MesginaG

StripMethodDesign Handbook

[Typethedocumentsubtitle]
Professor A.Hillerborg

11

 StripMethodDesignHandbook
JOIN US ON THE INTERNET VIA WWW, GOPHER, FTP OR EMAIL: WWW: http://www.thomson.com GOPHER: gopher.thomson.com A service of FTP: ftp.thomson.com EMAIL: findit@kiosk.thomson.com

StripMethodDesignHandbook
Professor A.Hillerborg Emeritus Professor at Lund Institute of Technology, Sweden

E & An Imprint of Chapman & Hall London Weinheim New York Tokyo Melbourne Madras

FN

SPON

Published by E & FN Spon, an imprint of Chapman & Hall, 26 Boundary Row, London SE1 8HN, UK Chapman & Hall, 26 Boundary Row, London, SE1 8HN, UK Chapman & Hall GmbH, Pappelallee 3, 69469 Weinheim, Germany Chapman & Hall USA, 115 Fifth Avenue, New York, NY 10003, USA Chapman & Hall Japan, ITP-Japan, Kyowa Building, 3F, 221 Hirakawacho, Chiyoda-ku, Tokyo 102, Japan Chapman & Hall Australia, 102 Dodds Street, South Melbourne, Victoria 3205, Australia Chapman & Hall India, R.Seshadri, 32 Second Main Road, CIT East, Madras 600 035, India First edition 1996 This edition published in the Taylor & Francis e-Library, 2003. 1996 ISBN 0-203-47467-8 Master e-book ISBN ISBN 0-203-23874-5 (OEB Format) ISBN 0 419 18740 5 (Print Edition)
2

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the UK Copyright Designs and Patents Act, 1988, this publication may not be reproduced, stored, or transmitted, in any form or by any means, without the prior permission in writing of the publishers, or in the case of reprographic reproduction only in accordance with the terms of the licences issues by the Copyright Licensing Agency in the UK, or in accordance with the terms of licences issued by the appropriate Reproduction Rights Organization outside the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publishers at the London address printed on this page. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. A catalogue record for this book is available from the British Library

TableofContents
TableofContents..........................................................................................................................................4 Notation........................................................................................................................................................7 Conversionfactors........................................................................................................................................8 Preface..........................................................................................................................................................8 CHAPTER1Introduction............................................................................................................................. 11 1.1Scope.................................................................................................................................................11 1.2Thestripmethod............................................................................................................................. 11 1.3Stripmethodversusyieldlinetheory........................................................................................... 12 1.4Stripmethodversustheoryofelasticity....................................................................................... 13 1.5Serviceability...................................................................................................................................13 1.5.1Cracking....................................................................................................................................13 1.5.2Deformations............................................................................................................................ 15 1.6Liveloads.........................................................................................................................................15 1.7Minimumreinforcement ................................................................................................................. 15 CHAPTER2Fundamentalsofthestripmethod...................................................................................... 16 2.1General.............................................................................................................................................16 2.2Therationalapplicationofthesimplestripmethod.................................................................... 17 2.3Averagemomentsinonewayelements....................................................................................... 20 2.3.1General......................................................................................................................................20 2.3.2Uniformloads........................................................................................................................... 21 2.3.3Loadswithalinearvariationinthereinforcementdirection..............................................22 2.3.4Loadswithalinearvariationatrightanglestothereinforcementdirection.....................23 2.3.5Elementswithashearforcealonganedge............................................................................ 25 2.3.6Elementswithaskewanglebetweenspanreinforcementandsupport.............................25 2.4Designmomentsinonewayelements.......................................................................................... 27 2.4.1Generalconsiderations............................................................................................................ 27 2.4.2Lateraldistributionofdesignmoments................................................................................. 27 2.5Designmomentsincornersupportedelements.......................................................................... 28 2.5.1Cornersupportedelements.................................................................................................... 28 2.5.2Rectangularelementswithuniformloads............................................................................. 29 2.5.3Nonrectangularelementswithuniformloadsandorthogonalreinforcement.................30 2.5.4Elementswithnonorthogonalreinforcement...................................................................... 31 2.5.5Elementswithnonuniformloads.......................................................................................... 31 2.6Concentratedloads......................................................................................................................... 32 2.6.1Onewayelements.................................................................................................................... 32 2.6.2Cornersupportedelements.................................................................................................... 33 2.7Strips................................................................................................................................................34 2.7.1Combiningelementstoformstrips........................................................................................ 34 2.7.2Continuousstripswithuniformloads.................................................................................... 35 2.8Supportbands.................................................................................................................................35 2.8.1General......................................................................................................................................35 2.8.2Comparisonwithcornersupportedelements ....................................................................... 36 2.8.3Applicationrules...................................................................................................................... 36 2.9Ratiosbetweenmoments............................................................................................................... 38 2.9.1Ratiobetweensupportandspanmomentsinthesamedirection.......................................38 2.9.2Momentsindifferentdirections............................................................................................. 38 2.10Lengthandanchorageofreinforcingbars.................................................................................. 39 2.10.1Onewayelements .................................................................................................................. 39 2.10.2Cornersupportedelements.................................................................................................. 40 2.10.3Anchorageatfreeedges........................................................................................................ 40 2.11Supportreactions............................................................................................................................ 41 CHAPTER3Rectangularslabswithallsidessupported........................................................................ 42 4

3.1Uniformloads..................................................................................................................................42 3.1.1Simplysupportedslabs............................................................................................................ 42 3.1.2Fixedandsimplesupports...................................................................................................... 43 3.2Triangularloads.................................................................................................................................45 3.3Concentratedloads......................................................................................................................... 49 3.3.1General......................................................................................................................................49 3.3.2Aconcentratedloadalone....................................................................................................... 49 3.3.3Distributedandconcentratedloadstogether........................................................................ 52 CHAPTER4Rectangularslabswithonefreeedge................................................................................. 54 4.1Introduction.....................................................................................................................................54 4.1.1Generalprinciples.................................................................................................................... 54 4.1.2Torsionalmoments.Cornerreinforcement........................................................................... 55 4.2Uniformloads ....................................................................................................................................55 4.3Triangularloads.................................................................................................................................58 4.4Concentratedloads......................................................................................................................... 61 4.4.1Loadsclosetothefreeedge.................................................................................................... 61 4.4.2Loadsnotclosetothefreeedge.............................................................................................. 62 CHAPTER5Rectangularslabswithtwofreeedges................................................................................. 64 5.1Twooppositefreeedges................................................................................................................. 64 5.2Twoadjacentfreeedges................................................................................................................. 64 5.2.1General......................................................................................................................................64 5.2.2Simplysupportededges,uniformloads................................................................................. 64 5.2.3Onefixededge,uniformloads................................................................................................. 66 5.2.4Twofixededges,uniformloads.............................................................................................. 69 5.2.5Nonuniformloads................................................................................................................... 70 CHAPTER6Triangularslabs.................................................................................................................... 71 6.1General.............................................................................................................................................71 6.1.1Reinforcementdirections........................................................................................................ 71 6.1.2Calculationofaveragemomentsinwholeelements............................................................. 71 6.1.3Distributionofreinforcement................................................................................................. 72 6.2Uniformloads..................................................................................................................................72 6.2.1Allsidessimplysupported...................................................................................................... 72 6.2.2Onefreeedge............................................................................................................................ 75 6.2.3Fixedandsimplysupportededges......................................................................................... 77 6.3Triangularloads.................................................................................................................................78 6.4Concentratedloads........................................................................................................................... 80 CHAPTER7Slabswithnonorthogonaledges............................................................................................ 80 7.1General..........................................................................................................................................80 7.2Fourstraightedges........................................................................................................................... 81 7.2.1Alledgessupported .................................................................................................................. 81 7.2.2Onefreeedge............................................................................................................................ 82 7.2.3Twooppositefreeedges.......................................................................................................... 85 7.2.4Twoadjacentfreeedges.......................................................................................................... 86 7.3Othercases.......................................................................................................................................89 7.3.1Circularslabswithauniformload.......................................................................................... 89 7.3.2Generalcasewithalledgessupported................................................................................... 91 7.3.3Generalcasewithonestraightfreeedge............................................................................... 96 7.3.4Generalcasewithtwoormorefreeedges............................................................................. 96 CHAPTER8Regularflatslabswithuniformloads................................................................................. 97 8.1General.............................................................................................................................................97 8.1.1Definitionofregular.............................................................................................................. 97 8.1.2Droppanelsandcolumncapitals............................................................................................ 97 8.1.3Determinationofspan............................................................................................................. 97 8.1.4Calculationofaveragedesignmoments................................................................................. 98 8.1.5Lateraldistributionofreinforcement ..................................................................................... 99 5

8.1.6Summaryofthedesignprocedure........................................................................................ 100 8.2Exteriorwallorbeamsupports................................................................................................... 101 8.2.1Onesingleinteriorcolumn.................................................................................................... 101 8.2.2Morethanoneinteriorcolumn............................................................................................. 103 8.3Exteriorcolumnsupports............................................................................................................. 105 8.3.1General....................................................................................................................................105 8.3.2Columnsupportatoneedge .................................................................................................. 105 8.3.3Columnsupportatacorner ................................................................................................... 107 8.4Slabcantileveringoutsidecolumns................................................................................................ 108 8.5Oblongpanelsandcornersupportedelements............................................................................. 109 CHAPTER9Regularflatslabswithnonuniformloads............................................................................. 111 9.1Introduction....................................................................................................................................111 9.2Uniformloadsinonedirection....................................................................................................... 111 9.3Differentloadsonpanels................................................................................................................ 113 9.4Concentratedloads......................................................................................................................... 115 CHAPTER10Irregularflatslabs................................................................................................................ 119 10.1General..........................................................................................................................................119 10.2Designprocedure.......................................................................................................................... 119 10.3Edgesstraightandfullysupported............................................................................................... 121 10.4Edgesstraightandpartlycolumnsupported................................................................................ 125 10.5Edgecurvedandfullysupported.................................................................................................. 128 10.6Edgecurvedandcolumnsupported............................................................................................. 131 10.7Slabcantileveringoutsidecolumns.............................................................................................. 134 CHAPTER11Lshapedslabsandlargewallopenings............................................................................... 145 11.1General..........................................................................................................................................145 11.2Reentrantcorner........................................................................................................................... 146 11.3Supportingwallwithalargeopening....................................................................................... 148 11.3.1Innerwall.............................................................................................................................. 148 11.3.2Wallalonganedge............................................................................................................... 150 11.3.3Slabcantileveringoutsidewall........................................................................................... 154 CHAPTER12Openingsinslabs................................................................................................................. 157 12.1General...................................................................................................................................157 12.2Slabswithalledgessupported................................................................................................... 158 12.2.1Rectangularslabs................................................................................................................. 158 12.2.2Nonrectangularslabs.......................................................................................................... 164 12.3Slabswithonefreeedge............................................................................................................. 166 12.3.1Openingnotclosetothefreeedge..................................................................................... 166 12.3.2Openingatthefreeedge...................................................................................................... 205 12.4Slabswithtwofreeedges........................................................................................................... 169 12.4.1Twooppositefreeedges...................................................................................................... 169 12.4.2Twoadjacentfreeedgesandsimplesupports..................................................................208 12.4.3Twoadjacentfreeedgesandfixedsupports.....................................................................171 12.5Cornersupportedelements .......................................................................................................... 211 CHAPTER13Systemsofcontinuousslabs................................................................................................ 174 13.1General..........................................................................................................................................174 13.2Systemsofrectangularslabs......................................................................................................... 175 13.3Rectangularslabsandconcretewalls........................................................................................... 223 13.4Othercases...................................................................................................................................223 CHAPTER14Joistfloors............................................................................................................................ 224 14.1General..........................................................................................................................................224 14.2Noncornersupportedfloors........................................................................................................ 224 14.3Floorswithcornersupportedelements....................................................................................... 228 CHAPTER15Prestressedslabs.................................................................................................................. 234 15.1General..........................................................................................................................................234 15.2Thesimplestripmethodfortendons............................................................................... 191 6

15.3Prestressedsupportbands........................................................................................................... 236 15.4Flatslabs ........................................................................................................................................237 References................................................................................................................................................240

Notation
a b ba c l l M Widthofreinforcementfordistributionofaconcentratedload,Section2.6.1. Widthofareinforcementbandforcarryingaconcentratedload,Section2.6.1. Average width of the elements which are supported by a support band, see Section 2.8.3 Lengthinthereinforcementdirectionfromasupporttothelineofzeroshearforcein an element, see Section 2.3. Indices are used to separate different lengths within an element,Section2.3,orlengthsbelongingtodifferentelementsintheexamples. Widthofanelement,seeSection2.3.Indicesareusedtoindicatedifferentpartsof the width. Additional length of a reinforcing bar for anchorage beyond the point where it can theoreticallybeended(Section2.10).InSection14.2ithasanothermeaning. DesignmomentinkNm.Apositivemomentisamomentwhichcausestensioninthe bottom reinforcement. Indices are used for the direction of the reinforcement correspondingtothemoment(xory,sometimesalsoz), forsupportmoment(s)orspanmoment(f), for the place where the moment is acting, e.g. a number of an element or a letter denotingasupportortwolettersdenotingthespanbetweentwosupports. DesignmomentperunitwidthinkNm/m.IndicesareusedinthesamewayasforM.In general m stands for an average moment on the width of an element. Examples of notationsare: Spanmomentintheloadbearingdirection. Designspanmomentforreinforcementinthexdirection. Designspanmomentinanelementdenoted1. DesignsupportmomentatsupportA. Design span moment for reinforcement in the xdirection for the span between supportsAandB. Design span moment for the span between A and B, the direction not necessarily followingacoordinateaxis. LoadorshearforceinkN/m. LoadperunitareainkN/m2. ReactionforceinkNorkN/m. Coordinates. Ratio between moment in the middle strip and the average moment in a corner supportedelement,Section2.5.2. Ratiobetweenwidthofsupport stripandtotalwidthofa cornersupportedelement, Section2.5.2. Factor for the determination of the length of support bars in cornersupported elements,Section2.10.2. Afreeedge. Asimplysupportededge. Afixedorcontinuousedge. Anopeninginaslab. Asupportingwall. Asupportingcolumn.
7

m mf mxf mf1 msA mxAB mAB Q q R x,y,(z)

Alimitedloadedarea. Adividinglinebetweenelements,asarulealineofzeroshearforce.Designmoments are,withafewexceptions,activeinsuchlines. Thepositionofasupportband.Itmayalsoshowthepositionofalineofzeromoment incaseswherethebottomandtopreinforcementhavedifferentdirections. Theloadbearingdirectioninaonewayelement.Iftwosignswithdifferentdirections areshownwithinthesameelementitmeansthattheloadisdividedbetweenthetwo directions. Thetwoloadbearingdirectionsinacornersupportedelement. Diagram showing the lateral distribution of a design moment. The lines within the diagram show the direction of the reinforcement and the values of the moments are writteninacorrespondingdirection. Diagramillustratingaloaddistribution.

Conversionfactors
TheSIsystemisusedthroughoutthebook.Allsizesaregiveninm(metres).Allloadsandforcesare giveninkN(kilonewtons),kN/m(kilonewtonspermetre)orkN/m2(kilonewtonspersquaremetre), depending on type of load or force. Bending moments designed M are always in kNm, bending momentsdesignedmarealwaysinkNm/m. SIunits USunits 1m 3.281ft 1kN 224.8lb. 1kN/m 68.52lb./ft 1kN/m2 20.89psf 1kNm 737.6ftlb. 1kNm/m 224.8ftlb./ft

Preface
In the early fifties design methods for reinforced concrete slabs were discussed within a Swedish concretecodecommittee,whereIwastheworkingmember,preparingtheproposals.Themainpoint of disagreement was whether the yield line theory was to be accepted in the code. Some of the committeememberswereagainsttheacceptanceoftheyieldlinetheorybecauseitisinprincipleon the unsafe side and may lead to dangerous mistakes in the hands of designers with insufficient knowledgeofitsapplicationandlimitations.Intheendtheyieldlinetheorywasacceptedwithsome limitations, but one of the committee members asked me if there did not exist any design method basedonthetheoryofplasticity,butwithresultsonthesafeside.TheanswerthattimewasNo. Towards the end of the committee work Professor Prager, the wellknown expert on the theory of plasticity,happenedtogiveaseriesoflecturesinSweden,whereIhadtheopportunitytogetbetter acquaintedwiththetwotheoremsofthetheoryofplasticity,theupperboundtheorem,uponwhich theyieldlinetheoryisfounded,andthelowerboundtheorem,whichbythenhadfoundnopractical application, at least not to reinforced concrete slabs. Both theorems were described as methods mainlyintendedtocheckthestrength ofagivenstructure,notin thefirstplaceasdesignmethods. Alsothe lowerboundtheoremwas mainlydescribed asabasisforcheckingthe strengthofagiven structureandtheconclusionwasthatitisnotverysuitableforthatpurpose.Thebackgroundtothis statementwasthatonlytheapplicationtohomogenousmaterialslikemetalplateswasdiscussed,not theapplicationtomaterialswherethebendingstrengthcanbevaried. It then struck me that the lower bound theorem could be used the other way round for reinforced concrete slabs, starting by seeking a statically admissible moment field and then arranging the reinforcementtotakethesemoments.Thiswasthebeginningofthestripmethod.Theideawasfirst
8

published in a Swedish journal (in Swedish) in 1956. The theory was called Equilibrium theory for reinforced concrete slabs. As a special case the assumption of strips which carried the load only by bendingmomentswasmentionedandcalledtheStripmethod.Thisiswhatwetodaycallthesimple stripmethod.Atthattimenosolutionexistedfordesigningcolumnsupportedslabsbymeansofthis equilibriumtheory. InthelatefiftiesitwasusualthatslabsinSwedishapartmentbuildingsweresupportedonwallsplus one interior column. No suitable design method existed for this case. I was asked by the head of the
design office of the Swedish firm Riksbyggen to propose a design method for this case. The result was a publicationinSwedishin1959,whichwaslatertranslatedintoEnglishbyBlakeyinAustraliaandpublishedin 1964underthetitleStripmethodforslabsoncolumns,Lshapedplates,etc.Thisextensionofthestripmethod haslaterbecomeknownastheadvancedstripmethod.

The first time the strip method was mentioned in a nonSwedish publication was at the IABSE congressinStockholmin1960,whereIpresentedashortpaperwiththetitleAplastictheoryforthe designofreinforcedconcreteslabs.Thispaperarousedtheinterestofsomeresearchers,whostudied theSwedishpublications(orunofficialtranslations)andwrotepapersandreportsaboutthetheory. ThusCrawfordtreatedthestripmethodinhisdoctoralthesisattheUniversityofIllinois,Urbana,in 1962andinacorrespondingpaperin1964. MuchearlyinterestforthestripmethodwasshownbyArmerandWood,whopublishedanumberof paperswherethemethodwasdescribedanddiscussed.Theyhaveplayedamajorroleinmakingthe methodinternationallyknown. IntheearlyseventiesIhadfoundthattheinterestinthemethodwassogreatthatitwastimetowrite abookwhichtreatedthe methodinagreaterdetail.Theresultwasabookwhichwaspublishedin Swedishin1974andinEnglishin1975withthetitleStripMethodsofDesign.Myintentionwiththat bookwastwofold.Iwishedtoshowhowmostdesignproblemsforslabscanbetreatedbymeansof the strip method in a rigorous way, but I also wished to give advice for its practical application. Whereas I think that the first goal was reached, the second was not. The book has rightly been regardedastootheoreticalanddifficultforpracticalapplication. From1973tomyretirementIwasaprofessorinbuildingmaterialsandhadtodevotemyinterestto othertopicsthantostructuraldesignproblems.DuringthisperiodIdidpracticallynothingaboutthe stripmethodexceptthecontactsIkeptwithinterestedpeople. Duringthelast20yearsthestripmethodhasbeenintroducedintomanytextbooksonthedesignof reinforcedconcrete.Inmostcasesthetreatmentismainlylimitedtothebasicideaandthetreatment ofsimplecasesbymeansofthesimplestripmethod,asthisiseasiesttoexplainandtoapply.Inmy opinion this is a pity, as the greatest advantage of the strip method is that it makes it possible to performarathersimple,safeandeconomicaldesignofmanyslabswhicharecomplicatedtodesignby meansofothermethods. Theinterestinthestripmethodthusseemstohaveincreased,butprobablythepracticalapplication haslaggedbehindbecauseofalimitedunderstandingoftheapplication.Thismademeconsiderthe possibilityofwritinganewbook,intendedforthepeopleindesignoffices.Aftermyretirementafew yearsagoIhavegotthetimeforwritingthebook.AgrantfromkeochGretaLisshedsStiftelsefor buyingacomputerandappropriateprogramsforthatpurposehasmadeitpossibleformetocarry throughtheproject. WhereasinmyearlierbookItriedtoshowrigorouslycorrecttheoreticalsolutions,thistimeIhave allowedmyselfsomeapproximationsandsimplificationswhenIhavegiventherecommendationsfor the practical application. This has been done in order to simplify and systematize the numerical analyses. As far as I can judge the resulting design is always on the safe side in spite of these approximations, which sometimes cannot be shown to formally fulfil the requirements of the lower boundtheoremofthetheoryofplasticity.Anywaythedesignisalwayssaferthanadesignbasedon the yield line theory. Checks by means of the yield line theory of slabs designed according to the recommendationsinthisbooknevershowthatitisontheunsafeside,atleastasfarasIhavefound.A formalexceptionisthatIhavedisregardedthecornerleverswhicharesometimestakenintoaccount in the yield line theory. Instead I have recommended lateral moment distributions where the influenceofthecornerleversisminimized.
9

Thebookisnotintendedtobereadrightthrough,buttobeusedindesignofficesasasupportforthe designerwhomeetsadesignproblem.Heshouldjustbeabletolookupthetypeofslabandstudythe relevantpagesinthebook. It should be pointed out that the approach in the book only gives the moments for the design of flexural reinforcement and the reaction forces, and does not give recommendations for the design withregardtoshearandpunching.Rulesfromrelevantcodeshavetobefollowedinthesecases. WhenIstartedwritingthebookIthoughtthatitwouldbeasimpleandstraightforwardtaskforme toshowhowtoapplythemethod.InpracticeitdidnotprovesosimplewhenItriedtofindsolutions whichweresimpleandeasilyexplainedinthemorecomplicatedcases.Inspiteofmyeffortsmaybe someofthesolutionsstillwillbelookeduponascomplicated.Itmusthoweverberememberedthat manyoftheslabsanalysedarestaticallycomplicated,e.g.flatslabswithirregularlyplacedcolumns, andthatitisnotrealistictohopeforverysimplesolutionsforsuchcases. The book contains thousands of numerical calculations. Although I have tried to check everything thoroughly there are certainly some errors left. As all authors know it is very difficult to observe mistakesinwhatyouhavewrittenyourself.Iaskthereadertoexcusepossiblemistakes. Iwishtoexpressmythankstoallmyfriendsandcolleaguesallaroundtheworldwhobytheirinterest and support through the years have encouraged me to decide to write this book. I refrain from mentioningnames,asthereisariskthatImightforgetsomeone. Itismysincerehopethatthebookwillproveusefulinthedesignoffices. Nykping,Sweden ArneHillerborg
10

CHAPTER1 Introduction
1.1 Scope
Thegeneralscopeofthisbookistogiveguidanceonthepracticalapplicationofthestripmethod. Thestripmethodisinprincipleamethodfordesigningslabssothatthesafetyagainstbendingfailure issufficient.Asopposedtotheyieldlinetheoryitgivesasafedesignagainstbendingfailure.Thestrip methoddoesnotinitselfleadtoadesignwhichisclosetothataccordingtothetheoryofelasticity, nordoesittakeshearorpunchingfailureintoaccount.Theadditionalrecommendationsgiveninthis bookhowevertakethemomentdistributionsaccordingtothetheoryofelasticityintoaccountinan approximatewayandgiveshearforceswhichcanbeusedinshearandpunchingdesign. ThestripmethodwasfirstdevelopedinthemidfiftiesandpublishedinSwedish.Sometranslationsin Englishwerepublishedinthesixties.Thesefirstpublicationsshowedthegeneralprinciplesandsome applications, but they were not very complete. A more complete publication in English appeared in 1975inthebookStripMethodofDesign.Thatbookhadadoublescope:todeveloptheoreticallywell founded rules for the application of the strip method to cases met with in practical design, and to demonstratetheapplication. The development of the rules for practical application involved in many cases rather complicated discussions and theoretical derivations, which were necessary in order to prove that the resulting practical rules rested on a solid theoretical basis. As a result the book has been looked upon as theoreticallycomplicatedanddifficulttoreadandapply.Thisimpres sionmayhavebeenincreasedby
itsdiscussionofmanyexamplesofdifferentalternativepossibilities.

Bearinginmindthisbackgroundandtheincreasinginterestinthestripmethod,thepresentbookhas beenwrittenwiththesingleobjectiveofdemonstratingtheapplicationtoagreatnumberofpractical examples, without discussing the theoretical background in detail. Those who are interested in the theoreticalbackgroundarereferredtothebookStripMethodofDesign. In order to make the application of the method to practical design as simple as possible some approximations have been used which have been estimated to be acceptable even though the acceptabilityhasnotbeenstrictlyproved.Evenwiththeseapproximationstheresultingdesignsare probably safer than many accepted designs based on yield line theory, theory of elasticity or code rules. Theintentionisthatadesignershouldbeabletoapplythestripmethodtothedesignofaslabmet with in practice without having to read the book but just by looking for the relevant examples and following the rules given in connection with the examples, including the references to the general guidelinesandrulesgiveninthetwointroductorychapters.

1.2 Thestripmethod
Thestripmethodisbasedonthelowerboundtheoremofthetheoryofplasticity,whichmeansthatit inprincipleleadstoadequatesafetyattheultimatelimitstate,providedthatthereinforcedconcrete slab has a sufficiently plastic behaviour. This is the case for ordinary underreinforced slabs under predominantly static loads. The plastic properties of a slab decrease with increasing reinforcement ratioandtosomeextentalsowithincreasingdepth.Withadesignbasedontherecommendationsin thisbook,includingtherecommendationsinSection1.5,thedemandontheplasticpropertiesofthe slab is not very high. The solutions should give adequate safety in most cases, possibly with the exceptionofslabsofveryhighstrengthconcretewithhighreinforcementratios.

11

Asthetheoryofplasticityonlytakesintoaccounttheultimatelimitstate,supplementaryruleshaveto be given to deal with the properties under service conditions, i.e. deflections and cracks. Such supplementary rules are given in Section 1.5, and the applications of these rules are shown and sometimesdiscussedintheexamples. Thelowerboundtheoremofthetheoryofplasticitystatesthatifamomentdistributioncanbefound which fulfils the equilibrium equations, and the slab is able to carry these moments, the slab has sufficientsafetyintheultimatelimitstate.Inthestripmethodthistheoremhasbeenreformulatedin thefollowingway: Find a moment distribution which fulfils the equilibrium equations. Design the reinforcement for thesemoments. Themomentdistributionhasonlytofulfiltheequilibriumequations,butnootherconditions,suchas therelationbetweenmomentsandcurvatures.Thismeansthatmanydifferentmomentdistributions arepossible,inprincipleaninfinitenumberofdistributions.Ofcourse,somedistributionsaremore suitable than others from different points of view. The reasons and rules for the choice of suitable distributionswillbediscussedinSection1.5.

1.3 Stripmethodversusyieldlinetheory
Theyieldlinetheoryisbasedontheupperboundtheoremofthetheoryofplasticity.Thismeansin principlethataloadisfoundwhichishighenoughtomaketheslabfail,i.e.thesafetyintheultimate limitstateisequaltoorlowerthantheintendedvalue.Ifthetheoryiscorrectlyappliedthedifference between the intended and the real safety is negligible, but there exists a great risk that unsuitable solutions may be used, leading to reduced safety factors, particularly in complicated cases like irregularslabsandslabswithfreeedges. Withthestripmethodthesolutionisinprinciplesafe,i.e.therealsafetyfactorisequaltoorhigher thantheintended.Ifunsuitablesolutionsareused,thesafetymaybemuchhigherthantheintended, leadingtoapooreconomy.Fromthepointofviewofsafetythestripmethodhastobepreferredto theyieldlinetheory. Astheyield linetheory givessafetyfactorsequal to or belowthe intended value,whereasthestrip methodgivesvaluesequaltoorabovetheintendedvalue,exactlytheintendedvaluewillbefoundin the case where the two solutions coincide. This gives the exact solution according to the theory of plasticity.Exactsolutionsshouldinprinciplebesought,asexactlytheintendedsafetygivesthebest economy.Howcloseastripmethodsolutionistotheexactsolutioncanbecheckedbyapplyingthe yieldlinetheorytothefoundsolution.Inmostoftheexamplesinthisbookacheckagainstyieldline theory shows that the difference is only a few percent, which means that the strip method leads to safetyfactorswhichareequaltoorjustslightlyabovetheintendedvalues. Whencomparingthestripmethodandtheyieldlinetheoryitshouldbenotedthatthestripmethodis adesignmethod,asamomentdistributionisdetermined,whichisusedforthereinforcementdesign. Theyieldlinetheoryisamethodforcheckofstrength.Whentheyieldlinetheoryisusedfordesign, assumptionshavetobemadeforthemomentdistribution,e.g.relationsbetweendifferentmoments. Inpracticethereinforcementisoftenassumedtobeevenlydistributed,whichasamatteroffactmay notbeveryefficient.Thestripmethodinmostapplicationsleadstoamomentdistributionwherethe reinforcementisheavieratplaceswhereitismostefficient,e.g.alongafreeedgeoraboveacolumn support.Asthestripmethodthustendstousethereinforcementinamoreefficientway,stripmethod solutionsoftengivebetterreinforcementeconomythantheyieldlinesolution,inspiteofthefactthat the strip method solution is safer. The reinforcement distribution according to the strip method solutionisoftenalsobetterfromthepointofviewofthebehaviourunderserviceconditions. Areinforcementdesigndoesnotonlymeanthedesignofthesectionsofmaximummoments,butalso thedeterminationofthelengthsofreinforcingbars,andthecurtailmentofthereinforcement.Asthe strip method design is in principle based on complete moment fields, it also gives the necessary information regarding the curtailment of reinforcement. With the yield line theory it is very complicated to determine the curtailment of reinforcement in all but the simplest cases. The result
12

from the application of the yield line theory may be either reinforcing bars which are too short or unnecessarilylongbars,leadingtopoorreinforcementeconomy,asthelengthinpracticeisbasedon estimations,duetothecomplexityofmakingtherelevantanalyses. Fromtheaboveitseemsevidentthatthestripmethodhasmanyadvantagesovertheyieldlinetheory asamethodfordesignofreinforcedconcreteslabs.Inasituationwherethestrengthofagivenslab hastobechecked,theyieldlinetheoryisusuallytobepreferred.

1.4 Stripmethodversustheoryofelasticity
Itissometimesstatedthatthestripmethodisnotaveryusefulpracticaldesignmethodtoday,aswe are able to design slabs by means of efficient finite element programs, based on the theory of elasticity.Thispointofviewisworthsomediscussion. A finite element analysis gives a moment field, including torsional moments which also have to be taken into account for the determination of the design moments for the reinforcement. The design moment field is usually unsuitable for direct use for the design of the reinforcement. The moments haveacontinuouslateralvariationwhichwouldrequireacorrespondingcontinuousvariationofthe distancesbetweenreinforcingbars.Thisisofcoursenotpossiblefromapracticalpointofview. Onesolutiontothisproblemistodesignthereinforcementforthehighestdesignmomentwithina certain width. This approach is on the safe side, but may lead to poor reinforcement economy, for example,comparedtoastripmethodsolution. A correct solution according to the theory of elasticity sometimes shows very pronounced moment concentrations.Forinstance,thisisthecaseatcolumnsupportsandsupportsatreentrantcorners.It isinpracticenotpossibletoreinforceforthesehighlocalmoments. In order to avoid poor reinforcement economy and high reinforcement concentrations the reinforcementmaybedesignedforanaveragedesignmomentoveracertainwidth.Asamatteroffact thisapproachisbasedonthetheoryofplasticity,althoughappliedinanarbitraryway.Itmayleadto resultswhichareoutofcontrolregardingsafety.Inthisaveragingprocesssomeoftheadvantagesof thetheoryofelasticityarelost.Inanefficientuseoffiniteelementbaseddesignsomepostprocessing procedurehastobeusedfortheaveraging.Theresult,e.g.regardingsafety,economy,andproperties intheservicestate,willdependonthispostprocessor. Efficient use of the finite element method, with due regard to economy and safety, may thus necessitatetheuseofrathersophisticatedprogramsincludingpostprocessors.Thecostofusingsuch programshastobecomparedtothecostofmakingadesignbymeansofthestripmethod.Inmost casesthetimeformakingastripmethoddesignbyhandcalculationissoshortthatitdoesnotpayto useasophisticatedfiniteelementprogram. Ahandcalculationbymeansofthestripmethodcanprobablyinmanycasescompetefavourablywith adesignbasedonfiniteelementanalysis. It should also be possible to write computer programs based on the strip method, although such programsdonotsofarseemtohavebeendeveloped.

1.5 Serviceability
1.5.1 Cracking
In discussing cracking and crack control it is important to take into account the importance of the cracksinarealisticway.Cracksneednotbeavoidedorlimitedunderallconditions.Wherethereisno risk of reinforcement corrosion, which is the case for most indoor structures, cracks are only to be limitediftheycauseavisibledamage.Theuppersurfaceofaslabisoftencoveredbysomeflooring, carpet, parquet etc. Then a certain amount of cracking is of no practical importance and the top reinforcementmaybeconcentratedinthepartswithinasectionwherethelargestnegativemoments may be expected under service conditions, whereas the parts with smaller moments are left unreinforced.
13

Incaseswherecrackinghastobelimitedtherehastobesufficientreinforcementinallsectionswhere themomentsarelargeenoughtocausecracks.Thisreinforcementmustnotyieldintheservicestate. The basic way of fulfilling this requirement is to choose solutions where the design moments are similar to those which may be expected according to the theory of elasticity. Some modifications of thisgeneralrulemaybeacceptedandrecommended. Itisnotnecessarytotrytofollowtheelasticmomentsindetailregardingthelateraldistributionina sectionwithmaximummoments.Thedesignmomentmaybeassumedtobeconstantoverquitelarge widths,evenifthemomentsaccordingtothetheoryofelasticityvaryinthatwidth.Themainthingis that the average moment in the section is close to the elastic value and that there is a general agreementbetweentheelasticandthechosendistribution. Whenthemomentsarecalculatedaccordingtothetheoryofelasticityitisgenerallyassumedthatthe slab has a constant stiffness, independent of the amount of reinforcement. This corresponds to an assumptionthatthereisadirectproportionalitybetweenmomentandcurvature.Intheplaceswhere the moments are largest the curvature is also largest. From this it follows that the stresses in the reinforcementarealsolargestwherethemomentsarelargest,evenifthereinforcementisdesigned for the theoretical moments. Yielding of reinforcement can be expected to ocur first in the sections wherethelargestmomentsoccurifthereinforcementisdesignedforthemomentsaccordingtothe theory of elasticity. In order to avoid yielding and large cracks more reinforcement than is needed according to the theory of elasticity should in principle be chosen for the sections with the largest moments and less reinforcement should be used in sections with smaller moments. The difference between large and small design moments should thus be exaggerated compared to the values accordingtothetheoryofelasticity. Asthesupportmomentsareoftenlargerthanthespanmomentsitmayberecommendedtochoosea higher ratio between the numerical values of the support and span moments than according to the theory of elasticity, or at least not a smaller ratio. This recommendation also leads to a good reinforcementeconomy. Themaincheckonsuitabilityofthedesignmomentsisthustheratiobetweenthenumericalvaluesof supportandspanmoments.Whereastripwithbothendsfixedhasauniformloadactingonitswhole lengthbetweenthesupportsthisratioshouldbeabout23fromthesepointsofview.Incaseswhere cracks are less important on the upper surface of the slab, e.g. where there is a floor cover, values downtoabout1.5maybeaccepted. Whereastripisloadedonlyneartheendsandunloadedinthecentralpartahigherratioispreferred. Bythechoiceofasuitableratiobetweensupportandspanmomentsregardcanalsobegiventothe relativeimportanceofcracksintheupperandlowersurfacesoftheslabforthestructureinquestion. Wheretheserulesarefollowedthedesignmomentsaccordingtothestripmethodmayprobablybe usedalsoforatheoreticalcrackcontrolaccordingtoexistingformulas. Specialattentionmayhavetobepaidtopartsofaslabwheretheloadiscarriedinaquitedifferent way from that assumed in the strip method. This is particularly the case where much of the load is carried by torsional moments, but the strip method disregards the torsional moments and assumes thatalltheloadiscarriedbybendingmomentsinthedirectionsofthecoordinateaxes.Thissituation occurs in the vicinity of corners, particularly where simply supported edges meet. It also occurs in slabswithfreeedges,whereitmayinsomecasesdominate. Cracking is best limited by reinforcement which is placed approximately in the directions of the principalmoments.Wherelargetorsionalmomentsoccurthesedirectionsdeviateconsiderablyfrom thoseofthecoordinateaxes.Wherethetorsionalmomentsdominate,thedirectionsoftheprincipal momentsareatabout45tothecoordinateaxes. Wheretwosimplysupportededgesmeetatacornerthestripmethodinitsnormalapplicationdoes not give any negative moment or top reinforcement. In reality there is a negative moment correspondingtoatorsionalmoment.Thismomentmaygivecracksapproximatelyatrightanglesto the bisector. The best way of limiting such cracks is by introducing some top corner reinforcement paralleltothebisector.Thedesignofthisreinforcementshouldbebasedonthetheoryofelasticity. Manycodesgivedesignrecommendations.
14

Corner reinforcement has nothing to do with safety and it is only needed for crack control. Where cracksontheuppersurfaceareunimportantthisreinforcementmaybeomitted. At corners where corner reinforcement may be needed the corner has a tendency to lift from the support. This should be taken into accounteither byanchoring ofthe corner,by arrangements that allowthecornertoliftwithoutdamagingtheadjacentstructure,orbymakinganintentionalcrackin theuppersurface.

1.5.2 Deformations
Thedistributionofreinforcementhasaverylimitedinfluenceonthedeformationsintheservicestate as long as it does not correspond to design moments which deviate appreciably from the moments accordingtothetheoryofelasticity.Aslongastherecommendationswithregardtocrackcontrolare followed,thedistributionofreinforcementmayberegardedasfavourablefromthepointofviewof deformations. Calculations of theoretical deformation values have to be based on the theory of elasticity. Most normalformulasandproceduresmaybeappliedtoslabsdesignedbythestripmethod.Suchanalyses willnotbediscussedorappliedinthisbook.

1.6 Liveloads
Asthestripmethodisbasedonthetheoryofplasticityitcanonlybeusedtogivethestructurethe intendedsafetyagainstcollapseunderagivenconstantloadsituation,whichisnormallyafullloadon allthestructure. Where the live load forms an important part of the total load on a continuous slab the moments at some sections may be increased by unloading some parts of the slab. Typically the increase in momentsinonepaneldependsontheremovalofloadsfromotherpanels. The strip method can be used for analysing the change in behaviour at ultimate load due to the unloading of certain panels. This is mainly a matter of changes in requirements for lengths of reinforcingbars. Where there is a repeated change in magnitudes and positions of live loads the stresses in reinforcement and concrete will vary. Such a variation leads theoretically to a decrease in safety against collapse through the effect known as shakedown. For most structures this effect is of no practicalimportance.Insomecasesitshould,however,betakenintoaccountThishastobedoneby anadditiontothedesignmoments,whichhastobecalculatedbymeansofthetheoryofelasticity. Even though the additional moments are calculated by means of the theory of elasticity the basic moments may be calculated by means of the strip method. The additional moments are generally rather small compared to the basic moments. Approximate formulas or estimates then give an acceptableaccuracy. In cases where the change in live load magnitude is large and is repeated a great many times with nearly full intensity there may be a risk of fatigue failure. Such structures should be designed by meansofthetheoryofelasticity.

1.7 Minimumreinforcement
Most codes contain rules regarding minimum reinforcement. These rules of course have to be followed.Therulesareverydifferentindifferentcountries.Themainreasonforthegreatdifferences seemstobethelackofwellfoundedjustificationfor minimum reinforcement.No accounthasbeen takenofminimumreinforcementintheexamples.
15

Whererulesforminimumreinforcementleadtomorereinforcementthanisneededaccordingtoan analysis,somereinforcementmaybesavedbymakingarevisedanalysis,wheretherelevantmoment isincreaseduptoavaluecorrespondingtotheminimumreinforcement.Thisrevisionwillleadtoa decrease in design moments at other sections with a corresponding reduction in the reinforcement requirement. A typical example is an oblong rectangular slab. The strip method will often lead to a rather weak reinforcementinthelongdirection.Throughachangeinthepositionsofthelinesofzeroshearforce thisdesignmomentcanbeincreasedwhilethedesignmomentintheshortdirectiondecreases.

CHAPTER2 Fundamentalsofthestripmethod
2.1 General
Goodintroductionstothestripmethodaregiveninmanytextbooks,e.g.byFerguson,BreenandJirsa, MacGregor, Nilson and Winter, Park and Gamble, and Wilby. For a more complete presentation see StripMethodofDesign.Hereonlyaveryshortintroductionwillbegivenandtheemphasiswillbeon rulesandrecommendationsforpracticalapplicationofthemethodtodesign. Thestripmethodisbasedonthelowerboundtheoremofthetheoryofplasticity.Thismeansthatthe solutionsobtainedareonthesafeside,providedthatthetheoryofplasticityisapplicable,whichisthe case for bending failures in slabs with normal types of reinforcement and concrete and normal proportions of reinforcement. As the theorem is usually formulated its purpose is to check the loadbearing capacity of a given structure. In the strip method an approach has been chosen which instead aims to design the reinforcement so as to fulfil the requirements of the theorem. The strip methodisthusbasedonthefollowingformulationofthelowerboundtheorem: Seekasolutiontotheequilibriumequation.Reinforcetheslabforthesemoments. Itshouldbenotedthatthesolutionhasonlytofulfiltheequilibriumequation,butnottosatisfyany compatibility criterion, e.g. according to the theory of elasticity. As a slab is highly statically indeterminatethismeansthataninfinitenumberofsolutionsexist The complete equilibrium equation contains bending moments in two directions, and torsional momentswithregardtothesedirections.Anysolutionwhichfulfilstheequationcan,inprinciple,be used for the design, and thus an infinite number of possible designs exist. For practical design it is important to find a solution which is favorable in terms of economy and of behavior under service conditions. Fromthepointofviewofeconomy,notonlyistheresultingamountofreinforcementimportant,but also the simplicity of design and construction. For satisfactory behaviour under service loading the designmomentsusedtodeterminethereinforcementshouldnotdeviatetoomuchfromthosegiven bythetheoryofelasticity. Torsional moments complicate the design procedure and also often require more reinforcement. Solutions without torsional moments are therefore to be preferred where this is possible. Such solutionscorrespondtothesimplestripmethod,whichisbasedonthefollowingprinciple: Inthesimplestripmethodtheloadisassumedtobecarriedbystripsthatruninthereinforcement directions.Notorsionalmomentsactinthesestrips. The simple strip method can only be applied where the strips are supported so that they can be treated like beams. This is not generally possible with slabs which are supported by columns, and special solution techniques have been developed for such cases. One such technique is called the advanced strip method. This method is very powerful and simple for many cases encountered in
16

practicaldesign,butashithertopresentedithashadthelimitationthatitrequiresacertainregularity inslabshapeandloadingconditions.Ithasherebeenextendedtomoreirregularslabsandloading conditions. Analternativetechniqueoftreatingslabswithcolumnsupportsorotherconcentratedsupportsisby meansofthesimplestripmethodcombinedwithsupportbands,whichactassupportsforthestrips, seeSection2.8.Thisisthemostgeneralmethodwhichcanalwaysbeappliedandwhichmustbeused where the conditions that control the use of other methods are not met. It requires a more time consuminganalysisthantheothermethods.

2.2 Therationalapplicationofthesimplestripmethod
Inthesimplestripmethodtheslabisdividedintostripsinthedirectionsofthereinforcement,which carry different parts of the total load. Usually only two directions are used, corresponding to the x and ydirections. Each strip is then considered statically as a oneway strip, which can be analyzed withordinarystaticsforbeams. Theloadonacertainareaoftheslabisdividedbetweenthestrips.Forexample,onehalfoftheload canbetakeninonedirectionandtheotherhalfinanotherdirection.Generally,thesimplestandmost economical solution is, however, found if the whole load on each area is carried by only one of the stripdirections.Thisprincipleisnormallyassumedinthisbook.Wecanthusformulatethefollowing principletobeappliedinmostcases: Thewholeloadwithineachpartoftheslabisassumedtobecarriedbystripsinonereinforcement direction. In the figures the slab is divided into parts with different load bearing directions. The relevant directionwithineachareaisshownbyadoubleheadedarrow(seeFig.2.2.1). The load is preferably carried with a minimum of cost, which normally means with a minimum amountofreinforcement.Asafirstapproximationthisusuallymeansthattheloadshouldbecarried inthedirectionthatrunstowardsthenearestsupport,asthisresultsintheminimummomentandthe minimumreinforcementarea.Fromthepointofviewofeconomy,thelengthsofthereinforcingbars arealsoimportant.Wherethemomentsarepositive,thelengthofthebarsisapproximatelyequalto thespanintherelevantdirection.Insuchcases,therefore,moreoftheloadshouldbecarriedinthe shortdirectioninarectangularslab. A consequence of these considerations is that a suitable dividing line between areas with different loadbearingdirectionsisastraightlinewhichstartsatacornerofaslabandformsananglewiththe edges.Fig.2.2.1showsatypicalsimpleexample,arectangularslabwithadistributedloadandmore orlessfixededges.Thedividinglinesareshownasdashdotlines.

Fig.2.2.1 Thedividinglinesarenormally assumedtobelinesofzeroshearforce. Alongtheselines the shear force is thus assumed to be zero (in all directions). The use of lines of zero shear force makes it possible to simplify and rationalize the design at the same time, as it usually leads to good reinforcement economy. For the choice of positions of the lines of zero shear force the following recommendationsmaybegiven.
17

A line of zero shear force which starts at a corner where two fixed edges meet may be drawn approximatelytobisecttheangleformedbytheseedges,butmaybealittleclosertoashortthantoa longedge. Alineofzeroshearforcewhichstartsatacornerwheretwofreelysupportededgesmeetshouldbe drawn markedly closer to the shorter edge. The distances to the edges may be chosen to be approximatelyproportionaltothelengthsoftheedges,forexample. Where afixed edge and a freely supportededge meet, the line of zero shear force should be drawn muchclosertothefreeedgethantothefixededge. The economy in reinforcement is not much influenced by variations in the positions of the lines of zero shear force in the vicinity of the optimum position. In cases of doubt it is fairly easy to make severalanalyseswithdifferentassumptionsandthentocomparetheresults. Theuseoflinesofzeroshearforceisbestillustratedonasimplestrip(orabeam).Theslabstripin Fig. 2.2.2 is acted upon by a uniform load q and has support moments ms1 and ms2 (shown with a positivedirection,thoughtheyarenormallynegative).Thecorrespondinglinesforshearforcesand momentsarealsoshown.Themaximummomentmfoccursatthepointofzeroshearforce. Thepartstotheleftandrightofthepointofzeroshearforcecanbetreatedasseparateelementsifwe knoworassumethepositionofthispoint.Theseseparatedelementsareshowninthelowerfigure. Thefollowingequilibriumequationisvalidforeachoftheelements: (2.1) withtheindices1and2deleted. ThebeaminFig.2.2.2canthusbelookedonasbeingformedbytwoelements,whichmeetatthepoint ofzeroshearforce.ThiscorrespondstostripsintheydirectioninthecentralpartoftheslabinFig. 2.2.1. Inmanycasestheloadedelementsofthestripsdonotmeet,butinsteadthereisanunloadedpartin between.ThisisthecaseforthinstripsinthexdirectioninFig.2.2.1.SuchastripisillustratedinFig. 2.2.3. It can be separated into three elements, the loaded elements near the ends and the unloaded element in between. The unloaded element is subjected to zero shear force, and thus carries a constantbendingmomentmf. The spanmomentsmf mustbethesameattheinnerendsofthetwo loadbearingelementsinordertomaintainequilibrium. Inarigorouslycorrectapplicationofthesimplestripmethodwewouldstudymanythinstripsinthe xdirection in Fig. 2.2.1 with different loaded lengths and different resulting design moments. This leadstoaveryunevenlateraldistributionofdesignmomentsandareinforcementdistributionwhich is unsuitable from a practical point of view. For practical design the average moment over a reasonablewidthmustbeconsidered.Inthefirstplace,therefore,theanalysisshouldgivetheaverage moments. To calculate these average moments we introduce slab elements, which are the parts of the slab borderedbylinesofzeroshearforceandonesupportededge.Eachslabelementis

18

 Fig.2.2.2 actively carrying load in one reinforcement direction, the direction shown by the doubleheaded arrow.ThustheslabinFig.2.2.1islookeduponasconsistingoffourslabelements,twoactiveinthe xdirectionandtwointheydirection.Theloadwithineachelementisassumedtobecarriedonlyby bending moments corresponding to the reinforcement direction. Such elements are called oneway elements.Eachonewayelementhastobesupportedoveritswholewidth.Theaveragemomentsand momentdistributionsarediscussedinSections2.35.Caseswhichcanbetreatedbymeansonlyof onewayelementscanbefoundbythesimplestripmethod. Even though the dividing lines between elements with different loadbearing directions, shown as dashdot lines, are normally lines of zero shear force, there are occasions when the analysis is simplifiedbyusingsuchlineswheretheshearforceisnotzero.Itisthenonthe

Fig.2.2.3 safesidetoundertaketheanalysisasiftheshearforcewerezeroalongsuchlines,providedthatthe stripofwhichtheelementformsparthasasupportatbothends. This is explained in Fig. 2.2.4, which shows a strip loaded only in the vicinity of the left end. The loaded part corresponds to an element. The righthand end of the loaded part corresponds to a dividinglinebetweenelements.Theuppercurveshowsthecorrectmomentcurve,whereasthelower
19

curveisdeterminedontheassumptionthattheshearforceiszeroatthedividingline.Themoments accordingtothelowercurvearealwaysonthesafesidecomparedtothoseaccordingtothecorrect curve. Where this approximation is used the economical consequences for the reinforcement are usuallyinsignificant.

Fig.2.2.4

2.3 Averagemomentsinonewayelements
2.3.1 General
The load on a oneway element is carried to the support by bending moments in one direction, correspondingtothedirectionofthemainreinforcement.Thisdirectionisshowninfiguresasashort linewitharrowsatbothends,seee.g.Fig2.2.1.Thistypeofarrowthusindicatesthattheelementisa onewayelementandalsoshowsthedirectionoftheloadbearingreinforcement. Aonewayelementissupportedonlyalong oneedge.Normallytheshearforce iszero along allthe otheredges.Theformulasgivenbelowrefertothiscase.Theedgeswithzeroshearforceareshownas dashdotlines. Inmostcasesthereinforcementdirectionis atrightanglestothesupportededge.Thiscasewillbe treatedfirst Theloadperunitareaisdenotedq. Theoretically,aonewayelementconsistsofmanyparallelthinstripsinthereinforcementdirection. The moment in each thin strip can be calculated and a lateral moment distribution can thus be determined. The principle is illustrated in Fig. 2.3.1, which shows a triangular element with load bearingreinforcementinthexdirectionandwithonesideparalleltothatdirection. The element carries a uniform load q and is divided into thin strips in the xdirection. Each strip is assumedtohavezeroshearforceatthenonsupportedend.Thelengthofastripisyc/l,andthesum ofendmomentsineachthinstripcanbewritten

Fig.2.3.1

(2.2) Thecorrespondinglateralmomentdistributionisshowninthefigure.Theaveragemomentisqc2/6. Formulas are given below for average moments in typical elements with distributed loads. These formulasareutilisedinthenumericalexamples.Thetheoreticalmomentdistributionsareillustrated. InpracticaldesignapplicationsthemomentdistributionissimplifiedasdiscussedinSection2.4.
20

2.3.2 Uniformloads
Inarectangularonewayelement,Fig.2.3.2,thesumofaveragemomentsis (2.3) Inatriangularonewayelement,Fig.2.3.3,thesumofaveragemomentsis (2.4) Inatrapezoidalonewayelement,Fig.2.3.4,thesumofaveragemomentsis

(2.5)

Fig.2.3.2

Fig.2.3.3

Fig.2.3.4 Inanirregularfoursidedonewayelement,Fig.2.3.5,thesumofaveragemomentsis
21

 (2.6) Thisformulacanbeexpressedinageneralwayforslabswithanarbitrarynumberofsidesandwith thenumberingoflengthsfollowingtheprinciplesofFig.2.3.5: (2.7) ForthecaseillustratedinFig.2.3.6theformulaissimplifiedto (2.8)

Fig.2.3.5

Fig.2.3.6

2.3.3 Loadswithalinearvariationinthereinforcementdirection
Theloadisassumedtovaryfromzeroatthesupporttoq0perunitareaadistancecfromthesupport. Inarectangularonewayelement,Fig.2.3.7,thesumofaveragemomentsis (2.9)

22

Fig.2.3.7 Inatriangularonewayelement,Fig.2.3.8,thesumofaveragemomentsis

(2.10) Inatrapezoidalonewayelement,Fig.2.3.9,thesumofaveragemomentsis

(2.11)

Fig2.3.8

Fig.2.3.9

2.3.4 Loadswithalinearvariationatrightanglestothereinforcementdirection
Theloadisassumedtovarybetweenzeroatthetopoftheslabsinthefigurestoq0atthebottomof theslab,i.e.withinadistancel. Inarectangularonewayelement,Fig.2.3.10,thesumofaveragemomentsis
23

(2.12)

Fig.2.3.10 Inatriangularonewayelement,Fig.2.3.11,thesumofaveragemomentsis

(2.13) Inatrapezoidalonewayelement,Fig.2.3.12,thesumofaveragemomentsis

(2.14)

Fig2.3.11

24

Fig,2.3.12

2.3.5 Elementswithashearforcealonganedge
Insomecasestheshearforceisnotzeroalongtheedgeofanelement.Atypicalcaseiswhereashear force has a linear intensity variation along an edge according to Fig. 2.3.13. The averagemoment is then (2.15)

Fig.2.3.13

2.3.6 Elementswithaskewanglebetweenspanreinforcementandsupport
Insomeslabsitisnaturaltohavedifferentdirectionsforsupportandspanreinforcement.Thisisthe case for triangular slabs and other slabs with nonorthogonal edges. The support reinforcement shouldnormallybearrangedatrightanglestothesupport,asthisisthemostefficientarrangement fortakingthesupportmomentandforlimitingcrackwidths.Spanreinforcementisoftenarrangedin twoorthogonallayers. Themostdirectwayoftreatingthecaseofdifferentdirectionsofsupportandspanreinforcementis throughtheintroductionofaline(orcurve)ofzeromoment.Ononesideofthislinethemomentis positive and on the other side it is negative. The positive moments are taken by the span reinforcement and the negative moments by the support reinforcement. The load is assumed to be carriedinthedirectionsofthereinforcement,thatisindifferentdirectionsoneachsideofthelineof zeromoment.Wecanmakeadistinctionbetweenspanstripsandsupportstrips. Alongthelineofzeromomentshearforcesareacting.Theseshearforcesoriginatefromtheloadon thespanstrips.Thelinesofzeromomentactasfreesupportsforthespanstrips.Thesupportstrips actascantilevers,carryingtheloadonthestripsandtheshearforcesfromthespanstrips. The shear force in a strip is normally expressed as a force Q per unit width at right angles to the reinforcementdirection.Whereaspanstripissupportedonasupportstripatalineofzeromoment thewidthsofthecooperatingstripsarenotthesame.UsingnotationaccordingtoFig.2.3.14,weget thefollowingrelationbetweentheshearforcesperunitwidth:

(2.16)

25

 Fig.2.3.14 Inthiswayitispossibletocalculatethepositiveandnegativedesignmomentsandtheirdistributions bymeansofthesimplestripmethod.ExamplesofsuchcalculationsaregiveninChapter6. Thisapproachisonlysuitablewherethespanstripsarecarryingalltheloadinonedirection.Inmany cases where the directions of the support and span reinforcements are different the span reinforcement in two directions cooperate in carrying the load on an element. Then the following generalapproachcanbeused,wheretheequilibriumofthewholeelementisconsidered,takinginto accountthemomenttakenbythespanreinforcementinbothdirections. ThelefthandpartinFig.2.3.15showsanelementwherethespanreinforcementisarrangedparallel to the x and yaxes, whereas the support reinforcement is arranged at right angles to the support, whichformsananglewiththexaxis.Thereinforcementinthexandydirectionscorrespondsto averagemomentsmxfandmyfandthesupportreinforcementcorrespondstoanaveragemomentms. Thetotalmomentsactingontheelementaregiveninthefigure.

Fig.2.3.15 The righthand part of Fig. 2.3.15 shows a corresponding element which is rectified so that the support is at right angles to the xaxis. The distances in the xdirection are the same for the two elements. The areas of the elements are the same. Each small area in the lefthand element has a corresponding area of the same size in the righthand element. The distance at right angles to the support for such an area in the lefthand element is sin times the distance at right angles to the supportintherighthandelement.Assumingthesameloadperunitareaatthecorrespondingpoints inthe twoelements,themomentwithrespecttothesupportfor thelefthandelementisthussin timesthemomentintherighthandelement.Ifwedenotetheaveragemomentcausedbytheloadin the righthand element m0, the total moment caused by the load in the lefthand element is thus m0lsin. Wecannowwritetheequilibriumequationforthelefthandelement,whichistheactualelementwe areinterestedin: (2.17)
26

whichcanberearrangedinto (2.18) Inthisequationm0istheaveragemomentfortherighthandelementinFig.2.3.15.Fortheelements treated in equations (2.3)(2.15) it corresponds to the righthand side of these equations. It can be inEq.(2.18). seenthattheseequationscorrespondto Ifthespanreinforcementisnotorthogonal,butwithonereinforcementdirectionparallelwiththex axis and the other reinforcement direction parallel to the support, the second term in Eq. (2.18) vanishes. This approach is not a use of oneway elements, as the load is carried in more than one direction. Unlike in the use of oneway elements it is in this case not possible to determine how the load is carried within the element. It is also not possible to determine a theoretical lateral distribution of designmoments.Anestimateofasuitabledistributionofdesignmomentscanhoweverbebasedon thedistributionsforonewayelementsofthesamegeneralshape.

2.4 Designmomentsinonewayelements
2.4.1 Generalconsiderations
Thestripmethodgivesinprincipleaninfinitenumberofpossiblepermissiblemomentdistributions. Forpracticaldesignasolutionshouldbechosenwhichsuitsourdemands.Themaindemandsare: 1. Suitablebehaviorunderserviceconditions. 2. Goodreinforcementeconomy,includingsimplicityindesignandconstruction. Indiscussingmomentdistributionstherearetwodifferenttypesofdistributiontotakeintoaccount, viz.distributionbetweensupportandspanmoments(distributioninthereinforcementdirection)and lateral distribution (distribution at right angles to the reinforcement direction). The ratio between supportandspanmomentsisdiscussedinSection2.9.1.

2.4.2 Lateraldistributionofdesignmoments
In the application of the simple strip method average moments in oneway elements are first calculated.Inarigorousanalysisusingthestripmethodthemomentisnotnormallyconstantacross thesection,butvariesduetothevaryinglengthsofthethinonewaystrips,andsometimesalsodueto varyingloadintensities.Theformalmomentvariationacrossthesectionisshownfordifferentcases inFigs2.3.113. Forarectangularelementwithauniformloadthemomentisconstantacrossthewidth.Inthiscase theaveragemomentcanbedirectlyusedfordesign.Inallothercasesthestrictmomentdistribution is not uniform, but decreases towards one or both sides. From a practical point of view it is not possible to follow these theoretical distributions in detail, and it is also not necessary, as the behaviouroftheslabisnotsensitivetolimitedvariationsinthelateralreinforcementdistribution.On the other hand the choice of an evenly distributed reinforcement corresponding to the average momentmayinmanycasesbetooroughanapproximation. Areinforcingbarisusuallymoreactiveandthereforemorebeneficialforthebehaviouroftheslabifit issituatedwherethecurvatureoftheslabinthedirectionofthebarishigh.Barswhichareparallel andclosetoasupportarenotveryactive,asthereispracticallynocurvatureintheirdirection.This fact should be taken into account in the distribution of design moments. It may even be rational to leavepartswithasmallcurvaturetotallywithoutreinforcementandassumeazerodesignmomenton acertainwidthoftheelement.Asthiswouldprobablynotbeacceptedbysomecodesthispossibility hasnotbeenappliedinthemajorityoftheexamples.

27

Where the theoretical strict moment distribution is uneven it is generally recommended that one design moment value is chosen for the part where the greatest theoretical moments occur and a smallerdesignmomentischosenoutsidethispart. Whereanaveragemomentmav,actingacrossawidthl,determinesthedesignmomentsmd1onwidth l1andthedesignmomentmd2onwidth(ll1),Fig.2.4.1,thefollowingrelationisvalid: (2.19) Withchosenvaluesoftheratiosmd2/md1andl1/l,thedesignmomentmd1canbecalculatedfromthe followingformula:

(2.20) The ratio md2/md1 is often chosen as 1/2 or 1/3 in order to achieve a simple reinforcement arrangement.Inthenumericalexamplesthevalue1/2isoftenused. Thesuitablechoiceofl1/ldependsontheshapeoftheelementandtheloaddistribution.Proposalsfor thischoicearegivenintheexamples. Whereanotherdistributionischosen,forexample,withthreedifferentvaluesofdesignmoments,the sameprinciplemayofcoursebeused.

Fig.2.4.1

2.5 Designmomentsincornersupportedelements
2.5.1 Cornersupportedelements
Acornersupportedelementisanelementwhichisonlysupportedatonecorner.Alongalltheedges the shear forces and the torsional moments (referred to the reinforcement directions) are zero. In figurestheedgesareshownasdashdotlines,indicatingzeroshearforces. The load on a cornersupported element has to be carried in two (or more) directions into the supportedcorner.Ithastohaveactivereinforcementintwo(ormore)directions.Itishereassumed thatthereareonlytworeinforcementdirections,whichareusuallyatrightanglestoeachother.The full load on the element has to be used for the calculation of moments in both directions. This is illustratedbycrossingdoubleheadedarrowsinthereinforcementdirections. Eachreinforcementdirectioncoincideswiththedirectionofoneoftheedgesoftheelement. Fig.2.5.1showsexamplesofcornersupportedelementswithasupportatthelowerleftcorner. Torsionalmomentsexistwithinacornersupportedelement,asitisnotpossibletocarryaloadtoone point without such moments. Both reinforcement directions cooperate in carrying the torsional moments. A certain amount of reinforcement is required for this purpose in addition to the reinforcementforcarryingthebendingmoments.Thetotalamountofreinforcementrequiredistaken into account in the rules given below, which express the required reinforcement as design bending moments.

28

Withtherules andlimitationsgivenbelow the maximumdesignmomentsoccurattheedgesofthe elements, which means that only the edge bending moments need to be calculated as a basis for reinforcementdesign.Withoutsuchrulesandlimitationshigherdesign

Fig.2.5.1 momentsmightoccurinsidetheelement,whichwouldcomplicatetheanalysesandbeuneconomical.

2.5.2 Rectangularelementswithuniformloads
Thedominatingtypeofcornersupportedelementinpracticaldesignistherectangularelementwith a uniform load. This case has therefore been investigated in more detail, leading to detailed design rules.TherulesandtheirbackgroundaregiveninStripMethodofDesign.Heretheruleswillbegiven inasimplifiedform,suitableforpracticaldesign.Therulesareonthesafeside,sometimesverymuch so.Minordeviationsfromtherulesforlateralmomentdistributionsmaybeaccepted. Fig.2.5.2illustratesarectangularcornersupportedelementwithauniformloadqperunitarea.The averagebendingmomentsalongtheedgeshaveindicesxandyforthecorrespondingreinforcement directions,sforsupportandfforspan(field).Theaveragemomentsmxs(supportmoment,negative sign)andmxf(spanmoment,positivesign)areactingontheelementwidthcy.Theequilibriumofthe elementwithrespecttotheyaxisdemandsthat

(2.21)

Fig.2.5.2 Thesemomentsareusuallydistributedontwostripswiththewidthscyand(1)cyrespectively.The stripwithwidthcy,closesttothesupport,iscalledthecolumnstrip,andthestripwithwidth(1)cy iscalledthemiddlestrip.Thesetermsarechoseninaccordancewiththenormaltermsusedforflat plates,whichareacommonapplicationofcornersupportedelements. Thedistributionofdesignmomentsbetweenthetwostripshastobesuchthatthenumericalsumof moments is higher in the column strip than in the middle strip in order to make the moment
29

distributionfulfiltheequilibriumconditionswithinthewholeelement.Thedistributionofmoments betweenthestripsisdefinedbythecoefficient: (2.22) Themomentdistributionhastofulfilthefollowingconditions (2.23) (2.24) Inmanycasesischosenas0.5,andthenwehave: (2.25) Forthesimplestpossiblearrangementofreinforcementallthesupportreinforcementisplacedwithin the column strip and the span reinforcement is evenly distributed. This moment distribution is illustrated in Fig. 2.5.3. As in this case there is no support moment in the middle strip the moment withinthisstripismxf.Thuswehave

(2.26) Combining(2.22)and(2.26)gives (2.27) or (2.28) Applyingthisto gives

(2.29) This simple moment distribution may be applied for all normally used ratios between support and spanmoments.Forsmallervaluesof,whichmeansmoreconcentratedsupportreinforcement,the . upperlimitofthemomentratioisreduced.Itisequalto2for Whetherthisreinforcementarrangementissuitabledependsonthedemandforcrackwidthcontrol. A reinforcement distribution more in accordance with Fig. 2.5.2 will presumably reduce maximum crackwidths,especiallyfortopcracksfarfromthesupport.

Fig.2.5.3

2.5.3 Nonrectangularelementswithuniformloadsandorthogonalreinforcement
30

Alsoinnonrectangularelementsitisappropriatetodividetheelementintotwostripsanddistribute the moments between them. It is not possible to give such detailed rules for these cases as for rectangular elements. It is, however, possible to give some general recommendations, which in practicewillleadtoasafedesign. Fig. 2.5.4 shows triangular cornersupported elements of three different arrangements of reinforcement in the xdirection. With the same definitions as for the rectangular element the followingrulesarerecommendedfortheseslabs: Casea: (2.30) Caseb: (2.31) Casec:

(2.32) Most cornersupported elements with orthogonal reinforcement are rectangular or triangular. In cases where other shapes occur they have a shape which is intermediate between rectangular and triangular, and suitable moment distributions may be estimated by means of the recommendations above.Itisimportanttorememberthateachreinforcementdirectionmustbeparalleltoanedgeof theelement.

Fig.2.5.4

2.5.4 Elementswithnonorthogonalreinforcement
Thedeterminationofthedesignmomentsinanelementwithnonorthogonalreinforcementisbased onasocalledaffinitylaw.Accordingtothislawthedesignmomentsarethesameasthemomentsina rectifiedelementwithorthogonalreinforcementandthesamelengthandwidthwithrespecttothe reinforcementinquestion. This rule is exemplified in Fig. 2.5.5, which shows a rhomboidal element and the corresponding rectangular elements for the calculation of the design moments in the x and ydirections. Thus the design moments and the resulting distributions for the reinforcement in the xdirection are determined from the slab in the x1y1system and for the reinforcement in the ydirection from the slabinthex2y2system.

2.5.5 Elementswithnonuniformloads
Thenumericalsumofdesignmomentsineachdirectioniscalculatedwiththesameformulasasfor oneway elements. In the distribution of design moments between column strip and middle strip regard nust be taken of the load distribution. If the column strip in one direction is more heavily loadedthanthemiddlestripahigherportionofthemomentshouldbetakenbythecolumnstripthan indicatedbytherulesforuniformloadandviceversa.Itisnotpossibletogiveexactrulestocoverall
31

cases,butastherulesforuniformloadarequitewide,itoughttobepossibletofindsafedistributions inmostcasesstartingfromtherulesforuniformloadandmodifyingthemwithrespecttotheactual loaddistribution. ThecaseofaconcentratedloadactingonacornersupportedelementistreatedinSection2.6.2. AnalternativetreatmentofcornersupportedelementsisdiscussedinSection2.8.

Fig.2.5.5

2.6 Concentratedloads
2.6.1 Onewayelements
Aconcentratedloadisaloadwhichhastoohighavalueperunitareatobetakendirectlybyaone way strip (or crossing oneway strips) without giving rise to too excessively high local moments. It maybeapointload,alineloadorahighloadonalimitedarea.Thegeneralwayoftakingcareofa concentrated load is by distributing it over a suitable width by means of specially designed distributionreinforcement Fig. 2.6.1 shows an example. A concentrated load F acts as a uniform load over a small rectangular areawithwidthb1inthexdirection.Itistobecarriedbyarectangular,simplysupportedslab.The loadiscarriedonastripintheydirectionwithawidthb,chosentohaveamomentinthestripwhich is not too high per unit width. Thus the load F has to be evenly distributed over a strip width b by means of a small strip in the xdirection. The bending moment in this strip is . This momentshouldbeevenlydistributedonasuitablewidtha,givingamoment (2.33)

32

 Fig.2.6.1 It is recommended that the width a should be chosen such that it is approximately centered on the concentratedloadandsmallenoughtofulfilthe(approximateandsomewhatarbitrary)requirements giveninthefigure.Thevalueofbischosensoastogetreinforcementinthestripwhichiswellbelow balancedreinforcementinordertoensureplasticbehaviour. Theloaddistributiononthewidthbismainlyofimportanceforthespanreinforcement.Ifthestripis continuous,thesupportmoment(reinforcement)maybedistributedoveralargerwidththanb,asthe supportinitselfactsasaloaddistributor. Inmanycasesitmaybesuitabletodividetheconcentratedloadovertwostrips,oneinthexdirection andoneintheydirection,inordertogetareinforcementdistributionwhichisbetterfromthepoint of view of performance under service conditions. In such a case it is recommended that the load should bedivided between the directions approximately in inverseproportion tothe ratio between thespanstothefourthpower,providedthatthesupportconditionsarethesameforbothstrips.Ifthe stripshavedifferentsupportconditionsthisshouldbetakenintoaccountsothattheloadtakenbya stripwithfixedendscanbeincreasedcomparedtotheloadtakenbyasimplysupportedstrip. Whenconcentratedloadsareactingonaslabtogetherwithdistributedloadsspecialloaddistribution reinforcementisnotnecessaryifthedistributedloadsaredominating.Thismaybeconsideredtobe thecaseifaconcentratedloadislessthan25%ofthesumofdistributedloads.Thentheconcentrated loadissimplyincludedintheequilibriumequationsfortheelements.Thedesignmomentsmayalso be redistributed so as to have more reinforcement near the concentrated load. If there are several concentratedloadsthissimplifiedproceduremaybeusedevenifthesumoftheconcentratedloadsis muchhigher,andparticularlyiftheconcentratedloadsarenotclosetoeachother. ApplicationsareshowninthenumericalexamplesinSection3.3.

2.6.2 Cornersupportedelements
Cornersupportedelementsalreadyhavereinforcementintwodirectionsandarethusoftenableto takecareofconcentratedloadswithoutanyspecialdistributionreinforcement. The point load F on the cornersupported element in Fig. 2.6.2 gives rise to bending moments . It is generally recommended that the numerical sum of span and support moments . should be distributed evenly over the width ly. Over this width the moment is then Somevariationsfromthisbasicruleareacceptable,e.g.adistributionoverasomewhatlargerwidth.

33

Fig.2.6.2 Eventhoughthenumericalsumofspanandsupportmomentshasauniformlateraldistributionitis recommended that the span moment should be concentrated in the vicinity of the load and the support moment in the vicinity of the supported corner. Such a distribution is in better agreement withthemomentdistributionunderserviceconditionsandwilllimitcracking. Whenlyisclosetozerothebendingmomentperunitwidthbecomestoolargeforthereinforcement in the xdirection according to the expression above. In such cases some extra distribution reinforcementmaybenecessary,designedaccordingtothegeneralprincipleforonewayelements. ForapplicationsseeSection9.4.

2.7 Strips
2.7.1 Combiningelementstoformstrips
Theelementsintowhichtheslabisdividedhavetobecombinedinsuchawaythattheequilibrium conditionsarefulfilled.Theseconditionsarerelatedtothebendingmomentsandtheshearforcesat theedgesoftheelements.Torsionalmomentsareneverpresentattheedgesoftheelements. Shear forces are in most cases assumed to be zero at those edges of the elements, which are not supported,butinsomeapplicationsnonzeroshearforcesmayappearatsuchanedge. Oneway elements with the same loadbearing direction are often not directly connected to each other, but by means of a constant moment transferredthrough elements with another loadbearing direction,cf.Fig.2.2.3.AtypicalexampleisshowninFig.2.7.1,ofarectangularslabwithauniform load, two simply supported edges and two fixed edges. The positive moments in elements 1 and 3 havetobethesame,andthismomentistransferredthrough2and4,whichinpracticemeansthatthe reinforcementisgoingbetween1and3through2and4.

Fig.2.7.1 The choice of cvalues is based on the rules in Section 2.2. It may in practice also be influenced by rulesforminimumreinforcementiftheoptimumreinforcementeconomyissought. In flat slabs cornersupported elements are often directly connected with each other and with rectangular oneway elements into continuous strips as in Fig. 2.7.2. Each such strip acts as a continuousbeam,andcanbetreatedassuch.TheslabinFig.2.7.2thushasacontinuousstripinthree spans with width wx in the xdirection and a continuous strip in two spans with width wy in the y direction.Thedesignoftheslabisbasedontheanalysisofthesetwostrips,takingintoaccountthe
34

rulesforreinforcementdistributionforcornersupportedelements.Theaveragedesignmomentsin the triangular elements near the corners of the slab are equal to onethird of the corresponding momentsintheadjoiningrectangularelements. MoredetailedrulesandnumericalexamplesaregiveninChapter8.

2.7.2 Continuousstripswithuniformloads
InacontinuousstripformedaccordingtoFig.2.7.2,forexample,eachpartofthestripbetweentwo supports,canbetreatedusingnormalformulasforbeams.Thecalculationofdesignmomentsusually startswithanestimateofsuitablesupportmoments.Atacontinuoussupportthemomentinastrip withauniformloadisnormallychosenwithrespecttowhatcanbeexpectedaccordingtothetheory ofelasticity,cf.section2.9.1.Afterthesupportmomentshavebeenchosen,thedistancec1inFig.2.2.2 canbecalculatedfromtheformula (2.34) andthespanmomentfromtheformula

(2.35)

2.8 Supportbands
2.8.1 General
Asupportbandisabandofreinforcementinonedirection,actingasasupportforstripsinanother direction.Bymeansofsupportbandsitispossibletomakedirectuseofthegeneralprinciplesofthe simplestripmethodforalltypesofslab.Itisthemostgeneralmethod,andthemethodwhichhasto beusedincaseswhereotherapproachescannotbeapplied. A reinforcement band of course has a certain width. In Strip Method of Design it has been demonstratedhowreinforcementbandsmaybeusedinastrictlycorrectway,takingintoaccountthe widths of the bands. In order to simplify the analyses the following approximation will be accepted hereforthecalculationofthedesignmomentinthesupportband: Inthenumericalanalysisthesupportbandisassumedtohavezerowidth. Thereinforcementforthemomentinthesupportbandisdistributedoveracertainwidth,whichis limitedbyrulesgiveninSection2.8.3.Iftheserulesarefollowedthesafetyattheultimatelimitstate canbeestimatedtobesufficientinspiteoftheapproximation.
35

2.8.2 Comparisonwithcornersupportedelements
A rectangular cornersupported element with a uniform load can be analysed by means of simple stripssupportedonsupportbandsalongthecoordinateaxes,whichareintheirturnsupportedatthe corner. Fig. 2.8.1 a) shows an element where half the load is assumed to be carried in each direction. This givesevenlydistributedmoments ,andareactionforceonthesupportbandalongthe xaxis equal to qcy/2. This reaction force on the support band gives concentrated moments ,correspondingtoanaveragemoment onthewholewidthcy.Thushalfthetotal moment isevenlydistributedandtheotherhalfisconcentratedintheassumedsupportbandof zerowidth.

Fig.2.8.1 ThemomentscalculatedfromanassumptionofsupportbandsinFig.2.8.1a)canbecomparedtothe moment distribution according to Fig. 2.8.1 b), which is acceptable according to condition (2.25). Fromthiscomparisonitcanbeconcludedthatinthiscaseitisacceptabletodistributeconcentrated momentsfromanassumedsupportbandofzerowidthoverhalfthewidthoftheelementandthatthe solution is still on the safe side. This conclusion is drawn for a rectangular element with a uniform load, but indirectly it can also be concluded that moments from an assumed support band of zero widthcanalwaysbedistributedoveracertainwidthoftheelement.Themostsuitablechoiceofthis widthdependsontheloaddistributionandtheshapeoftheelement.Ifmoreoftheloadisactingnear thesupportbandasmallerwidthshouldbechosen.Withsomecautionitisnotdifficulttochoosea widthwhichissafe.Detailedrecommendationsaregivenbelow. Itmayalsobenotedthataccordingtoconditions(2.23)and(2.24)themomentfromthesupportband maybeevenlydistributedoveranyarbitrarywidthbetweenzeroandly/2.

2.8.3 Applicationrules
Themomentsinthesupportbandaredistributedoveracertainwidthtogivethedesignmomentsfor thereinforcement.Thewidthofthereinforcementbandhastobelimitedsoitcanaccommodatethe momentswhichareconcentratedinabandofzerowidth.Itisnotpossibletoestablishgeneralrules forthemaximumacceptablewidthofthereinforcementbandbasedonarigoroussolutionaccording tothelowerboundtheoremofthetheoryofplasticity.Thefollowingrecommendationsarebasedon the comparison above with a cornersupported element and on estimates. They are intended for situationswheretheloadontheslabisuniform.
36

Thewidthofthereinforcementbandisbasedonacomparisonwiththeaveragewidthoftheelements whicharesupportedbythesupportband.Thisaveragewidthisdenotedba.Thewidthofthesupport reinforcement in the band may be equal to ba provided that the band is supported over its whole widthbyasupportwhichisnearlyatrightanglestotheband.Supportreinforcementovercolumns maybedistributedoverawidthofabout0.5ba.Spanreinforcementmaybedistributedoverawidth between0.5baandba,dependingontheimportanceofthebandforthestaticbehaviouroftheslab. The more important the band is, the narrower the width of the reinforcement band. The reinforcement should if possible be distributed on both sides of the theoretical support band in proportiontotheloadsfromthetwosides. Where a support band has a support with a strong force concentration certain rules have to be followedinordertopreventlocalfailure. For the case in Fig. 2.8.2 some minimum top reinforcement is required at the ends of the band to preventradialcracksfromthesupportpoint.Thisreinforcementisplacedatrightanglestotheband andisdesignedforabendingmoment ,determinedfromthefollowingcondition:

(2.36) whereRisthesupportreactionfromthesupportband,and mbisthenumericalsumofthespanandsupportmomentswhicharetakenbythereinforcementin thereinforcingbandatthesupportwhereRisacting. Thecorrespondingreinforcementshouldbepresentfromthesupporttoadistanceequaltoonethird ofthedistancetothepointofmaximummomentinthesupportbandandbewellanchored.Itdoes notneedtobeadditionalreinforcement,asthereinforcementarrangedforotherreasonsmaycover thisneed. Acheckaccordingtotheserulesmustbemadeforbothendsofthereinforcementband. Where the band is supported on a column, a check should be made that the following relation is fulfilled

Fig.2.8.2

(2.37) wherethespanandsupportmomentscorrespondtothereinforcementinthevicinityofthecolumn.If thespanmomentsaredifferentonthedifferentsidesofthecolumntheaveragevalueisused.Risthe reactionforceatthecolumn,Aisthesupportareaatthecolumnandqistheloadperunitareainthe vicinityofthecolumn. It is not normally necessary to perform this check, as the recommendations for the design are intendedtofulfilthisrelationautomatically. Inadditiontotheconcentratedreinforcementinabandsomeextrareinforcementmaybeneededfor crackcontrol. Although it cannot be strictly proven by means of the lower bound theorem that the above recommendationsarealwaysonthesafesidethedesigncanbeconsideredtobesafe.Inanycaseitis
37

alwayssaferthanadesignbasedontheyieldlinetheory:thiscanbecheckedbyapplyingtheyield linetheorytoslabsdesignedaccordingtotheserecommendations.

2.9 Ratiosbetweenmoments
2.9.1 Ratiobetweensupportandspanmomentsinthesamedirection
TherecommendationsbelowfollowthegeneralguidelinesinSection1.5. Thestripmethoddoesnotinitselfgiveratiosbetweensupportandspanmoments,astheequilibrium equationscanbefulfilledindependentlyofthisratio.Fromthepointofviewofsafetyattheultimate limitstatethisratioisunimportant.Theratiois,however,ofimportanceforbehaviourunderservice conditions and for reinforcement economy. These factors should therefore be taken into account in thechoiceoftheratiosbetweensupportandspanmoments. Itcan beshownthatthebestchoiceoftheratiobetweenthenumericalvalues ofsupportandspan momentsinanelementorstripisusuallyapproximatelyequaltotheratioaccordingtothetheoryof elasticityorsomewhathigher.Theratiomay,however,bechosenwithinratherwidelimitswithout anyeffectonsafetyandwithonlyaverylimitedeffectondeformationsunderserviceconditions.It mainlyhasaninfluenceonthewidthofcracks.Withahigherratiothecrackwidthsabovethesupport are somewhat decreased whereas the crack widths in the span are somewhat increased. If the intention is to limit the crack width on the upper side of the slab a large ratio between span and supportmomentsshouldthereforebechosen. Wheretheloadisuniformitisgenerallyrecommendedthat,forcontinuousstripsandforstripswith fixedsupports,aratiobetweenthenumericalvaluesofsupportandspanmomentsshouldbechosen around2,saybetween1.5and3.0.Fortriangularelements,suchasattheshortedgeofarectangular slab,highervaluesmaybeused.Whereastripiscontinuousoverasupporttheaverageofthespan momentsonbothsidesofthesupportisusedforthecalculationofmomentratio. SoforaslabsuchasthatinFig.2.7.1theratioc2/c4ischosenbetween1.6and2.0,correspondingtoa moment ratio of between 1.56 and 3.0, whereas c1/c3 may be chosen between 1.7 and 2.2, correspondingtoamomentratiobetween1.9and3.8. Whereitispossibletoestimatethemomentsaccordingtothetheoryofelasticitythisshouldbedone as a basis for the determination of the main moments. This is the case, for instance, where the advanced strip method is used for flat slabs. The support moment at a support where the slab is continuouscanbetakentobeapproximatelyequaltotheaverageofthemomentscorrespondingto fixededgesforthespansonthetwosidesofthesupport. For the irregular flat slabs discussed in Chapter 10 a special approach for the determination of support moments has been introduced in order to find support moments which are also in approximateagreementwiththetheoryofelasticityinthesecomplexcases.

2.9.2 Momentsindifferentdirections
In some applications there is no real choice of moment distribution in different directions. For example,thisisthecasefortheflatslabinFig.2.7.2,wherethedistributionisgivenbytheanalysisof thestrips.Thisisduetothefactthatwithinthemajorpartoftheslabtheloadiscarriedbycorner supported elements, which carry the whole load in both directions, and where the load is thus not distributedbetweenthedirections. Inslabswheretheloadiscarriedbyonewayelements,e.g.theslabinFig.2.7.1,itispossibletomake a choice of the direction of loadbearing reinforcement within those parts of the slab where the
38

elements with different directions meet. Thus, for the slab in Fig. 2.7.1, it is possible to increase or decreasec1andc3,leadingtoincreasedordecreasedmomentsinelements1and3andcorresponding decreasesorincreasesofthemomentsintheoppositedirection.Thechoiceofdirectionsofthelines ofzeroshearforcestartingatthecornershasbeendiscussedinSection2.2.

2.10 Lengthandanchorageofreinforcingbars
2.10.1 Onewayelements
Inprinciple,thelengthofreinforcingbarsisdeterminedfromthecurveofbendingmomentswhich showsthevariationofbendingmomentsinthedirectionofthestrip.Thiscurveiseasytodetermine forrectangular elementswithauniformload,butinmostothercasesthecurveisnotwelldefined. Thenthe necessary length ofreinforcing bars has to be determined by means of someapproximate rule,whichshouldbeonthesafeside. Foronewayelementswithauniformloadthefollowingrulesarerecommended: 1.Onehalfofthebottomreinforcementistakentothesupport.Closetoacornerofaslab,however,all thebottomreinforcementistakentothesupport. 2.Theotherhalfofthebottomreinforcementistakentoadistancefromthesupportequalto

(2.38) 3.Onehalfofthetopreinforcementistakentoadistancefromthesupportequalto

(2.39) 4.Therestofthetopreinforcementistakentoadistancefromthesupportequalto (2.40) Intheseformulascisthelengthoftheelement,showninFigs2.2.25,andlisanadditionallengthof the reinforcing bars for anchorage behind the point where the moment curve shows that it is not neededanymore.Thisadditionallength,whichdependsontheslabdepthandonthediameterofthe reinforcement,isgiveninmanynationalcodes.Ifthisisnotthecase,itisrecommendedthatavalueis usedequaltothedepthoftheslab. Itisnotpossibletogiveexactrulescoveringallpossiblesituationswithrespecttoloaddistributions. Incasesotherthanthosetreatedaboveitcanonlyberecommendedthatreinforcementlengthsare chosenwhicharejudgedtobeonthesafeside. Asanexample,theapplicationoftheaboveformulastothereinforcementinthexdirectionwillbe shown for the slab in Example 3.2. The value of l is assumed to 0.15 m. The moment values are ,whichgives . .Inthiselementhalfthebottomreinforcementistakento Forelement1inFig.3.1.3wehave thesupportandtheotherhalf,accordingtoEq.(2.38),toadistancefromthesupportequalto (2.41) One half of the top reinforcement is, according to Eq. (2.39), taken to a distance from the support equalto
39

 (2.42) The rest of the top reinforcement is, according to Eq. (2.40), taken to a distance from the support equalto (2.43) ,thesupportmomentiszeroandthereisnosupportreinforcement.Half Forelement3,with the bottom reinforcement is taken to the support and the other half to a distance from the support equalto (2.44) Thisisasmalldistanceandfromaneconomicpointofviewallthereinforcementmayaswellbetaken tothesupport.Itisgenerallyrecommendedthatallbottomreinforcementistakentothesupportina slabwhichissimplysupported. Thelengthofreinforcingbarswillnotbecalculatedinthenumericalexamples,asthecalculationis simpleandsometimesotherruleshavetobefollowedaccordingtocodes.

2.10.2 Cornersupportedelements
Thedistributionofdesignmomentswithinacornersupportedelementisextremelycomplex,dueto thefactthattorsionalmomentsplayanimportantpartincarryingtheloadtoacorneroftheelement. These torsional moments have to be taken into account in the design of reinforcement, which is formally done as an addition to the design bending moments which the section have to resist. It is hardly possible to find solutions which cover all situations in a theoretically correct way. The rules givenbelowarebasicallyonthesafesideatthesametimeasgivingapproximatelythesameresultsas commondesignmethodsforflatslabs. Inallcornersupportedelementsallbottomreinforcementshouldbetakenallthewaytotheedgesof theelement. Inrectangularcornersupportedelementswithauniformloadhalfthetopreinforcementshouldbe takentoadistancefromthesupportlineequalto (2.45) Therestofthetopreinforcementshouldbetakentoadistancefromthesupportlineequalto (2.46) Intheseformulaslhasthesamemeaningasin2.10.1,dependsontheratiobetweensupportand span moments according to Fig. 2.10.1 and c is the length of the element in the direction of the reinforcement.

2.10.3 Anchorageatfreeedges
Where a reinforcing bar ends at a free edge it must be satisfactorily anchored. In many cases a deformedbarmayjustendattheedge,butwheretheforceinthebarisexpectedtobehighcloseto theedge,thebarshouldbeprovidedwithanextrastronganchorage,e.g.accordingtoFig.2.9.2.This willoccurwherethereisashortdistancefromalineofmaximummoment(lineofzeroshearforce)to theedgeorwherelargetorsionalmomentsareactingintheslab,whichisthecaseinthevicinityof point supports or where point loads are acting close to a free edge, for example. Large torsional momentsalsoactinslabswithtwoadjacentfreeedges.

40

Fig.2.10.1

Fig.2.10.2

2.11 Support reactions

Inonewayelementstheloadiscarriedintheloadbearingdirectionintothesupport.Itisthussimple to determine the support reaction and its distribution. With a uniform load on an element the distributioncorrespondstotheshapeoftheelement.Theaccuratedistributionofthereactionshould betakenintoaccountinthedesignofsupportingstructureslikebeamsandsupportbands. Wherethesupportisnotparalleltooneofthetwospanreinforcementdirections,asinFig.2.3.15,itis not possible to use oneway elements and therefore not possible to determine an accurate distributionofsupportreactions.Insuchacasetheloadisassumedtobe carriedatrightanglestothe
support(theshortestdistancetothesupport).Thisassumptiongivesthecorrecttotalsupportreactionandthe resultingmomentsinasupportingbeamorbandwillinallpracticalcasesbeonthesafeside.Themaximum shearforceinasupportingbeamorbandmaybeslightlyontheunsafeside.

For cornersupported elements the whole load on the element is theoretically acting as a support reactionatthesupportedcorner.Inrealityitisofcoursedistributedinsomewayovertheareaofthe support.Theresultanttothesupportforcesisdeterminedbythetheoreticalforcesfromtheelements actinguponthesupport.
41

CHAPTER3 Rectangularslabswithallsidessupported
3.1 Uniformloads
3.1.1 Simplysupportedslabs
Example3.1 TheslabinFig.3.1.1issimplysupportedwithauniformloadof9kN/m2.Thelinesofzeroshearforce arechosenasshowninthefigureinaccordancewiththerulesgiveninSection2.2.Theaveragespan momentinthexdirectionis,Eq.(2.4): (3.1) Theaveragespanmomentintheydirectionis,Eq.(2.5): (3.2) The lateral distribution of design moments follows the general recommendations in Section 2.4.2. Thus, for example, we may choose to have moments in the side strips which are half those in the centralstrip(andwithdistancesbetweenbarstwiceaslarge).Ifwechoose

42

Fig.3.1.1 thewidthsofthesidestripsto1.1minbothdirections,wegetinthecentralstrips,Eq(2.20):

(3.3)

(3.4) ThismomentdistributionisillustratedinFig.3.1.2.

3.1.2 Fixedandsimplesupports
Example3.2 TheslabinFig.3.1.3hasthesamesizeandload(9kN/m2)asthepreviousslab,buttheupperandleft hand edges are fixed and there are negative moments along these edges. The numerical sum of momentsintheseelementsisthusgreaterthanintheelementsatthesim

Fig.3.1.2 ply supported edges. This is achieved by choosing suitable sizes of the elements, see Section 2.9.1. WiththechoiceofsizeasshowninFig3.1.3forelements1and3wegetfromEq.(2.4) (3.5) (3.6) (3.7)
43

Forelements4and2wegetfromEq.(2.5) (3.8) (3.9) (3.10)

Fig.3.1.3 Thenumericalratiosbetweenthesupportandspanmomentsare2.45inthexdirectionand2.07in theydirection.Thesevaluesareacceptable(seeSection2.9.1),butif,forexample,wewishtohavea slightlysmallerratiofortheydirectionwecanchangethedividingline(e.g.withcvaluesof2.7and and andtheratio 1.7insteadof2.8and1.6)andrepeattheanalysis.Wethenfind 1.49.Thisratioisevidentlysensitivetothechoiceofthepositionofthelineofzeroshearforce. Ifwechoosethesametypeofmomentdistributionasinthepreviousexamplewefindthedistribution ofdesignmomentsaccordingtoFig.3.1.4. Thedeterminationofthelengthsofreinforcingbarsfortheslabinthisexampleisdemonstratedin Section2.10.1. Example3.3 TheslabinFig.3.1.5hasaloadof11kN/m2.Thelefthandedgeisfixed,whereastherighthandedge iselasticallyrestrainedsothatthesupportmomentislower.Thisistakenintoaccountintheanalysis byassuminganonsymmetricalpatternoflinesofzeroshearforce,asshowninthefigure.

Fig.3.1.4

44

Fig.3.1.5 UsingEq.(2.1)wegetforelement1 (3.11) andforelement3

(3.12) ,whichgives Wecan,forinstancechoose FromEq.(2.4)wegetforelements2and4

(3.13) IncalculatingdesignmomentsaccordingtoEq.(2.20)thewidthsofthesidestripsarechosenequalto 1.0minthexdirectionand1.5mintheydirection.Thereasonwhyahighervaluehasbeenchosen fortheydirectionisthatreinforcementalongafixededgeisratherinefficientintakingupstresses under normal loads, as the slab has very limited deflections in these regions. The resulting distributionofdesignmomentsisshowninFig.3.1.6.

Fig.3.1.6

3.2 Triangular loads

Example3.4 The slab in Fig. 3.2.1 is simply supported along all edges. It is a vertical slab acted upon by water pressure,whichiszeroattheupperedgeandisassumedtoincreaseby10kN/m2permdepth.(This valueisanapproximationusedtomakeiteasiertofollowtheanalysis.Amorecorrectvalueisabout
45

10.2.) For reasons of symmetry . The value of conditionthatthespanmomentsinelements2and4mustbethesame.

is determined by the

Fig.3.2.1 For the different elements we have the following expressions for the moments, calculated from Eq. (2.13)forelements1and3,from(2.11)forelement2andfromacombinationof(2.5)and(2.11)for element4(evenlydistributedloadof34minusatriangularloadwiththevalue10c4attheupperline ofzeroshearforce):

(3.14)

(3.15)

(3.16) leadstoanequationofthethirddegree.Itisinpracticeeasier Todirectlyusethecondition it is found that to use iteration, trying different values until the condition is fulfilled. Using ,andwegetthefollowingresultforaveragemoments: Theratiobetweenthemomentsinthexandydirectionsshouldalsobecheckedinordertoachievea good reinforcement economy and satisfactory behaviour under service conditions. In this case this ratiomayseemsuitablewithrespecttothespansinthexandydirections.Theratiomaybechanged bymeansofanotherchoiceofthevalueofl1.However,neithereconomynorserviceabilityaremuch influencedbysmallchangesofthistype. Withthevaluesabove,thedistributionofdesignmomentscanbechoseninaccordancewithFig3.2.2. The distribution of reinforcement in the xdirection is chosen with less reinforcement in the upper panoftheslab,wheretheloadissmaller.

Fig.3.2.2 Example3.5 Fig.3.2.3showsaslabwithalledgesfixed,whichisacteduponbyaloadwhichis18kN/m2atthetop andincreasesby5kN/m2permetreto39kN/m2atthebottom.Achoiceoflinesofzeroshearforceis shown.Forallpartsoftheslabacombinationofuniformloadandtriangularloadhastobeused.

46

 Fig.3.2.3 Forelement1(identicaltoelement3)Eq.(2.4)isusedwith (3.17) Forelement2Eq.(2.5)isusedwith (3.18) Forelement4,Eq.(2.5)isusedwith

andEq.(2.13)isusedwith

andEq.(2.11)isusedwith andEq.(2.11)isusedwith

(3.19) We can now choose values of mxf and myf, which give suitable ratios between support and span moments.Thefollowingvaluesmaybechosen,forexample: Withthesevaluesandsidestripsinbothdirectionsequalto1.0mwegetthedistributionofdesign momentsshowninFig.3.2.4.Iftheratiosbetweenthedifferentmomentsarenotregardedassuitable, either the choice of span moments can be changed, or the pattern of lines of zero moments can be adjustedandthecalculationrepeated.

Fig.3.2.4 Example3.6 ThewallinFig.3.2.5isacteduponbyawaterpressurewithawaterlevel1.0mbelowtheupperedge. Thepressureistriangularwithanassumedincreaseof10kN/m2permetredepth.Theloadingcases on the elements do not directly correspond with those treated in Section 2.3. It is necessary to
47

combineseveralofthesecasesinordertocalculatetheaveragemoments.Thecombinationofloading casesforthecalculationofmomentsinelements1,2and3isindicatedinFig.3.2.5. Inelement1(and3)thisgivesthefollowingevaluation:

Fig.3.2.5

(3.20) ThefirsttermistheinfluenceofatriangularloadonthewholeelementaccordingtoEq.(2.13)and the second term is the influence of a negative uniform load on the whole element according to Eq. (2.4).Thethirdandfourthtermsshowtheinfluenceofloadsononlytheupperpartoftheelement, down to a depth of 1.0 m, where the height is 0.69 m. The average moment on the upper part is distributedonthewholeelementwidth4.2mthroughmultiplicationby1.0/4.2. Forelement2weget

(3.21) ThefirsttermistheinfluenceofatriangularloadonthewholeelementaccordingtoEq.(2.11)and the second term is the influence of a negative uniform load on the whole element according to Eq. (2.5).Thethirdandfourthtermsgivetheinfluenceofauniformloadandanegativetriangularloadon onlytheupper1.0moftheelement.Thewidthoftheelementatthatlevelis3.62m. Forelement4weget (3.22) The first term is the influence of a uniform load according to Eq. (2.5) and the second term is the influenceofanegativetriangularloadaccordingtoEq.(2.11). The following moments are chosen in order to achieve suitable ratios between support and span moments: A suitable distribution of design moments is proposed in Fig. 3.2.6. The reinforcement in the x direction is not symmetrical, but is concentrated downwards because of the very nonsymmetrical loaddistribution.

48

Fig.3.2.6

3.3 Concentratedloads
3.3.1 General
Aconcentratedloadisaloadwhichactsonlyonarelativelysmallpartofaslab.Itmaybeapointload, alineloadoraloadwithahighintensityonasmallarea. A concentrated load generally acts together with a distributed load. The way that the concentrated loadistakenintoaccountindesigndependsontherelativeimportanceofthisloadcomparedtothe distributedload.Iftheconcentratedloadisimportantitmustbeseparatelytakenintoaccountinthe design.Ifthedistributed loadisdominanttheinfluenceoftheconcentrated loadcanbeincludedat thesametimeasthedistributedload.Itissuggestedthataseparatecalculationisnotneededincases wheretheconcentratedloadislessthan about25%ofthesumofdistributedloads. Iftwoormore concentratedloadsarespreadovertheslabthisfiguremaybeincreased.

3.3.2 Aconcentratedloadalone
Example3.7 TheslabinFig.3.3.1isacteduponbyapointloadof150kN.Itshouldbedesignedtocarrythewhole loadinthedirectionoftheshortestspan,i.e.intheydirection.Thisistheleastexpensivewaytocarry thisload,butifonlytheresultingreinforcementisused,theslabwillnotbehavewellintheservice state. This example is only shown in order to demonstrate the principle used in design for concentratedloads.Thenextexampleshowshowaconcentratedloadcanbedividedbetweenstrips intwodirectionsinordertoachieveareinforcementwhichisbetterforthebehaviourintheservice state. Asshowninthefiguretheloadisassumedtobecarriedbyastripintheydirectionwithawidthof 2.0m.Theloadisdistributedoverthiswidthbymeansofreinforcementinthexdirectioninashort stripwithwidth1.5m.ThedesignmomentinthisstripiscalculatedbymeansofEq.(2.33): (3.23) Thetwoelements2and4meetatthepointload.Strictly,onepartF4ofthepointloadiscarriedby element2andtheremainingpartbyelement4.Wegetthefollowingrelations: (3.24)

49

Fig.3.3.1

(3.25) A suitable choice is , which gives ; . The resulting distribution of design momentsisshowninFig.3.3.2. The distribution reinforcement in the xdirection is mainly intended for the distribution of the bending moment over the width of the span reinforcement in the ydirection. The support reinforcementcanbeactiveoveralargerwidth,asthesupportactsasloaddistributorandforcesthe reinforcement to cooperate. The support moment can therefore be distributed over a larger width thanthewidthof2.0massumedforthedesign.ThishasbeenindicatedbyadashedlineinFig.3.3.2.

Fig3.3.2 Example3.8 This example is intended to illustrate both how a concentrated load is distributed on strips in two directionsandhowtotreataconcentratedloadintheformofahighloadintensityonasmallarea.A totalloadof150kNisuniformlydistributedonanareaof .Theloadisdistributedonone stripinthexdirectionwithwidth2.0m,andonestripintheydirectionwithwidth2.5m. Whentheloadisdistributedinthetwodirectionsitissatisfactorytodividetheloadapproximately inverselytotheratioofspanstothefourthpower.Inthiscasewewillhave40kNinthexdirection and110kNintheydirection. Foreachdirectionwecanassumealineofzeroshearforce(maximummoment)atrightanglestothe loadbearing direction, placed inside the loaded area. The coordinates of these lines are denoted x1 andy1.
50

ForthestripinthexdirectionwegetamomentmyfforloaddistributionfromEq.(2.33): (3.26)

Fig.3.3.3 Forthemomentsinthetwoelements1and3formingthestripwehave: (3.27)

(3.28) Different values of x1 are tried in these formulas until a suitable ratio between support and span momentsisfound.Thusfor wefind . Forthestripintheydirectionwegetamomentmxfforloaddistribution: (3.29) Forthemomentsinthetwoelements2and4formingthestripwehave (3.30)

(3.31) wefind ,whichhaveasuitableratio. For Themomentsfor loaddistributionhavetheir maxima atthecentreoftheloadedarea, whereasthe span moments in the strips have their maxima at x1 and y1. Adding together the moments for load distribution and the span moments in the strips will therefore give values slightly on the safe side. Performingthisadditionwegetthefollowingdesignmoments: This moment distribution is shown in Fig. 3.3.4. As pointed out in the previous example it is acceptabletodistributethesupportmomentsoverlargerwidths.

51

Fig.3.3.4

3.3.3 Distributedandconcentratedloadstogether
Example3.9 ThesimplysupportedslabinFig.3.3.5hasauniformloadof7kN/m2andapointloadof40kNinthe positionshown.Thetotaluniformloadis .Thepointloadisonly20%oftheuniform load,whichmeansthatitcanbetreatedinthesimplifiedwayrecommendedin2.6.1and3.3.1.Asitis difficult to estimate directly if it is to be taken only by element 2 or if it is distributed between elements2and4westartbyassumingthatthelineofzeroshearforcepassesthroughthepointload butthatthewholeloadiscarriedbyelement2.WethengetthefollowingaveragemomentsfromEq. (2.5)forthedistributedloadandasimplemomentequationforthepointload: (3.32) (3.33)

Fig.3.3.5 Fromthesevaluesitcanbeseenthatelement2takesalittletoomuchofthepointload.Itcaneasily beshownthattheconditionthatthetwovaluesshallbeidenticalisfulfilledifelement2carries37.39 .Iftheaboveanalysishadgivenm2<m4thelineof kNandelement42.61kN.Themomentis zeroshearforceshouldhavebeenmovedtoapositionfurtherdowninthefigureinordertomakethe momentsequal.

52

Asimplifiedwayofcalculatingmomentsinthiscaseistotreatthedistributedloadandthepointload separatelyandaddthemaximummomentsfromthesetwocases,inspiteofthefactthatthemaxima donotoccuratthesamesections.Suchanadditionisonthesafeside.Forsymmetryreasonstheline of zero shear force for the distributed load in this case is situated from the edge. The averagemomentis (3.34) This value is about 3% higher than the value when the distributed load and the point load were treatedtogether.Thissmalldifferenceisacceptablefromthepointofviewofreinforcementeconomy. Thissimplifiedapproachissatisfactoryinmanycases. ForthereinforcementinthexdirectionEq.(2.4)gives: (3.35) Theratiobetweenthemomentsinthexandydirectionscanberegardedasacceptable,butmaybea littlelow.Ifwewishtoincreasethisratiowemayincreasetheheightofthetrianglesandrepeatthe calculation. We may calculate the distribution of design moments approximately as in Examples 3.1 and 3.2. However,inthiscasemuchofthebendingmomentiscausedbythepointloadandfromthepointof viewofperformanceintheservicestateitisbettertoconcentratemoreofthereinforcementcloserto thepointload. Atthesametimeanotherpossibilitycanbedemonstratedhere,viz.zonesofzeroreinforcementalong the supports. The slab has zero curvature along the supports and thus the reinforcement takes no stresses, at least not until at a late ultimate state when severe cracks appear and membrane action takes place. Therefore the reinforcement along the supports is of little use and it is better to use it where curvature exists. In the authors opinion such unreinforced zones should be used along supportsinallslabs,butmostbuildingcodesdonotseemtoacceptthis,whichiswhymostexamples inthisbookareshownwithoutsuchzones. Inthisexampleawidthof0.5mhasbeenchosenfortheunreinforcedzonesandthereinforcement has been placed with a certain concentration around the point load. The proposed distribution of designmomentsisshowninFig.3.3.6.Intheydirectionthesmallermomentshavebeenchosentobe onethirdofthelargermoments.

Fig.3.3.6

53

CHAPTER4 Rectangularslabswithonefreeedge
4.1 Introduction
4.1.1 Generalprinciples
Asoneedgeisfree(unsupported)thestripsatrightanglestothatedgehavenosupportatthefree end.Insteadtheyhavetobesupportedinternally,i.e.byastripalongtheedge.Somesolutionsofthis typearegiveninStripMethodofDesign.Thesolutionsshownbelowaresomewhatdifferent,asthe designprocedureadoptedissimilartothatusedforslabswhereallthesidesaresupported. ThesupportalongthefreeedgeisassumedtobeasupportbandofthetypedescribedinSection2.8, i.e.abandwhichintheanalysisistreatedashavingzerowidth,butwheretheresultingreinforcement isdistributedoveracertainwidthalongtheedge.AccordingtotherecommendationsinSection2.8 thewidthoverwhichthereinforcementisdistributeddependsontheimportanceofthesupportband tothesafetyoftheslab.Typicallythiswidthcanbechosenratherarbitrarilyifthefreeedgeisinthe shorter direction of the slab, but it must satisfy some limiting rules if the free edge is in the longer direction of the slab. Some concentration of reinforcement along a free edge is generally recommended. Fig. 4.1.1 a) shows a slab with the upper edge free. Lines of zero shear force are shown exactly the sameasusedforslabswithfoursidessupported.Element2hasnorealsupport.Insteaditisassumed to be supported on asupport band along the free edge. This support band is loaded by the load on element 2. The bending moment Mx in the strip is determined by the load and its distribution on element2.

Fig.4.1.1 Forthetotalmomentmxintheslabwecanaddthemomentinthestriptothemomentsinelements1 and 3. Formally this can be done by analysing new elements 12 and 32, where the corresponding partsofelement2areincorporatedintoelements1and3.Thusthedesignoftheslabcanbebasedon theelementsandlinesofzeromomentsshowninFig.4.1.1b). A design based on Fig. 4.1.1 b) can only be used if . The values of c2 and c4 depend on the momentsmyfandmysandonthetypeofload.Forauniformloaditcanbeshownthatthisconditionis fulfilledif
54

(4.1)

Incaseswherethisrelationisvaliditcanbeassumedthatthewidthofthebandofreinforcementisof minorimportanceforthesafetyoftheslab.Thusinsuchacasenocheckofthepartofthemoment whichisattributedtothebandisneeded,norisanycalculationofasuitablebandofreinforcementor widthofsuchabandneeded.However,acertainconcentrationofreinforcementinthevicinityofthe freeedgeisstillrecommendedinsuchcases. Where the above relationship is not fulfilled, elements 2 and 4 will no longer be triangles, but trapezoids,Fig.4.1.2.Thelongerthelineofzeroshearforcebetweenelements2and4becomes,the moreimportantistheroleplayedbythesupportbandinthesafetyoftheslab.Themomentsinthe supportbandmustnowbecalculatedandthewidthofthereinforcementbanddeterminedaccording totheprinciplesdiscussedinSection2.8.

Fig.4.1.2

4.1.2 Torsionalmoments.Cornerreinforcement
The distribution of moments in slabs with an unsupported edge is more complicated than in slabs with all sides supported, particularly when the free edge is one of the longer edges. In such a case, torsional moments play an important role, whereas the bending moments corresponding to reinforcementatrightanglestothefreeedgeareofminorimportance.Inthestripmethod,solutions areusedwheretheloadiscarriedbybendingmomentsonly,withouttakingtorsionalmomentsinto account.Theresultingdesignisonthesafesideregardingtheultimatelimitstate,butthedistribution of reinforcement may be quite different from that given by an analysis based on the theory of elasticity. According to the theory of elasticity large torsional moments occur where two simply supported edgesmeet.Intheseareasreinforcementmaybenecessarytopreventunacceptablecracking.Many codes specify particular comer reinforcement that should be provided. To limit cracks in the upper faceoftheslabtheprovisionoftopreinforcementintheslabcornersparallelltothelinesbisecting the corners is recommended. The amount of such reinforcement may be based on code rules or on values of torsional moments obtained from tables of moment values determined by the theory of elasticity. Theimportanceofthetorsionalmomentsincreasesastheratiobetweenthelengthofthefreeedge andthelengthoftheperpendicularedges(a/binFig.4.1.1)increases.Theuseoftheformofthestrip metod presented here is not recommended for simply supported slabs if this ratio is higher than about2.AnalternativeapproachforsuchslabsisdemonstratedinStripMethodofDesign.

4.2 Uniform loads

Example4.1 TheslabinFig.4.2.1hasonefreeedge,onesimplysupportededgeandtwofixededges,andsupports a uniform load of 9 kN/m2. The shape is such that, with a relatively short free edge, the approach accordingtoFig.4.1.1b)maybeused.Theproposedlinesofzeroshearforceareshown.Forelements 12and32wegetfromEq.(2.5)
55

(4.2) (4.3) andthus .Forelement4wegetfromEq.(2.4)

Fig.4.2.1

(4.4) Forafixedshortedgeitissuitabletochooseahighratiobetweenthenumericalvaluesofthesupport and .Theconditionof andspanmoments.Forexample,inthiscasewemaychoose Eq.(4.1)iswellfulfilled. These average moments can be distributed as shown in Fig. 4.2.2. The distribution of the reinforcementisarrangedsothatthereisastrongbandofreinforcementalongthefreeedge.

Fig.4.2.2 Example4.2 The slab shown in Fig. 4.2.3 carries a uniform load of 11 kN/m2. It is of the type which cannot be designedbymeansofthesimplifiedapproachusedinthepreviousexample,asrelation(4.1)cannot befulfilledbyadoptingasuitableratiobetweenthemomentsinthexandydirections.Insteadthe slabhastobeanalysedasifitweresimplysupportedonasupportbandlocatedalongitsupperedge, afterwhichananalysismustbemadeofthesupportband.
56

WiththechosenpatternoflinesofzeroshearforceandbyapplyingEqs.(2.4)and(2.5)asshown,we getthefollowingaveragemomentsintheslab

Fig.4.2.3 (4.5) (4.6) (4.7)

(4.8) The load distribution on the support band corresponds to the shape of element 2. If we assume a sectionofzeroshearforceinthesupportbandsituatedatadistanceof2.3mfromtherighthandend, wefindthemoments (4.9)

(4.10) IftherecommendationsinSection2.8arefollowedstrictly,thespanmomentMfwillbeconcentrated onawidthcorrespondingtohalftheaveragewidthofelement2,whichcorrespondsto0.45m.The widthhasbeenchosento0.5mintheproposeddistributionofdesignmomentsshowninFig.4.2.4.It mightevenhavebeendistributedoveralargerwidthwithoutanyriskofinconvenience.Inanycaseit isnotpossibletoshowbyyieldlinetheorythatsuchadesignisunsafe.Themomentsmxfandmxshave beendistributedovertheremainingpartoftheslab.

57

Fig.4.2.4

4.3 Triangular loads

In practice, triangular load is usually a liquid pressure, an earth pressure or something similar. A rectangularslabwithonefreeedgeinsuchcaseshasthefreeedgeastheupperedge.Onlythiscase willbedemonstratedhere. Forthetreatmentoftriangularloadswithacombinationofloadingcases,seealsoExamples3.5and 3.6. Example4.3 TheslabshowninFig.4.3.1formsonesideofawatertankwiththehighestwaterlevel1.0mbelow the upper edge. The water pressure is assumed to increase by 10 kN/m2 per metre depth (a more correctvalueis10.2).Thedistributionofthewaterpressureisshowninthefigure.Theshapeofthe slabissuchthatthesimplifiedapproachofFig.4.1.1b)maybeused.Thatthisassumptioniscorrect canafterwardsbecheckedbystudyingatriangularelementoftype2inFig.2.1.1a)andestablishing thatitdoesnotreachelement4.

Fig.4.3.1 The average moments in the xdirection can be calculated by means of Eq. (2.14) with .Theaveragemomentfromthisequationisvalidforthedepth4.0m.Toget theaveragemomentforthetotaldepth5.0m,thisvaluehastobemultipliedby4.0/5.0. (4.11) TheaveragemomentsintheydirectionarecalculatedwithEq.(2.4)forauniformloadof40kN/m2 minusatriangularloadof18kN/m2atthetop,calculatedwithEq.(2.10). (4.12) Thefollowingvaluesofaveragemomentsarechosentoensuresuitableratiosbetweensupportand .Adistributionofdesignmomentsisproposed spanmoments: in Fig. 4.3.2. With respect to the load distribution, the mxmoments might have been more concentrated downwards, but on the other hand there is also the general rule that recommends concentrating some reinforcement close to the free edge. As a compromise the design moment has beengivenaconstantvalueoverthemajorpartofthedepth.

58

 Fig.4.3.2 Example4.4 The slab in Fig. 4.3.3 forms one side of a water tank with the water surface 0.4 m below the upper edge.Fromthatlevelthewaterpressureisassumedtoincreaseby10kN/m2permetredepth(amore correctvalueis10.2).TheshapeoftheslabissuchthatthesimplifiedapproachaccordingtoFig.4.1.1 b)cannotbeapplied.AnassumedpatternoflinesofzeroshearforceisshowninFig.4.3.3.

Fig.4.3.3 InordertobeabletousethestandardformulasinChapter2forcalculatingthemomentinelement2 the load has to be divided into the following cases, which are added to form the actual load (cf. Example3.6,Fig.3.2.5): 1. Atriangularload,zeroatthefreeedgeand22atthebottom. 2. Auniformload4overthewholedepth. 3. Auniformload+4ontheuppermost0.4 m. 4. Atriangularload,zeroatthefreeedgeand4at0.4mbelowthefreeedge. ForthefirstandfourthcasesEq.(2.11)isapplied,andforthesecondandthirdcasesEq.(2.5).Forthe thirdandfourthcases ,thewidthofelement2atthewaterlevel.

(4.13)

59

The moment in element 4 is calculated from the influence of a uniform load with intensity 42 accordingtoEq.(2.5)minusatriangularloadwithintensity24atthetopoftheelementaccordingto Eq.(2.11). (4.14) . Weget The average moments in elements 1 and 3 can, with satisfactory accuracy, be calculated with a triangular load with maximum intensity 46 according to Eq. (2.13) minus a uniform load with intensity4accordingtoEq.(2.4).Asmalldifferenceintheuppercornerisdisregarded. (4.15) ; . Wemaythenchoose Theloadonthesupportbandalongthefreeedgecorrespondstotheloadonelement2.Theloadis looked upon as consisting of onepart with a pyramidal shape and one part with aprismatic shape. Duetosymmetrythepointofzeroshearforce(maximumspanmoment)issituatedinthecentre.

(4.16) . Wecanchoose Todeterminethedesignmomentswehavetoaddthemomentsfromelement1andfromthesupport band.AdistributionofdesignmomentsisproposedinFig.4.3.4.Ifthedistributionsfromthedifferent parts had been followed we should have had a concentrated band of reinforcement along the free edge, and also heavier reinforcement below the centre of the slab, as the load is increasing downwards. Halfway between the top and the centre the reinforcement would have been weaker. Such an uneven distribution of reinforcement does not correspond to our experience of the real behaviourinservice.Thereforethedesignmomenthasbeenchosenwithauniformdistributionover mostofthedepth,withoutanyconcentration.

Fig.4.3.4

60

4.4 Concentratedloads
4.4.1 Loadsclosetothefreeedge
Aconcentratedloadactingclosetoafreeedgehastobecarriedbyreinforcementalongtheedge.If theloadisactingveryclosetotheedgeitwillcauseanegativemomentwithacertaindemandfortop reinforcementatrightanglestotheedge.Thisreinforcementwillinthefirstplacebeneededforthe distribution of the load to a strip of a certain width in the ultimate limit state, but it may also be needed to limit top cracks. In principle, the strip method only takes the ultimate limit state into account,andthisistheonlycasewhichwillbedemonstratedintheexamplebelow.Inordertolimit topcracksaminimumwidthoftheassumedstripalongtheedgeisrecommended,e.g.1/10to1/5of thelengthofthefreeedge. Inpracticetheconcentratedloadalwaysactstogetherwithadistributedload,suchastheloadarising fromtheweightoftheslab. Example4.5 TheslabinFig.4.4.1carriesapointloadof30kNataposition0.1minsidetheedge.Ifwelet10kNbe carriedtotherightand20kNtotheleftsupportthemomentintheedgestripis , . If we decide not to reinforce for a higher span moment than about 40 kNm/m for the point load (in addition to the moment for the other loads on the slab), we have to distributetheloadoverastripwidthofabout .Wecanassumeawidthof0.65m.

Fig.4.4.1 Apossibleloaddistributioninthestripintheydirectionisshowninthefiguretotheright.Inorder tomaintainequilibriuminthestripweneedadownwardforceatthelowerendofthestrip,i.e.atthe support.Fromsimplestaticsthisforceisfoundtobe .Inarigorousanalysisthis forcemaybetakenintoaccount,butasitissmallcomparedtotheactingload,wewilldisregardits influenceandsimplytakethedesignmomentforthereinforcementatrightanglestotheedgeasthe .This loadtimesthedistancebetweentheloadandthecentreofthestrip,thus analysis is quite accurate enough bearing in mind the approximate assumptions, which are conservativefromthepointofviewofsafety,ascanbecheckedbymeansofyieldlinetheory. The result in this case is thus that the point load gives span reinforcement corresponding to , distributed on a strip with a width of 0.65 m along the edge, and top reinforcement at right angles to the edge at the load. The support moment corresponding to canbedistributedoveragreaterwidth,asthesupportinitselfhasaloaddistributingeffect. ThemomentMycanbedistributedonquiteasmallwidth,say0.51.0minthiscase.
61

The use of the simplified approach which is demonstrated here is generally recommended for designingreinforcementtocarryaconcentratedloadclosetoafreeedge: 1.Calculatemomentsinthestripalongtheedge. 2.Determineasuitablewidthofthestriptogiveanacceptableconcentrationofspanreinforcement. Calculatethedesignmomentforthetopreinforcementatrightanglestotheedgeastheloadtimes 3.thedistancefromtheloadtothecentreofthestrip.Thisreinforcementisconcentratedonarather smallwidthneartheload.

4.4.2 Loadsnotclosetothefreeedge
Inmanycasesitissimplesttoassumethattheloadiscarriedbyastripwhichisparalleltothefree edge.Inthiscasethedesignisdoneexactlyasinaslabwithfoursidessupported,seeSection3.3. Whenthedistancefromtheloadtothesupportoppositetothefreeedgeisshort,theloadorpartofit isbestcarriedinthedirectionatrightanglestothefreeedge.Accordingtothegeneralprinciplesfor slabswithafreeedge,thestripatrightanglestothatedgeisassumedtohaveasupportattheedge, thuscausingmomentsinastripalongtheedge. Theexamplebelowdemonstrateshowthedesignisperformedforalargeconcentratedload,whichis thedominantloadontheslab.Wheretheconcentratedloadisonlyaminorpartofthetotalloadon theslabasimplifiedapproachmaybeusedaswasdemonstratedinExample3.9. ForthegeneraltreatmentofconcentratedloadsseeSections2.6and3.3. Example4.6 TheslabinFig.4.4.2carriesapointloadof40kNatapositionwhichisclosertothesupportopposite tothefreeedge.Thisisacasewheretheload(oragreatpartofit)isbestcarriedbyastripinthey direction. Ifthewholeloadiscarriedbyastripintheydirectionthisstripisassumedtobesupportedatthe freeedge.Wecanassume,forexample,that6kNiscarriedtothefreeedge.Thisgivesaspanmoment .Thesupportmomentisthen . Theedgestriphastocarrythereaction6kN.Ifweassumethat2kNiscarriedtotherightsupport, ; . themomentsintheedgestripare: If the whole load is instead carried by a strip in the xdirection we can then assume that 13 kN is in this strip and carried to the right support, which gives .

Fig.4.4.2 Ifwecomparethetwosolutionswefindthatthefirstsolutiongivesasumofspanmomentsequalto compared to 31.2 in the second, and a numerical sum of support moments equal to inthefirstand66.0inthesecondsolution.Asthereinforcementareaisproportionalto themomentthiscomparisonshowsthatthefirstsolutionistobepreferredfromaneconomicalpoint ofview.
62

Itisalsopossibletodividetheloadbetweenthetwosolutions,forexample,inordertogetabetter reinforcementdistributionwithrespecttothebehaviourunderserviceconditions.Iftheconcentrated loadisdominantsuchadivisionissuitable,butifthedistributedloadontheslabisdominantsucha divisionisunnecessary. For a complete solution some reinforcement for load distribution across the strip should also be provided. If we choose a strip width of 1.0 m for the main strip and 0.5 m for the load distribution reinforcement,thedesignmomentforthisreinforcementis ,Eq.(2.33).If wechooseawidthof0.5mforthestripalongthefreeedge,wegetthedesignmomentsaccordingto thefirstsolutionasshowninFig.4.4.3.Thesupportmomentshavebeendistributedovertwicethe widthsofthestrips.Thisisacceptableasthesupportsactasloaddistributors.

63

CHAPTER5 Rectangularslabswithtwofreeedges
5.1 Twooppositefreeedges
Whenarectangularslabhastwooppositesidessupportedandtheothertwoedgesfree,itactsasa series of oneway strips and causes no design problems, at least not as long as the loads are distributed. Designforconcentratedloadsismadeinthesamewayasforotherrectangularslabs,seeSection3.3 forloadsintheinterioroftheslabandSection4.4forloadsclosetoafreeedge.

5.2 Twoadjacentfreeedges
5.2.1 General
Slabswithtwoadjacentfreeedgeshavearathercomplicatedstaticbehaviour,particularlywhenthe other two edges are freely supported. Because strips parallel to the edges do not have supports at bothendstheloadhastobecarriedtoagreatextentbytorsionalmomentswithrespecttotheedges. There are at least two ways of coping with the problem. One possibility is to carry out an analysis whichincludestorsionalmomentsanddesignreinforcementparalleltotheedgestotakethetorsional moments.Theslabcanthenbereinforcedonlywithbarsparalleltotheedges,whichisanadvantage during construction. Solutions including torsional moments have been discussed in Strip Method of Design. These solutions are generally too complicated to be recommended for practical design. One suchsolutionwill,however,begivenasanalternativebelow. Another possible solution is to make use of a support band, which spans between the two corners where the free and supported edges meet. This method may lead to somewhat lengthy numerical calculationsandchecks.ItsapplicationisdemonstratedinChapter7onanonrectangularslab.Herea similarbutnumericallysimplermethodwillbeappliedinthefirstexample,usingcornersupported elements. The results of these two methods are similar, as the reinforcement is not parallel to the edges.Thisisadisadvantagefromthepointofviewofconstruction,butitsavesmuchreinforcement atthesametimeasitreducesdeflectionsandcracksasthereinforcementisactiveapproximatelyin thedirectionsoftheprincipalmoments.

5.2.2 Simplysupportededges,uniformloads
Thisisthemostcomplicatedcase,asnoloadcanbetakensolelybybendingmomentsparalleltothe edges. It can even be argued that this is an unsatisfactory type of structure. It is probably not very common,atleastnotforstructuresotherthansmallbalconies,forexample. Thecentreofgravityofauniformloadontheslabissituatedwherethediagonalscross.Itistherefore natural to assume that the slab is supported at point supports at the outer ends of the supported edges.Iftheslabisassumedtobesupportedalongthesupportededgessomeofthesupportreaction hastobenegativeinordertofulfilequilibriumconditions.
64

Apossiblesolutionistoassumethattheslabissupportedattheoutercornersofthesupportededges and to base the design on the use of cornersupported elements. Such a loadbearing system is illustratedinFig.5.2.1withcornersupportsatBandD.Thecornersupportedelementsaretriangles formedbytheedgesandthediagonals,whicharelinesofzeroshearforce.Inordertobeabletouse therulesfortriangularcornersupportedelementsthereinforcementdirectionshavetobeparallelto thesidesofthetriangles,inthiscaseparalleltothediagonals.Itisimpossibletomaintainequilibrium onalineofzeroshearforceifthereinforcementisparalleltoanedgeoftheslab,astheacceptable distributionsofdesignmomentsinthetwoelementswhichmeetatthelineofzeroshearforcecannot thenbemadetocoincide. Asthereinforcementdirectionsinthegeneralcasearenotatrightangles,theruleinSection2.5.4has tobeapplied.Thisrulesaysthatthecvaluesforboththereinforcementdirectionsareequaltohalf thelengthofthediagonal. The design of the reinforcement in direction BD is based on Fig. 2.5.4 c) as the ydirection in this figureisparalleltothesidewhichdoesnotgotothesupportedcorner.Asthereisnosupportmoment indirectionBD,wefindforthatdirectiontheaveragemoment

Fig.5.2.1

(5.1) ForthereinforcementindirectionACthereisnospanmoment.ThedesignisbasedonFig.2.5.4b). WefindfordirectionACtheaveragemoment (5.2) ForthedistributionofdesignmomentswehavetofollowtherulesgiveninSection2.5.3. FordirectionBDEq.(2.32)isvalid.Inthisdirectionwehavetotakeallthedesignmomentonlyinthe column strip as . It may be suitable to distribute the reinforcement as evenly as possible, i. e. with .Thismeansthatthedesignmomentiszerointheexteriorpartoftheelementand3times theaveragemomentinthecentralpart(columnstrip").Itisnotacceptabletoleavetheexteriorparts without reinforcement, so some reinforcement must also be provided in these parts. In Fig. 5.2.2 a designmomentisrecommendedequaltohalftheaveragemomentaccordingtoEq.(5.1)withinthese parts.Thischoicecanofcoursebediscussed. For direction AC Eq. (2.31) is valid. Here it is acceptable to use which means an evenly distributeddesignmomentequaltotheaveragemoment,whichisasuitablechoiceinthiscase. UsingthesevalueswegetthedesignmomentsaccordingtoFig.5.2.2.Itmustbenotedthattherules inSection2.4.3are basedontheassumption thatallreinforcingbarshave afulllengththrough the elementandarewellanchoredattheedges.Itisalsorecommendedtohaveatleastonebottombar andonetopbaralongeachfreeedge.

65

 Fig.5.2.2 Asthereinforcementbasedonthisdesigniscomplicated,involvingdifferentlengthsforeachbar,a design with bars parallel to the edges may be preferred, even if it leads to a greater amount of reinforcement.ApplyingasolutionincludingtorsionalmomentsofthetypediscussedinStripMethod ofDesign,Section4.3,itcanbeshownthatasafedesignwouldbetoreinforcethewholeslabparallel totheedgesforthepositivemoments (5.3) andthenegativemoments (5.4) Asthereinforcementisintendedtotaketorsionalmomentsithastobewellanchoredattheedges. In order to fulfil equilibrium conditions with this solution the inner corner A in Fig. 5.2.1 has to be anchoredbyaforceRequalto

(5.5) Thetotalamountofdesignreinforcementismorethandoubledwiththisdesign,butinspiteofthisit mayoftenberecommendedduetothesimplicityofconstruction.Dueregardhasalsotobetakenof the fact that the first solution may be regarded as unacceptable without some secondary reinforcement.

5.2.3 Onefixededge,uniformloads
If a support is fixed, a strip at right angles to that support can act as a cantilever, carrying load by meansofpurebendingmoments.Inprinciplealltheloadcanbecarriedinthiswayifatleastoneof the supports can take negative moments. If only one of the supports is fixed such a solution is uneconomical and it also gives a design which is not suitable for the behaviour of the slab under serviceconditions. Ifalltheloadiscarriedbycantileveractiontherewillbeonlytopreinforcementintheslab.Thisisnot acceptable, as it may result in wide bottom cracks. In the service state moments will give rise to tensioninthebottomfaceoftheslab.Asimplewayoftakingcareofthisproblemistoassumethat onepartoftheloadiscarriedwithoutcantileveraction,i.e.asiftheslabweresimplysupported.The moments caused by thatpart of the loadare then calculated by means ofequations (5.3)and (5.4). The amount of the load carried in this way is chosen to ensure that a suitable minimum amount of bottomreinforcementisprovided. Example5.1 The slab in Fig. 5.2.3 has one long fixed support and one short free support. The load is uniform 9 kN/m2.Inordertogetsomebottomreinforcementinbothdirectionsitisassumedthat20%ofthe loadgivesmomentsaccordingtoEqs(5.3)and(5.4).Thisratioshouldbechosensothattheamountof
66

bottom reinforcement is sufficient to prevent wide bottom cracks and to take at least some of the torsionalmomentsintheservicestate. The part of the load to be used in Eqs (5.3) and (5.4) is thus . This gives design moments,whichareactiveinthewholeslab: (5.6)

(5.7)

Fig.5.2.3 Therestoftheload,7.2kN/m2,iscarriedbyonewaystripsasshowninFig.5.2.3.Thestripinthex directionissupportedatitsrighthandendbythecantileverintheydirectionalongthedashedline, which is the limit between the x and ystrips. The line of zero shear force is halfway between the dashedlineandthelefthandsupport. Asthecantileverintheydirectionhastoactasasupportforthexstrip,ithastocarryalltheloadto the right of the line of zero shearforce. Theaverage supportmoment is calculated by means of Eq. (2.5) (5.8) TheaveragespanmomentinthexstripiscalculatedbymeansofEq.(2.4) (5.9) (notarealsupport Thetotalmomentsarethesumofthemomentsfromthetwocases,thus moment,butanegativemoment), ; ; .Thedesign momentsaccordingtocase1shouldbeevenlydistributed,whereasthedesignmomentsaccordingto case2shouldbedistributedwithrespecttotheshapeoftheelements,i.e.withmorereinforcement alongthefreeedges.Distributionsofdesignmomentsforbottomreinforcementinthexdirectionand top reinforcement in the ydirection are proposed in Fig. 5.2.4. The design moments for top reinforcementinthexdirectionandforthebottomreinforcementintheydirectionareconstantwith thevaluesabove.Thereinforcementcorrespondingtocase1shouldcoverthewholeslabareaandbe well anchored, whereas the reinforcement corresponding to case 2 may be curtailed according to normaldetailingrules.

67

 Fig.5.2.4 Example5.2 The slab in Fig. 5.2.5 has the same proportions and the same load, 9.0 kN/m2, as the slab in the previousexample,butitistheshortedgethatisfixedandthelongedgewhichissimplysupported. Thisleadstobehaviourexpectedtobeclosertothatofaslabwherebothedgesaresimplysupported. It is then natural to carry more of the load as if the slab were simply supported. Whereas in the previousexamplewechosetocarry20%oftheloadinthisway,herewechoosetotakehalftheload inthisway.Thus4.5kN/m2isassumedtogivemomentsaccordingtoEqs(5.3)and(5.4): (5.10)

(5.11)

Fig.5.2.5 Theremainingpartoftheload,4.5kN/m2,iscarriedbythestripsystemaccordingtoFig.5.2.5,which issimilartothesysteminFig.5.2.3 withtheexceptionthatthe dividingline betweentheprimarily loadbearingdirectionsisdrawntotheoutercorner.Ifithadbeendrawntoapositionmoretothe left, the strips in the ydirection would not have had a support on their whole width. If it had been drawntoalowerpositionattherighthandedgethemomentinthestripwouldhaveincreasedtoo much. InthesamewayasinthepreviousexamplethemomentsarecalculatedbymeansofEqs.(2.5)and (2.4)respectively: (5.12)
68

(5.13) and .Distributionsofthe Addingthesemomentstothoseofthefirstcaseweget correspondingdesignmomentsareproposedinFig.5.2.6.Inadditiontothesemomentsthesteelfor evenly distributed reinforcement should be arranged for design moments mxf and mys according to Eqs(5.10)and(5.11)above.

Fig.5.2.6

5.2.4 Twofixededges,uniformloads
With both supported edges fixed it is simplest and most economical to carry the whole load on cantileveringstripsfromthosesupports.Withthissolutiontheslabonlyrequirestopreinforcement, butthismaynotensuresatisfactorybehaviourunderserviceconditions.Inordertoprovideforsome bottomreinforcement inthedesignsomeoftheloadcaninstead becarriedas ifthesupportswere freelysupported,i.e.applyingEqs(5.3)and(5.4)forthatpartoftheload. The dividing line between the cantilevers in the two directions is best drawn approximately in the directionofthebisectorbetweenthesupportededges. Example5.3 The slab in Fig. 5.2.7 has the same proportions and the same load, 9 kN/m2, as the slabs in the previoustwoexamples,butbothsupportededgesarefixed. Inordertoget somebottomreinforcementwetake 20%of theload,1.8kN/m2,asiftheslabwere simply supported, applying Eqs (5.3) and (5.4). We can take the values directly from Example 5.3, wherethesameassumptionwasused: and . Therestoftheload,7.2kN/m2,iscarriedbythecantilevers.FromEqs(2.4)and(2.5)wederivethe averagesupportmoments (5.14) (5.15)

69

Fig.5.2.7 ; . Distributions of the negative design Adding the negative values above gives momentsareproposedinFig.5.2.8.Inadditiontothesenegativemomentsthewholeslabshouldbe reinforcedinbothdirectionsforpositivemomentswithintensity10.80.

5.2.5 Nonuniformloads
Thesameprinciplesasthoseabovecanbeappliedtoothertypesofdistributedloads.Theprincipleof Fig.5.2.1iscomplicatedtoapply,partlybecausetheshapesofthecornersupportedelementshaveto be modified in order to fulfil equilibrium conditions, and partly because the rules for acceptable momentdistributionsarenotknownindetail.IftheprincipleaccordingtoFig.5.2.1isappliedthishas tobedonewithcare,usingdesignmomentswhichareestimatedtobewellonthesafeside. Itisalwayspossibletousetheapproachwithasupportband,whichisdemonstratedinSection7.2.4. For slabs with fixed supports the principles demonstrated in Examples 5.13 can be applied to any type of distributed load. The part carried by applying Eqs. (5.3) and (5.4) is always chosen to be uniform. Concentrated loads on slabs with fixed supports can be treated with the methods demonstrated in Sections3.3and4.4.

Fig.5.2.8 AconcentratedloadFkNonaslabwithtwosimplysupportededgescanbeassumedtogiveriseto designmomentsequaltoF/2kNm/mwithinarectanglewithsidesalongthesupportededgesanda corneratthecentreoftheload,seeFig.5.2.9.

70

 Fig.5.2.9 As nonuniform loads on slabs with two adjacent free edges are rarely encountered, no numerical exampleisgiven,butitishopedthattheadvicegivenabovewillbesufficienttoenableasafedesignto becarriedoutinanysuchcases.

CHAPTER6 Triangularslabs
6.1 General
6.1.1 Reinforcementdirections
Atfixedorcontinuoussupportsthemostefficientuseofreinforcementiswiththebarsatrightangles tothesupport.Spanreinforcementissuitablyarrangedasarectangularmeshwithoneofthemesh directionsparalleltooneoftheslabedges.Iftheslabhasonefreeedgeitisnaturaltoplacethespan reinforcement parallel to that edge. In other slabs it is often suitable to choose one of the span reinforcementdirectionstobeparalleltotheshortestside.

6.1.2 Calculationofaveragemomentsinwholeelements
71

Whereas in a rectangular slab all reinforcement is normally arranged in only two orthogonal directions, parallel to and at right angles to the edges, this is not possible in triangular slabs. This poses particular problems. Reinforcement oriented in different directions has to cooperate. Some reinforcementformsaskewangletotheedgewherethecorrespondingelementhasitssupport. Thebasicmethodofanalysisisthesameasforrectangularslabs.Theslabisdividedintoelementsby meansofassumedlinesofzeroshearforcesothateachelementhasitssupportalongonesideofthe triangle. For each element the moment equilibrium is considered with respect to the support. The differencefromrectangularslabsisthatthereisnotonewelldefinedloadbearingdirectionineach element,astheloadhastobecarriedbymeansofreinforcementrunningindifferentdirections.For example, these may be one direction for the top reinforcement and one or two directions for the bottomreinforcement.Therearetwodifferentapproachesfordeterminingthedesignmomentsunder thesecircumstances,seeSection2.3.6. Oneapproachistointroducealineofzeromoment,withpositivemomentsononesideandnegative moments on the other. Span reinforcement is active within the part with positive moment and determines the loadbearing direction within that part, whereas the support reinforcement determines the loadbearing direction within the part with negative moments. Strips with span reinforcement are treated as supported on cantilevers with support reinforcement. With this approachthemomentdistributioncanbedeterminedinarigorousway.Thenumericalcalculations maybecomecomplicated,particularlywhenthesamesupportreinforcementfunctionstogetherwith thespanreinforcementintwodirections. Theotherapproachistowriteacompleteequilibriumequationwithrespecttothesupportforeach element, taking into account the moment vectors along all the boundaries of the element. The numericalcalculationsaregenerallysimplerwiththisapproach.Thedisadvantagesarethattheresult doesnotgiveaclearoverviewofthedetailsofhowtheloadiscarried,orofthetheoreticalmoment distributions which would be used as a basis for determining the distribution and curtailment of reinforcement.

6.1.3 Distributionofreinforcement
Whereareinforcingbarcutsoveracorner,thelengthofthebarwithintheslabisveryshortnearthe corner.Abarwithsuchashortlengthcannotbeexpectedtobeofanyuse.Itisthereforeproposed that,forslabsofnormalthickness(about0.15to0.2m)barsshorterthanabout0.5mshouldnotbe usedwithintheslab.Thedesignmomentswithinsuchpartsofaslabarethusreducedtozero. Elements in triangular slabs have triangular shapes. The theoretical moment distribution is more uneventhaninelementsoftrapezoidalshape,whichcarrymostoftheloadinrectangularslabs,see Figs. 2.3.3 and 2.3.4. In the proposed distribution of design moments in the examples the ratio betweenthemajorandminordesignmomentshasbeenchosentobe3,whereasforrectangularslabs thisratiowasoftenchosentobe2. Atcornerstheremaybeaneedfortopreinforcementrunningapproximatelyinthedirectionofthe bisectorofthecornerangleinordertoavoidtopcracks.Suchreinforcementcannotbederivedfrom thesolutionsgivenbythe stripmethod,buthastobedesignedbysomeothermeans,basedon the theoryofelasticity.

6.2 Uniformloads
6.2.1 Allsidessimplysupported
Example6.1 ThesimplysupportedtriangularslabinFig.6.2.1hasauniformloadof8kN/m2.Theassumedlinesof zeroshearforceareshown.

72

Fig.6.2.1 IthasbeenshowninStripMethodofDesignthatasuitableheightofelement3is0.2to0.25ofthe lengthoftheshortside.Inthiscaseithasbeenchosentobe . Firstanapproximatesolutionwillbeshown,whereweusetheapproximationdemonstratedinFig. 2.2.4forelement3(whichmeansthatthedashdotlinesborderingelement3arenotreallinesofzero shear force, though they are treated as such) and also disregard the influence of myf on the equilibrium of elements 1 and 2. This means that both elements 1 and 2 will cause the same span moment mxf, calculated from Eq. (2.4) with c being the horizontal distance from the point of intersectionofthelinesofzeroshearforcetotheedge,cf.Fig.2.3.15.Inordertogetthesamevalues ofmxfthedistancesc1andc2havetobeequal.Wethusfind (6.1) (6.2)

(6.3) Thisisevidentlyaverysimplewayofdeterminingtheaveragemoments.Theapproximationisonthe safeside.Itmaybenotedthatthehorizontalpositionofthepeakofthetriangledoesnotinfluencethe result. Foramorerigoroussolutionwehavetotakeintoaccounttheinfluenceofmyfontheequilibriumof elements1and2,andstudytheequilibriumwithrespecttothesupports.Fig.6.2.2showselement1 withtheactingtotalmomentsattheboundaries.ApplyingEq.(2.18)andEq.(2.4)weget

(6.4)

73

 Fig.6.2.2 Thelastterminthisequationisthereductioninmxfduetomyf.Inthesamewaywefindforelement2 (6.5) Wealsohavetherelation (6.6) we can solve these equations and find , and , which is a lower With value than the approximate value of 5.20 above. An estimation of the relative amounts of reinforcement with the two solutions can be made by comparing the sum of moments in the two directions, which gives the ratio . The simple approximate solution thus requiresabout8%morereinforcementbecauseitismoreconservative. In Fig. 6.2.3 a distribution of design moments has been proposed, based on the more accurate and economical solution, and taking into account the general rules for reinforcement distribution in Section6.1.3.Thetheoreticaldistributionofmomentsmxfisalsoshown.

Fig.6.2.3 Lastlyitshouldbementionedthataverysimplesolutionforasimplysupportedtriangularslabwitha uniform load is possible if the same average moments are used in two orthogonal directions. This averagemomentis

(6.7)

74

whereriistheradiusoftheinscribedcircle.Thepointofintersectionofthelinesofzeroshearforceis ,whichgives .Thetotal atthecentreofthecircle.Forthetriangleintheexample amount of reinforcement with this solution is a few percent higher than with the most accurate solutionabove,butlowerthanwiththesimpleapproximatesolution.

6.2.2 Onefreeedge
Example6.2 The slab in Fig. 6.2.4 has two orthogonal fixed supports and one free edge. It is acted upon by a uniform load of 7 kN/m2. The top reinforcement is at right angles to the supports and the bottom reinforcement is parallel to the free edge. Thus there is designed bottom reinforcement only in the directionparalleltothefreeedge.Theelementsaredenoted1and2,theaveragesupportmoments ms1andms2,andthespanmomentmxf. Acoordinatesystemwiththexaxisinthedirectionofthespanbottomreinforcement,paralleltothe freeedge,isintroducedinordertobeabletouseEq.(2.18).ApplyingthatequationandEq.(2.4)and usingthenotationinthefigurewehavetheequilibriumequations

(6.8)

(6.9) Wealsohavetherelation (6.10) Assumingvaluesofmxfandc1,thecorrespondingsupportmomentscanbecalculated.Thechoiceis madesuchthatsuitableratiosbetweenthemomentsresult.With and

Fig.6.2.4 thesupportmomentsare , .ThesemomentsareusedforthedesigninFig. 6.2.6. Inordertodemonstratetheuseofcooperatingstripsandhowatheoreticalmomentdistributioncan bedetermined,ananalysiswillalsobemadewiththismethod,usingthesameaveragemoments.For thispurposelinesofzeroshearforceandzeromomentsareshowninFig.6.2.5.Elements1and2are dividedintospanelementsIfand2fandsupportelements1sand2s.Eachoftheseelementsnowhas onlyoneloadbearingdirection. With and thecvalueofelementsIfand2fare,cf.Eq.(2.4):
75

(6.11) ThedistancesshowninFig.6.2.5arecalculatedbymeansofsimplegeometry. Theshearforcesatthelinesofzeromomentareproportionaltothedistancefromtheinnercornerof theslab.Wecanstudyathinspanstripalongtheedge.Attheendsofthisstriptheshearforceis (6.12) ThecorrespondingshearforcesattheendsofthesupportstripsarecalculatedbymeansofEq.(2.16). Insertingtherelevantsinvaluesfoundfromordinarytrigonometricalrelations,wefind:

Fig.6.2.5 (6.13)

(6.14) The average moments caused by these shear forces on the relevant parts of the supports are accordingtoEq.(2.15) (6.15)

(6.16) Theaveragemomentscausedbythedirectloadonthesupportelementsgive (6.17)

(6.18) Thesumsofaveragemomentsalongthesupportsare (6.19)

(6.20) Thesevaluesareidenticalwiththevaluesdeterminedforelements1and2above. Itmaybenotedfromtheanalysisthattheinfluenceoftheshearforcealongthelineofzeromoment dominatesincalculatingthesupportmoments.Thishastobetakenintoaccountindistributingthe designmoments. AdistributionofdesignmomentshasbeenproposedinFig.6.2.6.Inordertoillustratetheinfluenceof thedifferentpartsofthesupportmoment,acurvehasbeendrawnforthetheoreticaldistributionof

76

ms2accordingtothemodelwithcooperatingstrips.Itcanbeseenthatthismodelgivesasteepjump inthemomentintensity.Thisjumpdoesnothavemuchtodowiththerealbehaviouroftheslab,and thereisnopointintryingtoreinforceexactlyaccordingtothisdistribution.Thecurvedoes,however, indicatewhichpartoftheedgemostofthereinforcementshouldbeplacedin.

Fig.6.2.6 Regarding the span reinforcement it is always advisable to have a certain concentration of reinforcementalongafreeedge. In this case the slab has design bottom reinforcement in only one direction. Some secondary reinforcementmightbeintroducedatrightanglestothedesignreinforcementaccordingtorelevant coderules.Intheauthorsopinionitisquestionablewhethersuchreinforcementisofanyuse.

6.2.3 Fixedandsimplysupportededges
Example6.3 TheslabinFig.6.2.7hastwofixededgesandonesimplysupportededge.Theloadis9kN/m2.The linesofzeroshearforceareshownwiththecvaluesofthethreeelementsinsertedinthexandy directions respectively. The bottom reinforcement is arranged in the x and ydirections. From Eqs (2.18)and(2.4)wegettheequilibriumequations

Fig.6.2.7

77

(6.21)

(6.22)

(6.23) Wealsohavethegeometricalrelation (6.24) have been chosen all the corresponding moments can be After two cvalues and the ratio calculated. The choice is repeated until the ratios between the moments are estimated to be , gives , , ; , . acceptable.Asolutionwith Inthiscaseitiscomplicatedtomakeadesignbasedonthecooperationofstrips.Thepossibilityexists of using the approximate solution where element 3 is treated according to the approximation describedinSection2.2,Fig.2.2.4,butthisgivesasolutionwhichisunnecessarilyconservative.Based ontheshapeoftheelementsitishoweverpossibletoestimateasuitablemomentdistribution. Amomentdistributionbasedonthevaluesgivenaboveandtakingintoaccountthepointsofviewin Section6.1.3isproposedinFig.6.2.8.

6.3 Triangular loads

Example6.4 Fig.6.3.1showsawallintheshapeofatriangularslabwithonefixededge,onesimplysupportededge andonefreeedge.Thewallisacteduponbyanearthpressure,whichcanbeassumedtobezeroat thefreeedgeandtoincreaseby8kN/m2perverticalmetre.Thismeansthattheloadontheslabhasa linear variation in both the vertical and the horizontal directions, with a maximum intensity of attheinnercorner. The wall is reinforced for a positive moment with bars parallel to the free edge and for a negative momentatthefixedsupportwithbarsatrightanglestothatsupport.Theassumedlineofzeroshear forceisshowninthefigure.

78

 Fig.6.2.8 TheanalysisofthisslabfollowsexactlythesamelinesasExample6.2above.Theonlydifferenceis .WethususeEq.(2.18) thattheloadingcasesforbothelementscorrespondtoFig.2.3.11with togetherwith2.13,whichgivestheequations

(6.25)

(6.26) Wealsohavethegeometricalrelation

Fig.6.3.1

79

Fig.6.3.2 (6.27) and

(6.28) , wefind , .Adistributionofdesignmomentsisproposedin Choosing Fig. 6.3.2. As the load is zero along the free edge there is no reason to use a strong band of reinforcementalongtheedge,butontheotherhandthereinforcementshouldnotbereducedalong theedge.Besidethereinforcementwhichisdesignedforthesemomentssomeminimumtransverse bottomreinforcementmayberequiredbycodes,aswellasextratopreinforcementatthecorner.

6.4 Concentrated loads

Thedesignforaconcentratedloadonatriangularslabismadewiththesamemethodswhichareused forrectangularslabs,seeSections3.3and4.4.

CHAPTER7 Slabswithnonorthogonaledges
7.1 General
Thischaptertreatsslabswhichareneitherrectangularnortriangularandwherethecornersbetween edges have angles smaller than 180. Corners with an angle greater than 180 are called reentrant corners.SlabswithreentrantcornersaretreatedinChapter11. The general rules for analysis are the same as for rectangular and triangular slabs. Examples of analysesareonlygivenforuniformloads,assuchloadsarethemostcommon.Incaseswhereother typesofloadareactingonaslabmodificationstotaketheseintoaccountcanbebasedontherules andexamplesgivenforrectangularandtriangularslabs. Supportreinforcementismostefficientifitisarrangedatrightanglestothedirectionofthesupport. Thishasbeenassumedtoapplygenerally. Spanreinforcementisnormallyassumedtobearrangedintwoorthogonaldirections,paralleltothe chosen coordinate axes. In some cases one part of the span reinforcement is arranged in another direction.Thisappliesparticularlywithfreeedges,wheresomespanreinforcementisalwaysplaced alongthefreeedge. Asintheearlierchapterstheanalysisisbasedontheassumptionofapatternoflinesofzeroshear force. These lines should, in principle, be so arranged that the amount of reinforcement needed to carrytheloadisassmallaspossible.Thisgenerallymeansthatifthesupportconditionsarethesame (bothfreelysupportedorbothfixed)thedistancesfromalineofzeroshearforcetothetwonearest
80

supportsshouldbechosentobeapproximatelyequal.Wherethesupportconditionsaredifferentthe distancetothefixedsupportshouldbe1.5to2timesthedistancetothesimplesupport.Inadirection with longer spanreinforcement the distance shall be chosen to be shorter compared to a direction with a
shorter span reinforcement. The final choice is based on an estimate of the suitability of the ratio between differentmoments.

Bymeansofthelinesofzeroshearforcetheslabisdividedintoelements.Eachelementisboundedby a number of lines of zero shear force and one straight support. The equilibrium equation for each elementisestablished,e.g.bymeansofEq.(2.18). By means of the equilibrium equations and geometrical conditions the average moments can be calculated for an assumed pattern of lines of zero shear force. This pattern will often have to be modified a number of times before the ratio between the different moments is considered to be satisfactory. In the examples below only the final pattern, which is the result of such modifications, willbeshown. During the analyses the moments are generally treated as if they were uniformly distributed in the lateral direction. The final distribution of design moments is however chosen with respect to an estimated more correct theoretical distribution, with greater moments in the central parts. This procedureissatisfactoryfromthepointofviewofsafety.

7.2 Four straight edges

7.2.1 Alledgessupported
At each of the corners a line of zero shear force will start. There will usually also be one such line which does not start at a corner. The direction of this line is best chosen approximately in the directionofthebisectortothesupportsoftheelementswhichareseparatedbytheline.Inthisway thepatternoflinesofzeroshearforceisatthesametimeayieldlinepattern,whichshowsthatthe designisreasonablyeconomical. Example7.1 TheslabinFig.7.2.1hastwofixedandtwofreelysupportededges.Thetopreinforcementisatright angles to the supports and the bottom reinforcement is parallel to the x and yaxes. The load is 9 kN/m2. Astheslabhasaratherirregularshapethexandycoordinateshavebeengivenforthecornersand for the crossing points of the lines of zero shear force. The pattern of the lines of zero shear force follows the general rules given above. It is the result of a trial and error process until the ratios betweenthemomentsareestimatedtobeacceptable. Withthispatternoflinesofzeroshearforcewegetthefollowingcvalues: . Applyingequations(2.18),(2.4)and(2.6),andnotingthat ,

81

Fig.7.2.1 and (7.1) ,wegetthefollowingequilibriumequationsforelements1,2,3and4

(7.2) (7.3)

(7.4) . Solvingtheseequationswefind Inthiscasewefinduniquevaluesoftheaveragemomentswithagivenpatternoflinesofzeroshear force.Thisisbecausetherearetwofixededgesandthusfourmomentstodeter mine,equaltothenumberofequations.Iftherehadbeenthreefixededgesoneofthemomentvalues couldhavebeenchosenandiftherehadbeenfourfixededgestwoofthemomentvaluescouldhave beenchosen.If,ontheotherhand,fewerthantwoedgesarefixedthepatternoflinesofzeroshear forcehastobearrangedsothatallfourequationsaresatisfiedalthoughthenumberofunknownsis lessthanfour. AdistributionofdesignmomentsisproposedinFig.7.2.2.

Fig.7.2.2

7.2.2 Onefreeedge
The application to nonrectangular slabs with one free edge is a mixture of the principles demonstratedforrectangularslabswithonefreeedgeinChapter4andtheprinciplesshownabove forslabswithnonorthogonaledges. Dependingontheshapeoftheslabthechoiceofthemaindirectionsofbottomreinforcementandof thepatternoflinesofzeroshearforcewilldiffer.Iftheshapeisveryirregular,withallcornersnon orthogonal, the numerical analysis becomes laborious, even if in principle it is no more difficult. Thereforeonlyoneexamplewillbeshownwithaslabwhichisnottooirregular.
82

Example7.2 TheslabinFig.7.2.3hasthreefixededgesandonefreeedge.Thefixededgesformrightangles,but thefreeedgehasaskewdirection.Theloadis12kN/m2.

Fig.7.2.3 The main bottom reinforcement directions are parallel to the x and yaxes and the main support reinforcementisatrightanglestothesupports.Asupportbandisalsoarrangedalongthefreeedge. The analysis starts with the calculation of moments for the design of the main reinforcement. After thatthedesignofthesupportbandiscarriedout. Thefigureshowsapossiblepatternoflinesofzeroshearforceforthecalculationofdesignmoments forthemainreinforcement.Othershapesofthepatternarealsopossible,e.g.withelements2and4as trianglesand1and3astrapezoids,orwithallfourelementsastriangles,meetingatapoint.Thec values should be chosen so that the ratios between the different moments are estimated to be acceptable.Asuitablechoiceofvaluescanalsosimplifythenumericalanalyses.Thusinthisexampleit issuitabletochoose .Ofcoursethesumsofcvaluesintheydirectionhavetobeadaptedto thesizeoftheslab. Thefollowingnumericalvalueshavebeenusedintheanalysisbelow: . Applying equations (2.18), (2.4), (2.5) and (2.6) and noting that , we get the following equilibriumequationsforelements1,2,3and4 (7.5) (7.6) (7.7)

(7.8) Wehaveliveunknownmomentsbutonlyfourequations.Wethushavetochooseoneofthemoment values. It is in this case not appropriate to choose a value for myf or ms4, as the value of mxf is very sensitive to such a choice. The best value to choose is mxf. A suitable value may be , which . gives Thesupportbandalongthefreeedgehastocarrytheloadfromelement2.Thisloadisassumedtobe carriedatrightanglesintothesupportband.Themaximumloadperunitlengthofthesupportbandis
83

.Fig.7.2.4showstheloadonthesupportband.Themomentsinthebandcanbe determined withusualmethods. Wewillassume apointofzero shear forceat adistanceof3.80m from the lefthand end. With notation Mbf for the span moment and Mbs1 and Mbs3 for the support momentswegetthefollowingequilibriumequations: (7.9) (7.10) Ifwechoose wefind .

Fig.7.2.4 According to the recommendations in Section 2.8.3 the bottom reinforcement in the support band shouldbedistributedacrossawidthofabouthalftheaveragewidthofelement2. This average width is 0.7 m. A width of 0.4 m may therefore be chosen, which gives 60 kNm/m as designmomentforthebottomreinforcement. Thesupportmomentsinthesupportbandcorrespondtoareinforcementdirectionwhichisparallel to the direction of the free edge. This direction is not at right angles to the supports of the slab. Support reinforcement is, however, most efficient if it is arranged at right angles to a support, as expectedcrackswillbeparalleltothesupport.ItispossibletotakecareofthemomentsMbs1andMbs3 bymeansofreinforcementatrightanglestothesupport.Inthiscasethereinforcementsonlyneedto takethecomponents and respectively.Inthisexamplewechosetotake halfthesupportmomentsineachway.Thuswereinforcefor23.87and23.47respectivelyinthe direction of the edge and add 22.86 and 22.48 respectively to the support moments for reinforcement at right angles to the supports. The reinforcement in the direction of the edge is concentrated in a width of 0.4 m whereas the reinforcement at right angles to the edge is less concentrated,seetherecommendationsinSection2.8.3. Thetotalsupportmomentstobetakenbyreinforcementatrightanglestotheedgesareforsupport ,andforsupport .Thesetotalmomentshave tobedistributedwithrespecttotheiroriginswhichmeansthatthetotalmomentcontainsonepart fromtheelementandonepartfromthefree

Fig.7.2.5
84

edge; in approximate terms, the part from the band should be concentrated near the free edge. A suitabledistributionofdesignmomentsisproposedinFig.7.2.5. With the distribution shown, the concentration of reinforcement is higher at the upper end of the righthand support than at the upper end of the left hand support, as two layers of support reinforcement are crossing in the former case. This seems to be in agreement with the moment distributionwhichcanbeexpectedaccordingtothetheoryofelasticity.

7.2.3 Twooppositefreeedges
Theloadwill in principlebecarriedbystripsbetween the supported edges.This ismainlydoneby meansofonewaystrips,oftenwiththeaidofsupportbandsalongthefreeedges. Example7.3 OneofthesupportsfortheslabinFig.7.2.6isfixedandoneisasimplesupport.Theloadis7kN/m2. Thesizesoftheslabaregivenbymeansofcoordinates.

Fig.7.2.6 The load is assumed to be carried by reinforcement which is at right angles to the fixed support. Theoretically it would be slightly better to arrange the span reinforcement at right angles to the dividing line between elements 1 and 2, the line of zero shear force, but this would complicate the analysis and the gain is small. The line of zero shear force between elements 1 and 2 is chosen approximatelyasthebisectorbetweenthesupportdirections. Withthechosenshapesoftheelements,whichareshowninthefigure,wegettheaveragemoments inelements1and2bymeansofEq.(2.8) (7.11)

(7.12) andthus . Intheendpartsoftheslabtriangularelements36havebeenchosen.Otherpossibilitiesexist,which maybemoreefficient,butarealsomorecomplicated.Eq.(2.4)givesforthesetriangularelements (7.13) (7.14) andthus (7.15) .

85

Elements3and6aresupportedonsupportbandsalongthefreeedges.Astheelementsareofapure onewaytype,theloadisdirectlycarriedintheydirectionandgivestriangularloaddistributionson thebandswiththemaximumloadatoneend.Forthecalculationofmomentsinthesupportbands, ordinaryformulasforbeamsaccordingtothetheoryofelasticitycanbeusedinthiscase. Thesupportbandalongelement3carriesatotalloadof andhasaspanof6.40m. and a span moment of This gives a support moment of . The average width of the element is about 0.6 m and according to the recommendationsinSection2.8.3thespanreinforcementmaybedistributedoverawidthofabout 0.4m,asthiswidthisnotofgreatimportanceforthebehaviouroftheslab. Thesupportbandalongelement6carriesatotalloadof andhasaspanof3.52m. and a span moment of This gives a support moment of . The average width of the element is about 0.5 m and a corresponding distributionwidthofthespanreinforcementisabout0.4m. AdistributionofdesignmomentsbasedonthecalculatedvaluesisproposedinFig.7.2.7.Themain conditionisthatthetotaldesignmomentinasectionhastocorrespondtothesumofthecalculated moments. Thus, for example, the sum of span moments for the reinforcement in the ydirection is .Thelateraldistributionischosenwithrespecttothevariationof thedistancebetweenthesupportandthelineofzeroshearforce.

Fig.7.2.7 Atthelowerrighthandcornertwolayersoftopreinforcementhavebeenproposedwhichcrossover eachother,onelayeratrightanglestothesupportandonelayeralongthefreeedge.Thechoiceof this arrangement is influenced by the fact that a moment concentration can be expected at such a corner according to the theory of elasticity. The concentration of reinforcement prevents the formationoflargecracksatthisposition.

7.2.4 Twoadjacentfreeedges
IthasbeendemonstratedinSection5.2thatthedesignofarectangularslabwithtwoadjacentfree edgesmaybecomplicated.Ofcoursethedesignofaslabwithnonorthogonaledgesandtwoadjacent freeedgesisstillmorecomplicated.Solutionswiththetypesofelementwhichareusedinmostofthe othercasescannotbeappliedinthiscase.Wethereforehavetomakedirectuseofthebasicprinciples of the strip method with support bands in one direction supporting strips in other directions, as describedinSection2.8. Example7.4 TheslabinFig.7.2.8carriesaloadof7kN/m2.IthasonefixededgeAB,onefreelysupportededgeAD andtwofreeedgesBCandCD.

86

 Fig.7.2.8 AsupportbandisassumedtohaveitssupportsatBandD.AccordingtotherulesinSection2.8the supportbandisassumedtohavezerowidthwhenitactsassupportforcrossingstrips.Thesupport . The crossing strips, which are supported by the band follows a line with the equation supportband,areassumedtohaveadirectionatrightanglestothesupportband.Thisisoftenthe best assumption from the points of view of reinforcement economy and behaviour under service conditions. There may however be situations where another direction is preferable. This will be discussedattheendofthisexample. As one support is fixed, some top reinforcement should also be arranged at right angles to that support.Thisreinforcementcanbeassumedtocarryalltheloadwithinacertaindistancefromthe support.ItisassumedthatitcarriesalltheloadtotheleftofalinefromAtothepointontheedge withcoordinates(2.2/3.4).Thechoiceismadesuchthatthissupportmomentisofapproximatelythe same magnitude as the negative moment above the support band, which seems reasonable in this case. The line cuts the support band at point c with coordinates (1.88/2.50). The distance from (2.2/3.4)totheedgeABis1.99m.ApplicationofEq.(2.4)givesanaveragesupportmoment (7.16) Intheremainingpartoftheslabtheloadiscarriedbystripsatrightanglestothesupportband.Some typicalstripsarenowconsidered.

Fig.7.2.9 ,whichcutsthesupportbandatpoint ThestripthroughpointCisalinewiththeequation awithcoordinates(4.09/0.73).ThestripanditsloadareshowninFig.7.2.9a).Thesupportreactions actingatthesupportbandatpointaandatthesupportededgeare (7.17)

(7.18) Thereisthusanegativereactionatpoint(3.5/0),whichmeansthattheslabhastobeanchored. . This In the same way we can analyse a strip through point A, which has the equation stripcutsthesupportbandatpointbwithcoordinates(2.56/1.95)andtheedgeat(3.45/3.06).The stripanditsloadareshowninFig.7.2.9b).Wefindthesupportreactions
87

(7.19) (7.20) .This Wecancontinuetochoosethestripthroughpointc,whichhastheequation linecutsthefreeedgeat(2.53/3.31)andthefixededgeat(0.73/1.06).Thestripanditsloadisshown inFig.7.2.9c).ForsimplicityitisassumedthatthereisafreesupportatlineAB.Wefindthereaction atc (7.21) Atlastwelookatthestripthrough(2.2/3.4).Thisstripcutsthesupportbandat(1.63/2.69).Atthis pointtheloadonthesupportbandhasfallentozero. Fig.7.2.10illustratestheloadonthesupportband.Someadditionalpointshavebeencalculatedinthe samewayasaboveinordertobeabletodrawthecurve.Ontheotherhandaquitesatisfactoryresult wouldhavebeenachievedwithstraightlinesbetweenthecalculatedvaluesabove.

Fig.7.2.10 An analysis of the support band as a simply supported beam with the load according to Fig. 7.2.10 and . The maximum moment in the support band is gives the support reactions atadistanceof3.9mfromBand2.5mfromD. In order to determine a suitable width for the reinforcement band the recommendations in Section 2.8.3areused.Theaveragewidthofthepartsoftheslabwhicharesupportedbythesupportbandis notwelldefinedinthiscase,butitmaybeestimatedtobeabout2.5m.Asthesupportbandinthis case is of great importance for the behaviour of the slab a suitable width is . With this widthwefindthemaximummoment (7.22) As the slab is cantilevering outside the support band some minimum reinforcement for a negative moment must be introduced at the ends of the support band according to Eq. (2.36). At D the anglesaccordingtoFig.2.8.2are andweget (7.23) Thisisanegativemoment,requiringtopreinforcementwithinadistanceequalto2.5/3=0.83mfrom D. WealsohavetomakethesameanalysisforthesupportatB.Therewefind and and (7.24) The bending moments in the strips through a, b and c can easily be calculated from Fig. 7.2.9. The resultsare17.9,7.1and3.8,respectively.Thetotalmomentinthestripsrestingonthesupportband ismosteasilycalculatedbymeansofEq.(2.6),usingnumbersfromFigs7.2.9and7.2.10andnoting thatthedistancefromdto(2.2/3.4)is0.91m.

88

(7.25) Fig.7.2.11showsaproposeddistributionofnegativedesignmoments.Allthereinforcementatright anglestothesupportbandshouldbecarriedtotheedges.Thereinforcementforthesupportmoment 3.1 close to corner B can take a moment across the support band equal to , where 37.3istheanglebetweenedgeABandthesupportband.Thuscondition(7.24)isfulfilled.

Fig.7.2.11 across a width of 1.25 m The support band should be designed for a positive moment symmetrical to the support band. All this reinforcement should be carried to the edges and be well anchored. In addition to the band reinforcement, minimum bottom reinforcement should be introduced, preferably parallel to the band, and possibly also some bottom reinforcement at right angles. Asthereinforcementinthesupportbandissomuchheavierthantherestofthereinforcementitmay bequestionedwhetherthisisasuitabledesign.Couldwegetamoreevenreinforcementdistribution by using some other design method? If, instead, we apply the theory of elasticity in a theoretically correct way (e.g. using elements which are small enough in the finite element method), we would presumably geta still highermaximum momentclosetocorner D. Toapplytheyieldlinetheoryto thisslabinacorrectwayisextremelydifficult,aschecksofmanydifferentyieldlinepatternswould havetobecarriedout,includingfanshapedlocalpatterns,forinstanceinthevicinityofcornerD. Infactitisquestionablewhetherthisisasuitablewayofmakingtheslab,orwhetheritwouldnotbe bettertousearealbeaminsteadofasupportband. Insteadofarrangingthetopreinforcementatrightanglestothesupportbanditmaybearrangedat another angle. If, for example, a slope of 1/1.65 is used instead of 1/1.25 there will be no negative reaction along edge AB and the load on the support band will be lower. This would lead to smaller momentsinthesupportbandandacorrespondingreductionofreinforcement.Ontheotherhandthe reinforcementcrossingthebandwillincrease.Dependingontheanglesbetweentheedgesoftheslab thistypeofsolutionmaysometimesbethemostsuitable.

7.3 Othercases
7.3.1 Circularslabswithauniformload
Circular slabs with polar symmetric load and support can easily be analysed with the equilibrium equationexpressedinpolarcoordinates.SuchanalysesareshowninStripMethodofDesign,Chapter 3.Theresultsarepresentedasradialandtangentialmoments.Inpracticeitiseasiertoarrangethe bottomreinforcementasarectangularmeshofintwodirectionsatrightanglestoeachother.Itcan be demonstrated that the bottom reinforcement in both directions should then be designed for the tangentialmoment. Forasimplysupportedcircularslabwithradiusrandauniformloadq,reinforcedintwodirectionsat rightangles,theaveragemomentsaccordingtothisanalysisare
89

(7.26) Thecorrespondingreinforcementcanbeevenlydistributedandshouldinthatcasebecarriedtothe support.Itmay,however,bebettertoconcentrateitsomewhattowardsthecentre,e.g.asshownin Fig.7.3.1.

Fig.7.3.1 TheslabcanalsobetreatedbydirectuseofthesimplestripmethodaccordingtoFig.7.3.2,wherethe load is carried to the nearest support in one of the two reinforcement directions. As the support is curvedtheaveragemomentismostsimplycalculatedbymeansofanumericalintegration.Theresult shows an average moment which is about 16% lower than the value above, and thus theoretically moreeconomical.ThecorrespondingmomentdistributionisalsoshowninFig.7.3.2andcomparedto asolutionaccordingtothetheoryofelasticity(curvemarkedel.).Itcanbeseenthat,inthiscase,the momentdistributiondeterminedfromthedirectuseofthesimplestripmethodisnotverysuitable withrespecttoserviceconditions,asitdoesnotgiveanydesignreinforcementacrosstheouterpart of the diameter. The distribution of reaction force along the edge is very uneven according to the simplestripsolutionwithzeroshearforceinthe45direction.Thisalsoshowsthatthesolutionisfar from the reality of the service state. It is therefore recommended that values should be chosen according to the equation above and a distribution of design moments which is approximately accordingtoFig.7.3.1orisconstant. If the edges of the slab are fixed the support reinforcement is best arranged at right angles to the support,i.e.inaradialarrangement.Inthenormalcasethetopreinforcementisthusradialandthe bottomreinforcementisanorthogonalnet. Ifthesupportmomentismsthefollowingrelationfortheaveragedesignmomentsisvalidaccording tothepolarsymmetricsolution:

Fig.7.3.2

(7.27) Itmaybenotedthatthisvalueisthesameasforacircumscribedsquare,seeFig.7.3.3,wherecforthe elementsinthesquarecorrespondstorforthecircle.Thevalidformulafortheelementsis(2.4).Thus theaveragemomentsinacircularslabcanbetakenasthemomentsinacircumscribedsquareslab withthesidesparalleltothedirectionsofthebottomreinforcement.

90

Fig.7.3.3

7.3.2 Generalcasewithalledgessupported
Whereaslabissupportedaroundallitsedgesitisalwayspossibletoapplythesimplestripmethod, dividing the slab into narrow strips in the reinforcement directions, although the numerical computationsmaybecomelengthy.Alsotheresultmaybeadesignwhichisnotappropriateforthe behaviour in the service state, as the design moment may be zero within some parts where appreciable moments occur under service conditions. Some minimum reinforcement then has to be added. Onewayofsimplifyingtheanalysismaybetoapplythefollowinggeneralrule: It is always safe to design the slab for the moments in a circumscribed slab with the same bottom reinforcement directions and with all support reinforcement at right angles to the supports in both slabs. Ifthisruleisapplied,thereinforcementcanbecurtailedaccordingtotherulesgiveninSection2.9.1, providedthatthedistancetothesupportistakenasthedistancetotheedgeofthecircumscribedslab for the bottom reinforcement andto the edge of the real slab for the support reinforcement. The c valueistakenfromthecircumscribedslab. Example7.5 The elliptical slab in the upper part of Fig. 7.3.4 has fixed edges and carries a load of 9 kN/m2. The designmomentsarefirstdeterminedfromthecircumscribedrectangularslabinthelowerpartofthe figure.Thechoiceoflinesofzeroshearforceisshown.BasedontheselinesandEqs.(2.5)and(2.4) theaveragedesignmomentsare: (7.28)

(7.29) We can choose . A distribution of design moments based on these values is proposed in Fig. 7.3.5. The limit between the parts where mxs and mys are active is wherethecorrespondinglineofzeroshearforcecutstheellipseinFig.7.3.4.Thedistributionofmys mighthavebeenchosenwithsomeconcentrationtowardsthecentre,asthesupportmomentsinthe servicestatearehighestthere. Thiswayofusingthemomentsfromthecircumscribedrectanglegivesverysimplecalculationsanda conservative result. On the other hand the design may be regarded as too conservative and thus uneconomical. In order to get a more economical design a direct application of the strip method is possible, dividing the slab into narrow strips in the xand ydirections for span moments and cantileveringstripsatrightanglestothesupportforthesupportmoments.Wherethespanstripsrest ontheendsofthecantileversdueregardhasto

91

 Fig.7.3.4 be given to the difference in directions according to the rules in Section 2.3.6. Such an analysis is lengthyandtimeconsumingandwillnotbeshownhere. However, there also exists the possibility of circumscribing the ellipse with a polygon instead of a rectangle.Thisgivesresultswhicharenotsoconservativeandyetthecalculationsarenottoolengthy or complicated. Fig. 7.3.6 shows the ellipse circumscribed by a polygon shapedby introducing lines withtheslopes0.5,whicharetangentstotheellipse.Onlyonequarteroftheslabisshown,asitis symmetricalwithrespecttoboththexandyaxes.

Fig.7.3.5

92

Fig.7.3.6 Theassumedlinesofzeroshearforceareshowninthefigure,aswellastheassumeddistributionof spandesignmoments.Notethatmxfandmyfinthiscasedonotmeanaveragemomentsbutmoments intherelevantparts. Forelements1and3weget,bymeansofEqs(2.4)and(2.5), (7.30)

(7.31) Forelement2wecanapplyEq.(2.18)togetherwith(2.6).Thedirectionofedge2cutsthexaxisat 5.66.ThefollowingvaluesareintroducedintoEq.(2.6): ; We also have to introduce the average span moments on the width corresponding to edge 2. This givesusthefollowingrelation:

(7.32) Introducing

,therelationcanbewritten

(7.33) Ifwechoose and weget ofdesignmomentsisshowninFig.7.3.7.

.Thecorrespondingdistribution

Fig.7.3.7 This analysis reduces the total amount of reinforcement by about 14%, compared with the analysis basedonacircumscribedrectangle.Italsogivesamoresatisfactoryreinforcementdistribution Thedistributionofsupportmomentsandtheratiobetweenmomentsinthexandydirectionscanbe changedbychangingtheassumedpointswherethelinesofzeroshearforcemeetthexaxis. Example7.6 TheslabinFig.7.3.8hasthreefixedandtwofreelysupportededgesandauniformloadof7kN/m2.It is to be reinforced with bottom reinforcement in the directions of the coordinate axes and support reinforcementatrightanglestothesupports.

93

 Fig.7.3.8 Threedifferentwaysofperformingthedesignwillbediscussed. Thefirstmethodistousethemomentsfromthecircumscribedrectangle.Thismethodisbyfarthe simplest,butitmayleadtoaverycoservative,i.e.uneconomical,solution. ThecircumscribedrectangularslabisshowninFig.7.3.9.Inorderthatthisslabshallcorrespondto therealslabasfaraspossiblethetwolongedgesareassumedtobefixedonlyonthepartsadjacentto thefixededgesintherealslab.Assumedlinesofzeroshearforcearealsoshowninthefigure.Based on these lines we get the following relations from Eqs. (2.4) and (2.5). The moments ms are the averagemomentsonthefixedpartsoftheedges.

Fig.7.3.9 (7.34) (7.35) (7.36)

(7.37) andifwechoose wefind and . Weget Support BC in the real slab corresponds partly to edge 1 and partly to edge 2 in the circumscribed rectangularslab.Thecorrespondingpartsdependonwherethelineofzeroshearforcecutstheedge.
94

Wethenfindthat32%correspondstoedge1and68%toedge2.Ifweusethesepercentageswefind theaveragesupportmoment8.22. Asusualasuitabledistributionofthedifferentaveragemomentsshouldbechosen. Thissimpledesignisonthesafeside,butprobablytoomuchso.Thereforeanalternative,butmore laborious,solutionwillbedemonstrated,basedontheanalysisofequilibriumconditionsforelements, eachofwhichhasonesidealonganedgeoftheslab.Aconfigurationoflinesofzeroshearforce,which dividestheslabintosuchelements,isshowninFig.7.3.10.

Fig.7.3.10 In determining the pattern of lines of zero shear force some general rules may be followed. The elementsattheshorteredgesareusuallytriangles.Thelinesbetweentheelementsatthelongersides runapproximatelyinadirectiontowardsthepointwherethedirectionsoftheseedgesmeet.Inthis waythe generalshapemaybedetermined. Forthecompletedeterminationof theshapeatrialand errorprocedurehastobeusedinordertogetsuitablerelationsbetweenthedifferentmoments.In thisprocedurethepositionsofpointsa,bandcarethereforevaried.Oftenitissufficienttoexactly fulfil the equilibrium conditions only for the largest elements but a check should be made that the designmomentsarelargeenoughtomaintainequilibriuminthesmallerelements.Thisisonthesafe side,butwithonlyasmallinfluenceontheeconomyofthedesign. Before starting the equilibrium analyses the distribution of the span design moments should be decided,asthisdistributioninfluencestheequilibriumconditions.Anassumeddistributionhasbeen giveninthefigure. Inthiscasesuitablechoicesarethefollowingcoordinatesforthepoints: Thecompleteanalysiswillnotbeshown,asitisratherlaborious,butasanexampletheequilibrium condition for element 2 will be given. It is based on Eq. (2.18) combined with (2.6). The geometric valuestobeusedinEq.(2.6)arethefollowing: Using these values and the relevant average span moments we get the equilibrium condition for element2

(7.38) or,with

(7.39) Writing the corresponding equations for the other four elements and choosing values of the span momentswecanfindthefollowingmomentvalues:
95

 When we compare these values with the previous solution we find that they are on average about 20%lower.Wecanthus,inthiscase,saveabout20%reinforcementbyusingthismorecomplicated solution. This comparison must be take into account the fact that the latter solution may involve severalhoursworkbeforea suitable patternoflinesof zeroshearforceisfoundfromthetrialand error process, whereas the first solution can be made in a couple of minutes. The cost of reinforcementhastobecomparedtothecostoftime. A third possibility is a direct application of the principles of the simple strip method. Fig. 7.3.11 illustrates how this may be performed. The slab is divided into zones with different loadbearing directions.Thedashedlinesareinthiscaselinesofzeromoments.Thezonesbythefixedsupports arezoneswherethesupportreinforcementisactive.Thesesupportstripsactascantileverswithzero momentattheends.Intheremainingpartstheloadiscarriedbyspanstripsparalleltothebottom reinforcement, in the direction of the coordinate axes. The load within a certain part may also be dividedbetweenthetworeinforcementdirections.Thespanstripsrestonthecantileversandonthe supports.Themomentsare calculated innarrow stripsatsuitabledistances,whichgivesamoment distributionandanaveragemomentofeachtype.Asuitabledistributionoftheaveragemomentsis usedasthebasisfordesign.Theratiosbetweendifferentaveragemomentsdependonhowthelimits arechosenbetweendifferentparts. Where a span strip rests on a support strip due regard has to be paid to the change in direction accordingtotherulesinSection2.3.6.

Fig.7.3.11 WiththestripdirectionsinFig.7.3.11therewillbenodesignreinforcementintheydirectionnear the righthand side of the slab. Some minimum reinforcement can be arranged there, but it is also possibletogetdesignreinforcementbydividingtheloadinthisregionbetweenthetwodirections. This type of analysis is rather timeconsuming, as it takes time to analyse a number of strips. The analysis may also have to be repeated with other limits between the different zones, if the ratios betweendifferentmomentsprovetobeunsuitable.

7.3.3 Generalcasewithonestraightfreeedge
Whereonestraightedgeisfreeasupportbandhastobeassumedalongthisedge.Theanalysismay be based on a combination of the second approach in the example above and the method demonstratedinSection7.2.2.Asanalternative,adirectapplicationofthesimplestripmethodmay beusedasdiscussedaboveanddemonstratedinSection7.2.4.

7.3.4 Generalcasewithtwoormorefreeedges
This complicated case has to be analysed by means of the direct use of the simple strip method, includingsupportbands,inprincipleasdemonstratedinSection7.2.4.
96

CHAPTER8 Regularflatslabswithuniformloads
8.1 General
8.1.1 Definitionofregular
Inthischapterthewordregularmeansthatthesupportsformanorthogonalnetandthatallinterior supports are columns. The exterior supports may be walls or columns and the slab may cantilever outsidetheexteriorsupports.Thereshouldbenomajoropeningsintheslab. The advanced strip method was developed in the first place for the design of the regular flat slabs treatedinthischapter.Providedthatcertainapproximationsareaccepteditcan,however,beapplied toamuchwidergroupofslabs,treatedinotherchapters.

8.1.2 Droppanelsandcolumncapitals
Adroppanel,Fig.8.1.1,meansthattheslabisgivenagreaterdepthinanareaaroundacolumn.In this way the slab becomes stronger in the part where the moments and shear forces are largest, withoutmuchincreaseinthemomentscausedbythedeadweight. Wheredroppanelsareused,thewidthsofthecolumnstripsarebestchosenequaltothewidthsofthe droppanels,atleastregardingsupportreinforcementatthecolumns.Thewidthsofthecolumnstrips maybechosenlargerforthespanreinforcement. Asadroppanelincreasesthestiffnessoftheslabatthesupportitmaybeappropriatetoincreasethe supportmomentsomewhatcomparedtothecasewithoutadroppanel.

Fig.8.1.1 A column capital, Fig. 8.1.1, is intended to increase the punching strength and at the same time decreasethebendingmomentsintheslab. Acolumncapitaldecreasestheratiobetweensupportandspanmomentsaccordingtothetheoryof elasticity,comparedtotheratiowithoutacapital.

8.1.3 Determinationofspan
The main type of element in a flat slab is the cornersupported element, supported on the circumferenceofacolumnoracolumncapital.Theoreticallythesupportareaisinfinitelysmallwhich
97

givesan infinitelyhighstress.Inpractice,ofcourse,thesupportareaandsupportstresshave to be finite.Thismaybeinterpretedsothatthereactionforceactssomewhereinsidethecolumnandnot atsomepointatthecircumference. Ifwelookattheultimatelimitstate,acrackwillappearatthecircumferenceofaconcretecolumn, wherethusthecriticalsectionis.Forthedeterminationofdesignmomentsthespancanthereforebe assumedtobetakentothecircumferenceofthecolumn,andthetheoreticalpointofsupportforthe cornersupported element at that circumference. If there is a sufficiently strong column capital the pointofsupportmaybeassumedtobeatitscircumference. If we accept that a point on the circumference is chosen as the theoretical point of support, from which the span is calculated, the same point is, of course, valid for both directions of the element. Thus,ifwehavecircularcolumnswithradiusratequalcentreslcinbothdirections,thespansmaybe ,providedthatthepointsofsupportaretakeninthe45directionfromthecolumn takenas centres,seeFigs8.1.2and8.1.3.Bychangingtheassumedpointofsupportthespancanbesomewhat decreasedinonedirectionbutthenithastobeincreasedintheotherdirection. Itisimportanttonotethatthespanmaynotbetakenasthecleardistancel2r,whichisusedinsome codes. It can easily be demonstrated by means of the yield line theory that such an assumption is theoreticallyunsafe.

Fig.8.1.2 Inanalysisbythestripmethodthereisaonewaystripbetween,andactingtogetherwith,thecorner supportedelements,seeFig.8.1.2.Wherethecolumnsarecircularthespanofsuchaonewaystripis alittleshorterthanthatofthecornersupportedelements,closertol2r.Astheonewaystripsareof aminorimportanceforthetotalamountofreinforcement,theirspanscanbetakentocoincidewith thoseofthecornersupportedelements. Insummary,thefollowingruleisusedforthedeterminationofthespanwheretheslabissupported onaconcretecolumn: Thespanofastriprestingonaconcretecolumnismeasuredfromthesideofaninscribedrectangle withedgesparalleltothereinforcement. Ifthecolumnismadeofsomeothermaterial,e.g.masonry,thespanhastobemadealittlelonger.

8.1.4 Calculationofaveragedesignmoments
The elements are combined to form strips in the way which is exemplified in Fig. 8.1.3. The whole widthwxcanbetreatedasonewidestripinthexdirectionandthewholewidthwyasonewidestrip intheydirection.Thewholestaticanalysisislimitedtotheanalysisofthesetwostripsinthesame wayasacontinuousbeamisanalysed.Thisanalysisgivestheaveragedesignmoments. When the average design moments are known, the positions of the lines of zero shear force (maximummoments)canbecalculated.Thusthemomentsinthexdirectiondeterminethewidthsof theindividualstripsandelementsintheydirectionandviceversa.

98

important to minimize cracks on the bottom side the support moment can be somewhat decreased. Such a changeshouldnotbegreaterthanabout15%.Withinthelimitsgivenbytheserecommendationstheinfluence oftheratiobetweensupportandspanmomentshasnonoticeableinfluenceondeflectionsintheservicestate.

Fig.8.1.3 The determination of support moments in the strips can be based on the theory of elasticity and calculatedby ordinary methods for continuous beams. It is, however, not necessary to calculate the supportmomentsaccuratelyaccordingtothetheoryofelasticity,althoughtheratiobetweensupport and span moments should not deviate too much from that derived from the theory. Sufficient agreementwiththetheoryofelasticityisachievedifthefollowingrecommendationsarefollowed.See alsoSection1.5. Firstsupportmomentsarecalculatedforeachspanaccordingtothetheoryofelasticityassumingthat itisfixedatinteriorcolumnsupports,i.e.ql2/12iftheoppositeendisfixedandql2/8iftheopposite endis freely supported.Iftheoppositeendiscontinuouswithawall anintermediatevaluemaybe used,e.g.ql2/10.Ifthereisadroppanelwhichincreasesthestiffnessoftheslaboverthecolumnthe supportmomentsmaybeincreased.Ifthesupportfortheslabiswide,e.g.whereacolumncapitalis used,thesupportmomentsmaybedecreased.Thedesignsupportmomentistakenastheaverageof themomentsfromthespansmeetingatthesupport,providedthatnomomentisassumedtobetaken bythecolumn.Startingfromthisaveragevaluesomemodificationcanbemade.Ifitisimportantto minimizecracksonthetopsideoftheslabthesupportmomentcanbesomewhatincreasedandititis

Regard should be given to moment transfer between the slab and the columns, where this transfer maybeofimportanceforthebehaviorofthestructure.Forinteriorcolumnsthemomenttransfermay often be disregarded, but for exterior columns this is as a rule not to be recommended. How these momentsarecalculatedisoutsidethescopeofthestripmethod.Forthesakeofsimplicitytheyare thereforedisregardedintheexamples,withtheexceptionofExample8.4.Thisdoesnotmeanthatitis generallyrecommendedtodisregardthemoments.Relevantcoderuleshavetobefollowed.

8.1.5 Lateraldistributionofreinforcement
The strip method in itself only gives some limits for the distribution of design moments across the widthofcornersupportedelements,seeSection2.5.2.Someadditionalrulesmaybeneededforcrack controlunderserviceconditions. Thesimplestpossiblemomentdistributioniswithaconstantspanmomentoverthewholewidthand the support moment only within a limited width (the column strip) over the column, leaving the middle strip between the columns without top reinforcement. This choice leads to simple drawings andsimpleconstructionandisthusfavourablefromthepointofviewofeconomy.Itisalsosuitable
99

for limitation of deflections. It is, however, not the best design for crack limitation. It may lead to visuallyunacceptabletopcracksbetweenthecolumnsandisonlyrecommendedwheresuchcracks willbecoveredbysomefloorfinish. For the best possible crack control the design moment (reinforcement) should be distributed with someregardtothemomentdistributionaccordingtothetheoryofelasticity.Wherecrackcontrolon thetopsurfaceisessentialthedesignsupportmomentinthemiddlestripshouldbechosentobe30 50% of the average support moment. Where crack control is essential on the bottom surface the designspanmomentinthecolumnstripmaybeincreasedbyabout20%.Inbothcasestheaverage momentsarekeptunchanged. The theoretical moments are different in the cornersupported elements and in the oneway strip betweenthem,asthereisauniformmomentinthelatterelement.Whenlateralmomentdistributions arediscussedhalftheonewaystripisassumedtobelongtoeachofthecornersupportedelementsin ordertosimplifythediscussion. Thewidthofthecolumnstripmaygenerallybechosenashalfthetotalwidthoftheindividualstrip. (An individual strip is a strip with a width equal to the sum of the widths of two cornersupported elements supported on the same column and the oneway strip in between, see Fig. 8.1.3). Where thereisadifferenceinspanbetweenthexandydirectionsitmaybeappropriatetochooseasmaller widthforthecolumnstripintheshorterdirection,e.g.sothatithasthesamewidthinbothdirections. Thewidthofthecolumnstripshouldneverbechosenlargerthanhalfthetotalwidthoftheindividual strip. Where a moment transfer between a column and a slab is taken into account in the design, the reinforcementhastobearrangedinsuchawaythatthismomenttransfercantakeplace. Inthevicinityofanexteriorcolumnthereshouldbesometorsionalreinforcementalongthefreeedge oftheslabinordertolimittorsionalcracking. TypicalrecommendedlateralmomentdistributionsareillustratedinFig.8.1.4.Distributions1isthe normally recommended distribution of the support moment where limitation of cracks on the top surfaceisnotimportant,whereass2showsanexampleofasuitabledistributionforcracklimitation on the top surface. Distributions s3 and s4 are examples of corresponding distributions of support momentsintheshorterdirectionwherethespansaredifferentinthetwodirections.Distributionf1is the normally recommended even distribution of span design moments, whereas f2 may be used wherecracklimitationonthebottomsurfaceisparticularlyimportant. InFig.8.1.4thecolumnstripisshownwiththeeccentricitywhichresultsfromtheapplicationofthe cornersupported element. In practice the reinforcement may be placed centrally over an interior column. ThechosendistributionsshouldbecheckedtoensurethattheyfulfiltheconditioninEq.(2.24).This conditionisautomaticallyfulfilledifdistributions1ors2oranintermediatebetweenthemisusedfor thesupportmomentandf1forthespanmoment,andtheratiobetweensupportandspanmomentis between1and3,whichisthemostcommoncase. Wherethereisnosupportmoment,thespanmomentmustbemoreconcentratedinordertofulfilthe conditionaccordingtoEq.(2.24),whichmeansthatthemomentinthemiddlestripmustnotbelarger than0.7mf.Thismaybethecasewherethereisnomomenttransferfromanexteriorcolumn. Where there is no span moment, i.e. where a slab is cantilevering outside an exterior column, the support moment must be distributed so that the condition according to Eq. (2.24) is fulfilled. Distributionss2ands4areonthesafesidewithrespecttothiscondition.

8.1.6 Summaryofthedesignprocedure
The determination of the design moments according to the advanced strip method is made in the followingsteps: 1.Determinethelengthofthespansasdescribedin8.1.3. 2.Determinetheaveragesupportmomentsasdescribedin8.1.4.
100

3.CalculatethecvaluesbymeansofEq.(2.34)and,fromthese,calculatethewidthsoftheindividual strips. 4.CalculatetheaveragespanmomentsbymeansofEq.(2.35). 5.Chooselateralmomentdistributionsaccordingtotherecommendationsin8.1.5.CheckagainstEq. (2.24)ifthisisthoughttobenecessary.

Fig.8.1.4 6. Use these design moments for the design of the reinforcement. The design moments in the triangularelementsinthecorneroftheslab(seeFig.8.1.3)aretakenasonethirdofthemomentsin theparallelpartsofthestrips.Wherethereareexteriorcolumnsarrangetorsionalreinforcementat thefreeedge. 7.DeterminethelengthofreinforcingbarsaccordingtotherulesinSection2.10. 8. Calculate the support reactions at the columns and check for punching according to rules in handbooksorcodes.Thesupportreactionisequaltotheloadinsidethelinesofzeroshearforcein thespanssurroundingthecolumn. The examples will generally only show the application of points 14 above, as the choice of lateral momentdistributiondependsontherequirementforcrackcontrolintheactualsituation.

8.2 Exteriorwallorbeamsupports
8.2.1 Onesingleinteriorcolumn
Whereasmanymethodsforthedesignofflatslabsareonlyvalidifthereisatleastacertainnumberof spansineachdirection,thestripmethodisnotlimitedbyanysuchrule.Itmaythusbeappliedtoa slabwhichissupportedonwallsorbeamsallaroundandononesingleinteriorcolumn.Thiscaseis
101

easilytreatedwiththestripmethodalthoughitisrathercomplicatedwithothermethods,suchasthe yieldlinetheory,atleastiftheslabisnotsymmetricalwithrespecttothecolumn. Example8.1 The slab in Fig. 8.2.1 carries a uniform load of 8 kN/m2. Support A can be assumed to be fixed, whereas support D is only partly restrained and can be assumed to have a support moment corresponding to half the fixed end moment. The other two supports are freely supported. It is assumedthatnomomentistransferredfromthecolumntotheslab,i.e.thesupportmomentsonboth sidesofthecolumnarethesame. Thesupportmomentsareestimatedtobe (8.1) (8.2) (8.3)

Fig.8.2.1

(8.4) Ifwewish,wemaychoosesupportmomentswhicharesomewhatlargerorsmaller,butinthiscase wejustchoosethecalculatedvalues.WecannowcalculatethecvaluesfromEq.(2.34): (8.5)

102

(8.6) (8.7)

(8.8) WecannowcalculatetheaveragespanmomentsfromEq.(2.35): (8.9) (8.10) (8.11)

(8.12) Providedthatweusemomentdistributionschosenfroms1,s2andf1(correspondingto )inFig. 8.1.4wemaychoosethedistributionofsupportmomentsarbitrarilyexceptwithregardtospanAE, wheretheratiobetweensupportandspanmomentis .Ifwechoosedistributions1for ,whichistoolowaccordingtoEq.(2.25).Wehave thesupportmomentwefind assupportmomentinthemiddlestrip. totakeatleast AsthespanmomentinAEcomesoutsolowitmighthavebeenbettertochoosealittlelowervalueof the support moment at A, e.g. 12. The difference in behaviour of the slab with such a change is probablyquiteinsignificant. Thesupportreactionatthecolumnis .

8.2.2 Morethanoneinteriorcolumn
Inaregularflatslaballcolumnsshouldbesituatedatthecrossingpointsoflinesparalleltotheedges. Thedesignprocedurewillbethesameregardlessofthenumberofcolumns.Itwillbedemonstrated onaslabwithonlytwointeriorcolumns. Example8.2 TheslabinFig.8.2.2carriesaloadof11kN/m2.SupportsAandBarefixed,CandDfreelysupported. Thecolumnshavestrongcolumncapitalswithadiameterof2.0m.

103

 Fig.8.2.2 Thespansaredeterminedasiftheinteriorsupportsconsistofrectanglesinscribedintothecircular columncapitals.Aasquareisusuallychosen,andthiswillbedonehere.Aslightlybetterchoicewould havebeentousearectangle,forexample,withlength1.6minthexdirectionandwidth1.2minthe ydirection. This would have decreased the reinforcement in the xdirection by a few percent and increaseditintheydirectionbyafewpercentandgivenatotaldecreaseinreinforcementofbetween 1and2percent,whichisinsignificant. Withtheinscribedsquareasthesupportwegetthespansgiveninthefigure.Withthesespanswecan calculatetheapproximatemagnitudesofthesupportmomentsandchoosethedesignvaluesofthese moments: (8.13) (8.14) (8.15) (8.16)

(8.17) Thevaluesofsupportmomentsatthecolumnshavebeenchosensomewhatlowerthanthecalculated valuesbecauseofthelargecolumncapitals.ThevalueofmAhasbeenchosenalittlehigherthanthe calculated value because the moment at E is a little lower than the value for a fixed support. The choiceofvaluesmaybediscussed,butsmallvariationswillnotinfluencethebehaviouroftheslab. ThecvaluesarecalculatedfromEq.(2.34)andthecorrespondingspanmomentsfromEq.(235): (8.18) (8.19) (8.20)
104

(8.21) (8.22) (8.23) (8.24) (8.25) (8.26)

(8.27) In this case the ratios between support and span moments are such that the distribution of design moments can be chosen rather freely. As the oneway elements between the cornersupported elementsareratherwideitisacceptabletouseawidthofthecolumnstripwhichiswiderthanhalf thewidthoftheindividualstrip.Asthespansintheydirectionaremuchsmallerthanthespansinthe xdirectiontherelativewidthofthecolumnstripintheydirectionmaybechosensmaller,e.g.with thesamewidthasinthexdirection.

8.3 Exteriorcolumnsupports
8.3.1 General
In this section the emphasis is on the design of a slab where the edge of the slab is supported by columns.Inordertosimplifytheexamples,slabsareconsideredwithoutanyinteriorcolumn,asitis notdifficulttocombinethedesignproceduredemonstratedbelowwiththeprocedureabove. Whereitisassumedthatnomomenttransfertakesplacebetweentheexteriorcolumnandtheslabit muststillberememberedthatmostofthesupportreactionisassumedtoactatthecolumnface.This eccentricitycausesamomentinthecolumn.

8.3.2 Columnsupportatoneedge
Example8.3 The slab in Fig. 8.3.1 carries a load of 9 kN/m2. All supported edges are freely supported. It is also assumedthatthereisnomomenttransfertothecolumn. TheonlysupportmomentisthemomentmxBabovethecolumn.Thevalueisestimatedfrom (8.28)

105

Fig.8.3.1 Wemaychoose moments: (8.29) (8.30) (8.31) (8.32) (8.33)

.FromEqs.(2.34)and(2.35)wenowcalculatethecvaluesandthespan

(8.34) The ratio between support and span moments permits a rather arbitrary distribution of reinforcementinthexdirectionofthetypesshowninFig.8.1.4.Itisacceptabletoconcentratesome reinforcementtothevicinityofthefreeedge(withinthelimitsgivenbyEq.(2.24)),particularlysome span reinforcement. On the other hand the oneway element theoretically requires an evenly distributeddesignmoment,sothespanmomentshouldnotbetoounevenlydistributed. Intheydirectionthereisnosupportmoment,whichmeansthattheconditioninEq.(2.24)hastobe fulfilledbythespanreinforcementalone.Thusthedesignspanreinforcementinthemiddlestripmay notbelargerthan0.7mDEandtheremainingpartmustbeconcentratedinthecolumnstrip.Thisof coursemeansthatthereinforcementisnotevenlydistributedintheonewayelementatedgeD,but thisisofnopracticalimportance. Thereshouldbesomebarsalongthefreeedgeandsomestirrupsinordertoavoidtorsionalcracks. Anexampleofasuitabledistributionofthemomentsinthelefthandhalfoftheslabcanbeseenin Fig.8.3.3. Example8.4 The slab and load are the same as in the preceding example, but with a moment transfer to the column.Itisassumedthattheaveragemomentis8.0kNm/matthesectionthroughtheinneredge of the column, which gives a total moment of . How this moment is calculatedorestimatedisnotdiscussedhere.
106

Theanalysisisexactlyasaboveregardingthexdirection.Fortheydirectionweget: (8.35)

(8.36) It is recommended that the top reinforcement corresponding to the moment in the column is concentratedinasmallwidth.Ithastobeanchoredinsuchawaythatthecontinuityissecured.As thesupportmomentisrathersmalltheconditioninEq.(2.24)isnotfulfilledifthespanreinforcement is evenly distributed. The span design moment for the middle strip may not be larger than andtheremainingparthastobeconcentratedinacolumnstripofasuitablewidth. Thereshouldbesomebarsalongthefreeedgeandsomestirrupsinordertoavoidtorsionalcracks.

8.3.3 Columnsupportatacorner
Example8.5 Fig. 8.3.2 shows a slab which is similar to the slab in the preceding two examples except that the supportingwallCisexchangedforacolumnattheoutercorner.Ifweassumethattheslabisfreely supportedatthesideofthecolumnsthecvaluesandtheaveragemomentsareexactlythesameasin Example8.3.Theonlydifferenceisthedistributionofdesignmoments.

Fig.8.3.2 As the cornersupported element at the corner column has no support moment the span reinforcementhastobedistributedinsuchawaythatcondition(2.24)isfulfilled.Thismeansthatnot more than0.7mBC may be taken in the middle strip.In the xdirection it is, however, satisfactoryto takelessthanthisinthemiddlestripandmoreinthecolumnstrip,i.e.alongthefreeedge,andalsoto make the column strip smaller than half the strip width. In this way some reinforcement is concentrated along the free edge, which is to be recommended, particularly in this case where the stripspansbetweentwocolumns. As the slab is assumed to be simply supported at the corner column there is no design top reinforcement in that region. There will, however, be certain negative moments under service conditions, which may cause cracks, mainly in the direction of the bisector to the corner. It is advisabletohavesometopreinforcementatrightanglestothatdirection. Intheydirectionthedistributionofspanmomentsmustalsobechosensoastofulfilcondition(2.24), i.e.withnotmorethan0.7mDBinthemiddlestrip.InthiscasethecooperatingelementatedgeDisa oneway element, which means that it is not acceptable to have too uneven a distribution of span moment. For this direction the moment in the middle strip may thus be chosen to be 0.7mDB or somewhatsmaller.
107

Fig.8.3.3showsapossibledistributionofdesignmoments.

Fig.8.3.3 In most practical cases it is recommended to take into account the moment transfer to the corner columnaswellastotheedgecolumn.Theinfluenceofthesemomentsonthedesignisdemonstrated inExample8.4.

8.4 Slab cantilevering outside columns

Wheretheslabcantileversoutsideacolumnwegetcornersupportedelementswithoutanypositive moment corresponding to the reinforcement in the direction of the cantilever. Then the support designmomenthastohaveadistributionwhichsatisfiescondition(2.24). Example8.6 TheslabinFig.8.4.1carriesaloadof14kN/m2andissupportedoncolumncapitalswithadiameter of1.5m.Thesupportistakenastheinscribedsquare,whichhasasidelengthof1.06m.Thisgivesthe spanlengthsshowninthefigure.Nomomenttransferisassumedtotakeplacebetweentheslaband thecolumns.

Fig.8.4.1 Theaveragemomentscausedbythecantileversarestaticallydeterminate.Wethusget (8.37)

(8.38) ThesupportmomentmxBisstaticallyindeterminateandcanbeestimatedfrom
108

 (8.39) The reason why mxB has been chosen lower than the approximate value is partly the large column capital and partly the fact that mxC is larger than the moment corresponding to a fixed support for spanBC.Itmightevenhavebeenreducedfurther. WecannowcalculatethecvaluesandthecorrespondingspanmomentsfromEqs(2.34)and(2.35): (8.40) (8.41) (8.42) (8.43) (8.44)

(8.45) Fig.8.4.2showsapossibledistributionofdesignmomentsforthereinforcementinthexdirection.It .If,forexample,we fulfilscondition(2.24)andissuitableforcracklimitation.Itcorrespondsto checkthevaluetotheleftofcolumnBwefind (8.46) Thus we are close to the limit, and we could not have increased the support moment at the edge withoutdecreasingthecorrespondingspanmomenttoanequalextent. Inthiscaseitisnotrecommendedthatallthesupportmomentshouldbeconcentratedinthecolumn strip,asthiswouldleavetheedgewithouttopreinforcement.

Fig.8.4.2

8.5 Oblong panels and corner-supported elements

The advanced strip method always results in a design which is safe, which means that the slab has adequatesafetyagainstbendingfailure.Forthedesignitisalsoimportantthattheslabfunctionswell under service conditions and that the reinforcement economy is good. For most slabs met with in
109

practice this can be expected to be the case if the rules above regarding moment ratios and lateral momentdistributionsarefollowed. If, however, the panels are very elongated the advanced strip method may lead to a moment distributionwhichistoodifferentfromtheexpecteddistributionintheservicestate.Thismightlead tounacceptablecracksinplaceswherethemomentshavebeenunderestimated.Insuchcasesitmay be better to assume that the slab is composed of ordinary rectangular slabs supported on support bands. Fig.8.5.1demonstratesanexample.Fig.8.5.1a)showselementsforadesignbasedontheadvanced stripmethod.Thepanelsstudiedinthisapproach,e.g.ABCD,areveryelongated,aswellassomeof thecornersupportedelements. Fig.8.5.1 b)showselementsforadesignofthesameslabassumingrectangularslabssupportedon support bands. Thus the rectangular slab ABEF is assumed to be supported along all sides and continuousatADF.ThesupportsalongAFandBEconsistofsupport

edgeofarectangularslabissupportedononeormorecolumns.Inthiscasetheapproachaccordingtob)is normallytobepreferred,althoughtheapproachaccordingtoa)mayformallybeapplied.

Fig.8.5.1 bands which are continuous over columns D and C respectively. The design follows the principles demonstratedinChapter4. Ifthedesignmomentsobtainedfromthetwoapproachesarecompareditisfoundthata)giveslarger momentsmxandmysbutsmallermomentsmyfthanb).Theriskwiththeapplicationoftheadvanced stripmethodthusliesinanunderestimationofmyfleadingtopossiblecrackingintheundersideofthe slab.Thisriskincreasesthemoretheshapesofthepanelsandthecornersupportedelementsdeviate fromthatofasquare.Itisnotpossibletogiveanygeneralruleforwhentheriskhastobetakeninto account. In practice this will probably seldom be the case, as the column supports are normally arrangedsothatthespansinthe twodirectionsarenottoounequal.Apracticalcasemaybewhereafree

In cases where there is a doubt whether approach a) gives an acceptable result it is recommended thatbothapproachesareusedandtheresultscompared.Ifthedifferencesaregreatitisadvisableto useapproachb)ortotakesomekindofweightedaveragebetweentheresults.
110

Itmustberememberedthatthechoiceofapproachisunimportantwithregardtotheultimatelimit stateandthatithasonlytodowithcrackcontrol.

CHAPTER9 Regularflatslabswithnonuniformloads
9.1 Introduction
Wheretheloadonaflatslabisnotuniformtheadvancedstripmethodinitsbasicformcanonlybe usediftheloadisuniforminoneofthemaindirections.Acceptableapproximatesolutionswiththe methodcan,however,befound. The reason why the advanced strip method in its basic form is not generally applicable for non uniformloadsisthatthelinesofzeroshearforcearenotcontinuous,sothewidthsofthestripswill vary.

9.2 Uniform loads in one direction

If the loadon a flat slab is uniform in one of the main directions the advanced strip method can be applied, as the lines of zero shear force are straight and continuous. The only differences from the analysisofflatslabswithauniformloadarethattheloadvariesinonedirectionandthemomentsin parallelstripsintheotherdirectiondiffer,i.e.theyhavedifferentloads.Themomentsinthesestrips arechosensothatthemaximumspanmomentsoccuralongthesameline. Example9.1 TheloadontheslabinFig.9.2.1is9kN/m2betweenlinesaande,and17kN/m2betweenlineseand i. The edge along line 1 is fixed, whereas the other edges are freely supported. All columns are .

111

 Fig.9.2.1 The strips in the xdirection have different loads on the spans ae and ei. An approximate elastic analysisshowsthatsuitablevaluesofsupportmomentsare: . BasedonthesesupportmomentsthecvaluesgiveninthefigurehavebeencalculatedbymeansofEq. (2.34).Eq.(2.35)givesthespanmoments: . Thestripsintheydirectiontotherightoflineehaveauniformloadof17kN/m2.Anapproximate elasticanalysisshowsthatthesupportmomentscanbechosentobeequal: . Thecorrespondingcvaluesaregiveninthefigure.Thespanmomentsare: . Thestripstotheleftoflineehaveauniformloadof9kN/m2.Themomentsaresimplytakenas9/17 ofthosetotherightoflinee: . The distributions of design moments follow the recommendations in Section 8.1.5. One special question in this case is the support moments for the reinforcement in the ydirection above the columnsinlinee.Themostcorrectdistributionistofollowthecalculatedmomentsstrictlyanduse different values to the left and right of the column centre. This is what is recommended in the first place. Suitable distributions of design moments along lines 1, 2 and 3 are shown in Fig. 9.2.2. The distributionscorrespondto ,whichmeansthatthewidthsofthecolumnstripsareequaltohalf thetotalstripwidths.Otherdistributionsarepossible,dependingontheneedforcrackcontrol.

112

 Fig.9.2.2 AcheckshouldinprinciplealwaysbemadethattherulesforlateralmomentdistributionsinSection 2.5.2arefollowed.Withthenormallyrecommendeddistributionofsupportandspanmomentssuch checksneed,however,onlybemadeiftheratiobetweenthenumericalvaluesofthesupportandspan momentsisaboveabout3orbelowabout1.Forthisslabsuchacheckonlyhastobemadeforthe momentsmxinelementde,wherethisratiois .Ifwechooseanevendistributionofspan momentandconcentrateallthesupportmomentinthecolumnstripweget .This isnotanacceptablevalueaccordingtoEq.(2.24).Wehavetotakesomeofthesupportmomentinthe middlestrip.Wemay,forexample,take0.4msinthemiddlestrip,ashasbeendoneinFig.9.2.2.Then weget ,whichissatisfactory. Fromthepointofviewofconstructionitmaybeadvantageoustouseonlyonedesignvalueforthe support reinforcement over each column. Instead of the values 39.0 and 73.6 a weighted mean valuemightbeusedoverthewholewidthforthecolumnstripinlinee.Thisweightedmeanvalueis (9.1) Thesafetyandbehaviouroftheslabwouldprobablynotbemuchinfluencedbychoosingtheaverage valueinsteadofthetwovaluesinFig.9.2.2,butinordertoavoidanyrisksitisrecommendedthatthe twovaluesgiveninthefigureareused.

9.3 Different loads on panels

Theadvancedstripmethodcanbeappliedtothesituationwheredifferentloadsareappliedtopanels inanapproximatebutsafewaybynotkeepingtheregularityofthenetoflinesofzeroshearforcein thespansandacceptingthatthelinesmaychangepositionwheretheypassfromonestriptoanother. This means that the strips change width wheretheypass a support. The strips on both sides of the support do not fit together. Even if they have the same average moments there is not perfect equilibriumoverthesupportasthetotalmomentsaredifferent. The simplest way of treating this problem is to reinforce for the largest total moment over the support, i.e. to reinforce for the support moment on the width of the widest strip. In this way the design is on the safe side. In practice the difference in width is seldom great, which means that the economicalconsequenceofthissimpleapproachisunimportant. Example9.2 Fig.9.3.1 shows aslabwithdifferentloadsondifferentareas. The slabrests onlarge columnheads andisfreelysupportedalongalledges. The strips which are used for the design are shown in Fig. 9.3.2, where all the cvalues which have resultedfromtheanalysisaregiven. Inthexdirectionthestripbetweenlines2and3hasonespanwithload8kN/m2andtwospanswith 18 kN/m2. An approximate elastic analysis of the strip shows that suitable support moments are and . With these support moments we get the cvalues given in the figure from Eq. (2.34)andspanmoments fromEq.(2.35).

113

Fig.9.3.1

Fig.9.3.2 , In the same way, for the strip between lines 3 and 4 we find the moments . . Forthestripintheydirectionbetweenlinesbandcwefind Forthestripbetweenlinescandfwefind . Figs9.3.3and9.3.4showproposeddistributionsofdesignmomentsforthereinforcementinthex andydirections,respectively.

114

 Fig.9.3.3 Thespanmomentshavebeenchosentobeuniformwithineachstripandtohavethevaluesabove. Forthedistributionofsupportmomentsthecolumnstripsandmiddlestripshavebeenassumedto . The support moments in the middle strips have been have equal widths, corresponding to takenas0.4times,andinthecolumnstrips1.6times,theaveragemomentinthestriptowhichthey belong.Thedistributionofsupportmomentsthuscorrespondstocases2inFig.8.1.4. Justasinthepreviousexamplethesupportmomentsinthecolumnstripsmightbeevenlydistributed andequaltoaweightedaverage,butinthefirstplaceitisrecommendedthatthedistributionsshown inthefiguresareused. AcheckagainsttherulesinSection2.5.2hasinthiscasetobeperformedforthemoments and , where the ratio between the numerical values of the support and span moments is greaterthan3.WiththedistributioninFig.9.3.3wehave ,whichisacceptable accordingtoEq.(2.24).

Fig.9.3.4

9.4 Concentrated loads

Where more or less concentrated loads are acting on a flat slab they are taken into account in the determinationoftheaveragesupportandspanmomentsinthestripsjustlikeuniformloads. Iftheconcentratedloadissmallcomparedtothetotaluniformloadonthesamepanel,saylessthen 10%,thenormallateraldistributionofdesignmomentsmaybeused.Itmaybesomewhatbetterfor the behaviour under service conditions to make some redistribution of moment to the part of the panel where the concentrated load is acting, and thus to a column strip if the concentrated load is actingmainlywithinthatstrip. Iftheconcentratedloadisnotsmallcomparedtothetotaluniformloadonthepanel,theanalysisis bestmadeseparatelyfortheuniformloadandfortheconcentratedloadandthemomentsarethen
115

added.Thisleadstosomeoverestimationofthespanmoments,asthemaximummomentsfromthe twoloadingcasesdonotappearinthesamesections. Theprocedurefordeterminingthedesignmomentscausedbyaconcentratedloadisbestillustrated bymeansofanexample. Example9.3 Fig.9.4.1showsaninteriorpanelwithinaflatslabwithaconcentratedload actingoveran area withitscentreat(3.0/1.0)inthecoordinatesystemshown.

Fig.9.4.1 Westartbydeterminingthetotalmomentscausedbytheconcentratedload.Thesupportmoments maybetakenasabouthalfthemomentsaccordingtothetheoryofelasticityforabeamwiththesame loadandfixedends.Wemaychoosethevalues , whereindex1indicatestheendsonthecoordinateaxes. Basedonthesevalueswefindthesupportreactions (9.2) (9.3) .Thepositionsofthe Theloadpermlengthinthestripsinbothdirectionsisequalto linesofzeroshearforcearedeterminedfrom (9.4) (9.5) Thespanmomentsare (9.6) (9.7) Thelinesofzeroshearforcedividestheloadbetweenthedifferentelementsanddifferentpartsofthe strip.Thus1.05/1.20oftheloadiscarriedonthestripwithwidthcy1.Thecorrespondingpartsofthe momentsaredistributedonthewidthsofthestripsinquestion.Wegetthefollowingaveragedesign moments: Inthestripwithwidthcy1:
116

(9.8) (9.9) (9.10)

(9.11) Inthesamewaywefind: Inthestripwithwidth Inthestripwithwidth Inthestripwithwidth Itcanbeseenthatthedominantmomentsarethoseinthestripwithwidthcy1.Thesemomentsare ratherhighandconcentratedinanarrowband.Itmaybebettertodistributetheminawiderband. Thiscanbedonethroughtheintroductionofdistributionreinforcementintheydirection,applying theprincipledemonstratedinSection2.6.1.Letusdistributetheloadoveradistanceof inthe . Eq. (2.33) gives a design moment for the distribution ydirection with a bandwidth reinforcement (9.12) . One part of this area is outside the The load is now acting on the slab over an area of studied cornersupported elements and acts on the parallel oneway element in the xdirection (or slightly on the other side of that element). The part of the load acting on the cornersupported elementis ,andithasitscentreat(3.0/1.25). ThetotalmomentsMx(includingtheonewayelement)andthevalueofcx1areunchanged.Itproves satisfactorytochoose .Withthesemomentswefind . Asonly2.5/3.0ofthewidthintheydirectioniswithinthecornersupportedelementonly2.5/3.0of themomentsMxbelongtotheseelements.Forthestripwithwidthcylwethusgettheratiobetweenm andM: (9.13) Inthiswaywegetthefollowingaveragedesignmoments: . The parallel oneway element has the In the strip with width samemoments. . Inthestripwithwidth . Inthestripwithwidth Inthestripwithwidth . Now the largest moments per unit width are reduced to about 55% of the earlier values at the expenseofsomedistributionreinforcement. ThedistributionofmfmsforapointloadwillbeuniformaccordingtoSection2.6.2.Byintegration overtheloadedareaauniformdistributionforapointloadcorrespondstoatriangulardistribution onthewidthofauniformloadandaconstantvalueontheremainingpart,accordingtoFig.9.4.2.This distribution may be transformed to stepped constant values. It is suitable to distribute ms with a certainconcentrationtowardsthesupportedcornerandmfsomewhattowardstheload.Forthefinal distributionofdesignmomentssomefurtherredistributionmaybeacceptedinordertosimplify.

117

 Fig.9.4.2 Fig. 9.4.3 shows a possible distribution of design moments based on the second solution. These momentsaretobeaddedtothemomentscausedbytheuniformloadontheslab.Thereinforcement shouldnot,inprinciple,becurtailedwithinacornersupportedelementinthiscase.

Fig.9.4.3 Thesupportmomentshavetobetakencareofintheadjacentpanels.Thiswillcauseachangeinthe position of the lines of zero shear force in the spans. The span moments may thus be reduced. The increaseinlengthofreinforcingbarsshouldbetakenintoaccount.
118

CHAPTER10 Irregularflatslabs
10.1 General
Anirregularflatslabmayhavecolumnsplacedquitearbitrarilyandedgesatanyangletoeachother. It is of course not possible to find simple general methods for the design of all types of such slabs, automaticallyleadingtosuitableresultsfromallpointsofview,particularlyregardingreinforcement economyandbehaviourunderserviceconditions. Whateverapproachisusedwithintheframeworkofthestripmethod,itismostlynecessarytousean iterationprocessinordertogetaresultwithasatisfactorydistributionofreinforcement.Estimating whetherthereinforcementissatisfactoryismainlybasedonacomparisonbetweensupportandspan momentsandonsomefeelingforthebehaviourofaslab. One possible approach might be the use of support bands between the columns, supporting strips whichintheirturncarrytheloads.Thisapproachisgenerallycomplicatedbecausetherearesomany choices to be made, and therefore so many iterations. A major problem is that the support bands whichmeetoveracolumnhavedifferentdirectionsofsupportmoments,whichhavetobetakenby straightreinforcementbars.Thisapproachisthereforenotgenerallysatisfactory. Analternativeapproachwillberecommendedherewhich isbasedontherestricting conditionthat mostreinforcingbarsareplacedintwoorthogonaldirections.Thisrestrictionisvalidforspanaswell assupportreinforcement.Anexceptionisreinforcementalongfreeedges,whichisbestplacedalong the edge. Another exception is that support reinforcement at a support where the slab is fixed or continuousisbestarrangedatrightanglestothesupport. Themainadvantageofthisapproachisthatitgivesthepossibilityofformulatinggeneralrulesforthe design, which can be carried through in a systematic way. Another advantage is that it leads to a simplereinforcementarrangement. The main disadvantage is that the reinforcement directions may deviate considerably from the directionsoftheprincipalmoments.Wherethisisthecasethereinforcementislessefficient,which means that the crack control is not asgood as whenthe reinforcement directions coincide withthe principalmomentdirections.Itmayalsomeanapoorerreinforcementeconomy.Ontheotherhand, reinforcementwhichfollowsthedirectionsoftheprincipalmomentsbetterismorecomplicatedand maybemoreexpensive. Over a column the moments often tend towards a polar symmetry. Where this is the case the reinforcement directions are not important, provided that there are two orthogonal reinforcement directions. From a practical point of view the advantages of the approach are probably greater than the disadvantages. Theapproachisnotgenerallyapplicabletoslabswithfreeedges.Thiswillbediscussedinconnection withsomeexamples.

10.2 Design procedure

Theproposedapproach,whichisbasedontheconditionthatallreinforcingbarswithintheinteriorof theslabareplacedintwodirectionsatrightangles,paralleltothecoordinateaxes,normallycontains thefollowingsteps:
119

1. Determine suitable directions of the coordinate axes, corresponding to the reinforcement directions. One axis may, for instance, be chosen parallel to an edge of the slab or to an important directionbetweencolumncentres. 2. Determine theoretical column profiles. The theoretical column profile for each column is an inscribedrectanglewithsidesparalleltothecoordinateaxes. 3.Determinethelinesofzeroshearforcewherethespanmomentshavetheirmaxima,thespanlines, inthefollowingway.First,placelinesofzeroshearforcebetweentheedgesandthecolumnsnearest theedges.Theselinesaredrawnatabouthalfthedistancebetween thecolumnandtheedgeifthe edgeisfixed,andatabout0.625ofthisdistancefromthecolumnsiftheedgeisfreelysupported.If theedgeisonlypartlyrestrained,useintermediatevalues.Thendrawlines(thesearenotlinesofzero shearforce)betweenadjacentcolumncentres.Startingfromthecentresoftheselinesandrunningat rightanglestothem,drawlinesofzeroshearforce.Drawtheminbothdirectionsuntiltheyintersect thecorrespondinglinesfromotherpairsofcolumnsorotherlinesofzeroshearforce.Incaseswhere twoormoreintersectionsareclosetoeachotheradjustthelinessothattheyhaveacommonpointof intersection. 4.Drawthelinesofzeroshearforcewherethesupportmomentshavetheirmaxima.Thesesupport linesareacontinuationofthesidesofthetheoreticalcolumnprofilesandaredrawninthedirections ofthecoordinateaxesuntiltheymeetthespanlines.Therearefoursupportlinesaroundeachcolumn (except for edge and corner columns), two in each direction. There is also a support line at each supportonawall,beamorsupportband. 5.Calculatethetotalmoments(spanminussupportmoment)withrespecttoeachofthesupportlines at the columns. Each moment is caused by the load on the element formed by the support line in questionandthenearestspanlines.Thespanmomentsarebestexpressedasaveragemoments(m) butthesupportmomentsatcolumnsastotalmoments(M).Calculatealsothecorrespondingaverage moments in the oneway elements supported on walls, beams or support bands. Where such a supportisnotparalleltoacoordinateaxisthespanmomentmwithrespecttothesupporthastobe takenbydesignspanmomentsmx andmy inthereinforcementdirections.The followingrelationis valid: (10.1) whereistheanglebetweenthesupportandthexaxis.Oftenitissuitabletoassumemx=my=m.Itis alsopossibletoincreaseoneofthedesignmomentsanddecreasetheotheraccordingtotheequation. 6.Choosesupport momentsMxandMyateachcolumnwheretheslab is continuousin therelevant direction.Valuesapproximatelyequaltoonethird(between0.30and0.37)ofthesumoftherelevant totalmomentsonbothsidesofthecolumnwillbesuitable.Thesevaluesareusuallycloseenoughto the support moment, according to the theory of elasticity. Choose (or calculate, if statically determined) support moments where the slab is supported by a wall, beam or support band. If the slab is fixed at such a support the support moment can be taken as approximately twothirds (between0.60and0.75)ofthesumoftherelevanttotalmomentsintheelement. 7.Calculatethespanmomentsfromthevaluesaccordingtosteps5and6.Thiswillgivetwovaluesof eachspanmoment,onefromeachsideofthespanline.Oftentherewillbeagreatdifferencebetween thesevalues.Thenthespanlinehastobemovedinthedirectionwhichwilldecreasethedifference.A suitabledistanceltomovethespanlinecanbeestimatedbymeansofthefollowingrelation (10.2) wherem1andm2arethetwospanmomentsandlistherelevantspan.Thengobacktostep4and maketheanalysiswiththenewpositionsofspanandsupportlines.Whenthedifferencebetweenthe spanmomentsatallspanlinesissmalltheresultisaccepted.Thespanmomentmightbetakenasthe averagemomentfromthetwovalues.Thisaveragemomentwillprobablygiveanacceptablelevelof safety,butinordertobeonthesafesideahighervaluemaybepreferred.Tousethehigherofthetwo values is always on the safe side, but may be unnecessarily conservative. A compromise is to add
120

threequarters of the higher value and onequarter of the lower value. This compromise is recommended and will be used in the examples. An acceptable difference between the higher and lower value will be about 0.4 times the higher value in situations where this compromise is used. Wherethespanmomentissmallcomparedtothesupportmoments(e.g.ashortspanbetweenlong spans)largerrelativedifferencesmaybeaccepted. 8.Basedonthecalculatedsupportandspanmoments,thedesignmomentsforthereinforcementare determined.Thedesignspanmomentsmaybetakendirectlyasthecalculatedaveragemoments or redistributed towards places wherethe curvature isestimated tobe larger. The supportmoment is suitablydistributedwith70to80%overthecentralhalfofthesupportlineandtherestdistributed ouside this part, see the examples. A distribution with the whole moment concentrated within the central part may be accepted only if the ratios between the numerical values of support and span momentsareeverywherebetweenabout1.5and2.5. Adesignbasedontheserulesistheoreticallynotassafeasthedesignofregularflatslabsaccordingto the rules given inChapters 8and 9, as the limitations on moment distributions used there are only validforrectangularelements.However,thedesignisprobablyquitesafe,particularlyifthechoiceof momentdistributionsisalsobasedonsomefeelingforthebehaviourofslabs.Anexceptionisslabs withfreeedges,whichhavetobehandledwithcare.Thisisdiscussedinsomeexamples. Ahandcalculationaccordingtotheserulestakesaratherlongtime,asmanysimpleoperationshave to be performed. It should not be too difficult to write a computer program which more or less automaticallycarriesthroughthecalculations,whichmay,perhaps,includesomeinteractionwiththe designer.

10.3 Edges straight and fully supported

Example10.1 The slab in Fig. 10.3.1 has a uniform load of 12 kN/m2. It has three freely supported and one fixed edge and is supported on three columns. The column capitals have diameters of 1.0 m. The coordinatesforthecornersandthecolumncentresaregiveninthefigureinthechosencoordinate system,whichhasbeenassumedtogivesuitablereinforcementdirections,paralleltothecoordinate axes.

Fig.10.3.1 ThetreatmentoftheslabaccordingtotheprocedureaboveisillustratedinFig.10.3.2.Thetheoretical column support profiles are chosen as inscribed squareswith edgesparallel to the coordinateaxes. Thesesquareshavesides0.70m.Thetypesofelementareindicatedwitharrowsintheusualway.
121

First,theassumedspanlinesofzeroshearforcearedrawn,startingwiththelinesabcdalongthe edges.Usingtherulesinpoint3abovewefindsuitablecoordinatesforthesepoints: a(4.15/14.83),b(20.75/8.23),c(21.62/2.50),d(6.53/1.32) Next,thelinesbetweenthecentresofthecolumnsaredrawn(notshowninthefigure)andspanlines atrightanglestotheselinesfromtheircentres.Thesespanlines(ofzeroshearforce)areextended until they meet other span lines. In this way we find the following coordinates for the points of intersection: e(13.83/1.89),f(5.57/6.78),g(12.61/7.72),h(14.15/10.85) Nowwehavethecompletepatternofspanlinesofzeroshearforce.Next,wedrawthesupportlines ofzeroshearforcealongthesidesofthetheoreticalcolumnsupportprofilesuntiltheymeetthespan lines.Thetotalnumberofpointsofintersectionbetweensupportandspanlinesis8foreachcolumn andthus,inthiscase,24inall.Thesecoordinatevalues

Fig.10.3.2 arenotallgivenhere.Onlysomeexampleswillbegiveninconnectionwithmomentcalculations. Wealsodeterminethedistancesatrightanglesfrompointsa,b,c,d,e,fandhtothesupports.These distancesareshownaslinesofzeroshearforceinFig.10.3.2. Nowwecanwritetheequilibriumequationsforthedifferentparts(elements)oftheslab.Thereare four such equations for each column and seven for the edge supports in this case, a total of 19 equations.Notalloftheseequationswillbedemonstratedindetail,butonlyafewtypicalexamples. Fortheotherelementsonlytheresultingrelationswillbeshown.Allequilibriumequationsarebased onEq.(2.7)withspecialcases(2.6)and(2.8). The element between span line af and the support has at a and at f. This gives the followingmomentequation: (10.3) ThesupportlinetotheleftofcolumnEhasendcoordinates(8.40/7.16)and(8.40/13.14)andthusa length of 5.98 m. The corresponding element has its corners at f and a. In writing the equilibrium equationitmustbenotedthatthespanmomenthasanactivewidthequaltothelengthofthesupport line, as the reinforcement outside the ends of this line works in both directions and thus gives no contributiontothemomentequation.Intheanalysisitisassumedthatthespanmomentisuniformly distributed within each element. If another distribution is used for the design, this difference will normallymakethedesignsafer,asanunevendistributionofreinforcementismadeinthedirection whereitismoreefficient. WiththevaluesabovewegetthefollowingequationforthesupportlinetotheleftofcolumnE:
122

 (10.4) Continuinginthesamewaywefindthefollowingrelationsforthereinforcementinthexdirection: (10.5) (10.6) (10.7) (10.8) (10.9) (10.10) (10.11) Basedontheserelationswecancalculatesuitablesupportmomentsbymeansoftheruleinpoint6in Section10.2. , . Introducingthesevaluesintotheequationsabovewegettwovaluesofeachspanmoment.Thesetwo values are given below within parentheses as well as the weighted average according to the rule proposedinpoint7inSection10.2.AsthedirectionsofedgesAandCarenotfarfromthatofthey axisitis,onthesafeside,assumedthat inEq.(10.1).Thus etc. , . Performingthecorrespondinganalysisfortheydirectiongivesthefollowingresults: (10.12) (10.13) (10.14) (10.15) (10.16) (10.17) (10.18) (10.19) (10.20) (10.21) , ,
123

. If we check the resultwith respect to the rule proposed in point 7in Section 10.2 we find that it is acceptable.ThemostquestionablevalueismyGE,wherethelowervalueismorethan40%lowerthan the higher value. On the other hand this moment is small compared to the other moments and this difference might be accepted. However, it will be demonstrated how the accuracy of the solution is increasedbychangingthecoordinatesforthepointsofintersectionofthespanlines. The principles for the change of these coordinates is to move the points of intersection in such a directionthatthedifferencesaredecreased.Thuse.g.pointg(orf)ismoveddownwardsinorderto decreasethedifferenceinmyEG.Pointgisalsomovedtotheleftinordertodecreasethedifferencein mxGF. Some of the other points are also moved in order to decrease other differences. It should be noted that points h, e and f, which were originally situated on the straight lines ab, cd and da, respectively,maybemovedwithouttakingintoaccountthesestraightlines. Thefollowingcoordinatesarechosenforthesecondanalysis: a (4.45/14.50), b (20.75/8.50), c (21.62/2.40), d (6.40/132), e (13.70/1.89), f (5.55/6.78), g (12.30/7.50),h(14.00/10.85). Itmaybenotedthatthechangesincoordinatesareatmostintheorderofabout0.3m,whichisabout 5%oftheaveragespan. Ananalysiswiththenewcoordinatesforthepointsofintersectionofthespanlinesgivethefollowing momentvalues: . , . . , . Acomparisonwiththepreviousanalysisshowsthatthedifferencesarerathersmall.Thedesignmight well have been basedon the first analysis. The practical effect onthe behaviour ofthe slab and the safetyarecertainlynotnoticeable. We now know all the moment values which we need for the design of the reinforcement. Possible designmomentdistributionsusingthesevaluesareshowninFig.10.3.3forreinforcementinthex direction and support moments at the fixed edge, and in Fig. 10.3.4 for reinforcement in the y direction.Thedistributionofsupportmomentsatthecolumnsfollowsthegeneralrecommendations.

Fig.10.3.3 In addition to the design moments shown some minimum reinforcement according to the relevant code should also be provided, as well as some corner reinforcement if prevention of top cracks is important.

124

 Fig.10.3.4 Wecanalsocalculatethereactionforcesonthecolumns.Thecolumnforceistheloadactingonthe area within the span lines of zero shear force surrounding the column. So, for example, the area belongingtocolumnEis45.5m2andthecorrespondingreactionforceis546kN.

10.4 Edges straight and partly column supported

Thetreatmentinthiscaseissimilartotheprecedingonewiththeexceptionthatreinforcementhasto be arranged along the columnsupported edges. Support bands are introduced along these edges, carryingthesupportreactions. Where a column is situated at an obtuse corner of a slab, case b) in Fig. 10.4.1, the situation is intermediatebetweenthatofacolumnatarightangledcorner,casea),andastraightedge,casec).In case a) there is no negative support moment, provided that the slab is assumed to be simply supportedonthecolumn.Incasec)thereisalargesupportmoment,correspondingtocontinuityof the slab over the column. In the intermediate case b) there may also be some support moment, particularlywhentheangleapproaches180.Thesupportmomentincreasesfromzerofor to thefullsupportmomentcausedbycontinuityfor .Probably,itincreasesslowlyatsmallangles andfasterasapproaches180.Thefollowingapproximateruleisproposedforthedesignsupport momentMs: (10.22) (10.23) whereMs,contisthesupportmomentcorrespondingtofullcontinuity.ThereinforcementforMshasto bewellanchored.

Fig.10.4.1 ThemomentMsatanobtusecornerhasdifferentdirectionsforthetwosupportbandsmeetingatthe corner.ForequilibriumtheremustalsobeamomentMbwithitsvectorperpendiculartothebisector: (10.24)

125

Thismomentmaybetakenbybottomreinforcementintheslaborbymomentinthecornercolumn.It is not necessary to arrange extra reinforcement to take Mb, but a check must be made that the reinforcementarrangedforotherreasonsissufficienttotakethatmoment. InordertopreventtopcracksinthedirectionofthebisectorthesupportreinforcementforMsshould bedistributedoveracertainwidth,selectedaccordingtotherulesforsupportbandsinSection2.8.3. Also,where ,somereinforcementforthispurposemaybeneeded,preferablyatrightanglesto thebisector. Example10.2 TheslabinFig.10.4.2isthesameasinthepreviousexamplewiththeexceptionthatedgesBandCare supportedoncolumnswithadiameterof0.4m.Theloadisalsothesame.Theslabisassumedtobe simplysupportedatthecolumns.WecanthususetheresultsfromExample10.1andneedonlymake thecalculationsforthesupportbands. Forthedesigncalculationthecircularcolumnsareexchangedforsquareinscribedcolumnswithsides 0.28maccordingtoFig.10.4.3.Forthecornercolumnaninscribedfigure

Fig.10.4.2 is used with edges at right angles to the directions of the support bands. In this case a triangle is suitable. The positions of the columns have been chosen so that the support band along the inside of the columnsinFig.10.4.3followsthesupportsinExample10.1.Wecanthustaketheloadonthesupport bandsfromthelinesofzeroshearforceinFig.10.3.2,assumingthattheloadiscarriedatrightangles intothesupportband.TheselinesareshowninFig.10.4.3.Inaddition,thereistheloadonthestrip betweenthecolumns,whichhasawidthof0.34m,correspondingtoaloadof4.0kN/m. Theloadonthesupportbands,includingtheloadonthestripbetweenthecolumns,isshowninFig. 10.4.4. As the angle at L is less than 135 the bands are assumed to be simply supported at that column.Thecalculationofthemomentscanbemadebymeansofordinarymethodsaccordingtothe theoryofelasticity.Here,instead,thesameiterativemethodwillbeappliedasfortheslabaccording totherecommendationinSection10.2. WestartbyanalysingbandHIJKL.Pointsofzeroshearforceareassumedatthecentresofallspans except span KL, where this point is assumed to be at a distance of from K. The analysisisrathertrivialandwillnotbeshownindetail.As

126

 Fig.10.4.3 anexamplethecalculationforthepartnexttoH willbeshown.Theloadatthepointof zero shear forceinspanHIis14.2kN/m.Theequilibriumequationis (10.25) Inthesamewaywefind (10.26) (10.27) (10.28) (10.29) (10.30)

Fig.10.4.4 (10.31) (10.32)

127

AccordingtotherulesinSection10.2wefindthefollowingdesignmoments: . . With we get from Eq. (10.25) as the support moment at the fixed edge. This is an acceptable value althoughalittlelow.IfwehadusedtherecommendationinSection10.2,abettervaluemighthave been ,givingadecreaseinMHIto20.2. AllthevaluesareacceptableaccordingtotherecommendationsinSection10.2 ThemaximumwidthofthereinforcementbandischeckedbymeansoftherulesinSection2.8.3.Thus the width on the inside ofthe support band may correspond to about half theaveragewidth of the elementwhichcausestheloadontheband.Thesebandwidthsarefoundtobeabout0.4mforspan HI,0.8mforspanIJ,0.9mforspanJK,and0.9mforspanKL.Tothesevaluesshouldbeaddedthe widthofthebandwhichisdirectlycarriedonthecolumns.Thiswidthis .Wheretwo differentvaluesarefoundonbothsidesofacolumntheaverageisapplied. A corresponding analysis for band LMN with the points of zero shear force according to the recommendationsatdistances3.15and3.11mrespectivelyfromcolumnMgives (10.33) (10.34) (10.35) (10.36) . ThedifferencebetweenthetwovaluesofMMNisnotquiteacceptable.Wethereforehavetomovethe pointofzeroshearforceinspanMNclosertoM.Thenewdistanceischosenas2.87m.Ifweonly make this change we will get a greater difference in the values of MN. We therefore also move the pointofzeroshearforceinspanLMclosertoM,choosing3.07m. Withthesenewvalueswefind . Itwillbeseenthatthedifferencebetweenthetworesultsisnotveryimportant.Thefirstcalculation might well have been accepted in spite of the different values of MMN. The latter calculation gives a savinginreinforcementofabout5%.Itmaybeworthwhilemakingasecondanalysisinordertosave reinforcement,butithardlyinfluencesthesafetyorthebehaviour. The resulting design moments in the support bands are shown in Fig. 10.4.5, given as moments in kNm/m in the band widths shown. At the obtuse corner a design moment for an extra top reinforcement is shown, which has been estimated to be about onequarter of the moment for full continuity. The complete design moments for the slab are the moments according to Figs 10.3.3, 10.3.4 and 10.4.5.

10.5 Edge curved and fully supported

Wheretheedgesarecurvedintheanalysistheymaybeexchangedforacircumscribedpolygon,which gives a safe design. The polygon is as a rule best chosen so that each side corresponds to the supportedsideofananalysedelement. ThegeneralrulesinSection10.2apply.

128

 Fig.10.4.5 Example10.3 The slab in Fig. 10.5.1 has a simply supported edge in the shape of a quarter of an ellipse. The two straight edges are fixed. There are two square interior columns with 0.3 m sides. The load is 10 kN/m2.Allreinforcementisassumedtobearrangedinthexandydirections. TheassumedcircumscribedpolygonandthelinesofzeroshearforceareshowninFig.10.5.2.Itmay benotedthateachsideofthepolygonformsthesideofanelementwheretheothersidesarelinesof zeroshearforce.Thepatternoflinesofzeroshearforcehasbeendeterminedaccordingtotherulesin Section 10.2 and adjusted after a first analysis, which is not shown here, in order to reduce the differencesbetweenvaluesofspanmoments. Thecoordinatesofthepointsofintersectionbetweenthespanlinesofzeroshearforceare: a(2.8/8.6),b(10.0/7.3),c(15.7/5.3),d(17.6/1.5),e(9.0/2.4),f(2.8/3.2). Performing the analysis in the same way as demonstrated in Example 10.1 we find the following relations:

Fig.10.5.1

129

Fig.10.5.2 (10.37) (10.38) (10.39) (10.40) (10.41) (10.42) (10.43) (10.44) (10.45) (10.46) (10.47) (10.48)

(10.49) (10.50) Applyingtherulesinsection10.2andwiththenotationusedinExample10.1wefurtherget . . , , . ItshouldbenotedthatthevaluesformxHD,myGBandmyHCareonlyvalidif withintheareasclose tothesupports,cf.Eq.(10.1).Itiseconomicallymoreefficienttohavedifferentvaluesofmxandmy. Thus to reinforce for mGB we may use a value of mx which only corresponds to the reinforcement
130

designedforthexdirection.Withtherulesusedformomentdistribution,thisvaluemaybetakenas . At B, . Applying Eq. (10.1) we the lower of onethird of mxAG or mxGH, which is find (10.51) giving forthatelementand . ApplyingthesameprinciplesformHDandmHCwefind (10.52) ,and giving (10.53) giving . ItisonlymxHDwhichisappreciablyinfluencedbythisanalysis.Thereasonisthattheanglebetween theedgeandthexaxisiscloseto45inthiscase.Wherethisangleiscloserto0or90theinfluence issmaller. Figs10.5.3and10.5.4showpossibledistributionsofdesignmomentsforreinforcementinthexand ydirections respectively. The general recommendations regarding distribution have been followed. The width over which the support reinforcement in the xdirection has been distributed is approximatelyequaltothelengthoftherelevantsupportlineofzeroshearforceinFig.10.5.2.The distributionwidthforthesupportreinforcementintheydirectionhasbeenchosenthesameasinthe xdirection,cf.Section8.1.5. Thewidthsofdistributionofspanreinforcementarebasedonthepatternofspanlinesofzeroshear forceinFig.10.5.2andarealwayschosentobeonthesafeside.

Fig.10.5.3

Fig.10.5.4

10.6 Edge curved and column supported

Example10.4
131

TheslabinFig.10.6.1isthesameasinthepreviousexamplewiththeexceptionthatthecurvededge issupportedbysquarecolumns situatedwiththeirinneredgesatthelineofsupportinthat example.Thecoordinatesofthecentresofthecolumnsaregiven.Theslabisassumedtobesimply supportedonthecolumns. Thepatternoflinesofzeroshearforceinsidetheslabistakenunchangedfromthepreviousexample. Fig.10.6.2showsthespanlinesandalsothesupportbandswhichcarrythesupportreactionstothe columns,shownasdashedlinesbetweencolumncorners.Thesupportbandsareassumedtomeetthe fixededges0.3mfromthecorners. The shapes of the elements closest to the free edge are not exactly the same as in the previous example. So, for example, the moment on span line ab comes from two elements with slightly differentdirectionsofthesupportlines.ThemomenttotheleftofBis .With forthat support line we also find . For the part to the right of B we find . With and (cf.Eq.(10.52))weget .Theaverage onlineabis andusingtheother value fromthepreviousexampleweget

Fig.10.6.1

Fig.10.6.2

. In the same way we find and . These values are slightlylowerthaninthepreviousexample,whichmaybeexpectedasthedimensionsoftheelements aresmaller. The load on the support bands corresponds to the load on the elements outside the line abcd supplemented by the lines from a and d to the intersections between the support bands and the supports.Theloadisassumedtobecarriedatrightanglesto thesupportbands.Theresultingload distribution on the support band is shown in Fig. 10.6.3, where a series of linear approximations is used.
132

 Fig.10.6.3 The calculation of moments in the support bands can be performed according to the theory of elasticityorinthesamewaythatisusedfortheelementsinsidetheslab,i.e.withassumedpositions ofpointsofzeroshearforce.Inbothcasesthesupportmomentsshouldbereducedaccordingtothe rules in Eqs (10.22) and (10.23). The method with assumed lines of zero shear force will be demonstrated. The angles between meeting support bands are 170.0 at B, 165.3 at C and 147.5 at D. The correspondingreductionfactorsforsupportmomentsare0.778atB,0.673atCand0.278atD. We can start by assuming that all points of zero shear force are at the centres of the spans, even thoughitcaneasilybeestimatedthatthesearenotthecorrectpositions.Simplestaticsthengivesthe followingrelations (10.54) (10.55) (10.56) Page195 (10.57) (10.58) (10.59) (10.60) (10.61) ApplyingtherecommendationsinSection10.2andthereductionfactorsgivenabovewecancalculate thefollowingmoments: , , . , . There are unacceptably large differences between the two values of span moments for MAB and for MDE. The positions of the points of zero shear force are changed in these two spans, while the positionsarekeptunchangedintheothertwospansforthenextcalculation.Anestimatebymeansof Eq.(10.2)showsthatitmaybesuitabletomovethepointofzeroshearforceinspanAB0.26mtothe rightandinspanDE0.35mtotheleft.Afterthischangeweget: . . ThedifferenceinspanmomentsisstilltoolargeinspanDE.Thepointofzeroshearforceismoved 0.28mfurthertotheleft.AtthesametimethepointofzeroshearforceinspanCDismoved0.12mto therightinordertodecreasethedifferenceinthatspan.Afterthesechangeswegetthemoments: .
133

, . Thesevaluesarequiteacceptableandmaybeusedforthedesign. The difference in direction of moment vectors on both sides ofthe columns gives rise to a moment according to Eq. (10.24), which has to be balanced by reinforcement. Thus at B, for instance, this moment is . At C the moment is 7.5 and at D 4.8. These moments should be takenbybottomreinforcementatrightanglestotheedge. Fig.10.6.4showsthemomentsinthesupportband.ThevaluesareforMinkNm,notforminkNm/m as in most other corresponding figures. The moments have been shown distributed on a width of about0.6m,whichmaybesuitable.Alternativewidthsandpositionsareofcoursealsopossible.At the columns it has been indicated that there are two different directions of the reinforcement corresponding to the support moments. Two different systems of reinforcement bars may be arranged,whichshouldbewellanchored,orbarsmaybehorizontallybentoverthecolumns.Inthe latter case some secondary reinforcement might be needed to avoid the risk of splitting due to the radialpressureagainstthebend.

Fig.10.6.4

10.7 Slab cantilevering outside columns

Wheretheslabcantileversoutsidecolumnsatleasttwodifferentapproachesmaybeused.Eitheruse canbemadeofsupportbandswhichsupportcantileversapproximatelyatrightanglestothesupport bands,orthetreatmentcanbebasedonthedesignprocedurewithreinforcementintwodirections, described in Section 10.2. In Example 10.5 both approaches are discussed and an analysis with the latter approach is demonstrated, whereas in Example 10.6 three different approaches are demonstratedindetailandcompared ItmustbenotedthattheapproachaccordingtoSection10.2doesnotautomaticallygiveresultson thesafeside,astherulesforacceptablemomentdistributionsmaybeclearlyviolated.Inparticular, theremaybesomenegativecantilevermomentswhichareunderestimatedwiththismethod,leading toariskofuncontrolledcracking.ThiswillbecommenteduponinExample10.6.However,thesafety againstcollapsewillprobablyalwaysbeadequateiftherecommendationsarefollowed. A particular problem which has to be taken into account is the case of reinforcing bars, normally bottom bars, which reach a free edge soon after crossing a line of zero shear force. As the reinforcementisassumedtobeabletoyieldatsuchalinethereneedstobeacertaindistancetothe freeedge,wherethesteelstressiszero.Inordertomakesurethatthereinforcementisfullyefficient inthelinesofzeroshearforceitisrecommendedthatthefollowingruleshouldbeapplied:
134

Areinforcingbarisonlytakenintoaccountinthedesignifithasalengthfromthelineofzeroshear force tothe point where it is terminated ata freeedge which is at least equal to l/8, where l is the lengthofthespanwherethebarisactive. Thisruleisnotveryprecise,asitisnotalwayspossibletodefinetherelevantvalueofl,butitcanbe expectedtogiveresultsonthesafesideinmostsituations.Incaseofdoubtavalueonthesafesideis chosen. Example10.5 TheslabinFig.10.7.1hascolumnsplacedasinthepreviousexample,buttheslabcantilevers1.5m outsidethecolumnswhichstoodalongtheedgeinthatexample.Thecolumnshavethesamecentres, but have their sides along the directions of the coordinate axes. All columns have a square section withsides0.3m.Theloadis10kN/m2,asinthepreviousexamples.Thefreeedgehastheshapeofan ellipse.

Fig.10.7.1 Fig.10.7.2showstheprinciplesoftheapproachwithasupportbandalongtheouterrowofcolumns. The treatment is very similar to that in Example 10.4 and the numerical calculations w ill not be shown. The main difference from Example 10.4 is that the cantilevering parts give rise to negative momentsoverthesupportband(correspondingtotopreinforcementperpendiculartothefreeedge), whichhave to be taken intoaccount in writing the equilibrium equations for the oneway elements inside the support band. It is recommended that the top reinforcement is arranged with a certain concentration towards the columns. The support band will take a substantially higher load than in Example10.4.

Fig.10.7.2 Fig.10.7.3showslinesofzeroshearforcewhentheslabisdesignedaccordingtoSection10.2,withall reinforcement parallel to the coordinate axes. The system of lines of zero shear force has been
135

determined according to the principles in Section 10.2. After some adjustments of the positions the followingcoordinatesofthepointsofintersectionhavebeenchosen: a (2.80/7.80), b (8.80/7.80), c (9.60/5.40), d (14.60/6.60), e (16.10/1.80), f (9.00/2.40), g (2.80/3.20),h(3.00/11.69),i(9.45/10.63),j(15.51/8.29),k(20.76/3.60). When establishing the equilibrium relations for the elements next to the free edge it is not evident howmuchofthespan(bottom)reinforcementshouldbetakenintoaccount.Tobeonthesafeside,it hasbeenassumedthatalltheloadiscarriedbythesupportreinforcementincaseswherethereisa smallanglebetweenthesupportlineandtheedge,whichis

Fig.10.7.3 thecaseforreinforcementinthexdirectionatcolumnDandintheydirectionforallthreecolumns B, C and D. This is done in order to eliminate the risk of too little top reinforcement in the cantileveringpart. For the elements where a free edge forms a small angle with the reinforcement direction the recommendationabovehasbeenapplied.Therelevantcasesarereinforcementinthexdirectionnear points h, i and j. In all three cases the value of l can be taken to be approximately 6.0 m and the distancefromthelineofzeroshearforcetothefreeedgenotmorethanabout0.75m.Dependingon theslopeofthefreeedgethereductioninwidthonwhichthereinforcementisactiveisfoundtobe 0.06math,0.20matiand0.36matj.Thesereductionshavebeenusedforthedeterminationofthe activewidthsforthemvaluesintheequationsbelow. Thefollowingequilibriumrelationshavebeenfound: (10.62) (10.63) (10.64) (10.65) (10.66) (10.67) (10.68) (10.69)
136

(10.70) (10.71) (10.72) (10.73) (10.74) (10.75) (10.76) (10.77) (10.78) (10.79) (10.80) (10.81) (10.82) (10.83) (10.84) (10.85)

(10.86) From these relations we can calculate the moments according to the recommendations in Section 10.2: , , . , , , . Fig.10.7.4showspossibledistributionsof design momentsforreinforcement inthe xdirectionand Fig.10.7.5fortheydirection.Forthereinforcementinthexdirectionsomesimplificationshavebeen usedformxA,wherethelargervaluemxA1onlyhasbeenused,formxABandmxAI,wherethevalue16.0 hasbeenusedforboth,andformXBC,mxICandmxIH,wherethevalue22.0hasbeenusedforallthree. Thechosenvaluesaresomewhatarbitrary,butonthesafeside.Somecorrespondingsimplifications mighthavebeenmadealsofortheydirection,butthedifferencestherearegreater.

137

 Fig.10.7.4 The support moment at E has been given a concentration towards the end where it meets the free edge.Thereasonisthatithasbeenestimatedthattheslabhasasmallcantileveractioninthatpart. The support moments over the columns have been distributed according to the general recommendationsonapproximatelythewidthofthecorrespondingsupportlinesofzeroshearforce. Itis,however,onlymeaningfultoarrangereinforcementwherethereisasuffi

Fig.10.7.5 cientlengthofreinforcingbar.ThisisthereasonwhythedistributionisunsymmetricalforMxD,MyC andMyD. In addition to the reinforcement for the design moments at least one top bar and one bottom bar shouldbeplacedalongthefreeedge.Atleastsomeofthetopreinforcementshouldbebentaround thesebarsaccordingtotheprincipleillustratedinFig.2.10.2. Example10.6 The triangular slab in Fig. 10.7.6 has all edges free, and is supported on three square columns .Itissymmetricalwithrespecttothexaxis.Thepositionsofthecornersandthecentresof thecolumnsaregivenascoordinates.Theloadis9kN/m2. Three differentpossibilities for designing this slab will be discussed. In all the solutions use will be madeofthesymmetrysothatonlytheupperhalfoftheslabisanalysed. Approach1 ThefirstapproachisadirectapplicationofthedesignprocedureaccordingtoSection10.2,withall reinforcementparalleltothecoordinateaxes.Thecorrespondingpatternoflinesofzeroshearforceis showninFig.10.7.7.Theonlypointswhichhavetobechosenareaand

138

 Fig.10.7.6 b.Accordingtotherecommendationsthelinebetweenthesepointshasadirectionapproximatelyat right angles to the line between columns A and B. The positions have been determined so that the span moments mxf from the two elements on both sides of the line areapproximately equal. In this waysuitablecoordinateshavebeenfoundtobe: a(4.8/2.4),b(4.0/0). ThesupportlineoftheelementtotheleftofcolumnAhasalengthof3.6m.ApplyingEq.(2.5)wefind (10.87) The support line of the element to the right of column A has a length of 3.5 m. As the span reinforcement meets the edge at a skew angle the recommendation on page 197 is applied. The relevantlengthlis6.3minthiscaseandtheminimumlengthofreinforcingbarsfromthelineofzero thereductioninactivewidthis0.27m shearforcetotheedgeisabout0.8m.Withtheslope andtheactivewidth .FromEq.(2.6)wethenfind

Fig.10.7.7
139

(10.88) Asweanalysetheupperhalfoftheslabweincludehalfthewidthsoftheonewayelementsinthex direction at column B. For the element to the right of column B we then find that the length of the relevantsupportlineis1.3m.FromEq.(2.4)wefind (10.89) Itshouldbenotedthatthisisonlyhalfthemomentabovethecolumn. FortheelementtotheleftofcolumnBthelengthofthesupportlineis1.4m.Theactivewidthofthe spanreinforcementistakentohavethesamevalueasabove,2.13m.WefindfromEq.(2.6) (10.90) . FromEqs(10.87)(10.90)wefind ThelengthofthesupportlineoftheelementabovecolumnAis4.78m.FromEq.(2.6)weget (10.91) ThelengthofthesupportlineoftheelementbelowcolumnAis4.68m.FromEq.(2.5): (10.92) . andso ThelengthofthesupportlineoftheelementabovecolumnBis7.50m.FromEq.(2.4): (10.93) Wenowknowallthedesignmomentsandjusthavetochooseasuitabledistributionaccordingtothe generalrecommendations.SuchadistributioncanbefoundinFig.10.7.10below. Approach2 InthisapproachuseismadeofasupportbandbetweencolumnsAandB,cantileveringtotheleftofA, Fig.10.7.8.Thesupportbandisdividedintotwobands,eachpassingthroughacomerofeachcolumn and passing through the column section. In this way the static system becomes clear. A somewhat more economical assumption might have been to let the bands rest on column corners without passingthroughthecolumnsections. TheparttotherightofcolumnBisassumedtobebalancedbyacorrespondingparttotheleftofthe column.Theloadswithintheseareasaretakenbyreinforcementinthexdirection.Intheremaining part of the slab the load is assumed to be primarily carried by reinforcement in the ydirection in stripswhichrestonthesupportbands. It would have been somewhat more economical and statically correct to use reinforcement perpendiculartotheedgestotakethemomentsinthecantileveringpartsoftheslab.Thisispossible, butitleadstoamorecomplicatedstaticsystembecausethestripswhichhavetobalanceeachother insidetheslabwillformananglewitheachother,causingaresultingmomentwhichhastobetaken intoaccount. ThemomentMxsBcanbetakenfromthecalculationinApproach1above,andisthus29.66onthe upperhalfoftheslab. The support reaction from the triangular elements on the oneway strip in the ydirection is for .Thisgivesamomentof13.69kNm.Tothisshouldbeaddedtheinfluenceof thedirectloadontheonewaystrip,whichgivesamomentofapproximately .The . totalmomentisthus

140

 Fig.10.7.8 Theloadonthewholeslabisprimarilycarriedintheydirectionforx<3.9.For3.9<x<7.8theloadis primarilycarriedintheydirectionfory>(x3.9)/3.Thecantilevermomentsinthesestripsare (10.94) (10.95) Thepositivemomentsintheystripsare (10.96)

(10.97) Themaximumvalueofmyfis22.23kNm/mandthetotalmomentis28.90kNm. Theloadsonthesupportbandsare (10.98) (10.99) Theanalysisofthebandsissimpleinprinciple,astheyarestaticallydeterminatewithzeromomentat B,buttheanalysisisratherlengthy.Theresultisasupportmoment andamaximumspan moment . AdistributionofdesignmomentsbasedonthisapproachcanbefoundinFig.10.7.10. Approach3 This approach follows the recommendations in Section 10.2 with the exception that the two reinforcementdirectionsarenotatrightanglesbutareinsteadparalleltothefreeedgesinthemain partoftheslab.Onedirectionisparalleltotheyaxis.Theotherdirectionisdifferentintheupperand lower half of the slab, as it follows the nearest free edge. The upper half will be analysed and the reinforcementdirectionstherefollowthezandyaxesinFig.10.7.9,whichalsoshowsthepatternof linesofzeroshearforce.AbovecolumnB,however,thereinforcementdirectionsfollowthedirections ofthexandyaxesinordertoavoidhavingthreecrossinglayersofreinforcement.Astheelementsto therightofthatcolumncarryalltheloadasacantilevertheloadbearingdirectionsintheseelements followthexandydirections,asindicatedinthefigure. Thetheoreticalsupportareaistakenastheinscribedareawiththesidesparalleltothereinforcement directions.Otherapproachesarealsopossible,whichmaydecreasethemomentssomewhat,butthe differenceissmallandthechosenapproachisinaccordancewiththegeneralrecommendations.
141

It is possible to use directions of support reinforcement over columns A and C which are at right anglestotheedgeinoneorbothdirections.Thiswouldmeanasmallreductionintotalreinforcement, butwouldhavethedisadvantagesthatthereinforcementwouldnotbeasactiveincontrollingcracks atrightanglestotheedgeandwouldnotreachtheverycorner. In applying the formulas from Section 2.3 there are two possibilities when the reinforcement directionsarenotatrightangles.Onepossibilityistousethecvaluesinthedirectionoftherelevant reinforcement and the lvalues at right angles to that direction. Another possibility is to take the momentequationaroundthesupportlineandusetherelevantcomponentsofthedesignmomentsto balancethismoment.Bothapproachesrequirethattheotherreinforcementdirectionisparalleltothe support line. This requirement is not fulfilled at column B, where three different reinforcement directionsmeet.Thisproblemwillbediscussedlater. It can be established that suitable coordinates for points c and d are the same as for a and b in Approach1,thus c(4.8/2.4),d(4.0/0). FortheelementtotheleftofcolumnAitissimplesttotakethemomentaroundthesupportline.The valueofthemomentisexactlythesameasinApproach1above,sowecanusethatvalue.

Fig.10.7.9

(10.100) we get , which is the same value as the support moment in the support With bandinApproach2. FortheelementtotherightofcolumnAwecanusethesameapproach.Theactivewidthofthespan reinforcementistakenas2.5m. (10.101) . whichgives Thesupportmomentis onthewidthoftheupperhalfoftheslabjustasintheprevious approaches.FortheelementtotheleftofcolumnBwemayagaintakethemomentaroundthevertical support line and make use ofthevalue fromApproach 1. It should be noted that mzf andMxsB have differentdirections. (10.102) . Using also the value 35.55 from above we can, according to the which gives . recommendation,calculatethedesignvalue
142

IncalculatingthemomentsforthereinforcementintheydirectionfortheelementabovecolumnAit is simplest to take the cvalue in the ydirection and the lvalues in the xdirection. The projected lengthinthexdirectionofthesupportlineis4.43m.FromEq.(2.5)weget (10.103) For the element below column A in the figure, the projected length of the support line is 4.37 m. Assumingthattheeffectivewidthofthespanreinforcementis4.0m,Eq.(2.6)gives (10.104) . fromwhich AtcolumnBthesituationissomewhatcomplicatedforcalculatingthemomentforthereinforcement intheydirection,astherearemomentscorrespondingbothtothexandzdirections.Wecanstartby analysingtheinfluenceoftheloadonthetriangletotherightofthelinethroughtheleftsideofthe column (this triangle is composed of the oneway strip and the cornersupported triangle), which givesnoproblem.Thisgivesrisetothemoment (10.105) For the trapezoidal element to the upper left of column B we take the moment around the support line. This equilibrium is influenced by the part of the moment MxsB which acts along the side of the element. It is not quite clear how this moment influences the equilibrium, as some of the corresponding reinforcement also passes the support line around which the equilibrium is studied, and helps in taking some of the support moment in this line. A comparison can be made with the solution according to Approach 2, where the element to the right of column B is balanced by a correspondingelementtotheleft.Inthatcaseevidentlysomereinforcementinthexdirectionpasses thesupportbandandhelpsintakingasupportmomentaroundthesupportband. Anassumptiononthesafesideistoassumethatthereinforcementinthexdirectiondoesnottake anypartin carryingthesupportmomentaroundthesupportline.Thepart ofMxsB whichhastobe taken into account comes from the triangle to the upper right of the column and is . In applying Eq. (2.5) around the support line we note that and : (10.106) .Addingthetwopartsweget . Thisgives ThereisnopositivespanmomentmyfintheelementtotheleftofcolumnB. ApossibledistributionofdesignmomentsisshowninFig.10.7.10. Fig. 10.7.10 shows the proposed distributions of design moments according to the three different approaches. As well as the reinforcement for the design moment there should also be some reinforcementtominimizetheriskoflargecracks.Thus,thereshouldbesomereinforcementalong thefreeedgeswheresuchreinforcementisnotgivenbythedesignmoments.OnlyApproach3gives such reinforcement along the sloping edge in the figure, but this approach does not give any such reinforcement beside column B. It may also be appropriate to put minimum reinforcement in some areaswherethedesignmomentsdonotrequirereinforcement. Comparison AcomparisonbetweentheresultsshowsthatApproach1givesratherlittlesupportreinforcementin theydirectioninthelefthandpartoftheslabandmuchintherighthandpartandleavesapartin betweenwithoutsupportreinforcement.Italsogivesspanreinforcementinthexdirectionwhichis toounsymmetricalwithrespecttocolumnA.Forboththesereasons,someadditionalreinforcement hastobeusedinordertominimizetherisksofcrackingandmaybeeventoensureadequatesafety.
143

ThesolutiongivenbyApproach1isnottoberecommended,atleastnotwithoutintroducingsome additional reinforcement, based on an estimate of the behaviour of the slab. One reason why the approachisnottoberecommendedhastodowiththefactthatthereisnotmuchredundancyinthe structure,asthesupportmomentsarestaticallydeterminate,unlikeinthepreviousexamples. WhencomparingtheresultsofApproaches2and3,whichbothcertainlygiveadequatesafety,itmay benotedthatApproach3givesareinforcementdistributionwithacertainconcentrationaroundthe columns, but also with some reinforcement over a greater width. Approach 2 gives a heavy concentrationalongthesupportbandbutnoconcentrationintheydirectionatcolumnA.Ifthistype ofsolutionisusedindesignitisrecommendednottofollowthistheoreticaldistributionstrictly,but toredistributesomeofthereinforcementinthesupportbandoveralargerwidthandtoconcentrate someofthesupportreinforcementintheotherdirectionoverthecolumns. ItwouldseemthatApproach3is,inthefirstplace,toberecommended.

Fig.10.7.10

144

 CHAPTER11 Lshapedslabsandlargewallopenings
11.1 General
Theslabsinthischaptercanbetreatedbymeansofsupportbands,buttheapproachusedherewith cornersupported elements is simpler, more systematic and safer, as the wellestablished rules for suchelementscanbeapplied. However,the static system is more complicated than in a regular flat slab, for example. There is no element to take the support moment from the cornersupported element, but this support moment hastobetransferredtothewallinthedirectionofthereinforcement.Thusforelement3inFig.11.1.1 theforcesinthesupportreinforcementinthexdirectionhavetobebalancedbyarathercomplicated momentdistributionwithinelement2,carryingtheforcestothesupportforelement2.Ithasbeen demonstrated in Strip Method of Design that it is possible to find such complete moment distributions,whichareonthesafeside,butthatitissufficienttoanalyseonlytheelementsshownin Fig.11.1.1. The moment distribution in the service state in the cornersupported element in Fig. 11.1.1 is different from that in a regular flat slab. The support moment can be expected to be more concentratedtowardsthesupport,whereasthenegativemomentattheotherendofthesupportline israthersmall.Morereinforcementshouldbeconcentratedinthevicinityofthesupportedcorner.It isrecommendedthatadesignsupportmomentdistributionshouldbeused,forexample,accordingto Fig.11.1.2,wheremavistheaveragesupportmoment.Thelengthofthereinforcingbarsintoelement 2inFig.11.1.1maybetakentobethesameasintoelement3,whicharedeterminedfromtherulesin Section2.10.2.

145

 Fig.11.1.1 Thesupportmomentinthecornersupportedelementunderserviceconditionsissmallerthanthat correspondingtoafixededge,asitisonlyfixedatthecorner,andmayrotatearoundthesupportline further away from the corner. The support moment should thus be chosen lower than that corresponding to a fixed edge. In the examples the support moment has been taken as 75% of the momentcorrespondingtoafixededge,whichhasbeenestimatedtobeasuitablevalue. In the examples the theoretically calculated support moments have been used. As always it is permissibletoincreaseordecreasethesevalues,e.g.inordertolimitthecrackwidthsinthetopor bottomsideoftheslabortomakeuseofminimumreinforcement.

Fig.11.1.2 The triangular end elements 1 and 5 in Fig. 11.1.1 have no direct counterpart to take the span moments,butinpracticethesemomentscanbetakencareofintheslab,andthedesigncanbebased ontheequilibriumoftheseelements,calculatedfromEq.(2.4).Someparticularproblemsrelatedto thistypeofelementarediscussedinExample11.4. Thedesignmomentsinthetriangularcomerelementsinparallelwithelements2,9,4and6inFig. 11.1.1 may be taken as onethird of the moments in the latter elements, cf. Section 8.1.6. This also holdsforthecornerbetweenelements7and8.

11.2 Reentrant corner

146

Example11.1 Fig.11.1.1showsaslabwithareentrantcornerandtheelementsusedforitsdesign.Theslabhasfour fixedandtwosimplysupportededges.Theloadis9kN/m2. Thesupportmomentsatthefixededgesarecalculatedbymeans ofstandardformulasaccordingto the theory of elasticity. The support moments for the cornersupported element 3 is reduced as describedabove.Asstrip37isnotfullyfixedattheleftendandalsomayhaveadeflectionatthat end,thesupportmomentforelement7isintermediatebetweenthatforstripswiththeoppositeends fixedorsimplysupported.Thecvaluesforelements2and3arecalculatedfromEq.(2.34)andthe corresponding span moments from Eq. (2.35). The cvalues for the triangular elements 1 and 5 are estimatedusingtothesameprinciplesasforrectangularslabs.Thus and areestimated tobesatisfactoryvalues.Becauseofsymmetry . (11.1) (11.2) (11.3) (11.4) (11.5) (11.6) (11.7) (11.8) (11.9) (11.10) (11.11) (11.12) (11.13) .Thegreatvalueoftheratioischosenbecausetheelementisanend

(11.14) Wemaychoose elementinalongstrip.

147

Fig.11.2.1showsthedistributionofdesignmoments.Someofthemomentsaresosmallthatrulesfor minimum reinforcement may give higher values. If this is the case the values of c1 and c5 may be increasedinordertomakeuseofthereinforcement.

Fig.11.2.1 Thesupportreinforcementinelements2and4maybemoreconcentratedtowardsthecorner,where theriskofcrackingishighest.Theoretically;theseelementsmaybetreatedas cornersupportedatthe With regard to checking permissible moment distributions in the cornersupported element see Example11.3below.Theconditionsaremetquitesatisfactorily.
corner,whichgivessuchaconcentrationofreinforcement.Inthenextexamplesuchaconcentrationhasbeen chosen.

11.3 Supportingwallwithalargeopening
Where a supporting wall has a large opening the situation is similar to that at a reentrant corner regarding design moments for reinforcement in the direction of the wall. For the reinforcement at rightanglestothatdirectionthesituationforthecornersupportedelementismoresimilartothatin acolumnsupportedslab.

11.3.1 Innerwall
Example11.2 Fig.11.3.1showsaslabwithaninnersupportingwallthatextendstolessthanhalftheslabwidth.The loadis9kN/m2.Theelementsarealsoshown. Elements 5, 1, 2, 3, 4 and 8 form a continuous strip. The average moments in this strip may be calculatedinthesamewayasforaflatslab: (11.15) (11.16)

148

(11.17) (11.18) (11.19) (11.20)

Fig.11.3.1 Elements1,3(withintermediateonewaystrip)and9formastrip.Withtherecommendationgiven inSection11.1theaveragemomentsinthestripcanbecalculated: (11.21) (11.22) (11.23) Ifwechoose weget

(11.24) Fig.11.3.2showsa possibledistributionofthedesignmoments.ThedistributionofmyAfollowsthe recommendation in Section 11.1 with a slight modification for the oneway element between the cornersupported elements. For the distribution of mxA first 2mxA has been distributed on half the widthofthecornersupportedelements.Thesamemomentintensityhasalsobeenusedonthefirst 0.6moverthewallandmxA/3overtheremainingpartofthewall.Suchadistributioncanbesaidto correspond to an assumption that elements 2 and 4 act as comersupported. It may also be better fromthepointofviewofcrackingthanadesignmomentmxAonthewholewidth(1.5m)ofelements 2and4.
149

Fig.11.3.2

11.3.2 Wallalonganedge
Example11.3 TheslabinFig.11.3.3hasthreefixededgesandoneedgewhichispartlysupportedbyawall.Theslab is assumed to be freely supported at the wall. If it had not been assumed to be freely supported elements2and3wouldhavehaddifferentlengthsintheydirectionandelement5wouldhavehadto bedividedintotwoelements.Theloadontheslabis9kN/m2.

Fig.11.3.3 Asuitablevalueofc1isestimatedtobe3.0m.Thisgives

(11.25) Becausethisisanendelementinalongslabtheratiobetweensupportandspanmomentsischosen tohaveahighvalue.Suitablevaluesare . Accordingtotherecommendationsaboveweget (11.26)

150

(11.27) (11.28) (11.29) (11.30) (11.31)

(11.32) Based on these values a distribution of support moments has been proposed in Fig. 11.3.4. For the supportmomentmxs3adistributionwitharatioof2betweenthehigherandthelowermomenthas been chosen, which means a little less concentration than in the previous example. The difference betweenthesedistributionsprobablyhasnopracticaleffect. For the cornersupported element 3, a check should be made that the conditions in Section 2.5.2 regarding moment distributions are met. In the xdirection twothirds of the support moment are taken with and onethird with . Taking a weighted average, this means that the lower .Withthedistributionshownweget .The limitationis conditionismetquitesatisfactorily.Wemightevendecreasethespanmomentsomewhat(to7.2)and putabandofreinforcementalongthefreeedge,whichmaybeadvantageousforcracklimitation. Intheydirectionwehavenosupportmomentinelement3andwethushavetofulfi1thelimitations withthedistributionofspanmoment.Withthedistributionshownwehave for ,whichis withinthelimits.Alsothespanmomentinelement2hasbeengivenacorrespondingdistribution. Thedistributionofsupportmomentsalongtherighthandedgehasbeengiveninaccordancewiththe general rules. This distribution is, however, not in a good agreement with the expected moments accordingtothetheoryofelasticity.Iftheedgehadbeenfreely

Fig.11.3.4 supportedtheangleofrotationwouldhavebeengreateratthecentreoftheedgethanattheupper end, as the slab deflects more in the central part than at the end of the supporting wall. Thus the supportmomentforthefixededgeisgreateratthecentrethanattheupperend,whichisoppositeto the moment distribution shown. For that reason it might be better to make a redistribution of the
151

designmoments,e.g.asshowntotherightinthefigure,particularlyifcracksatthesupportaretobe limited. Example11.4 TheslabinFig.11.3.5isthesameasinthepreviousexamplewiththeexceptionthatthewallopening hasbeenmoved1.0mtotheleftandalengthofwallhasbeenintroducedattherighthandendwhich is only 1.0 m long. Such a short wall length beside an opening introduces the problem that the triangular element 6 has to have such a small height c6 that the moments in that element are unrealistically small. It is necessary to find a procedure for dealing with this case, which forms a continuouslinkbetweenthecasescorrespondingtothelefthandpartandtherighthandpartinthe previous example, i.e. with a long wall and no wall respectively.The following procedure is recommended. Theloadqonelement4isdividedintotwoparts.Onepartkqiscarriedonlyinonedirection,justlike thewhole loadinthepreviousexample.Thestripcarryingthisload issupportedon therighthand supportandhasaspan .Theotherpart(1k)qis

Fig.11.3.5 carriedonelement4actingasacornersupportedelement.Thestripcarryingthisloadissupported . ontheendofthewallandhasaspan Asuitablevalueofkis (11.33) wherecisthelengthofthewallpart(correspondingtothepossibleheightc6ofthetriangularelement 6)andc0istheheightofthetriangularelement,whichwouldhavebeenchosenifthewallparthad beenlong.Inthiscasethevalueofc0correspondsto ,whichgives .Thisisofcoursenot anexactvalue,butitwillbeusedfortheanalysis. Asanapproximation,thesupportmomentsaregiventhefollowingvalues: (11.34) (11.35) (11.36) ThereactionatsupportAinkN/mis

152

(11.37) whichgives (11.38)

(11.39) Thesupportmomentmxs4cannotfalltozerojusttotherightofpointB.Ifc6issmallsomepartofthat momentistransferredtoC.ItisproposedtoassumethatthefollowingvalueisaddedtomCtotake intoaccountthateffectwhen : (11.40) Wealsohavetotakeintoaccountthemomentsinelement6: (11.41) Justasforelement1,aboutthreequartersofthisvalueistakenasthesupportmoment,whichgives .ThetotalaveragemomentatC(onthewidth )isthus (11.42) Theaveragemomentintheremainingpartoftherighthandsupportcanbetakenas . Theformalmomentdistributionattherighthandsupportisveryunevenaccordingtothisanalysis, with21.49atCbutonly2.26onthewidthcorrespondingtoc5.Inorderto limit cracking at the support the distribution of support moments has to be changed. A suitable distributionisproposedinFig.11.3.6. Element7iscooperatingwithelement4incarryingthepart(1k)ofthetotalload.Thispartofthe loadgivesmomentsmywhichare(1k)timesthemomentsinelement5.Fortheremainingpartofthe loadthesemomentsareasusualtakenasonethirdofthemomentsinelement5.Thetotalmoments my in element 7 are thus times the moments in element 5. The moments in element5arethesameasinthepreviousexample. (11.43) (11.44) Themomentmxf7is,accordingtonormalrules,takenasmxf3/3. AsuitabledistributionofdesignmomentsisproposedinFig.11.3.6.Itisbasedonthevaluesabove andvaluesfromthepreviousexample,wheretheseareunchanged.

153

Fig.11.3.6

11.3.3 Slabcantileveringoutsidewall
Example11.5 TheslabinFig.11.3.7cantilevers2.0moutsidethewallwiththeopening.Otherwisetheslabisthe sameasinthepreviousexample.Theloadisalsothesame,9kN/m2.Thecantileveringparthasafixed edgetotheleftinthefigure,whereastheothertwoedgesarefree.

Fig.11.3.7 Theelementsarealsoshowninthefigure.Thevalueofc1hasbeenchosensomewhatlowerthanin thepreviousexamplebecausethereisasupportmomentatthewallwhichdecreasesthedeflection andalsomakesitlesseconomictocarryloadinthexdirection.Theequilibriumequationforelement 1is (11.45) Asuitablechoiceis .

154

As c1 is assumed to correspond to c0 in Eq. (11.33) we get a different value of k from that of the previousexample: (11.46) (11.47) (11.48) (11.49)

(11.50) (11.51) (11.52)

(11.53) .AddingthedifferentmomentsatCweget Asinthepreviousexample,wehave (11.54) The average moment in the remaining part of the righthand support can be taken as . Inthecantileveringpartthemomentinelement10isthesameasin1.Forelements12and13the supportmomentsarecalculatedjustasinthepreviousexamples: (11.55) Becauseofsymmetrythecxvaluesforelements12and13are2.6m,whichisthesamevalueascx3 . fromthepreviousexample.Thespanmomentisthesameasinthepreviousexample, Thecantileveringslabgivesanaveragesupportmoment,validforelements1114: (11.56) According to the theory of elasticity, this support moment causes a change of +9.00 of the support momentforelement5,whichthusbecomes(seeExamples11.3and11.4) (11.57) Thisgives (11.58) (11.59)

155

As in the previous example, the moments in element 7 are taken as the moments in element 5 multipliedby(12k/3),whichinthiscaseis0.76: (11.60) (11.61) ApossiblemomentdistributionisshowninFig.11.3.8.Forclaritythedistributionofmomentsover thesupportingwallanditsopeninghasbeenshownseparated. Itshouldparticularlybenotedthatthereisnospanmomentmyinthecornersupportedelements12 and13.Thismeansthat,accordingtotherulesinSection2.5.2,someofthesupportmomenthastobe distributed on the whole width of the element. In this case much of the moment has been concentratedneartheendofthewalls,correspondingto ,whichrequires>0.31.This conditionismetquitesatisfactorily. Asalways,itisbesttohavemoreratherthelessreinforcementalongthelongfreeedge.Thishasbeen takenintoaccountintheproposeddistribution.

Fig.11.3.8

156

 CHAPTER12 Openingsinslabs
12.1General
How the design of a slab has to be modified because of an opening depends on the shape, size and positionoftheopening.Themostsignificantquestionishowmuchtheopeningisestimatedtochange thestaticbehaviouroftheslab,inparticularthestaticbehaviourinthevicinityoftheopening. Where the static behaviour of the slab is only slightly changed by the opening, the design may be basedontheanalysisoftheslabwithoutanopening.Thereinforcementwhichwouldbecutbythe openingmustbearrangedalongitsedgesandproperlyanchored. It is not possible to establish any simple wellfounded rules for when this approximate approach is applicable.Thefollowingsimpleruleisrecommended: Theapproximateapproachmaybeusediftheopeningcanbeinscribedinasquarewithsideequalto 0.2timesthesmallestspanintheslab.Itmay,however,notbeusedforcornersupportedelements, see Section 12.5, or for openings close to a free edge or in areas where torsional moments play an importantrole. Itcannot be provedby means ofthestripmethod thatthisrule alwaysleadsto safe results.Onthe other hand, it is probably never possible to prove by means of yield line theory that it may lead to unsafe results. From a practical point of view it seems that the rule is satisfactory. It is also in agreementwiththeacceptedpracticeofallowingsmallholeswithoutanystrengthanalysis. Fig.12.1.1a)showsasimplysupportedslabwithauniformloadandtheelementswhichwereused foritsdesign.Fig.12.1.1b)showsalimitingopeningforapplicationoftherecommendationabove.

Fig.12.1.1 Fig.12.1.1c)showsalargeropeningwithsupportbandsintroducedalongitsedges.Thelinesofzero shear force have been rearranged in order to simplify the numerical analysis. Some strips are supportedatoneendbyasupportband.Thesupportbandsareeitherassumedtobesupportedatthe edgesoftheslabor,alternatively,tobesupportedbysupportbandsintheperpendiculardirection. Fortheslabinthefigureitisreasonabletoassumethatthesupportbandsintheshortspandirection aresupportedattheedgesoftheslab,whereasthesupportbandsinthelongspandirectionmaybe partlyorcompletelysupportedbythesupportbandsintheshortdirection.Oftenamixtureofthetwo possibilitiesmaybeusedinordertogetasuitablerelationbetweentheamountsofreinforcementin thetwodirections. Intheanalysesthesupportbandsareassumedtohavezerowidthandtobesituatedexactlyalongthe edgesoftheopening(ortouchingtheopeningwheretheyarenotparalleltoanedge).Inpractice,the designmomentsinthesupportbandaredistributedoveracertainwidthwithinareasonabledistance fromtheedge,e.g.withinonethirdofthedistancetothenearestparallelslabsupport.Althoughsuch
157

adistributionisnotinstrictagreementwiththebasisofthestripmethod,itcanberegardedassafe.It isnotpossibletoshowbymeansofyieldlinetheorythatitisunsafe. In Fig. 12.1.2 the slab has fixed edges. In a) the slab is without any opening, in b) with an opening whichcutsmuchofthemainspanreinforcement.Astheedgesoftheslabarefixed,theloadcanbe carried on cantilevers from the edges. As the edges of the opening deflect, this deflection is mainly prevented by the cantilevering action. In this case the static performance is thus changed so that certain support moments increase and take care of most of the load, whereas much of the span reinforcementcanbeomitted.Inthefirstplacemuchofthespanreinforcementintheshortdirection maybeomitted,butsomespanreinforcementisneededalongtheloweredgeoftheopeninginthe figure. A support band is also shown along this edge, which takes care of the reaction from the triangularelementonthewidthoftheopening.

Fig.12.1.2 TheopeningsinFigs12.1.1and12.1.2havetheiredgesparalleltothemainreinforcementdirections. If this is not the case a choice must be made between support bands in the main reinforcement directionsorparalleltotheedgesoftheopening.Asarule,thefirstpossibilityisrecommended,asit givesthesimplestcalculationsandexecution. Independentoftheresultoftheanalysis,thereshouldalwaysbeatleastonebottomandonetopbar alongeachedgeofanopeninginordertodecreasecracking,particularlyatthecorners. Openingswithinthecolumnbandofacornersupportedelementposeparticularproblemsduetothe high torsional moments within this area. This is treated in Section 12.5. Openings at free edges are treatedinSections12.3and12.4.

12.2 Slabswithalledgessupported
12.2.1 Rectangularslabs
Example12.1 The slab in Fig. 12.2.1 is the same as in Example 3.1, but with a rather large opening ABCD. As in Example3.1theloadis9kN/m2.Thereisnolineloadalongtheedgesoftheopening. Thechosenlinesofzero shearforceareshown inthefigure,aswellastheassumed supportbands alongtheedgesoftheopening.

158

Fig.12.2.1 Thesupportreactionsfromstrips1,3,4and5onthesupportbandsarefoundbysimplestaticstobe (12.1) (12.2) (12.3)

(12.4) Thesupportbandsinthexdirectioncanbeassumedtobesupportedeitheronthesupportbandsat right angles or at their ends on the slab supports. A combination of these possibilities may also be assumed.Withtheproportionsintheactualcaseitseemsmostsuitabletoassumethatallormostof the load is carried by the support bands at right angles. It may well be assumed that all the load is carriedinthisway,butinordertodemonstratethemoregeneralcaseitwillbeassumedthat10%of theloadiscarriedtotheslabsupports. The static analysis of the support bands acted upon by these loads is trivial and only the resulting momentswillbegiven.Thefirsttermisthemaximummomentfromtheloadearnedbythesupport bandsatrightangles(span2.5m)andthesecondtermisthemaximummomentfromtheloadcarried totheslabsupports.Thesetwomaximaarenotsituatedatexactlythesamepoint,butitisalwayssafe toaddthem. (12.5) (12.6) The support bands in the ydirection are loaded by the loads RAD and RBC respectively, by atAandB,andby atCandD.Theresultingmomentsare (12.7) (12.8) Themomentsinthedifferentonewayelementsare (12.9) (12.10) (12.11) (12.12)
159

Thetriangularcornerelementswith haveaveragemoments3.38. With values of the widths of the support bands chosen partly with respect to the values of the moments,adistributionofdesignmomentsisshowninFig.12.2.2. Example12.2 TheslabinFig.12.2.3isthesameasinExample3.2,butwithanopening.Itisalsothesameasinthe previousexample,butwiththeupperandlefthandedgesinthefigurefixed. The fact that the edges are fixed changes the static behaviour of the parts of the slab next to these edges,asmuchoftheloadiscarriedbycantileveraction.Forthisreason,nosupport

Fig.12.2.2

Fig.12.2.3 bandisneededalongAB,asalltheloadonelement3canbecarriedsatisfactorilybythiscantilever action.Thisgivesasupportmoment (12.13) Themomentsinelements4and5arethesameasinthepreviousexample.Thisalsoholdsfortheir supportreactionsonthesupportbands, . Forthelefthandpartoftheslab,twopossibilitieswillbediscussed.Thefirstpossibilityisshownin themainfigure.Theloadsonelements1and7arecarriedinthexdirection,giving (12.14)

(12.15) Element 2 may be treated like element 3, giving the same support moment and no span moment. Element6may,asasafeapproximation,beassumedtohaveaspanmoment (12.16) onsupportbandAD.Onthesameassumptionasin Element7givesasupportreaction thepreviousexamplethat90%oftheloadonsupportbandCDiscarriedbythesupportbandsatright atthecrossingpoints,andthemomentinbandCDis angles,thesehavetocarry the same as in the previous example, , provided that the band is treated as if it were hingedatitsends.
160

Thesupportbandsintheydirectionareassumedtohavesupportmomentsatthefixededgeofthe slab.Thesesupportmomentsmaybechosentohavethevalue (12.17) Usingthisvaluewefind (12.18) (12.19) andthusasuitableratiobetweensupportandspanmoments. AnalternativesolutionforthelefthandpartoftheslabisshowninthelefthandpartofFig.12.2.3.In thatcaseelement1isassumedtocarryalltheloadasacantilever,giving (12.20) whichhappenstobethesamevalueaswiththefirstsolution.Fromthestripformedbyelements2 and6weget (12.21)

(12.22) As element 1 now does not give any reaction on support band AD this is only acted upon by the reactionforce7.59kNfromsupportbandCD.Choosing (12.23) wefind (12.24) whichmaybeacceptedasasuitableratiobetweensupportandspanmomentwithregardtothefact thattheloadisactingatsomedistancefromthesupport. Incomparingthetwosolutions,wefindthattheresultsarerathersimilar.Forthereinforcementin thexdirectionthesupportmomentisthesame.Thereisasmallspanmomentaccordingtothefirst solution but no span moment according to the second. With due regard to minimum reinforcement requirements,thisdifferenceis,inpractice,negligible.Forthereinforcementintheydirectionwecan compare the sum of design moments in the strip 26 and the support band AD. We then find the numerical sum of support moments to be in the first case and inthesecondcase.Forspanmomentswefind inthefirstcase inthesecondcase.Thesedifferencesareunimportant.Thesmallincreaseintotal reinforcement in the second case is compensated for by the lack of span reinforcement in the x direction. Either of the two approaches may be used without any noticeable difference in the behaviouroftheslab. Fig.12.2.4showsadistributionofdesignmomentsbasedonthefirstsolution.Thesupportmomentin asupportbandmaybedistributedoveragreaterwidththanthespanmoment,asthesupportactsas amomentdistributor.Thishasbeentakenintoaccountinthefigure.Asusual,itisrecommendedthat atleastonebottomandonetopbarareprovidedalongeachsideoftheopening. Example12.3 TheslabinFig.12.2.5isthesameasinExample12.1,butwithatriangularopeningABCinsteadofa rectangularone.Theloadisstill9kN/m2.Thesamesupportbandsinthedirectionsofthecoordinate axesareintroduced,butwithsupportbandsalongtheedgesACandBCaswell. TheloadonthetrianglesACaandBbCcanbeassumedtobecarriedineitherthexorydirection,or dividedbetweenthem.Inthiscase,theydirectionisthemainloadbearingdirection,becauseofthe shapeoftheslab.Thereforeithasbeenchosentocarryalltheloadonthesetrianglesinthatdirection.
161

Thus element 5 also includes these triangles and is spanning between the edge of the slab and the supportbandsACandBC.Elements1and4aresupportedonthesupportbandsintheydirection. The support reaction from element 5 on the support bands varies linearly from permetrewidthinthexdirection.Theaveragevalueis

Fig.12.2.4

Fig.12.2.5 10.80kNpermetrewidthandthedifferencebetweentheendsis8.1kNpermetrewidth.Thesupport reactionsfromthesesupportbandsare (12.25) (12.26) RAandRBarecarriedbythesupportbandsintheydirection.RCiscarriedbythesupportbandab, either only to a and b,where supported on the support bands in the ydirection, or with some part carryingtotheedgesoftheslab.Itis,asinthepreviousexamples,assumedthat10%iscarriedtothe edgesoftheslaband90%topointsaandb,andweget (12.27) (12.28) SupportbandAaisacteduponatAby7.59kNfromsupportbandACand5.57kNfromsupportband AB,seeExample12.1,ataby5.31kNfromsupportbandabandbetweenAandaby4.05kN/mfrom and the maximum element 1, see Example 12.1. These loads give the moments . moment In the same way, support band Bb is acted upon by 13.16 kN at B, 5.31 kN at b and 6.75 kN/m between B and b. These loads give the moments and the maximum moment . ThelengthofsupportbandsACandBCis
162

 (12.29) Theloadpermetrelengthofthesebandsisequalto1.25/ltimesthereactionspermetrewidthinthe .Themomentinthesupport xdirectiongivenabove.Theaverageloadisthus bandsis,withsufficientaccuracy(within2%) (12.30) Theaveragemomentinelement5canbecalculatedbymeansofEq.(2.8),takingintoaccountthefact that the cvalues in that formula correspond to half the lengths of the element, as the maximum momentissituatedatthecentreofthestrip. (12.31) As the lengths of the strips forming element 5 vary considerably from the centre to the sides the designmomentsshouldbechosenwithacorrespondingdistribution. Fig.12.2.6showsapossibledistributionofdesignmoments. ThisfigureshouldbecomparedtoFig. 12.2.2,fromwhichsomeofthemomentsaretaken. It is interesting to compare the amount of reinforcement in the slab with an opening according to Examples12.1and12.3withthatforthesameslabwithoutanyopening,Example3.1.Theamountof reinforcementisproportionaltotheaveragemomentoverthe

Fig.12.2.6 wholewidthoftheslab.Inthefirstplace,thereinforcementwhichisbesidetheopeningiscompared. FortheslabinExample12.1wefindtheaveragemoments (12.32) (12.33) FortheslabinExample12.3wefindtheaveragemoments (12.34) (12.35) The average moments for the slab without any opening are, according to Example 3.1, .Acomparisonshowsthatwewouldhavehadenoughtotalreinforcementbesidethe opening if we had used the simple rule of just moving the reinforcement which would have passed throughtheopeningtothesideofit.Thedistributionwouldhavebeensomewhatdifferent,butthisis hardlyofanyimportanceforthebehaviouroftheslab. The simple rule of moving the reinforcement does not, however, tell us anything about the reinforcementinthedirectionbetweentheedgeoftheslabandtheedgeoftheopening(elements1, 3, 4 and 5 in the examples), nor about reinforcement along the edges of the opening which are not parallel to the main reinforcement directions. For this reason the complete analysis is needed for openingsaboveacertainsize. Example12.4
163

TheslabinFig.12.2.7hasalargecircularopening.Supportbandsareintroducedinthedirectionsof thecoordinateaxesandinall45directions,allbandsformingtangentstothecircle.Theanalysisis similartothatinthepreviousexample,withthe45bandsformingsupportsforstripsinoneorboth directions. These bands are supported on the bands in the directions of the coordinate axes. As the analysisissimilartothatinthepreviousexamplenonumericalcalculationsareshown.

Fig.12.2.7 Asusual,itisrecommendedthatatleastonebottombarandonetopbarareprovidedalongtheedge oftheopening.

12.2.2 Nonrectangularslabs
The principles for treating openings in nonrectangular slabs are the same as for rectangular slabs. The main difference is that the numerical calculations are more timeconsuming because the static systemandtheelementsaremoreirregular.Insteadofusingsimpleformulasitisoftennecessaryto performnumericalsummationsandintegrationsofseriesofparallelstripsandofsupportbands.Only onesimpleexampleisillustrated,atriangularslab. Example12.5 ThetriangularslabinFig.12.2.8isthesameasinExample6.3,butwithalargeopeningABCD.The loadis9kN/m2. Theproposeddividinglinesbetweenelementsandsupportbandsareshowninthefigure.Between elements2and3thereisalineofzeroshearforce(maximummoment).

Fig.12.2.8 Betweenelements3and4thereisalineofzeromoment.Elements5and7areassumedtofunctionas purecantilevers. Element1issupportedbytheedgeoftheslabandthesupportbandalongAD.Thedashdotlinein thiscaseisjustaborderoftheelement,notalineofzeroshearforce.Themomentintheelementand thesupportreactiononthesupportbandcannotbecalculatedbymeansofsomesimpleformulas,but

164

anumericalanalysisofaseriesofstripsinthexdirectionhastobeperformed.Theaveragemoment intheelementis2.1kNm/m.ThesupportbandalongADisassumedtobesupportedonthesupport bandalongABandfixedattheedgeoftheslab.(Itmightalsohavebeenassumedtobeextendedand supportedatitsupperendinthefigureonthesupportoftheslab.)Thenumericalanalysisshowsthat asupportreactionof4.4kNatAgivesasupportmomentof2.94kNmandaspanmomentof1.50 kNminsupportbandAD. The average span moment between elements 2 and 3 is . The reaction from . This causes a support moment of 18.25 kNm, which is element 3 on 4 is distributedonthewidthoftheelement,whichis3.92m,givinganaveragemomentof4.66kNm/m. The direct load on the element gives an average moment of 1.62, so the total average support momentinelement4is6.28. Theaveragesupportmomentinelement5,calculatedwithEq.(2.6)is,3.09. Thesupportmomentinelement7is4.5. Elements6and8areanalysedbymeansoftheapproximationaccordingtoFig.2.2.4andtheresulting totalmomentisdistributedwithtwothirdsassupportmomentandonethirdasspanmoment.The resultis . ThesupportbandalongABisassumedtoconsistofapartwithpositivemoments,supportedatitsleft endattheedgeoftheslabandatitsrightendonacantilever,whichisperpendiculartothefixededge. Thelengthofthepartwithpositivemomentsis3.0mandthelengthofthecantileveris1.04m.The reaction4.4kNfromsupportbandADisacting1.36mfromtheleftend.Theresultingspanmomentis 3.27kNmandthesupportmoment2.07kNm. AdistributionofdesignmomentsisproposedinFig.12.2.9.Thespanmomentsinthesupportbands aredistributedoverawidthsof0.5m.Thesupportmomentsfromthesupportbandshavejustbeen addedtoandincludedinthesupportmomentsfromelements8respectively4and5,inthelattercase with about half in each. There is no reason to have a concentration of support moment at the theoreticalcantilevercorrespondingtothesupportband,asthemomentissmallandthesupportacts asadistributorofthemoment.

Fig.12.2.9 Thedesignmomentsclosetotheacutecornershavebeenassumedtobezero,asthecorresponding reinforcementwouldhaveaninsufficientlengthwithintheslabandthusbeuseless. Asusual,thereshouldbeatleastonebottomandonetopbarprovidedalongeachsideoftheopening inordertopreventlargecracks.InthiscasethereisprobablyapositivemomentatcornerCunder serviceconditionsandariskofacornercrackifthereisnoedgereinforcement. The moments in this slab with an opening may be compared to those in the same slab without an opening,Example6.3.Thisshowsthat,withtheopening,agreaterproportionoftheloadiscarriedin
165

thexdirection.Thisseemsreasonable,consideringtheactualsizeandpositionoftheopening.Inthis comparison it must, however, be remembered that the ratio between loads carried in the different directionsisbasedonamoreorlessarbitrarychoice.

12.3 Slabswithonefreeedge
12.3.1 Openingnotclosetothefreeedge
Thiscaseisanalysedwithmethodswhichareamixtureofthosedescribedaboveandthemethodsfor slabswithonefreeedgedescribedinChapter4. Example12.6 TheslabinFig.12.3.1isthesameasinExample4.2,butwithalargeopeningABCD.Theloadis11 kN/m2.Theassumedlinesofzeroshearforceareshown,aswellasthesupportbandsattheopening. Element5isassumedtocarrytheloadasacantilever,sonosupportbandisneededalongDC.The width of elements 2 and 6 is chosen so as to get suitable moments in element 1. A support band is assumedalongthefreeedge. WiththesupportreactionatthesupportbanddenotedRADwefind,forelement1, (12.36) With wefind Wefurtherfind (12.37) (12.38) (12.39)

,andthesevaluesseemacceptable.

Fig.12.3.1 (12.40) (12.41) (12.42) (12.43)

166

ThesupportbandalongABcaneitherbesupportedonthesupportbandsalongADandBCoronthe slabsupports.Inthiscasewewillassumethatitissupportedontheslabsupportsforthewholeload. This choice is made because it gives more reinforcement along the opening and more evenly distributed reinforcement in the xdirection. With the load acting between A and B and the reactionRattherighthandendwehavethefollowingequilibriumequations (12.44) (12.45) ,whichisasuitableratiointhiscase.Correspondinganalyses WithR=5wefind for bands AD and BC give , and the reactions on the supportbandalongthefreeedgeequalto1.5kNfromADand3.5kNfromBC. Thesupportbandalongthefreeedgeisacteduponby fromelement2,5.5kN/mfrom element3,triangularloadsneartheendsandthepointloadsfromsupportbandsADandBC.Suitable momentsarefoundtobe . The triangular elements at the corners give average moment contributions on the relevant widths: . The ratios between these moments do not follow the normal recommendations, but this is of no importance, as these moments are small compared to the other momentsintheslab.

Fig.12.3.2 Fig. 12.3.2 shows a proposed distribution of design moments. The constant mxf value 37.1 on the wholewidthbetweentheopeningandthefreeedge, calculatedas8.77/1.0+25.2/1.0+3.10,canbe questioned.Maybeitwouldbesomewhatbettertoconcentratethereinforcementmoretowardsthe freeedge,e.g.with49.0onhalfthewidthand25.2ontheotherhalf. ThesupportmomentsfromthesupportbandsADandBChavebeendistributedsothatthemomentis not too uneven along the edge. This type of distribution can be varied within rather wide limits withoutanysignificantinfluenceonthebehaviouroftheslab.Maybeinthiscaseitwouldhavebeen slightlybettertoplacemoreofthereinforcementclosertothelefthandsideoftheslab.Thesupport reinforcement according to Fig. 12.3.2 is probably too small by about one metre from the left hand cornerfromthepointofviewofcrackprevention.

12.3.2 Openingatthefreeedge
Wherethereisanopeningatthefreeedgethemainsupportbandisplacedjustinsidetheopening. Besidetheopeningareplacedsupportbands,whichcantileveroutandtakecareoftheloadcloserto thefreeedge. Example12.7 The slab in Fig. 12.3.3, which has an opening at the free edge, is again basically the same as in the previousexampleandExample4.2,withaloadof11kN/m2.Theproposedsupportbandsandlinesof zeroshearforceareshown.

167

 Fig.12.3.3 In this case it is rather difficult to estimate the behaviour and moment distribution under service conditions. The opening makes the slab weaker, so that it deflects more than a slab without an opening.Thislargerdeflectioncanbeexpectedtoleadtolargersupportmoments,particularlyalong theshortfixededge.Thepositionofthelineofzeroshearforcebetweenelements5and7determines thesupportmomentalongthelongfixededge.Ithasbeenplacedsomedistanceawayfromtheedge inordertogetsupportmomentswhichareestimatedtobelargeenough. ApplyingEqs(2.3)(2.5)andtheequationforasimplysupportedbeamweget (12.46) (12.47) (12.48) (12.49) (12.50) (12.51)

(12.52) and the load from element 3 on BC The load from element 2 on support band AD is .Thecorrespondingdesignmomentsinthesupportbandsare (12.53)

(12.54) Thesesupportbandsarefixedatthelowerendsinthefigure.Itisdifficulttoestimatehowlargethe supportmomentsare,andeventhesignsofthemoments.Itisthereforeassumedthatthemoments arezero,correspondingtohinges.Thisassumptionisofnoimportanceforthebehaviouroftheslab, asthemoments inthesesupportbandsaresmallcomparedtotheothermoments intheslab.With thisassumptionwegetsupportreactionsfromthesesupportbandsonthemainband (12.55) (12.56)

168

TheloadonthemainsupportbandalongCDconsistsofthesetwoforcesandtheloadonelement5. Ananalysisofthesupportbandwiththeassumptionthatthepointofmaximummomentis1.9from therighthandendgives (12.57)

(12.58) whichcanbeconsideredtobeasuitablemomentratiowithregardtothetypeofload. Fig.12.3.4showsaproposeddistributionofdesignmoments.Thedistributionalongtheshortfixed edge has been chosen to be somewhat different from the theoretical distribution according to the analysis.Agreatdealofthesupportmomentfromthemainsupportbandhasbeenmovedcloserto thefreeedgeoftheslab.Thisismotivatedbythefactthatitmaybeestimatedthatthemomentinthe servicestateincreasestowardsthecorneratthefreeedge.Ifthefixededgehadbeenfreelysupported the angular displacement along the edge would probably have increased towards that corner. The total sum of moments along the edge of course corresponds to the calculated values. Possibly the design momentshouldhavebeenvariedinmoresteps.

Fig.12.3.4

12.4 Slabswithtwofreeedges
12.4.1 Twooppositefreeedges
Theanalysisofthiscaseisquitesimilartothatofaslabwithonefreeedge,withtheonlydifference beeingthattwosupportbandsalongthefreeedgesareintroduced.

12.4.2 Twoadjacentfreeedgesandsimplesupports
Thediscussionbelowfirstconsidersrectangularslabs. Where the other two edges are simply supported, no solution can be found with the simple strip method if the reinforcement is placed in the directions of the edges, see Chapter 5. This is because therehavetobetorsionalmomentstofulfiltheequilibriumconditions.Suchsolutionsareoutsidethe scope of this book and are not treated, with the exception of some particular standard solutions, mainlyforcornersupportedelements. Theonlypossibilityfortreatingaslabwithtwoadjacentfreeedgesandtwosimplysupportededges bymeansofthestripmethodistoarrangethereinforcementinskewdirections,onereinforcement direction forming a support band along the diagonal between the edges where free and simply supportededgesmeet.Thissupportbandactsassupportforstripsinadirectionparalleltotheother diagonaloratsomeotherangletothefirstdiagonal.Thereinforcementinthesupportbandisbottom reinforcementandintheotherdirectionitistopreinforcement.Forreinforcementdistributionssee Section5.2.2andExample7.4. Analysisofaslabwithanopeningtendstobecomenumerically complicated,andasthesituationis notcommon,itwillonlybediscussedinprinciple,withoutanynumericalexample.
169

Iftheopeningissituatedwelloutsideorinsidethesupportbandtheanalysisisnottoocomplicated, as the support band is unbroken. With the opening outside the support band the approach is illustrated in Fig. 12.4.1. The main support band AC is unbroken. Secondary support bands are introducedonbothsidesoftheopeninginthedirectionofthediagonalBD,asinthefigure,oratsome otherangletothemainsupportband.Theloadbetweenthesebandsiscarriedtothebandsbymeans of bottom reinforcement parallel to AC or parallel to the edges of the opening. The corresponding momentscanbecalculatedratherapproximately,astheyaresmall.Thesecondarysupportbandsare supportedonthemainsupportbandandonthesimplysupportededges.

Fig.12.4.1 If the opening is situated so that it cuts the diagonal AC the analysis is more complicated. One possibility is illustrated in Fig. 12.4.2. From each of the corners A and C two support bands are introduced, tangents to the opening. These support bands together take over the role of the main supportbandinFig.12.4.1.Secondarysupportbandsareintroducedpastthecornersoftheopening. ThesebandsaresupportedonthepartsofthemainbandsbetweentheopeningandthecornersAand C respectively. Each secondary band is thus supported by two main bands. The load is divided betweenthesebandsina suitable proportion, e.g.sothatthe reactionof the secondary bandatthe simplysupportededgeisclosetozero.

Fig.12.4.2 The band Ad has no real support at d. Another secondary band has to be introduced to take that reactionandtransfertheforcetothemainbands.ThesamealsoholdsforbandCbatb.Pointdisnot necessarilysituatedat thefree edge.Itmaybeassumedtobesituatedcloserto theopening,which maygiveabettermomentdistribution. The procedure is numerically complicated, but with a suitable choice of moment distribution it probablyleadstoasounddesign.Inadditiontothedesignedreinforcementthereshouldbebottom and top bars along the edges of the opening and along the free edges as well as some minimum reinforcement. A design with the main reinforcement in the directions of the diagonals is advantageous for the behaviour of the slab and economical regarding the amount of reinforcement, as the reinforcement directions are close to the principal moment directions. However, it has the disadvantage that the reinforcingbarshavevaryinglengths,whichcomplicatesconstructionwork.
170

12.4.3 Twoadjacentfreeedgesandfixedsupports
Whereatleastoneedgeisfixed,thestaticbehaviourmaybebasedoncantileveraction,ashasbeen shown in Chapter 5. Then it may also be possible to analyse a slab with an opening and the reinforcementparallelwiththeedgesbymeansof the stripmethod,providedthatthelongedge or bothedgesarefixed. Example12.8 TheslabinFig.12.4.3isthesameasinExample5.1,butwithanopeningABCD.Theloadis9kN/m2.A formal analysis can be performed with the support bands along AD and BC, cantilevering from the fixedsupport,andtheelementsshown.Theelementscarryingloadinthexdirectionareassumedto besimplysupportedonthesupportbands.

Fig.12.4.3 Wegetthefollowingmomentsintheelements (12.59) (12.60) (12.61) (12.62) (12.63) (12.64) onsupportbandAD.Elements3aand3bgivereactions Element2bgivesareactionof onbothbands.Theseloadsgivethefollowingsupportmomentsforthebands (12.65) (12.66) AdistributionofdesignmomentsbasedonthesevaluesisshowninFig.12.4.4.Thesupportmoments inthesupportbandshavebeendistributedoncertainwidthsofthesupport.Ithastobenotedthatall reinforcement in the bands has to pass beside the opening, as the moment in the bands decreases ratherslowly.

171

 Fig.12.4.4 Thereinforcementdeterminedfromthisanalysisissufficientfromthepointofviewofsafety,butitis not sufficient to make the slab behave well in the service state, particularly regarding cracks. The analysishasbeenbasedontheassumptionthattheloadiscarriedbybendingmomentsinthexand ydirections.Inrealitytheloadinthistypeofslabgivesrisetolargetorsionalmoments.Therehasto be torsional reinforcement to prevent the formation of large torsional cracks. In the analysis of the correspondingslabwithoutanopening,Example5.1,someoftheloadhasbeenassumedtobecarried by means of a solution including torsional moments, which gives at least some torsional reinforcementintheslab.Nocorrespondingsolutionexistsfortheslabwithanopening.Theamount oftorsionalreinforcementshouldbeatleastaslargeintheslabwithanopeningasintheslabwithout anopening.Forarealisticestimateofthetorsionalmomentssomekindofanelasticanalysishastobe performed. ThedistributionofsupportmomentinFig.12.4.4isalsoquestionable.Thesuddenjumpinvalueat thesupportbandsisnotrealistic.Inordertopreventwidecrackssomeextrareinforcementhastobe introduced in these regions. The ratio between the support moments to the left and right of the openingisalsodifferentfromwhatcanbeexpectedfromanelasticanalysis.Themomenttotheright oftheopeningiscertainlylargerthantotheleftunderserviceconditions,asthemaximumdeflection of the slab is in its outer free corner. A better solution from this point of view can be obtained by supposingthattheuppermost1.5mofelement2issupportedonsupportbandBCinsteadofAD.With this change in assumption, the numerical value of MAD decreases by . The supportreactionatbandBCincreaseswith4.00kN/monthecorrespondinglength,whichgivesan increase in the numerical value of MBC of . The total amount of support moment decreases and the distribution is more satisfactory with such an assumption. At the same timethespanmomentintheuppermost1.5mofelement2increasesto10.9kNm/m.Somepartof thismomenthastobeaddedtom3,astheforceistransferredthroughelement3.Themaximumvalue ofthespanmomentwithinelement3becomes9.4kNm/m. It is evident that this example demonstrates a case where the strip method must be applied with caution, supplemented with at least some estimates of moment distributions in the service state, whicharedifferentfromthestripmethodsolution.

12.5 Corner-supported elements

In a cornersupported element, much of the loadbearing action takes place by means of torsional moments, which have been taken into account in the rules for the design of such elements. The importanceofthetorsionalmomentsishighestclosetothesupportedcorner.Anopeningchangesthe behaviour of the slab in a way which is not easy to estimate and to compensate for. Therefore no attempt will be made to give a complete recommendation for the analysis of cornersupported elementswithopenings,butonlysomegeneralpointsofviewandrecommendationsforlesscrucial cases.Onlyrectangularelementswithadistributedloadwillbediscussed.

172

Thetypeofdesigntobeusedforacornersupportedelementdependsonthesizeandpositionofthe opening. Regarding positions at least three cases can be identified. The following recommendations aregivenwithreferencetothesedifferentsituations. Within area A in Fig. 12.5.1, where the two column strips cross, the torsional moments are very important.Ifopeningsofasizelargerthanoftheorderofthedepthoftheslabaremadewithinthis areaitisrecommendedtorefrainfromthefulluseofthestripmethodandatleasttosupplementitby an elastic analysis. With smaller openings the reinforcement which would be cut by the opening is arrangedalongitsedges..

Fig.12.5.1 WithinareasBinFig.12.5.1,whereacolumnstripandamiddlestripcross,openingswhichcanbe inscribed inasquarewiththesidesparalleltothereinforcementdirectionsand notlarger than0.2 timesthesmallestcvalueoftheelementmaybetreatedinthefollowingway: Thereinforcementwhichshouldhavepassedthroughtheopeningiskeptinpositionandjustcutat theopening.Acorrespondingamountofreinforcementisplacedalongtheedgesoftheopeningand givenasufficientlengthtoensureasafeforcetransferbetweenthetworeinforcingsystems. WithinareaCinFig.12.5.1,wherethetwomiddlestripscross,openingswhichcanbeinscribedina squarewiththesidesparalleltothereinforcementdirectionandnotlargerthan0.3timesthesmallest cvaluemaybetreatedinthesameway. The application of these recommendations will lead to slabs with adequate safety against bending failure,andprobablyalsowithacceptablebehaviourunderserviceconditions. Where the openings are larger than those treated in these recommendations it may sometimes be possible to make an analysis based on the assumption of support bands, particularly for openings totallywithinareaCinFig.12.5.1orwithinareasB,butnotclosetothecoordinateaxesinthefigure. Such an analysis tends to be numerically complicated and to require qualified estimates of suitable reinforcement distributions with regard to the service state. In such cases it is probably better to makeanelasticanalysiswithacomputerprogram.
173

CHAPTER13 Systemsofcontinuousslabs
13.1 General
Mostofthepreviouschaptershavetreatedoneslabatatime.Theonlyexceptioncanbesaidtobeflat slabs which are continuous over the column supports. In this chapter the situation will be treated whereslabsarecontinuousoverwalls.Themainproblemisthechoiceofsuitablesupportmoments. The basic approach is to start by calculating support moments for the slabs which meet over the support as if they were fixed. The support moment is then calculated as a weighted average of the momentsonbothsidesofthesupportingwall.Possiblyalsosomeofthemomentdifferencemaybe taken by the wall. The weighted average is determined with regard to the estimated stiffness ratio between the slabs in accordance with normal methods applied in structural mechanics. Even if the estimateofthestiffnessratioisratherapproximatetheresultisasaruleacceptable.Itis,however, generallyrecommendedthatthesupportmomentischosensomewhatonthehighratherthanthelow side. Thechoiceofsupportmomenthasnoimportanceforsafetybuthasaninfluenceonthebehaviourin the service state. Within reasonable limits it has practically no influence on the deformations, so in practice in the first place it is the risk of wide cracks that should be taken into account. Enough reinforcement must be provided in sections where it is important to avoid visible cracks. Thus the supportmomentmustbechosenonthehighsideifitismainlyimportanttoavoidwidecracksonthe top surface, whereas the support moment may be chosen somewhat on the low side if it is more importanttoavoidwidecracksonthebottomsurface.Thischoicemaybedeterminedbythetypeof floorcovering,forinstance. Fig.13.1.1showsasituationwherethesupportmomentfromthelargeslabisbalancednotonlyby thesupportmomentsfromthesmallerslabs,butalsobythemomenttakenbythewallatrightangles, which presents a short fixed part of the support of the large slab. Close to the wall, the support moment for the large slab will be larger than if the edge had been fixed all along. In this case, the averageslabmomentsonbothsidesofthewallarethusnotthesame,assomemomentistakenbythe wallatrightangles.

Fig.13.1.1 Fig. 13.1.2shows another situation where the slab moments on the two sidesof the wall cannot be assumed to balance each other. The small slab has two free edges and the support moment distributionissuchthatitisconcentratedtowardstheupperendinthefigure,cf.Example5.1.The support moment from the load on the large slab, which is supported all around, has its maximum closertothecentreofthesupport.Thesetwomomentdistributionsthuscannotbalanceeachother completely, at least not in the service state. Some of the moment from the small slab has to be balancedbymomentstakenbytheupperwallatrightanglesinthefigure. Fig 13.1.3 shows a slab where there is continuity only along a certain part of an edge whereas the remaining part is simply supported. In analysing this situation we may start by assuming that the wholeedgeisfixed.Ifwethenreleasethemomentfromthesimplysupportedpart,themomentinthe remaining part will increase. It cannot, however, increase so much that the total moment along the
174

edge increases. With the notations in the figure, the average moment along the fixed part a0 of the edgeamustthusliebetweenthevaluesm0andam0la1,wherem0isthemomentwhenthewholeedge isfixed.Whichisthebestvaluebetweentheselimitshastobeestimated.Withtheproportionsshown inthefigurethemomentwillbelargestattheendofthecontinuouspart.

Fig.13.1.2

Fig.13.1.3

13.2 Systems of rectangular slabs

Inordertolimitthenumericalcalculations,theexampleschosenbelowareassimpleaspossible.The intentionisonlytodemonstratetheprinciples.Withaproperunderstandingofthegeneralapproach thereshouldbenoprobleminanalysingmorecomplexslabsystems. Example13.1 TheslabsysteminFig.13.2.1consistsofonlytwoslabs,whichmeetoverasupportingwall.Allother edgesaresimplysupported.Theloadis9kN/m2.Linesofzeroshearforceareshown,basedonthe generalprinciplesgiveninSection2.2.

Fig.13.2.1 Westartbycalculatingthesupportmomentsasiftheslabshadfixedsupportsatthewall.Applying Eqs(2.5)and(2.4)anddirectlysubtractingthespanmomentsfromthenumericalvaluesofsupport momentswefindtheaveragemoments

(13.1)

175

(13.2) Suitable choices of cvalues (with regard to ratios between span and support moments) are: ,whichgives .Asthesmallslabisstifferthanthelarge . slabthefinalmomentshouldbechosencloserto7.61thanto4.73,say We now have to adjust the c values to conform to the new value of ms. We can assume that the dimensionsinthedirectionsatrightanglesareunchanged.(Inmorecomplexslabsystemsofcourse changes in these directions may also have to be taken into account.) We can then apply the above equations. Forthesmallslabwehavetotakeintoaccountthat .Wefind .Forthelarge slab we may keep the value of c4, as the change in the support moment in this case influences the momentinelement4verylittle.Wefind . Example13.2 TheslabsysteminFig.13.2.2isthesameasinthepreviousexamplewiththeexceptionthatthesmall slab has a long free edge. This of course increases the support moment. It is now more difficult to estimate the pattern of lines of zero shear force for the small slab. It seems reasonable to assume whenthesupportisfixed,althoughasmallervaluemightbechosen.Theequilibriumequation gives

(13.3)

Fig.13.2.2 Forthelargeslab,Eq.(13.2)isvalid,andforthecasewithafixedsupportweusethesamevaluesas . above,i.e. Because of the free edge the small slab is now less stiff than the large slab. A suitable value of the support moment may be 8.10. Applying the same procedure as above we find . Theresultinthiscaseisverysensitivetothechoiceofc1.This,however,doesnotinfluencethesafety, onlythebehaviourintheservicestate. Example13.3 TheslabsysteminFig.13.2.3isagainthesame,butwiththeexceptionthatnowtheupperedgeofthe smallslabisalsofree.ForthatslabweapplythesametypeofsolutionasinExample5.1,startingwith anassumptionthatthecontinuousedgeisfixed.Mostoftheloadiscarriedbythestripsinthefigure, but part of it is carried by a solution including torsional moments, as discussed in Chapter 5. As in Example 5.1, we assume that 20% of the load is carried in this way. If we assume , the 2 remainder,7.2kN/m ,givesrisetoasupportmoment (13.4)

176

 Fig.13.2.3 This moment is unevenly distributed along the support, with the largest value in the vicinity of the uppercorner,seeFig.5.2.4.Thewholeofthemomentsms2cannotbeusedtobalance ,as theirdistributionsaresodifferent.Onepartofms2hastobecarriedbytheperpendicularsupportfor thelargeslab.Wecan,forexample,assumethat20%istakeninthatway.Theremainingpart,15.0, mightbetakenintoaccountindeterminingthevalueofms. Asthesmallslabisnowmuchlessstiffthanthelargeslab,asuitablesupportmomentmightbeabout 13.However,thislargeincreaseinms3correspondstoanappreciableangulardisplacementoverthe support.This angulardisplacementwillcausemoreofthesupportmomentms2 tobecarriedatthe uppercorner,andalsomoretobecarriedbytorsionalmoments. Thesituationisevidentlyverycomplex,anditisdifficulttoknowhowthemomentsandthetypeof static function can be best estimated. It is proposed that the part of the load taken by the solution includingtorsionalmomentsshouldbeincreasedto30%andc1shouldbeincreasedto1.5m.Wethen .Ifwefurtherassumethat30%iscarriedattheuppercornerwefind .This find valueseemsreasonablewithrespecttothelargeslab,andgives . Thissolutionistoaratherlargeextentevidentlybasedonestimates.Itis,however,safeaslongasthe assumed moment distributions are such that the corresponding reinforcement can function in an appropriateway.Thechosenmomentdistributionhastocoveranappropriatemomentdistribution for the large slab as well as for the small slab. This will be the case if we distribute the support moment according to the diagram shown to the left in the figure. A normal distribution, with two thirdsofmsoutsidethequarterpointsandfourthirdsinsidethequarterpointsforthelargeslabis covered,aswellasthetotalmomentforthesmallslabwithastrongconcentrationtowardstheupper corner, where the slab is locally fixed. The solution also seems to give an acceptable reinforcement distributionwithregardtotheservicestate. Example13.4 TheslabsysteminFig.13.2.4consistsofthreeslabsA,BandC,supportedonwallsbetweentheslabs butsimplysupportedalongallotheredges.Alltheslabshaveauniformloadof9kN/m2. ForslabAwecancalculatetheaveragesupportmomentbyassuming ,whichgives (13.5) ForslabBwecanassume ,whichgivestheaveragemoment

177

Fig.13.2.4

(13.6) SlabChasanedgewhichisonlypartlycontinuous.Wethenhavetostartbyassumingthatitisfully weget fixed.With (13.7) . From the shape of The total moment corresponding to this average moment is element9itcanbeseenthatthegreatestpartofthismomentistheoreticallycausedbystripswhich endatthepartoftheedgewhichisinrealitysimplysupported.Whenthewholeedgeisfixedthus, only a minor part of the total moment falls within the distance ab. When the moment on the freely supportedpartisreleased,however,themomentonabwillincrease,withaconcentrationclosetoa. Itmaybeestimatedthatbetween30and60%ofthetotalmomentwillremainwithinab,say,45%or about24kNm.Mostofthistotalmomentwillbetakenclosetoaandtransformedtothewallatright anglesata.Wemayassumethatthemomentistakenasadistributedmoment andthe remainingpartasaconcentratedmoment ata. Indeterminingthesupportmomentsalongthesupportabcwehavetotakeintoaccountthatms3is the largest and will cause an angular displacement of the slab over the support. This angular displacement will be prevented at b by the wall at right angles. This causes a concentrated support moment at b. An estimate may be that the moment difference on some 20 to 30% of the support lengthwillbetakenbythatconcentratedmoment.Themomentdifferenceisabout15kNm/mandthe concentratedmomentcanthusbetakentobe . is mainly taken at bc. We can assume and The remaining total moment ,whichgivesapproximatelytherighttotalmoment. Using the above moments, and averaging with an approximate regard to the stiffnesses, gives the designmoments , . Afterthesemomentshavebeendeterminedwecancalculatethenewcvalues.Thuswiththeaverage designsupportmoment weget

178

 (13.8) . whichleadsto InslabBwemaykeepc7unchangedandequalto1.3,andweget (13.9) . whichgives In slab C we may keep c11 unchanged and equal to 1.8. The average moment ms9 is andweget (13.10) . whichgives With the changed positions of the lines of zero shear force we can now calculate the average fixed edgesupportmomentsoverthewallbetweenslabsBandC,assuming (13.11)

(13.12) SlabBisslightlystifferthanCandasuitabledesignmomentis13.5. The remaining part of the analysis follows normal procedures. In determining the distribution of design moments, the concentrated moments at a and b should be distributed on small widths and added to the other moments. In slab C the span moment from elements 9 and 11 must be rather unsymmetricallydistributedinordertocompensatefortheunsymmetricaldistributionofthesupport moment. Example13.5 Awatertankhasarectangularbottom andadepthof4.0m.Itrestsoncolumnsalongthe edgesofthebottom.Itmustbedesignedforthecasewhereitisfilledwithwater.ThebottomBofthe tankandtwoofthesidesAandCareshowninFig.13.2.5.Patternsoflinesofzeroshearforcearealso shown.Becauseofsymmetry etc. The water pressure is assumed to increase by 10 kN/m2 per metre depth (a more correct value is 10.2).Itisthus40kN/m2atthebottom. Assuming , and taking the symmetry into account, we get the followingaveragemoments,applyingtheequationsinSections2.3.24: (13.13) (13.14) (13.15) (13.16)

179

Fig.13.2.5 (13.17) (13.18) (13.19)

(13.20) . Withregardtosuitablemomentratios,wemaychoose Elements2and11aresupportedbysupportbands.Themomentsinthesesupportbandsare (13.21)

(13.22) With regard to load distributions and suitable moment ratios we may choose as supportmomentswhentheendsarefixed. Ifwecomparethemomentsonbothsidesofsupportswefindthatthemomentdifferencesarequite smallinallcasesexceptforthesupportbands.Inthefirstplacewethereforetakethisdifferenceinto account. AsslabCissmallerandstifferthanA,asuitabledesignsupportmomentforthesupportbands,based on the above values, is 17 kNm. (Application of the recommendation in Section 10.2 gives ,thusthesameresult.)Suchavaluewillmeanthatthespanmomentintheband alongelement11isnegativeandthebandisbentinwards.Ontheotherhand,thebandalongelement 2willdeflectoutwards.Thesedeformationswillleadtoadecreaseinthenumericalvalueofms9and
180

anincreaseinms4.Takingthisandthemomentdifferencesattheedgesintoaccount,wecantrynew cvalues,whicharenowregardedasfinal. We start by assuming new values for slab C, . This leads to , . wecanfindthatasuitablevalueofc7is1.75,whichgives . Basedon Ifwefurtherassume and ,wefindthatasuitablevalueofc3is2.5,which . gives . A We have now determined all cvalues and we can calculate suitablevalueofthesupportmomentforthesupportbandsis15,whichgives ,a smallnegativevalue. wefind ,whichisacceptable. Applying Becauseofthetriangularloaddistributionandtheshapeofthetank,strongmomentconcentrations cannotbeexpectedatthefreeedges.Thedesignmomentdistributionmightthereforebechosentobe rathereven.Ontheotherhand,theanalysisisbasedontheassumptionofsupportbandsalongthe freeedges.ThemomentdistributionproposedinFig.13.2.6forslabAisacompromisebetweenthese pointsofview.

Fig.13.2.6 The result of this analysis is a tank with adequate safety and a reinforcement distribution which is suitablefromthepointofviewofdeformations.Itmaynotbethebestreinforcementarrangementto limitcracking.Thistypeofdesignisthereforenotrecommendedwheretheconcreteisintendedtobe watertight.Forsuchacasethedesignshouldbebasedonthetheoryofelasticity,takingdueaccount oftorsionalmoments.

13.3 Rectangular slabs and concrete walls

Concretewallscanberegardedasconcreteslabswithzeroload.Thesupportmomentscantherefore, as a reasonable approximation, be calculated with the same methods as above, with the support moments for fixed edges equal to zero for the walls, taking into account the ratios between the stiffnessesoftheslabs.

13.4 Other cases

Systems including nonrectangular slabs or slabs with cornersupported elements can be analyzed accordingtothesameprinciplesasabove.Ofcoursesuchcasestendtogivemorelaboriousnumerical calculationsandmoreintricateestimates.Thereforenonumericalexamplesareshown.
181

CHAPTER14 Joistfloors
14.1 General
In a joist floor Joists or ribs in one or two directions interact with a rather thin slab to take the moments.ForpositivemomentsthejoistsandtheslabactasTbeamswithtensionreinforcementin thebottomofthejoistsandcompressivestressesintheslab.Joistfloorscanbemadetotakemuch higherpositivethannegativemoments.Inplaceswherelargenegativemomentsoccur,thejoistfloor isreplacedbyasolidslabwiththesamedepth. Ajoisthasaverylimitedabilitytotaketorsionalmomentsinthedirectionoftheribs.Thedesignhas tobemadeontheassumptionthatonlybendingmomentsoccur.Thislimitsthepossibilityofusing joist floors in situations where torsional moments are essential for carrying the load. Thus, for example,ajoistfloorwithjoistsparalleltotheedgescannotbeusedforarectangularslabwithtwo adjacentedgesfreeandtheotheredgessimplysupported.Whereacornersupportedelementforms partofajoistfloortheremustbeasolidpartinthevicinityofthesupportedcorner. Asjoistfloorsaredesignedontheassumptionthattheytakenotorsionalmoments,butonlybending momentsinthedirectionsofthejoists,theapplicationofthestripmethodisbasedonthesimplestrip method. A joist floor has a limited ability to redistribute moment in the lateral direction. The lateral redistributionwithinwidelimits,whichmaybeacceptedforasolidslab,cannotbeacceptedwithina joistfloor.Eachjoisthas,inprinciple,totakethemomentfromtheloadwhichhasbeenassignedtoit. Where there is a system of crossing joists some lateral moment redistribution may however be accepted.

14.2 Non-corner-supported floors

The simple strip method may in principle be applied directly. Just as for solid slabs, the floor is normally divided into areas where the load is carried in the different joist directions. The dividing linesareoftenlinesofzeroshearforceandofmaximumpositivemoments.Theloadwhichisassigned toeachjoististhatontheareabetweenthecentrelinesbetweentheparalleljoists. As the lines of zero shear force often form an angle to the direction of a joist, the theoretical load distribution on a joist from a uniform load on the floor has a shape according to Fig. 14.2.1. The momentsinthejoistarethendeterminedfromtheequation (14.1) Thesecondtermcanoftenbedisregarded.

182

 Fig.14.2.1 Example14.1 ThesimplysupportedjoistfloorinFig.14.2.2hasjoistsat0.9mcentresinbothdirections.Theloadis 8 kN/m2, which corresponds to 7.2 kN/m for the loaded parts of the joists. An assumed pattern of linesofzeroshearforceisshown. Joistahas and .Wethusget (14.2) Inthesamewaywefind (14.3)

Fig.14.2.2 (14.4)

(14.5) Noticethatlissmallerforjoistd. Joistehas etc.Joisthhas (14.6) (14.7) (14.8)

.Weget

(14.9) Acertainlimitedredistributionofmomentsbetweenadjacentparallelstripsmaybeperformed.
183

Thissolutiongivesaratherunevenreinforcementdistribution.Itisalsopossibletomakeananalysis ontheassumptionthattheloadisdividedbetweenthetwodirectionswithincertainareas.Wecan, forexample,assumealoaddistributionaccordingtoFig.14.2.3,wherehalftheloadistakenineach directionwithinthecornerregions.Withthisdistributionweget (14.10) (14.11) (14.12)

Fig.14.2.3 This analysis is simpler and gives a more even reinforcement distribution. It gives a little more reinforcement. The sum of all moments is for the first analysis and for the second analysis , which is about 5% higher, corresponding to 5% more reinforcement. This small difference is in practicemoreorlesscompensatedforbyagreaterneedforminimumreinforcementinthefirstcase. Inmanycasesasolutionofthesecondtypeistobepreferred.Ofcourse,otherloaddistributionsthan onewithhalfineachdirectionmaybeused. Example14.2 ThejoistfloorinFig.14.2.4hasonefreeedgeandthreefixededges.Thejoistspacingis1.2m.Along thefixededgesthereisasolidslabofwidthequaltoonejoistspacingplushalfthewidthofajoist. Alongthefreeedgethereisabeamwithawidthequaltohalfthewidthofajoistplus0.2m.Theload isassumedtobeauniform8kN/m2.Thehigherdeadweightofthesolidpartsisthusdisregardedin ordertosimplifythenumericalcalculations. A simple approach has been chosen with dividing lines parallel with the edges and half the load carriedineachdirectiononareas . at3.0mfromeachend. Westartbyanalysingjoistb,whichhasaloadof

184

Fig.14.2.4

(14.13) Due to the higher stiffness near the supports and to the load distribution we may choose .Themomentsinjoistcmaybetakenasonehalfofthesevalues, . Stripd,whichisasolidslabstrip,musttakeloadfromawidthof1.8mandwillthustakeamoment whichis1.5timesthatinjoistc, .Thereinforcementforthesemomentsmustmainly beplacedfarfromtheparallelsupport. Joist f has a load of 4.8 kN/m at 3.0 m from the fixed end and is supported on beam a, where the supportreactionisdenotedRfa.Theequilibriumequationis (14.14) Suitable values may be , which gives the span moments . The ratio between moments is acceptable with respect to the higher stiffness near the fixed support and the loaddistribution.Themomentsinstripecanbetakenas1.5timesthesevalues, andthecorrespondingreaction ,acting0.9mfromthesupport. Joistghasaloadof9.6kN/monthewholelength.Theequilibriumequationis (14.15) Suitablevaluesmaybe Theedgebeamahastocarryauniformloadof and endandthereactions

. actingalongalengthof3.0mfromeach .Theequilibriumequationis

(14.16) Suitablevaluesmaybe . Thelargestpositivejoistmomentis13.35kNmandthelargestnegativejoistmomentwhereitmeets the solid slab can be shown to be about 6.4 kNm. The most stressed member is the edge beam. Whetherthedesignisacceptablecanonlybejudgedafterthereinforcementhasbeendesignedand theconcretestressescalculated.Itmayhavebeenadvantageoustochooseawideredgebeam.

14.3 Floors with corner-supported elements

In a cornersupported element the moment field in the vicinity of the support is nearly polar symmetric with negative tangential moments and positive or numerically smaller negative radial moments. This moment field contains large torsional moments with respect to the reinforcement
185

directions.Thesetorsionalmomentscannotbecarriedbyajoistfloor.Inthevicinityofthesupport therehastobeasolidslab,withtheabilitytotaketorsionalmoments. A cornersupported element which forms part of a joist floor thus has to have a solid slab in the vicinity of the supported corner, whereas the rest of the element has crossing joists. As a rule, the solidparthasarectangularshapeandhasitsedgesasacontinuationofthejoists.Onlythiscasewill betreatedhere,butofcoursethesameprinciplesmaybeusedevenwherethesolidparthasanother shape. Acornersupportedelementthushascrossingjoistswhichcarrytheloadonlybymeansofbending moments.Thesejoistsareintheirturncarriedbythesolidpartwhichalsotakestorsionalmoments andwhichcarriestheloadintothesupport. Fig.14.3.1showsacornersupportedelementwhichhasasolidpartnexttothesupportedcorner.The element is divided into column strips and middle strips. The boundary lines between column and middlestripsarehalfwaybetweenthesolidpartandthenearestjoist.Thewidthofthecolumnstrip inthexdirectionisdenotedycY. Withintheareawherethemiddlestripscrosstheloadisdividedbetweenthetwostrips.Thesimplest assumptionistodividetheloadintotwoequalparts,andonlythiscasewillbetreatedhere.Ofcourse itispossibletomakeotherassumptions,butthisdoesnotseemtogiveanyadvantageinthiscase. Amiddlestripissupportedonthecrossingcolumnstrip.Intheanalysiswemayuseaveragemoments and forcesper unit width. From these average moments and forces we later calculate the moments and forces in the joists and in the solid part. The forces per unit area on the middle strip in the x direction are shown in Fig. 14.3.2. The reaction on the column strip is assumed to be uniformly distributed.Forthemomentsperunitwidthinthemiddlestripinthexdirectionwefind,withindex mformiddlestrip (14.17) Itcanbedemonstratedthatitissuitabletohaveequalnumericalvaluesforthetwomoments,thus (14.18) Thevalueofthesupportmomentmayhavetobemodifiedinordertomakeitagreewiththevalue fromtheelementontheothersideofthesupport.Insuchacasethevalueofthespanmomenthasto bemodifiedsothatEq.(14.17)isfulfilled.

Fig.14.3.1

186

 Fig.14.32 TheloadfromthemiddlestripinthexdirectiononthecolumnstripintheydirectionisgiveninFig. 14.3.2.Thecolumnstripalsohastocarrythedirectloadq.Thetotalloadonthecolumnstripwithin thejoistedpartis (14.19) Thesumofallmomentsalongtheedgesoftheelementhastofulfiltheequilibriumequationforthe element.Whenweknowthissumandthemomentstakenbythemiddlestripwecaneasilycalculate themomentsinthecolumnstrip. In determining the average total moments for the element we may take into account the higher stiffness of the solid part. This can be done by choosing a somewhat higher numerical value of the support moment. A suitable increase is in the order 1015 % compared to the values for a normal solidslab. Thesolidparthasahigherdeadloadthanthejoistedpart.Asthisloadisactingclosetothesupportit mainlyinfluencesthesupportmoment.Itcanthereforejustbeaddedtothesupportmomentwithin thecolumnstnp. Where two cornersupported elements meet at a line of zero shear force in a regular system, the lateralmomentdistributionscanbemadetoagree.Inothercasestheremaybedifferenttheoretical lateral distributions. A typical case is where a cornersupported element meets a oneway element, whichissupported onitswholewidth.Thetheoretical distribution ofspanmomentsinthecorner supported element is rather uneven, with higher moments in the column strip, whereas the distributionintheonewayelementisuniform. Thiscasecanbeanalysedontheassumptionthattheelementwithasupportonitswholewidthisnot atrueonewayelement,butthatthemiddleandcolumnstripscontinueintothispart.Halftheloadon the middle strip is transferred to the column strip, just like in the cornersupported element. The reinforcementinthejoistsatrightanglesisthesameasintheparalleljoistsinthecornersupported element.Itcanbeshownthatthisanalysisleadstoanacceptabledistributionofspanmoments,even ifitisnotalwaysrigorouslycorrect.Arigoroussolutioncanalwaysbeachievedbyassumingthata smallerpartoftheloadthanonehalfiscarriedinthedirectionofthemiddlestrip.Thisassumption leadstoanincreaseinmomentsinthejoistsatrightangles,whichhardlyseemsnecessaryinpractice. As the joists are supposed to take only bending moments, the lengths of reinforcing bars can be calculated by ordinary methods. In the solid part, where torsional moments are also acting, it is recommended that all the support reinforcement is carried to theboundary of this part.No bottom reinforcementistheoreticallyneededwithinthesolidpart,unlessthereisapositivemomentatthe boundary. In this case the corresponding reinforcement is carried to the support line. Fig. 6.3.7 in StripMethodofDesignshowsthetheoreticaldesignmomentdistributionforthecase . Aswillbedemonstratedintheexamplebelow,thisanalysisleadstoaratherunevenlateralmoment distribution, with small moments in the middle strips. The resulting design is safe with respect to ultimateload.Theremaybeariskofunacceptablecracksinthemiddlestrips,particularlyonthetop surface. The moment distribution is a consequence of the assumption that the joists can take only bendingmomentsandnotorsionalmoments.Alateralredistributionofdesignmomentsasforsolid slabs cannot be made without the introduction and acceptance of torsional moments in the joists. Maybesomesuchredistributioncanbeaccepted.Here,however,onlythetheoreticalsolutionwithout torsionalmomentsinthejoistswillbeused.
187

Example14.3 The joist floor in Fig. 14.3.3 has joists at 0.6 m centres. It has solid parts corresponding to 6 filled modulesineachdirection.Thetotalwidthsofthecolumnstripsarethus ineachdirection. .Theouterboundariesaresimplysupported. Thesupportingcolumnshavesquaresections Thetotalloadonthemajorpartsoftheflooris11kN/m2.Withinthesolidpartstheexcessloadis5 kN/m2. We start by determining the average support moments in order to calculate the cvalues, which we need for the detailed analysis. Asstated above, the solid partsare stifferthan the joistedparts,and thisisheretakenintoaccountasafactor1.15forthesupportmoments.Theexcessloadonthesolid partisnottakenintoaccountatthisstage. (14.20) (14.21) ThecvaluesarecalculatedwithEq.(2.34)andtheaveragespanmomentswithEq.(2.35) (14.22) (14.23) (14.24) (14.25)

(14.26) Notingthatthewidthofthecolumnstripbelongingtoeachelementisthedistancefromthecornerof thecolumntothedividinglinebetweencolumnandmiddlestrip,wefindthat (14.27)

Fig.14.3.3
188

(14.28) (14.29)

(14.30) Forthesupportmomentinthemiddlestripinthexdirectionwefindsomewhatdifferentvaluesfrom thetwospansfromEq.(14.18).Weusetheaverageofthesevalues (14.31) ApplyingEq.(14.17)wefindthecorrespondingspanmomentsinthemiddlestrip (14.32)

(14.33) Wecannowcalculatetheaveragemomentsinthecolumnstripinthexdirectionbysubtractingthe parttakenbythemomentsinthemiddlestripsfromthetotalaverageanddistributetheresultonthe widthofthecolumnstrip.Forthesupportmomentwenowaddthemomentcausedbytheexcessload onthesolidpart.Notingthatthisexcessloadisonlyactingonawidthof1.55mofthetotalwidth1.85 mofthecolumnstripwefindthattheaverageexcessmomentis (14.34) (14.35) (14.36)

(14.37) Itisalsoofinteresttocalculatethemomentattheboundarybetweenthesolidpartandthejoists.For thispurposewecalculatetheloadonthejoistpartofthecolumnstripfromEq.(14.18) (14.38) Themomentsattheboundariesclosesttosections1and2respectivelyare (14.39) (14.40) Asthesemomentsarenegativesometopreinforcementisrequiredinthejoistsinthevicinityofthe solidpart.Nobottomreinforcementistheoreticallyneededinthesolidpart. FortheydirectionwegetthemomentsinthemiddlestripsfromEq.(14.18) (14.41) Themomentsinandtheloadonthecolumnstriphavesomewhatdifferentvaluesonbothsidesofthe column,aswehavedifferentxvalues.Wemaydirectlycalculatetheaveragevalue (14.42) (14.43) (14.44) Themomentattheboundaryofthesolidpartis
189

 (14.45) In addition to the cornersupported elements, there is a thin oneway element at each side of the column. The moments in this element can be calculated separately and the corresponding reinforcementadded.Itissimpler,andslightlyonthesafeside,toassumethatthemomentsinthis elementarethesameasinthecolumnstrips. ThejoistsmustbereinforcedformomentsinkNmcorrespondingtotherelevantmvaluesinkNm/m multipliedby0.6m.ThesolidpartsshouldbereinforcedfortotalmomentsinkNmcorrespondingto the above support moments multiplied by 4.2 m, which is the total width of the support strip, including the oneway element. This reinforcement may be uniformly distributed over the width of the solid part, which is 3.6 m plus the width of one joist, or with some concentration towards the support.

CHAPTER15 Prestressedslabs
15.1 General
The strip method is based on the theory of plasticity, and the main relevant plastic property of reinforced concrete slabs is the yielding of reinforcement at ultimate load. Prestressed slabs are mainly assumed to function elastically and their plastic properties are limited. The strip method as applied to slabs with ordinary reinforcement cannot be generally accepted as a design method for prestressedslabs.Thereexist,however,situationswherethestripmethodmaybeuseful. One possiblewayofapplyingthestripmethodis to usethemostbasicprinciple ofthesimplestrip methodanddividetheslabintoanumberofnarrowstrips,eachstripcontainingonetendon,which carriestheloadonthestrip.Inthiscasetheplasticpropertiesoftheslabareunimportant. Itisalsopossibletouseamixeddesign,e.g.withprestressingtendonsasasupportbandalongafree edge,whereastherestoftheslabhasnormalreinforcement.Inthiswaythedeflectioncanbelimited. When the strip method is applied in connection with prestressing it seems most natural to use the principleofloadbalancing,inwhichtheloadonastripisbalancedbytheforcecausedbythechange indirectionofacurvedtendon.Onlythisapproachwillbediscussedbelow,andthereaderisexpected to be familiar with it. However, the analysis is often made in terms of bending moments
190

corresponding to these forces. The idea of load balancing is of importancefor the arrangement and shapesofthetendons. The examples of applications below are also simplified by assuming only a constant load and disregarding prestress losses due to creep, shrinkage and relaxation, and other effects which are normally taken into account in the design ofprestressed structures. The intention of this chapter is only to show some possible ways of making use of the strip method for the design of prestressed slabs,nottogivecompleteguidanceforsuchadesign.

15.2Thesimplestripmethodfortendons
Thedirectapplicationofthesimplestripmethodforaslabwithprestressingtendonsissimilartoits applicationtojoistfloors.Inajoistflooreachjoistcarriestheloadfromacertainpartofthefloor.The joists are at equal spacings and the amount of reinforcement varies depending on the load. In a prestressedslaballtendonsareoftenequalandinordertoutilisethemthewidthsofthestripshave tovary.Insimplecasesthewidthsofthestripsarechosentogivethemequalmoments. Example15.1 The slab in Fig. 15.2.1 is simply supported. It has a load of 11 kN/m2 and is to be prestressed with tendons,abletoprovideeachforamomentof15kNm. According to the principles of the simple strip method the slab is divided by means of lines of zero shearforceintoareaswhichcarrytheloadindifferentdirections. Thetotalmomenttobetakenbythetendonsinthexdirectionis,accordingtoEq.(2.4), (15.1) Thenumberoftendonstotakethismomentisdenotednx.Weget (15.2) Thetotalmomenttobetakenbythetendonsintheydirectionis,accordingtoEq.(2.5), (15.3) Withthenumberoftendonstotakethismomentdenotedny,weget (15.4) Wecanthusassumeavalueofnxandcalculatecxfrom(15.2)andnyfrom(15.4).Ifweassume we get . The theoretical total number of tendons is 19.9. It proves that we get approximatelythesametheoreticaltotalnumberoftendonsifweassume or8.Ifweassume wegetatheoreticaltotalnumberof20.1. Inthiscaseitseemssuitabletochoose .

Fig.15.2.1
191

Suitablepositionsoftendonsaredeterminedinthefollowingway.Thefirsttendoninthexdirection takestheloadonawidthy1.Themomentrelationgives (15.5) .Withthetriangularloaddistributionwithinthatpartitisreasonabletoplacethefirst Wefind .Amoreexactdeterminationofthepositionisnotimportantforthebehaviourof tendonat theslab. Wecontinuebydeterminingavaluey2fortwotendonsbyreplacingy1withy2anddoublingtheleft side in the equation. This gives . The second tendon thus takes the load between and . It may be placed at . The third tendon is placed at and the fourth and fifth symmetrictothesecondandfirst. Inthesamewaywefindthatthetendonsintheydirectionmaybeplacedatx=1.3,2.1,2.5,2.8,3.1, 3.4,3.7,4.0etc. Theshapeofthetendonsshouldbedeterminedfromtheloadthattheyareexpectedtocarry.Thus, thetendonsshouldbestraightwithinthepartswheretheloadistakenbythetendonsinthedirection atrightangles.Thefirsttendoninthexdirectionshould,forinstance,becurvedapproximatelyinthe first1.3mfromtheedgeandthenstraightuntilthesamedistancefromtheoppositeedge. This design is the most economical with respectto the number oftendons. If, for some reason, it is thoughtbettertohavesome otherdistributionof tendons, theloadmaybeassumedtoactinsome otherway,e.g.dividedbetweenthetwodirectionswithincertainpartsoftheslab.

15.3 Prestressed support bands

Inordertolimitthedeflectionofaslabandpreventexcessivecracking,itmaybeadvantageoustouse prestressing tendons in a support band whereas the rest of the slab has normal reinforcement. A typicalcaseiswherealargeslabhasonelongfreeedgeandtheotheredgessimplysupported.With only normal reinforcement the deflection may be unacceptably large. According to the theory of elasticity, muchoftheloadiscarriedbytorsional moments,which maycauselarge cracksinunder service conditions, as reinforcement parallel to the edges is less effective for limiting skew cracks. Withaprestressedsupportbandalongthefreeedgetheslabmaybemadetoactasifitissupported alongalsothisedgeandtheloadiscarriedmainlybybendingmoments,forwhichthereinforcement iseffectiveforcracklimitation. Example15.2 TheslabinFig.15.3.1carriesaloadof11kN/m2.Ithasonelongfreeedgeandtheotheredgesare simply supported. A support band with prestressing tendons is arranged along the free edge. The widthofthissupportband,whichdependsonthetypeandnumberoftendonsandontheconcrete stresses,isassumedtobe0.6m. Theslabisassumedtobesupportedalongthecentrelineofthesupportband.Thisassumptionisnot thesameasthatnormallyusedinthisbook.Thereasonforthisdifferenceisthatthestripmethodis basedonthetheoryofplasticity,whichisnotapplicablefortheprestressedsupportband.Inorderto be on the safe side in the case of a prestressed support band, this different approach is used. The differenceismainlyoneofprincipleandhasaverysmallinfluenceontheresultingdesign. Theslabisthusassumedtohaveaspanof3.7mintheydirectionandtheordinaryreinforcementis designedaccordingtothemethodsinChapter3. The support band has to carry the load from the slab plus the direct load outside the centre of the band.BymeansofEqs(2.3)and(2.4)wefindthemomentinthesupportband (15.6)

192

 Fig.15.3.1 Theshapeoftendonsshouldbedeterminedfromtheloaddistribution.Asthisisnearlyuniformthe tendonsmaybegivenapurelyparabolicshape. Even though thetendons balancethe forces,this doesnot guarantee thatthe edge does not deflect. Theunevencompressivestressdistributionmaycausecurvature.Thiscanbeneutralizedbysuitable choicesofanchorage,prestressforceandshapeofthetendons.Thereaderisreferredtotheliterature onprestressedconcrete.

15.4 Flat slabs

Theanalysisofprestressedflatslabsmay,inprinciple,bemadewiththesameapproachasforjoist floors. If the column strip can be made so narrow that it is not much wider than the supporting columns,suchadesigncanberegardedasrigorouslycorrect.Withwidercolumnstripsthesolutionis less rigorous, as torsional moments appear within the area where the support strips cross. These torsional moments are necessary to bring the load to the support.The wider the column strips, the moreimportantarethetorsionalmoments. The torsional moments correspond to principal moment directions which form an angle to the directionsofthetendons.Inthevicinityofthesupporttheprincipalmomentsaremainlypolar,with thetangentialmomentshavingthelargestnegativevalues,whereastheradialmomentshavesmaller negative values or even positive ones. From equilibrium conditions it will be found that there is always a difference between radial and tangential moments close to the support of a corner supportedelementequalto2R/,whereRisthesupportreactionfromtheelement.Theslabhasin principletobedesignedforbothmoments. Theprestressedtendonscauseamomentfieldintheslabwithprincipaldirectionscorrespondingto the directions of the tendons. The resulting moment field acting on the concrete section is thus a combination of two fields having different principal directions. This resulting moment field has principaldirectionswhichdifferfrompointtopoint,andateachpointithastwoprincipalmoments whicharedifferent.Thesectionshouldbecheckedwithrespecttoboththesemomentsatallpoints withinthesupportarea. Thestresssituationisevidentlyextremelycomplex.Acorrectsolutionshouldbebasedonamoment field from the acting load calculated with the theory of elasticity, which unfortunately gives very unevenmomentdistributions,notdirectlyapplicablefordesign.Asanapproximationonthesafeside regardingultimateload,thedesignmaybebasedonthemomentfieldforacornersupportedelement accordingtothestripmethod.Forsuchanapproachthefollowingproceduremaybeused: 1.Determinevaluesforthecornersupportedelementsfromthechosenwidthsofthecolumnstrips. Thesewidthsdependontheacceptableconcretecompressivestress. 2.Calculatethemomentsinthemiddlestripsbythesamemethodasforjoistfloors. 3. Calculate the loads on the column strips. The column strip in one direction is regarded as the supportforboththemiddleandcolumnstripsinthedirectionatrightangles. 4. Determine the moments for the column strips based on these loads and the assumption that the stripsaresupportedatthecolumn.
193

5. Starting from these curves, determine design moment fields by adding the negative and positive mvaluesaccordingtoFig.15.4.1,whichisvalidforthexdirection.Thisresultsintwocurveswithin thesupportarea.Bothcurvesaretobetakenintoaccountinthedesign. Thisdesignissafewithregardtotheultimatelimitstate.Maybeitistooconservative.Itisprobably alsoacceptablefromthepointofviewofcracking,althoughitisnotbasedonmomentsaccordingto the theory of elasticity. The ratio between span and support moments should be based on the principlesofthetheoryofelasticity. Example15.3 Fig.15.4.2showsacornersupportedelementwith .The widthsof thepartsofthe columnstripswithintheelementare1.0minbothdirections.Thiscorrespondsto . Theloadis12kN/m2. ApplyingEq.(14.17)wefindforthemiddlestrips (15.7)

Fig.15.4.1

(15.8) ApplyingEq.(14.19)wegettheloadsonthepartsofthecolumnstripswheretheycrossthemiddle strips (15.9) (15.10) Themomentsinthecolumnstripscausedbytheseloadsare (15.11)

(15.12) Theloadsonthecolumnstripswithintheareawherethecolumnstripscrossare (15.13) (15.14)

194

Fig.15.4.2 Themomentswithinthisareaare (15.15)

(15.16) ThecurvesforthemomentsinthecolumnstripsaccordingtotheserelationsareshowninFig.15.4.2. Thesumsofmomentsinthecolumnstripsare (15.17) (15.18) To the values according to the above relations are to be added the values from Fig. 15.4.1. The resulting curves are shown in Fig. 15.4.2. Both the upper and lower curves have to be taken into accountindesign.Theconcretestressesresultingfromthesemoments,combinedwiththeinfluence ofprestressingtendons,havetobechecked.

References
Armer,G.S.T.:Thestripmethod:anewapproachtothedesignofslabs.Concrete,Sept.1968,358363. Crawford,R.E.:Limitdesignofreinforcedconcreteslabs.JournalofEngineeringMechanicsDivision, Proc.ASCE,Oct.1964,321342. Ferguson,M., Breen,J.E.andJirsa, J.O.:ReinforcedConcrete Fundamentals,5thed.1988,JohnWiley andSons. Hillerborg,A.:Equilibriumtheoryforreinforcedconcreteslabs(inSwedish).Belong1956.171182. Hillerborg, A.: Strip method for slabs on columns, Lshaped plates, etc. Translated from Swedish by F.A.Blakey,CSIRO,DivisionofBuildingResearch,Melbourne1964. Hillerborg, A.: A plastic theory for the design of reinforced concrete slabs. IABSE Sixth congress, Stockholm1960,PreliminaryPublication,177186. Hillerborg,A.:StripMethodofDesign.AViewpointPublication,C&CA1975.E&FNSpon. MacGregor,J.G.ReinforcedConcrete:MechanicsandDesign,2nded.1992,PrenticeHall. Nilson,A.H.andWinter,G.:DesignofConcreteStructures,11thed.1991,McGrawHill,Inc. Park,R.andGamble,W.L.:ReinforcedConcreteSlabs,1980,JohnWiley. Wilby, C.B.: Structural Concrete, 1983, Butterworth & Co. Revised as Concrete Materials and Structures,1991,CambridgeUniversityPress.
195

Wood,R.H.andArmer,G.S.T.:Thetheoryofthestripmethodforthedesignofslabs.InstitutionofCivil Engineers,Proceedings,Oct.1968,285311.

196