be reinforcedto satisfy the recommendations of Sections 5.1.5(a) or5.1.5(b).

(a) A beam should be provided having depth equal toor greater than the slab depth and having longitudinalreinforcement and closed stirrups designed to resist thetorsion transmitted from the discontinuous slab edge.The transverse reinforcement should extend a distancenot less than four times the slab thickness from bothsides of the support and should be spaced at not morethan 0.5d-, where dk, is the beam effective depth,except it need not be spaced less than 0.75 times theslab effective depth.

(b) An effective beam formed within the slab depthand reinforced by slab reinforcement should be pro-vided. For this effective beam, within a distance notless than two slab thicknesses on both sides of the sup-port, the top reinforcement perpendicular to the edgeshould be spaced not more than 0.75 times the slab ef-fective depth and should have a 1X-Ldeg hook with ex-tension returning along the bottom face of the slab adistance not less than I,, as defined in Section 5.4.5. Inlieu of hooked bars hairpin bars of diameter not lessthan that of the top slab bars may be inserted alongthe edge to overlap the top bars. At least four bars, ofdiameter not less than the diameter of the main slabbars, should be placed parallel to the discontinuousedge as follows: Two of the bars should be top bars,one along the slab edge and one not less than 0.75 c,nor more than c, from the slab edge. The other twobars should be bottom bars, placed so one bar is di-rectly below each of the two top bars.

At discontinuous edges, the use of spandrel beams isencouraged to insure adequate serviceability and tor-sional strength. Where spandrel beams are absent, theslab edge should be reinforced to act as a spandrelbeam. The recommended slab edge reinforcement is in-tended to control cracking. It is not intended that the

slob edge without spandrel beums be designed for tor-sion. Additionally, it is noted that the recommendededge reinforcement may be inadequate to act as a dia-phragm chord or strut tie. Typical examples of rein-forcement at edge connections are shown in Fig. 5.3.

For edge connections without beams, the bars run-ning parallel to the slab edge should be placed (wherepracticable) within the bars perpendicular to the edge orwithin the stirrups* if present.

5.24tecommendationr, for the joint5.2.1 Column longitudinal reinforcement-Column

longitudinal reinforcement passing through the jointshould satisfy Sections 10.9.1 and 10.9.2 of AC1 318.Offsets that satisfy requirements of AC1 318 are per-mitted within the joint.

In addition, the column reinforcement for Type 2joints should be distributed around the perimeter of thecolumn core. The center-to-center spacing between ad-jacent longitudinal bars should not exceed the larger of8 in. or one-third of the column cross-sectional dimen-sion in the direction for which the spacing is being de-termined.

Researchers have pointed out the need for we&d&tributed longitudinal reinforcement to confine con-crete.‘7 The recommendations for distribution of iongi-tudinai reinforcement for Type 2 connections are in-tended to insure adequate column ductility byimproving column confinement.

5.2.2 Transverse reinforcement5.2.2.1 Type I connections-Transverse reinforce-

ment is not required for interior connections. For exte-rior connections, horizontal transverse joint reinforce-ment should be provided. Within the depth of the slabplus drop panel, the reinforcement should satisfy Sec-tion 7.10 of AC1 318, with the following modifica-tions.

(a) At least one layer of transverse reinforcementshould be provided between the top and bottom levelsof slab longitudinal reinforcement.

(b) If the connection is part of the primary system forresisting nonseismic lateral loads, the center-to-centerspacing of the transverse reinforcement should not ex-ceed 8 in.

5.2.2.2 Type 2 connections-Column transversereinforcement above and below the joint should con-form to requirements of Appendix A of AC1 318.

For interior connections, transverse reinforcement isnot required within the depth of the joint. For exteriorconnections, as defined in Section 2.2.1, the columntransverse reinforcement should be continued throughthe joint, with at least one layer of transverse rein-forcement between the top and bottom slab reinforce-ment. Maximum spacing of transverse reinforcementwithin the slab depth should not exceed the smallest of(a) one-half the least column dimension, (b) eight timesthe smallest longitudinal bar diameter, or (c) 8 in. Allhoops should be closed with hooks at their ends of notless than 135 deg. Where rquired, crossties should beprovided at each layer of transverse reinforcement, and

DESIQN OF SLAB-COLUMN CONNECTIONS 3!i2.1&13

(a) Connection With Spandrel Beam

lw,dag.hook

otc mlc,< 0 < c,

90 dog. h

A

(b) “Boomleas’ Edge Connbctlon

Fig. 3.3-Typical details at discontinuous edges

each end of a crosstie should engage a perimeter longi-tudinal bar. Single-leg crossties should have a 135 degor greater bend on one end, and the other end mayhave a standard !Wdeg tie hook as defined in Section7.1 in AC1 318. If 90deg hooks are used, the hooksshould be placed at the interior face of the joint withinthe slab depth. All 135-deg hooks should have mini-mum extensions not less than the greater of 6 tie bardiameters and 3 in.

For Type I connections, joint confinement by trans-verse reinforcement is advised for exterior connectionswhere at least one face of the joint is not confined bythe slab. Because the joint may be thin in elevation, therequirements of ACI 318 are modified to recommend atleast one layer of transverse steel within the joint. Anadditional requirement is made for the more severeloading case where the slab resists lateral loads.

For Type 2 connections, the recommendations fortransverse reinforcement are the same as those given by

AC! 318 for columns in frames that are not part of thelateral force resisting system in regions of high seismicrisk, and for frames in regions of moderate seismicrisk, as appropriate.

For interior connections, adequate confinement isafforded by the slab. Reinforcement above and belowthe slab should conform to the recommendations.

Within the depth of the joint of exterior connec-tions, column longitudinal bars should be restrainedlaterally by spirals or by ties as required in Section7.10.5.3 of ACI 318 and as modified here.

5.3~Structural integrity reinforcementReinforcement as specified in 5.3.1 and 5.3.2 should

be provided to increase the resistance of the structuralsystem to progressive collapse.

5.3.1 Connections without beams-At interior con-nections, continuous bottom reinforcement passing

252.lR-14 MANUAL OF CONCRETE PRACtlCE

’ tbttanba&.I ; Iof Soot0 tha hotlmntd

Fig. S.4-Model of connection during punching failure

within the column cage in each principal directionshould have an area at least equal to

in which A,,,, = minimum area of effectively continu-ous bottom bars or mesh in each principal directionplaced over the support, w, = factored uniformly dis-tributed load, but not less than twice the slab servicedead load, I, and 1, = center-to-center span in eachprincipal direction, f, = yield stress of steel A,, and 4= 0.9. The quantity of reinforcement A, may be re-duced to two thirds of that given by Eq. (5-l) for edgeconnections, and to one-half of that given by Eq. (5-l)for corner connections. Where the calculated values ofA, in a given direction differ for adjacent spans, thelarger value should be used at that connection.

Bottom bars having area A, may be considered con-tinuous if (1) they are lap spliced outside a distance 21,from the column face with a minimum lap splice lengthequal to I,; (2) they are lap spliced within Phe columnplan area with a minimum lap splice length of I,,; (3)they are lap spliced immediately outside the columnwith a minimum lap splice of 2l,, provided the lapsplice occurs within a region containing top reinforce-ment; or (4) they are hooked or otherwise anchored atdiscontinuous edges to develop yield stress at the col-umn face.

Catastrophic progre&ve collapses have occurred inslab-column structures. ld Many of the failures have oc-curred during construction when young, relatively weakconcrete was subjected to heavy construction loads.Procedures for considering the effects of constructionloads have been described.a*J7*2d

For Type I connections, the minimum bottom rein-forcement given by Eq. (5-I) should be continuous overthe columns to reduce the likelihood of progressive col-lapse. Although not presently required by ACI 318,such reinforcement is frequently called out by manydesign offices.

For Type 2 connections, the design loading condi-tions may result in genera1 yielding of the top and/orbottom slab reinforcement at the connection. Experi-mental data” indicate that under such conditions thepunching shear strength may be reduced considerably

below the nominal value of 4flkpermitted by ACI318, thereby reducing the margin of safety against col-lapse. Thus, minimum continuous bottom reinforce-ment as specified by Eq. (5-I) is recommended to sup-port the slab in the event of a punching shear failure.

Eq. (5-l) was developed using the conceptual mode1of Fig. 5.4. In the model, the slab is supported afterpunching by bottom reinforcement draped over thesupport in the two directions. If the bottom reinforce-ment is considered to assume an angle of 30 deg withrespect to the horizontal, reinforcement having an areaequal to that given by Eq.. (5-I) will be capable of sup-porting the load w, within a tributary area equal to I,&.Identical expressions have been obtained by other in-vestigators using different interpretations of the basicmechanism.‘d~42 The adequacy of Eq. (5-l) has beendemonstrated by numerous experiments.“*42 The reduc-tions permitted for corner and edge connections resultin an area of reinforcement providing the same theo-retical resistance as provided for interior connections.For these exterior connections, 1, and 1, are intended tobe the full span dimensions, not the tributary area di-mensions.

It is noted that only bottom reinforcement is capableof significant post-punching resistance. To perform asintended, the bottom reinforcement must be effectivelycontinuous, and it must be placed directly over the col-umn and within the column cage. As depicted in Fig.5.4, top reinforcement is less effective than bottom re-inforcement because it tends to split the top concretecover.

The minimum recommended value of w, equal totwice the slab dead load is based on Reference 8, which

indicates that the total load resisted by a connectionduring construction may be approximately twice theslab dead load. Where detailed calculations and fieldmonitoring of construction loads indicate lower loads,the design may be based on the lower loads.

5.3.2 Connections with beams5.3.2.1 If the beam depth is less than twice the slab

depth at the support, the provisions of Section 5.3.1should be followed in both directions.

5.3.2.2 If the beam depth is at least equal to twicethe slab depth, adequate integrity is provided if provi-sions of AC1 318 are followed for the transverse beams,including minimum embedment of bottom bars in thesupport .

Progressive collapse has not been a prominent prob-lem in structures having beams between supports.Nonetheless, the value of well anchored bottom bars asprovision against collapse should not be overlooked.

5.4.Anchorage of reinforcement5.4.1 General recommendations-Reinforcement

should be anchored on each side of the critical sectionby embedment length or end anchorage. At connec-tions, the critical section for development of reinforce-ment is at the location of maximum bar stress. At con-nections in structures having rectangular bays, the crit-

DESIGN OF SLAB-COLUMN CONNECTIONS 352.1 R-l 5

ical section may be taken along a line intersecting thejoint face and perpendicular to the direction of the mo-ment .

5.4.2 Recommendations for Type I connections-Reinforcement at connections may be developed by us-ing hooked bars according to Section 5.4.4, by usingstraight bars passing through the connection accordingto Section 5.4.5, or by using straight bars terminatingat the connection according to Section 5.4.5.

5.4.3 Recommendations for Type 2 connections-Reinforcement at connections may be developed by us-ing hooked bars according to Section 5.4.4, except allbars terminating in the joint should be hooked withinthe transverse reinforcement of the joint using a Meghook. Alternately, anchorage may be provided bystraight bars passing through the connection accordingto Section 5.4.6. Straight bars should not be termi-nated within the region of slab comprising the connec-tion.

5.4.4 Hooked bars terminating at the connection-The development length Id,, of a bar terminating in astandard hook is

V-2)

with the following modifications:(a) The develo&ment length should be increased by 30

percent for all-lightweight and sand-lightweight con-crete.

(b) If transverse reinforcement in the joint is pro-vided at a spacing less than or equal to three times thediameter of the bar being developed, f& may be re-duced by 20 percent within the joint.

(c) For Type 1 connections, if side cover normal tothe plane of the hook is not less than,2’/1 in., and coveron the bar extension is not less than 2 in., I& may bereduced by 30 percent.

(d) For Type 1 connections, if reinforcement in ex-cess of that required for strength is provided, Id& may bereduced by the ratio AJrequired)/A,(provided).

In no case should the length I,, be less than thegreater of 6 in. or 8d,.

For most Type I and all Type 2 exterior connections,bars terminating at a connection will be anchored usinga standard hook as defined by ACI 318. The tail exten-sion of the hook should project toward the midheightof the joint. The development length given by Eq. (S-2)is similar to that required by ACI 318 and is evaluatedmore fully in work done by AC! Committee 408.‘9 Themodifications are to be applied concurrently.

The same length is specified for Type I and Type 2connections, based on the assumption that the effectsof load: reversals for Type 2 connections will be offsetby more stringent recommendations for joint confine-ment. These confinement recommendations are equiv-alent to the benefits from increased concrete cover overthe hook; hence, the modification of Section 5.4.4(c) isnot applicable to Type 2 connections. In addition, given

that yield is generally anticipated in Type 2 connec-tions, the modification of Section 5.4.4(d) is not to beapplied for the Type 2 connection.

Where significant strain hardening of reinforcementis anticipated due to inelastic deformations, 1.25 f,should be substituted for fY in Eq. (5-2).

5.4.5 Straight bars terminating at the connection-The development length id for a straight bar terminat-ing at a Type 1 connection should be computed as

1, = f,Ab ) o.o004d,fy25 47

(5-3)

provided the bar is contained within the core of thecolumn, with the following modifications:

(a) The length Id should be increased by 30 percentfor bars not terminating within the core of the column.For bars anchored partially within the column core, anyportion of the embedment length not within the con-fined core should be increased by 30 percent.

(b) The length I,, should be increased by 30 percent ifthe depth of concrete cast in one lift beneath the barexceeds 12 ln.

(c) The length I, should be multiplied by 1.33 for all-lightweight concrete, or by 1.18 for sand-lightweightconcrete.

(d) The length id may be reduced for Type 1 connec-tions by multiplying by the factor A, (required)/ A,(provided) where reinforcement is provided in excess ofthat required for strength.

The recommended development length ti similar tothat required by ACI 318.

Where the bar is not contained within the core of thecolumn, 1, should be increased as recommended in Ref-erence 59 to account for the greater tendency towardsplitting when concrete cover is small.

For Type 2 connections, straight bars should not ter-minate in the region of slab comprising the connection.

5.4.6 Bars passing through the joint-For Type 2connections, all straight slab bars passing through thejoint should be selected such that

hi/d, ) 15 (5-4)

where h, is the joint dimension parallel to the bar. Nospecial restrictions are made for column bars or forType 1 connections.

Straight slab bars are likely to slip within a joint dur-ing repeated inelastic lateral load reversals (ACI 352R,Reference 40). In slabs.of usual thickness, slip of rein-forcement can result in significant reduction of lateralload stiffness.6’ The purpose of the recommended ratiobetween bar size and joint dimension is to limit, but noteliminate, slippage of the bars through the connection.The recommended ratio is intended to avoid unusuaUylarge diameter slab bars and will not influence propor-tions in typical designs.

262.lR.16

.MANUAL OF CONCRETE PRACTICE

CHAPTER G-REFERENCES6.1 -Recommended referencesAmerican Concrete Institute318-83 (1986) Building Code Requirements

for Reinforced Concrete3181-83 Commentary on Building Code Re-

quirements for Reinforced Concrete423.3R-83 Recommendations for Concrete

Members Prestressed with UnbondedTendons

352 R-85 Recommendations for Design ofBeam-Column Joints in MonolithicReinforced Concrete Structures

American National Standards InstituteANSI A.58.1-82 Building Code Requirements for

Minimum Design Loads in Buildingsand Other Structures

International Conference of Building OfficialsU B C - 1 9 8 5 Uniform Building Code

6.2~Cited rekrencos1. “Flat Slab Breaks from Columns in Building,” Engineering

News Record, Oct. 11. 1956. pp. 24-25.2. “Building Collapse Blamed on Design, Construction,” Engi-

neering News-Record, July 15. 1971, p. 19.3. “Collapse Kills Five and Destroys Large Portion of 26Story

Apartment Building.“Engineering News-Record, Mar. 8, 1973, p.13.and subsequent articles on Mar. 15, 1973, p. 12, May 31, 1973, p. 13.and June 14, 1973, p. IS.

4. Leyendecker, Edgar V., and FattaI, !I. George, “Investigation ofthe Skyline Plaza Collapse in Fairfax County, Virginia,” BuildingScience Series No. 94, National Bureau of Standards, Washington,DC., Feb. 1977.88 pp.

5. ACI-AXE Committee 426, “The Shear Strength of ReinforcedConcrete Members-Slabs,” Proceedings, ASCE, V. IfJO, STE, Aug.1974. pp. 1543-1591.

6. Rosenblueth. Emilio, and Meli, Roberto, “The, 1985 Earth-quake: Causes and Effects in Mexico City.” Concrete International:Design & Construction. V. 8, No. 5, May 1986. pp. 23134.

7. Lew. H. S.; Carino, N. J.; and Fattal, S. 0.. “Cause of theCondominium Collapse in Cocoa Beach, Florida.” Concrere Inter-

national: Design & Construction, V. 4. No. 8. Aug. 1982. pp. 64-73.Also, Discussion, V. 5, No. 6. June 1983. pp. 58-61.

8. Agarwal, R. K., and Gardner, Noel J., “Form and Shore Re-quirements for Multistory PIat Slab Type Buildings.” AC1 JOURNAL,Proceedings V. 71, No. I 1, Nov. 1974, pp. 559-569.

9. Sbarounis, John A., “Multistory Flat Plate Buildings,” Con-crete International: Design & Construction. V. 6, No. 2. Feb. 1984.pp. 70-77.

10. Lew, H. S.; Carino, N. J.; Fattal, S. G.; and Batts, M. E.,“Investigation of Construction Failure of Harbour Cay Condomin-ium in Cocoa Beach, Florida,” Building Science Series No. 145 , Na-tional Bureau of Standards. Washington, D. C., Aug. 1982, 135 pp.

11. ACI-ASCE Committee 426, “Suggested Revisions to ShearProvisions for Building Codes,” AC1 JOURNAL, Proceedings V. 74,No. 9. Sept. 1977. pp. 458-468.

12. Meli. Robert, and Rodriguez, Mario, “Waffle Flat Plate-Col-umn Connections Under Alternating Loads,” Bulletin d’it@ormationNo. 132, ComitC Euro-International du B&on, Paris, Apr. 1979, pp.45-52.

13. Park, Robert, and Gamble, William L,. ReiMorc& ConcreteSlabs, John Wiley & Sons, New York, 1980,618 pp.

14. Vanderbilt, M. Daniel, “Shear Strength of Continuous Plates,”promdings, ASCE, V. 98. STS, May 1972. pp. 961-973.

15. Hawkins. N. M.. “Shear Strength of Slabs with MomentsTransferred to Columns,” Shear in Reinfomed Concrete. SP-42.

American Concrete Institute, Detroit, 1974, pp. 817-846.16. Rangan, B. Vijaya, and Hall, A. S.. “Moment and Shear

Transfer Between Slab and Edge Column,” AC1 JOURNAL, &owed-ings V. 5, No. 3, May-June 1983, pp. 183-191.

17. Hawkins, N. M., and Corley, W. 0.. ‘Moment Transfer toColumns in Slabs with Shearhead Reinforcement,” Shear in Rein-forced Concrete, SP42. American Concrete Institute, Detroit, 1974,pp. 847-879.

18. Corley, W. Gene, and Hawkins, Neil M., “Shearhead Rein-forcement for Slabs,” AC1 JOURNAL, promdings V. 65, No. 10, Ckt.1968, pp. 81 l-824.

19. Paulay. T., and Taylor, R. G., “Slab Coupling of EarthquakeResisting Shear Walls,” AC1 JOURNAL, Proceedings V. 78, No. 2,Mar.-Apr. 1981, pp. 130-140.

20. Schwaighofer, Joseph, and Collins, Michael P., “An Experi-mental Study of the Behavior of Reinforced Concrete CouplingSlabs,” AC1 Journal, Proceedings V. 74, No. 3, Mar. 1977. pp. I23-127.

21. Hawkins, Neil M., “Lateral Load Resistance of UnbondedPost-Tensioned Flat Plate Construction,” Jouma/, Prestressed Con-crete Institute, V. 26, No. 1, Jan.-Feb. 1981. pp 94-115.

22. Bums, Ned H., and Hemakom, Roongroj, “Test of Post-Ten-sioned Flat Plate With Banded Tendons,” Journal of Structuml En-gineering, ASCE. V. Ill. No. 9. Sept. 1985. pp. 1899-1915.

23. Kosut. Gary M.; Burns, Ned H.; and Winter, C. Victor, “Testof Four-Panel Post-Tensioned Flat Plate,” Journal ofStmctuml En-gineering, ASCE. V. 111. No. 9. Sept. 1985, pp. 19161929.

24. Smith. Stephen W., and Bums, Ned H., “Post-Tensioned FlatPlate to Column Connection Behavior,” Journal, hestressed Con-crete Institute, V. 19, No. 3, May-June 1974. pp.7691.

25. Design of Post-Tensioned Slak, Post-Tensioning Institute,Glenview. 1977.52 pp.

26. Trongtham. N., and Hawkins, N. M., “Moment Transfer toColumns in Unbonded Post-Tensioned Prestressed Concrete Slabs,”Report No. SM-77-3. Department of Civil Engineering, University ofWashington, Seattle, Oct. 1977.

27. Liu. Xi-La; Chen. Wai-Fah; and Bowman, Mark D., “Con-struction Loads on Supporting Floors,” Concmte International: De-sign & Construction, V. 7, No. 12, Dec. 1985, pp. 21-26.

28. Grundy, Paul, and Kabaila. A., “Construction Loads on Slabswith Shored Formwork in Multistory Buildings,” AC1 JournaI, Pro-ceedingsv. 60. No. 12. Dec. 1963. pp. 1729-1738.

29. Johansen, K. W., Yield Line Theory, Cement and ConcreteAssociation, London, 1962, 181 pp.

30. Hatcher, David S.; Sozen. Mete A.; and Siess, Chester P.,“Test of a Reinforced Concrete Flat Plate,” promdings, ASCE, V.91. STS, Oct. 1965, pp. 205-231.

31. Criswell, M. E., “Design and Testing of a Blast Resistant R/CSlab System.” Report No. N-72-10. U. S. Army Engineer Water-ways Experiment Station, Vicksburg, Nov. 1972.

32. Vanderbilt, M. Daniel. and Corley, W. Gene, “Frame AnaIy-sis of Concrete Buildings,” Concnrre International: Design d Con-struction, V. 5, No. 12, Dec. 1983, pp. 33-43.

33. Moehle, Jack P., and Diebold. John W., “Lateral Load Re-sponse of Flat-Plate Frame,” Journal of Structural Engineering,

ASCE. V. 111, No. 10. Oct. 1985, pp. 2149-2164.34. Mulcahy, J. F.. and Rotter, J. M., “Moment Rotation Ck-

acteristics of Flat Plate and Column Systems,” AC1 JOURNAL, Pro-ceedings V. 80. No. 2. Mar.-Apr. 1983, pp. 85-92.

35. Darvall, Peter, and Allen, Fred. “Lateral Load Effective Widthof Flat Plates with Drop Panels.” AC1 JOURNAL, Proceedings V. 81.No. 6. Nov.-Dec., 1984. pp. 613-617.

36. Regan, P. E.. and Braestrup. M. W.. “Punching Shear inReinforced Concrete,” Bulletin d’Information No, 168, ComitC Euro-International du B&ton, Lausanne. Jan. 1985, 232 pp.

37. Cracking, Deflection, and Ultimate Load of Concrete SlabSysrems, SP-30. American Concrete Institute, Detroit, 1971. 382 pp.

38. Shear in Reinforced Concrete, SP-42. American Concrete In-stitute, Detroit, 1974. 949 pp.

39. Alexander, Scott D. B.. and Simmonds. Sidney H., “UltimateStrength of Slab-Column Connections,” ACI Structuml Journal, V.84, No. 3, May-June 1987, pp. 255-261.

DESIGN OF SLAB-COLUMN CONNECTIONS 352.1 R-17

40. Simmonda, Sidney H., and Alexander, Scott D. B., “Tru88Model for Edge Column-Slab Connections,” ACI Structural JOW-nal, V. 84, No. 4, July-Aug. 1987, pp. 296-303.

41. Park, Robert, and Islam, Shatiqul, “Strength of Slab-ColumnConnections with Shear and Unbalanced Flexure,” Proceedings,AXE, V. 102. ST9, Sept. 1976, pp. 1879-1901.

42. Hawkins, N. M., and Mitchell, D., “Prolprsrive Collapse ofFlat Plate Structura,” AC1 JOURNAL, Mings V. 76, No. 7, July1979. pp. 775-808.

43. Dilger. Walter H., and Ohali, Knin. “Shear Reinforcement forConcrete Slabs,” Mfngs, ASCB, V. 107, ST12, Dec. 1981. pp.2403-2420.

44. Ivy, Charles B.; Ivey, Don L.; and Buth, Eugene, “Shear Ca-pacity of Lightweight Concrete Flat Slabs,” AC1 JOURNAL, Proceed-ings V. 66, No. 6. June 1969, pp. 490-493.

45. Zsutty, Theodore C., “Beam Shear Strength Predictions byAnalysis of Existing Data.” AC1 JOURNAL, Proceedings V. 65, No.11. Nov. 1968, pp. 943-951.

46. Moehle, Jack P., “Strength of Slab-Column Edge Connec-tions,” ACI Stmctuml Joumof, V. 85, No. 1, Jan.-Feb. 1988, pp. 89-98.

47. Pan, Aurtin, and Moehle, Jack P., “Lateral Diiplacementbuctllity of Reinforced Concrete Slab-Column Connections.” to bepublished in AC1 Strwtuml Journal.

48. Morrison. Denby G.; Hirasawa, Ikuo; attd Sozen. Mete A.,“Lateral-Load Tests of R/C Slab-Column Connections,” Joumul ofStructural Engineering, AXE, V. 109. No. 11, Nov. 1983, pp. 26%-2714.

49. Hawkinr, Neil M., “Seismic Response Constraints for SlabSystems.” Earthquake-Resistant Reiqforcvd Concmte Building Con-struction, University of California. Berkeley, 1977. V. 3. pp. 1253-1275.

50. Hanson, Norman W., and Hanson, John M., “Shear and Mo-ment Transfer Between Concrete Slabs and Columnr,” Jouma/, PCAResearch and Development Laboratories, V. 10. No. 1. Jan. 1968,pp. 2-16.

51. Islam, Shafiqul, and Park, Robert, “Tests on Slab-ColumnConnections with Shear and Unbalanced Flexure.” Proceedings,

AXE, V. 102, ST3. Mar. 1976, pp. 549-568.52. Zce, H. L., and Moehle, J. P., “Behavior of Interior and Ex-

terior Flat Plate Connections Subjected to Inelastic Load Reversals.”Report No. UCB/EERC-84/07. Earthquake Engineering ResearchCenter, University of California, Berkeley, Aug. 1984. 130 pp.

53. Collins, Michael P.. and Mitchell, Denis. “Shear and TorsionDesign of Prestrnsed and Non-Prestressed Concrete Beams.” Jour-nal. Prestressed Concrete Institute, V. 25, No. 5, Sept.-Ott. 1980. pp.32-100.

54. Hanson, John M., “Influence of Embedded Service Ducts onStrength of Flat-Plate Structures.” Research und Development But-&in No. RWM.OlD, Portland Cement Association, Skokie. 1970.16PP.

55. hghbol. E. Ramzy F., and de Paiva, H. A. Rawdon, “Testsof Flat-Plate Corner Column-Slab Connections,” Proceedings,ASCE, V. 99, ST3, Mar. 1973, pp. 551-572.

56. Rangan, B. Vijaya, and Hall, A. S., “Moment Redistributionin Flat Plate Floors,” AC1 Journal, Proc&ings V. 81, No. 6, Nov.-Dec. 1984, pp. 601-608.

57. Sheikh, Shamin A., and Uzumeri, S. M., “Strength and Duc-tility of Tied Concrete Columns,” Proreedings, AXE, V. 106. STS.May 1980. pp. 1079-l 102.

58. Mitchell, Denis. and Cook, William D., “Preventing Progres-sive Collapse of Slab Structures,” Journal of Structural Engineering,

ASCE. V. 110. No. 7. July 1984. pp. 1513.1532.59. AC1 Committee 408. “Suggested Development, Splice. and

Standard Hook Provisions for Deformed Bars in Tension,” (AC1408.1R-79). American Concrete Institute, Detroit, 1979, 3 pp.

60. fkrtuo, V. V.; Popov, E. P.; and Forzani, 9.. “Seismic Be-havior of Lightweight Concrete Beam-Column Subassemblages,”AC1 JOURNAL, Proceedings V. 77, No. 1, Jan.-Feb., 1980. pp. 44-52.

61. Hawkins, N. M., “Lateral Load Design Considerations for FlatPlate Structures,” Nonlinear &sign of Concrete Structures, StudyNo. 14. University of Waterloo Press, 1980, pp. 581-613.

EXAMPLES*

Example 1 - Dcsiga of III edge connection subjectedto gravity loading

-8” Slab

1 4 - 15’ 1

t = 40 p8f i,D = 115psf u

Type 1 connection

Design forcesU = 1.4D + 1.7Lv, = 38.6 kips

(2.2.2)

M* = 580 kip-in. at centroid of slab critical section

Check shearAssume P4 bars each way, %-in. cover

d = (7 + 6.5)/2 = 6.75 in.b. = 16 + 6.7s’ + 2(12 + 6.75/2) = 53.5 in.

A, = b.d = 53.5 x 6.75 = 361 in.v. = v, = 4A-c = 4 x 361 x &i-%6

= 91,300lb = 91.3 kips

(2.1)(4.2.1.1)

v. = cp V, = 0.85 x 91.3 = 77.6 kipsVJV, = 38,6/77.7 = 0.50 c 0.75, therefore, OK (4.2.1.2b)

Check moment transferc, + 2, = 16 + 2 x 12 = 4Oin.

+M. = +bd%(I - 0.5%f,&‘) (5.l.lb)) M. = 580 kips-in.which requires p = 0.0058; A, = 1.62 in.’

Use nine #4 bars

Reinforcement detailsTop reinforcement perpendicular to the slab edgespacing < 0.75d = 5.1 in.Development length of hooksL = v;dd45Od?)

(5.1.5b)

= (60,000 x O.S)/(SOm) = 9.5 in. (5.4.4)Structural integrity reinforcement (5.3.1)w, = greater of (1.4D + I .7L) and (20) = 0.230 ksf

A, = (++)(0.5w.l,l>)/(+L)I= (n)(O.5 x 0.230 x 22.5 x 15)/(0.9 x 60)= 0.48 in’.

Use two #5 bottom bars each way passing through columncage.

*Numbers in parentheses refer lo sections of this report.

MANUAL OF CONCRETE PRACTICE

Final design - Example 1

Column Wx16’

A-q

Exam8

lo 2 - Design of a corner connection sub.jecte to gravity loading

~Pd

Y Column l&l8

t igColumn-. i

MX

x M! Y b

f-43- slob

LPanel!

6

. - .

11.25'r

LColumn

Plan

Tamp. ond

f: = 4000 psif, = 60.000 psi

L = MPfD = 115psf

Type 1 connection

Design forces

( 2 . 2 . 2 )

u = 1.40 + 1.7L

VW = 19.3 kipM* = 290 kigin.Mb = 190 kip-in.

at centroid of slab critical section

Check shearAssume #3 bars each way, %-in. coverd = ( 7 . 0 6 + 6.69)/2 = 6 . 8 8 in.b, = 2(16 + 6.8812) = 38.9 in.A , = 3 8 . 9 x 6 . 8 8 = 268 in.

0.1)( 4 . 2 . 1 . 1 )

V” = v, = 4 A , fl = 4 x 2 6 8 x &i% = 6 7 , 8 0 0 lb = 67.8kipsV, = &V. = 0 . 8 5 x 6 7 . 8 = 5 7 . 5 k ipsVW/V, = 19.307.5 = 0.34 < 0.75. therefore, OK

Check qoment transferc, + c, = 16 + 16 = 32 in.+M, ) M* = 290 kipin.

which requires p. = 0.0035; A, = 0.78 in.’Use eight #3 bars

CbM, ) M,,, = 190 kip-in.which requires pr = 0.0025; A,. = 0.54 in.’Use five #3 bars

Reinforcement detailsTop reinforcement perpendicular to the slab edgespacing < 0.7Sd = 5.1 in.

Development length of hooksI,, = (JdJmOfl) = 7 . 1 i n .

Structural integrity reinforcementW* = 0.230 ksf as in Example I

A, = ww.w,4Y(~~~= (‘A)(03 x 0 . 2 3 0 x 2 2 . 5 x 15)/(0.9 x $0)

= 0 . 3 6 in?

( 5 . 4 . 4 )

(5.3.1)

Use two W4 bottom bars each way passing through column cage.

(4.2.1.2b)

(5.1.lb)

(5.1.5b)

IDESIQN OF SLABCOLUMN CONNECTIONS 952.1 R-19

Flnal design - Example 2

Column lS”%llc

J-Lo

1-A

APlan

Section A-A

Note: #4 bottom bara placed

t h r o u g h c o l u m n f o r p r o t e c t i o n

against progresstve collapse shall

have stand&d hooks (not shown).

Example 3 - E2

e connection subjected tobiaxial moment ue to gravity and wind loading

Loadcombination- -(1) 1.q +

I Column 12’~24~

Yt

e Pmel

I X

1;1 = 4000 psif, = 60.000 psi

L = 40 psfD = 115psfType 1 connection (2.2.2)

Design forcesU = 1.40 + 1.715

) 0.75(1*44 + 1.7L + 1.7w)) 0.9 + 1.3w

Check shearBasic data

d = 6.75 in.ba = 24 + 6.75 + 2(12 + 6.7512) = 61.5 in. (2.1)

Winddirection ki$Zt.

40.5 - 572 -394 - -413

x 3 0 . 4 -743 -298 690 -310

Y 3 4 . 8 -429 -610 - -625

I I I I IN o t e s : M, and M

M* and A?are flexural moments in the slab cohunn strip.

are moments transferred to the connection at the een-troid of the sla% critical section. For moment in the x-direction, no signis indicated. For moment in the y-direction, vahtcs are positive if thetransfer moment tends to place bottom slab steel in tension.

V. varies depending on whethn the wind is considered along the pas-itive or negative direction of the x- or y-axes. Only the larger value foreach load combination is tabulated.

A, = 61.5 x 6.75 = 415 in.*v.= Vc=4~A,-4x x 415 = 105 kips

(4.2.1.1)v. = &V. = 0.85 x 105 = 89.3 kips

Moment transfer in x-directiott:The maximum transfer moments in the x-direction occur for

loading cases (2) and (4). Loading case (4) must he checked he-cause it has the larger moment, and loading case (2) must hechecked because it has the larger shear. Both cases involve bi-ial moment transfer. Section 4.2.1.2(c) is followed.

For loading case (2)V, b V. + a WJ. + M,,)/b.

= 30.4 + 3.5(690)/61.5 = 69.7 kips, OK

For loading case (4)V, 2 18.3 + 3.5(703)/61.5 = 58.3 kips, OK

a521 R-20 MANUAL OF CONCRETE PRACTICE

Final design-Example 3

Temp. and Shrhkogatop and bottomuithln column atr@

Note: hf.,,, is taken equal to xero in the preceding becauseV./V. < 0.75.

Moment transfer in y-direction:The maximum moment transfer in the y-direction occurs un-

der uniaxial moment transfer. According to Section 4.2.1.2(b),effects of moment transfer on shear are ignored because V./V,< 0.75.

FCheck flexure

Reinforcement in x-direction:The column strip (5 1 in.) is designed to carry the total column

strip flexural moment M,,, requiring eleven 114 top and 14 at 12in. bottom (temperature and shrinkage).

For moment transfer

Y. = 1 - l/(1 + O&JB-,) = (4.2.1 Aa)t I - I/(1 + o.saJm) = 0.49Y,J%# = (1 - yJo(O3) = 359 kip-in. (S.i.lb)Transfer width = c, + 1.5h = 24 in.The column strip bars already in place will suffice if dis-tributed uniformly in column strip.

Note: SM,,Jb.V, > 0.6, therefore place temperature and shrink-age reinforcement at the bottom. (5.1.3b)

Reinforcement in y-direction:The entire moment M4 is to be resisted in flexure (5.1.lb)

by reinforcement within transfer width quai to

C, + 2e, = 24 + 2 x 12 = 48 in.

For MtiY = - 623 k&in., provide nine 14 top.For M,,,, = + 146 kip-in., provide #4 @ 12 in. (temperatureand shrinkage).Check spacing C 0.756. OK (5.1.5b)Check development, I, = 9.5 in., OK (5.4.4)

Structural integrity reinforcement (5.3.1)

= (%)(0.5 x 0.229 x 22.5 x 15)/(0.9 x 60)= 0.40 in.’

Use three #4 bottom bars each direction through column cage.

ExamR

lo 4with s

- Design of an interior connectionear capital

20’a _-

Cdumn 24”x24’ !R

f: = 4000 psif, = 60,000 psi

L = 250 psfD = 20 psf plus self weight

Type 1 connection

Design forces

(2.2.2)

U = 1.40 + 1.7LSlab reinforcement #4 bars each way.

DESIGN OF SLAB-COLUMN CONNECTIONS 352.1 R-21

Check shear Check shear(a) Around column

V” = 233 kips, M., = 300 k-in.A&/v. = 1.29 in. < O.zd, therefore, ignore moment transfer

(3.2.2)d = 10.75 in.b, = 4 x (10.75 + 24) = 139 in.A, = 139 x 10.75 = 149Oin.’ (4.2.1.1)V,a Vc=4flA,=4x 149Oxm=377kipsV, = 9 V, = 0.85 x 377 = 320 kipsV, < V., therefore, OK (4.2.1.1)

(b) Around shear capital

Assume #4 bars each way. %-in. cover.d = 6.75 in. (as per Example 1)b. = 4 x (22 + 6.75) = 115 in. (24A, = 6.75 x I15 = 776 in. (4.2.1.1)

For nonseismic loadsV” = V, = 4 A, fl = 4 x 776 x &O@r = 1% kipsv, = 6 v, = 0.85 x 1% = 167 kips > 97 kips, OK

For seismic loadsv. = C. V, = 0.75 x 1% = 147 kips

(C!. = 0.75 for seismic loads as per Table 1) (4.2.1.1)v, = 6 v. = 0.85 x 147 = 125 kips

Check moment transferV, = 225 kipsd = 6 . 7 5 in.b. - 4 x (48 + 6 . 7 5 ) = 219 in.b/d = 21916.7s = 32.4, therefore, C, = 0.75 (Table 1)A, = 219 x 6.75 = 1480in.’ (4.2.1.1)v, = C.V. 7 0.75 x 4 x 1480 x m = 281 kipsV, P 6 V, I 0.85 x 281 = 238 kips .

V, < V,, therefore, OK (4.2.1.1)

v, + P (hf.,, + M&/b, = 7 3 + 5 x 7801115= 107 kips < V., OK (4.2.1.2c)

Check maximum permitted vertical shearv, = 0.4 v, = 0.4 x 1% = 78 kips > 73 kips, OK

(4.2.1 Ski)

Reinforcement requirementsColumn strip flexural strength requirements are met by placing14 18 bars uniformly across the 10 ft wide column strip.

Reinforcement detailsProvide slab flexural steel to resist total slab moments as per AC1318.

Moment transfer strength is checked as follows7, = 1 - ‘y. = 0.6

No rquirements for moment transfer.Provide structural integrity reinforcement as per &CtiOn 5.3.1and as illustrated in previous examples.

Steel is required within c, + 3h = 46 in. to raist flexuralmomem of value y,M, = 0.6 x 780 = 470 kip-in.Reinforcement placed for total column strip moment (as perprevious paragraph) is adequate.

Example 5 - Design df an interior connection re.slsting seismic loads

Check if bottom steel is required for moment transfer (5.1.3b)SM,,/b.V. = 5 x 780/(115 x 41) = 0.83 > 0.6Therefore minimum reinforcement requirements must be met.Provide U 4 at 16 in. within width c, + 4 Ir = 54 in., result-ing in p’f, = 111 psi > 100 psi, OK.

I- 8’ Slob

f: = dOO0 psiL = 6O.OOOPsi

L = 50 psfD = 115 psf

Type 2 connection

Design forcesU = 1.40 + l.fL

> 0.75(1.40 + 1.7L. + 1.8fE)) 0.9D + 1.43W

Q.l.la)

Structural integrity reinforcement (5‘3.1)A* = (0.5wn4M~~~

= (0.5 x 0.246 x 20 x 20)/(0.9 x 60) = 0.91 in.’Use three #5 bottom bars each way passing through column

cage.

Check maximum reinforcementWithin c, + 3h = 46 in., p + p’ r; 0.75 pu, OK

Check maximum bar spacingS, = 2h = 16 in., OK

(5.1.4)

(5.1.2)

Final design - Example 5

(2.2.2)

Load combinationM.

kip-in.(1) 1.40 + 1.7L 1 4 5 0(2) 0.75(1.40 + 1.7~5 + 1.87E) E 1 4 4 0 780(3) 0.9D + 1.43E 41 960 780

Notes: M. = column strip total moment.M* = transfer moment.

NOTATIONcross-sectional area of reinforciw bar, in.’cross-sectional area of the slab critical section, in.’total area of steel at a cross section, in.’minimum area of effectively continuous bottom slabbars iu each principal direction placed over the support for resistance to progressive collapse. in?

= beam width. in.

MAN&k OF CONCRETE PRACTICE

perimeter of the slab critical section, in.dimension of the column transverse to the direction Ofmoment transferred to the column, in.dimension of the column transverse to the direction ofmoment transferred to the column, in.distance from the inner face of the column to the slabedge measured perpendicular to the edge, but not toexceed c,product of ail appropriate modification factors in Ta-ble 4.1slab effective depth, taken as the avera&te of the depthsfrom extreme concrete compression fiber to tensionsteel in two orthogonal directions. in.diameter of slab reinforcing bar, in.effective depth of transverse beam at connection, in.concrete compression strength, psidesign yield stress of slab reinforcement, psislab thickness. in.joint dimension in direction parallel to that of astraight slab bar passing through the joint, in.development length of straight bar, in.development length of hooked bar, in.center-to-center spans in each principal directionmoment transferred to the columnsimultat~us moments transferred to the column andacting in the two principal directions about the geo-metric centroid of the slab critical section

If, =

v, =

v, =

v, =

W. =

I; =

8” =

0’ =

# =Yf =

Y. =

This report was submitted to letter batlot of the c6tttmittee and vu; l p-prowl in accordance with AC1 balloting procedures.

basic shear strength of concrete without modificationsin Table 4.1, lbnominal shear strength in the absence of momenttransfer, lbdesign shear strength in the absence of moment trans-fer. lbfactored direct shear force acting on slab critical sec-tionfactored ultimate load, but not less than twice the slabdead load, to be considered for resistance to progm-sive collapsecoefficient to 3Z.q. (4.4)ratio of long to short dimensions of the column crosssectionratio of length of the slab critical section measuredparallel and transverse to the direction of momenttransfer, respectivelysteel ratio for bottom slab steel in one direction at theconnectionstrength reduction factorfractin of transfer moment at slab-column connectionthat is to be carried by slab flexure, same as AC1 318definition of 7,fraction of transfer moment at slab-column connectionthat is to be carried by eccentric shear stresses on theslab critical section. same as AC1 3 18 definition of y.

CONVERSION FACTORS1 in. = 25.4 mm1 psi - 6895 N/m*lIb=4.448N

1 kigin. = 0.113 kN-m

WERE PUBLISHED IN THE JULY-AUGUST 1999 ACI Structum~THE REPORT ACI 352.1 R-99, BUT ARE PROVIDED AS ADDI-

Recomrnendstkns for Design of SlabColumn Connections in Monolithk ReinbrcedConcrete Structures. Report by ACI-ASCE Committee 352

Discussion by Amin Ghali, 8. Vijaya Ranga( and Committee

By AMlN GHALIMewtber Amnjeon comma htlittlte, Profi?s9or of civil El@Karl~,Univemiiy of Ca&av. Calgary. Albtna. Canada

Section 4.2.1.2 of the report recommends three alter-nate methods for calculating the strength of slab-col-umn connections transferring shearing forces andbending moments. Method (a) is general and applies toany critical section at interior, edge, or corner col-umns. In this method, a fraction y. of the moment isassumed transferred by shear stress, which varies lin-early about the centroid of the slab critical section.Method (c) uses Eq. (4-4) which “emulates, in alge-braic form, the eccentric shear model” adopted inMethod (a). Thus, it can be expected that Method (a)and Eq. (4-4) give the same result. Method (b) ignoresthe moment transfer in corner and edge connectionsand considers that they have adequate strength whenthe shear stress caused by V, does not exceed 75 per-cent of VJkl,.

The assumption involved in Method (a) leads to thefollowing equation for the shear stress at any point onthe critical section

Mv=-l+Fy+ --‘X

I,(4-5)

I

where V, M,, and M, are the shear force and the mo-ments about centroidal principal axes x and y of thecritical section; A, I,, and I, are the area and secondmoments of area about the same axes.

The positive directions of the coordinates x and y andthe forces V, AU,, and M, are indicated in Fig. A. Thearrows represent a force and moments exerted by thecolumn on the critical section. Equal and opposite forceand moments representing the effect of the critical sec-tion on the column* exist but are not shown in Fig. A.

The symbols M, and M, represent the fraction of themoments transferred by shear; that is 7. multiplied bythe moment transferred between column and slab.

When using Eq. Q&5), it should be noted that x andy are the critical section centroidal principal axes, whichare not necessarily parallel to the slab edges or to theprincipal ax& of the column cross section. This will bethe case for the critical section at a corner column or atany column adjacent to nonsymmetrical openings.

The basic mechanics Eq. (4-5) is derived from the as-sumption of linear variation of v over the critical sec-tion and the conditions that Y has stress resultants qua1

*The doubkheadcd arrows shown on the plans of the slabs in Exatn~k~ 2and 3 of the report do not indiiate the moment directions unless a mention ismade that the arrow represent the action of the column on the critical sectionsor the effect of critical section on the column.

352.1 R-D2 MANUAL OF CONCRE’I’E MACT’ICE

V, M, A N D My REPftESENtEFFECTS Of COLUMN ON

ISLAB CRITICAL SECTION

I7

Fig. A - Positive directions of coordinates x and y andof V, M, and M,

to V, M,, and M,. Eq. (4-5) differs from the equationin Section 11.12.2.4 of the AC1 318-83 Commentary6*in that the critical section property J,, which the Com-mentary describes as “analogous to polar moment ofinertia,” is here replaced by the second moment ofArea I. The reasons for the change are given in Refer-ence 63, where it is shown that the Commentary equa-tion gives erroneous results when x and y are not cen-troidal principal axes, and when they are, the equationgives slightly smaller stresses than the stresses by Eq.(4-5). For the remainder of the present discussion, Eq.(4-S) will be considered applicable in Method (a) ofSection 4.2.1.2. of the Committee 352 report.

Eq. (4.4) of the report implies that the maximumshear stress at the critical section can be determined by

where b, is the perimiter of the critical section, and A&,,,and Mub2 are the factored moments transferred to thecolumn about the centroidal principal directions at thecentroid of the critical section. The values of a recom-mended are 5 for interior and 3.5 for edge connections.No value is given for corner connections, which prob-ably means that Eq. (4-6) does not apply in this case.In fact, when Eq. (4-6) is used for a corner column,with (r = 3.5, it gives a substantially different resultfrom Eq. (4-S).

The Committee report states that Eq. (4-6) does notapply when the long-to-short cross-sectional dimen-sions of the column are greater than two. There areseveral other cases not mentioned in the report forwhich Eq. (4-6) cannot possibly give the correct maxi-mum shear stress because the equation does not includethe necessary parameters. Examples of such cases are:columns with nonrectangular cross-sections, nonsym-metrical critical sections due to the presence of open-ings, and edge connections with slab overhang.

Eq. (4-5) is basic and general and does not need to besimplified by Eq. (4-6), which has so many limitations.

Method (b) is based on tests on edge connections thathave indicated that the slab strength in transfer of mo-ment perpendicular to slab edge is not influenced sig-nificantly by the shearing force. This phenomenon canmean that the fraction of moment transferred by shearis smaller for exterior columns than for interior col-umns. Thus, in Method (a), different values of thecoefficient yV should be used for interior, edge, and

FORCE AND MOMENTS TRANS- FORCES AND MOMENTSFERRED F R O M C O L U M N T O SLAO RESISTED BY SHEAR

Fig. B - Top views of a corner-column connection ex-ample

corner columns. It is also expected that, for edge andcomer columns, the coefficient yV becomes zero whenthe critical section is sufficiently far from the columnfaces. However, research is needed before an adjust-ment of yI can be made.

Method (b) allows substantially higher force andmoment transfer compared to Method (a), as will beshown below by a numerical example of a corner col-umn connection. Method (b) can lead to unsafe designbecause it extends the results of a test series of edgeconnections to corner connections without sufficientexperimental evidence..

E X A M P L EA corner column of cross section 16 x 16 in2 is con-

nected to an 8-in. slab with d = 6.88 in. The factoredforce and moments transferred from the column to theslab are indicated in Fig. B.* It is required to deter-mine, using Method (a), a multiplier I), which, whenapplied to the transferred force and moment, will makethe connection just safe. Repeat the design usingMethod (b) to determine a corresponding multiplier Q.Assume that normal weight concrete having f: = 4000psi is used, that flexural yielding in the slab is not an-ticipated, and that the connection is of Type 1.

Method (a) - The principal axes of the slab criticalsection are inclined 45 deg to the slab edges. The prop-erties of the critical section are

A = 2(6.875)( 19.44)

= 267 in.* I, = 124 (13.74Y = 4208 in.’

I,, = 327.49y = 16,800 in.’

The transferred moments are multiplied by yy = 0.4and replaced by equivalent components in the principalx- and y-directions. The shear in the critical section areto be determined for V = 19.3 kips; M, = 136 kips-in.,MY = 28 kips-in.

*The d a t a for this exampk are tk same as for Exmnpk 2 of tk Committeereport. with the exception of the directions of tk transferred m-1,. Herethe directions of the transferred moments are chosen such tkt they produce.tensile stress in tk top slab fikr in directions prpendicular IO the inner facesof the columns. This represents the common case in practice where tk mo-ments are caused by gravity forces on the slab.

SLAB-COLUMN CONNECVONS 352.1 R-03I

The maximum shear stress occurs at Point A, whosex and y coordinates are (0 and 6.87). Maximum shearstress by Eq. (4-5)

vlmv = s + $ff! (6.87) + ?&EjO)

= 72 + 222 + 0 = 294 psi

None of the modification factors of Table 4.1 apply;thus C, = 1 .O. The connection will be just safe when q,multiplied by v, is equal to VJA, = 6(4a)

294l),= 0.85 (4m); thus I), = 0.73

Method (b) - According to this method, the con-nection will be just safe when qb multiplied-by 72 psi,which is the shear stress due to the a vertical force, isqua1 to 0.75 1$(4fl). This gives

72 ,)b = 0.75(0.85) (4+/400@; thus ?)b = 2.24

From the example just given it can be seen thatMethod (b) considers the connection to be safe underload more than three times the load allowed by Method(a). Method (b) can hardly be considered an alternate toMethod (a).

CONCLUSIONIn view of the preceding example, it is suggested that

only Method (a) be retained, with the maximum stresscalculated by Eq. (4-5). A mention may be made thatthe value of y. can be smaller than the value given byEq. (4-3) when the connection is of an exterior column.

REFERENCE62. AC1 Committee 318, “Commentary on Building Code Rc

quiremcnts for Reinforced Concrete (AC1 316R-83):’ AmericanConcrete Institute, Detroit, 1983, 155 pp., and 1986 Supplement.

63. Ghali. Amin. Discussion of Section 11.12.6.2 of “ProposedRevision to: Building Code Requirements for Reinforced Concrete(AC1 318-83) (Revised 1986);‘ reported by AC1 Committee 318, ACIStructural Journal , V. 86, No. 3, May-June 1989. p. 329.

By B. VIJAYA RANQANFACI, Associate Profess and Head, Department of Structuml Enginuring,University of New Sourh Wales, Kensington. New South Waks, Ausrmlia

The members of Committee 352 should be congratu-lated for their report- This discussion deals mainly withSections 4.2.1.2(b), and Example 1. The design methoddescribed in these sections of the report is based on thework of Professor Moehle.e I am concerned that thismethod would lead to overconservative designs in prac-tice. The following points support my concern:

1. I have reworked Example 1 using AC1 318-83.According to the AC1 Building Code method, the mo-ment transferred by direct flexure is 375 kips-in., andtherefore 580 - 375 = 205 kips-in. is transferred as

torsion. The combined maximum shear stress due di-rect shear and moment transfer is found to be 211 psi,which is less than @fl = 4 x O.SSJ;fiTbii = 215 psi.The shear strength of the slab is therefore adequate. Totransfer a moment of 375 kips-in. by direct flexure, ad-equate area steel must be provided in the vicinity of thecolumn over a width of c, + 3h = 16 + (3 x 8) = 40in. This requires p = 0.0039; A, = 1.05 in. whichshould be compared with p = 0.0058; A, = 1.62 in.given in Example 1. In other words, the proposedmethod requires over 50 percent more steel than thatneeded by the AC1 Code method within the same slabwidth of 40 in. The AC1 method has been in use formore than 20 years, and I am not aware of any evi-dence showing that it is not adequate. With the adventof microcomputers and programmable calculators, verylittle effort is required to check a slab for adequateshear strength using the AC1 method. For these rea-sons, I fail to see the necessity for the proposed methodthat would lead to overconservative designs. Also, thesupporting data for limiting the spacing of bars to amaximum of 0.75d is not given in the report.

2. The overconservative nature of the proposedmethod is further supported by the results obtainedfrom a slapb specimen tested recently at the Universityof New South Wales. The test specimen is similar to theone I have tested earlier,16 except that there are noclosed ties in the slab at the edge.

The test specimen is a half-scale model of an edgeconnection with the following details: slab thickness =100 mm (3.94 in.), d = 82 mm (3.23 in.), c, = 250 mm(9.84 in.), c, = 200 mm (7.87 in.), fJ = 48.3 MPa(7004 psi), and slab steel perpendicular to the free edgeconsisted of 6.3 mm (0.25 in.) diameter bars at spac-ings of 100 mm (3.94 in.) at the top and 115 mm (4.53in.) at the bottom, In addition, two 8 mm (0.31 in.) di-ameter bars and two 6.3 mm (0.25 in.) diameter barswere also placed at the top within the column width.The yield strength of 6.3 mm bar is 460 MPa (66.7 ksi)and that of 8 mm bar is 535 MPa (77.6 ksi). The spec-imen failed in punching shear and the measured forcesat failure are V, = 108.2 kN (24.4 kips) and I&, = 27.9kNm (247 kips-in.). For this specimen, 6, = 864 mm(34.0 in.), V, = 864 x 82 x 0.34a = 167.4 kN(37.7 kips) and VU/V, = 108.2/167.4 = 0.65 < 0.75.According to the proposed method, therefore, thestrength of this edge connection is given by the mo-ment transfer strength of the slab flexural steel withinthe width of c, + 2c, = 200 + (2 x 250) = 700 mm(27.6 in.), which is found to be 14.0 kNm (124 kips-in.). The ratio of test strength/predicted strength =27.9D4.0 = 2.0.

I have calculated the strength of this test specimenusing the AC1 Building Code method. The predictedmoment transfer strength is 20 kNm (177 kips-in.) andtherefore test value/calculated value = 27.9120.0 =1.40.

I have also calculated the strength of this connectionusing the simple formula given in the Australian Stan-dard.” The predicted shear strength is 77.3 kN (17.4kips) and the test/calculated ratio is 108.2/77.3 = 1.40.

952.1 R-04 MANUAL W C0NCRETE Pf4ACtlCE

REFERENCE61. “Australii Standard for Concrete Structures,” (AS 360&1988),Standards Association of Australia, North Sydney, Mar. 1988, 108

P P .

COMMITfEE CLOSUREThe committee thanks Professors Ghali and Rangan

for their discussions of the recommendations. TheCommittee will consider seriously the points made inthe discussions in its future deliberations. Response totheir comments follows.

The committee agrees with Dr. Ghali that the threemethods of Section 4.2.1.2 for determining connectionshear and moment transfer strength do not produceidentical results. Method (a) of that section is the fa-miliar shear and moment transfer method of the AC1Building Code. Because this method has been success-ful for design for many years, the committee did notattempt to modify this method. The committee cannotcomment on the modifications to this method that wereproposed by Dr. Ghali. Those modifications and theirbases were submitted as discussion to the Committee3 18 proposed revisions and are not available to thecommittee.

Method (b) of Section 4.2.1.2 is not intended to pro-duce designs that are exactly the same as those pro-duced by Method (a); it is an alternative that has beenfound to match experimental data better than doesMethod (a). Recent comparisons with experimentaldata65 indicate that Method (b) is applicable to cornerconnections. Dr. Ghali suggests in his example that theshear stress for a corner connection should be calcu-lated about an axis 45 deg relative to the column prin-cipal axes. The committee will consider this recommen-dation further. However, the committee notes thatconnections tested in the laboratory typicaily displayyield lines across the slab prior to punching of the con-nection.

The committee thanks Dr. Ghali for pointing outthat Method (c) of Section 4.2.1.2 is applicable only torectangular interior connections and rectangular exte-

rior connections transferring moment parallel to theedge. For columns having other cross sections, Method(a) of Section 4.2.1.2 should be used.

The committee agrees with Dr. Rangan that the rec-ommendations may result in more reinforcement nearexterior columns than is required by the AC1 BuildingCode. The concentration of reinforcement is recom-mended to improve performance of the connection. Asreinforcement is spread over wider distances, the con-nection becomes more flexible, and, especially underlateral loading, may not be able to develop its strengthwithin reasonable deformation limits. Further, it shouldbe noted that the recommendations, although requiringconcentration of reinforcement near the exterior col-umn, do not require use of more reinforcement in to-tal. According to either the AC1 Building Code or theCommittee 352 recommendations, the total quantity ofreinforcement at the edge is determined by the totalslab moment at the edge.

The additional experimental data provided by Dr.Rangan are welcome and will be studied further. Asnoted in the background to the recommendations* theeccentric shear stress model of Section 4.2.1.2(a) of therecommendations usually is more conservative than isthe model of Section 4.2.1.2(b).

The committee agrees that computers and calculatorsfacilitate design but disagrees that such equipment ob-viates simplified techniques. When simplified tech-niques provide insight into the proportioning and de-tailing process, or simply aid conceptual design, theybecome tools as valuable as the computer or calculatorthat are purported to replace them.

REFERENCES65. Twang, S.-J., “An Experimental Study of Flat-Plate Strut-

tures Under Vertical and Lateral Loads,” graduate thesis, Universityof California, Berkeley, Jan. 1989. 271 pp.

66. Moehle, Jack P.; Kreger. Michael E.; and Leon, Roberto,“Background to Recommendat ions for Design of Reinforced Con-crete Slab-Column Connections.” ACJ Stmctural Journal, V. 85. No.6. Nov.-Dec. 1988. pp. 636-644.