You are on page 1of 18

Received on April 2J. 1987: Revised on November 10,. 1987.


I he papcr rcviewa h i d p litcraturr pertaining to thc ultlmarc strength. dellcctions and crackmg of reinforced coocrele rcctang~lnr.skew and circuidr slabs and gives a II\I of 136 pobiiwtiwx oa lhc topic. Key words: Concrete (remforcedl. sldna ( t w w v a y l . rncmhrdne action. ddlectim, cmckwidth, leinforcrment


Reinforced concrete slab is one of the structural elements which is extensively used in civil engineering colistruction. Twcl-way structural actiun, post-cracking nun-linear behaviour and development of membrane action complicate the anaiysis and estimation ofthe ultimate strength of the slab. Design of rhc slabs baseti on limit htates also rcquires a knowledge ot their deflections and cracking behnviour. A number of researchers have investigated the flexural strength, deflection and cracking of two-way slabs and these aspecls are briefly reviewed here. 2. Methods of slab analysis

The elastic analysis of slabs as thin piate requires the solution of biharmonic equation for specified boundary conditions. Solutions for diffcrcnt shapes of slabs and boundary This t . analysis is useful to conditions arc found in '1 imoshenko and ~ o i n o w s k ~ - K r i e g e r the elastic behaviour of thc slab, but the factor of safety against collapse cannot be assessed. Also it is complicated for slabs of irregular shape, which can however be arlalyssd more easily using yicltl-line analysis.



2.2 Yield-line theory

In 1921, the Danish engineer Aage 1ngerslev2proposed a method of calculation based upon the constant bending moments acting along the yield lines. Since 1931, K. W, ~ ~ has donehpioneering ~work in this ~ field, in~Denmark.~ His work ~ has heen ~ translated into ~nglish'in 1962. The yield-line theory assumes that an increase in load causes concentration of strain in steel and concrete along lines of maximum moment. These lines are called yield lines and they spread into a pattern which divides the slab into segments. Near failure, the elastic deformations of each segment are assumed negligible compared to plastic deformations at the yield lines. Consequently all curvatures in the slab at failure are assumed to be concentrated at the yield lines. The collapse load based on assumed mechanism is calculated from the energy equation. As the theory satisfies mechanism and the equilibrium condition, it provides a theoretical upper-bound solution. This theory can be effectively used for the analysis of any irregular type of slab having different type of boundary conditions and loadings. In 1962, European Committee for Concrete9 published a report that contains formulae for the analysis of slabs of various shapes subjected to uniformly distributed load, line loads and point loads. Some investigator^'^-^^ have reported design aids forplabs using yield-line theory. The yieldline theory is suitable for the ultimate load analysis with uniformly distributed (isotropic and orthotropic) reinforcement, but it is difficult to handle if non-uniform or curtailed reinforcement is utilized. Unless the correct mechanism is found the resulting design may be unsafe and the reinforcement volume is uneconomical. This method does not indicate the loadings on edge beams. Moreover it does not provide information regar. ding the distribution of bending and twisting moments in the interior of the slab away from the yield lines and thus the distribution of reinforcement becomes difficult. How. ever, this theory is finding acceptance due to the fact that the ultimate loads of slabs as determined in testd3 are much higher than those predicted by yield-line theory. This differencebetween yield-line loads and experimental ultimate loads is due to the development of membrane stresses, and this effect is considered in detail separately.

2.3 Lower-bound approach

In the lower-bound approach the solution should be determined to satisfy equilibrium and boundary conditions and provide a statistically admissible moment field without violating the yield conditions anywhere in the slab. This provides a safe solution because the collapse load may be greater than or equal to the calculated value. This solution provides information on the required distributionof positive and negative reinforcement. Some lower-bound solution^'^-'^ are available for simply supported, continuous, one shortedge free, rectangular isotropic and orthotropic slabs and circular slabs with isotropic and oahotropic reinforcement. Though most of these solutions do not coincide with upper-bound solutions, they are satisfactory as the differences between the upperand lower-bound solutions are small. In 1956, HillerborgZ0 developed the strip method b*d on the lower-bound approach. He proposed to design several types of slabs b y assuming the load to be carried only by bending moments in two perpendicular

directions. He neglected torsional moments in developing the method. This method is very much suitable for slabs with free, simple or fixed boundary conditions as the slab can be divided into some sinlple combinations of cantilever and simple beam strips. This is a design method and economy is achieved hy varying the reinforcement in diffcferent portions of the slab. Wood and Arrner" have examined the strip method and concluded that, it provides an exact solution ii the reinforcement provided is in accordance wlth the strip moments with an unlimited number of simultaneous modes. Fernando and em^*' attempted to find whether Wood and Armer's conclusion was true for a general case. Their examination shows that strip method does not providc unique solution always, but the exceptions are rare. Harrop2%tudicd the application of strip method to aimply supported dab-bean1 panels and found that the minimisation of reinforcement in slabs was counter-balanced by increase in beam rcinforcement. Subsequently ~ a r r o p ~ % p p l i c dthe strip method to the design of skew slabs. H e established an a h i t y relationship between skew and reclangula~ slabs which enables the design of skew slabs from rectangular slabs. Thakkar and Sridhar RaoZ5 introduced modifications on Hillerborg's method and prcscnted moment coefficicnts for rectangular slabs with different boundary conditions. It was claimed that their meihod could result in a saving of 15 to 20% of reinforcement over IS 456-1564 method. Test results of ~rrner'" Iyengar et a12' and Rajanna and Chandrasekhara" on rectangular, skew and square slabs (tested under uniiormly distributed load) confirm that the slabs designed by strip method hehave in a safe and satisfactory way. However, test results of other shapes and boundary conditions arc rcquircd to support the above finding. RajuzG3' gave moment coefficients for rectangular slabs with different boundary conditions. In 1.555, Hillerb~rg'~ extended the simple strip method to complex slabs which include a column support in the interior of the slah. The theoretical treatment is complicated and hencc was ca!led the Advanced Strip Method by Crawford". Alternatively, Kemp35 presented a generalisation of strip method of design suitable for slabs with patch loads and concentrated loads using appropriate distribution of the vertical shear forces instead of applied loads. Mall~ckand Gupta'" gave tentative proposals for the application of the method. Test results are required to know the behaviour o i slabs designed by the above method. Fernando and Kcmp" dcvcloped a strip-deflexion method using an elastic theory approach. They introduce compatibility of deflexion criteria of strips. Choosing a torsionless elastic moment field, they demonstrated their method lor square and rectangular slabs. This method provides information about the bending moments of htcral beams. They indicated that the method could be extended for point and patch loads. Wilby" has developed computer programmes using the strip deflexion method which also take into account individual point loads and any support conditions. Subsequently Wilby'' presented design tables for the slabs carrying uniformly distributed load for different boondar); conditions including slabs with free edges. Hillerborg" has reported the advanced strip method which includes supper% like interior, exterior corncr colmm and reentrant corners. Gurleydl has adopted the inversion technique to Hillerborg's simple strip method. Instead of assigning the strip loads he has assigned the moment fields at the critical locations so that the moments are in bimoment equilibrium with the applied loads. He has introduccd the stress-resultant bimoment and developed four equilibrium equations. But the extra equation is a material-dependent eqnilibr~um

equation and not as afourth independent rquation. Subsequently GurleyJ2 has presentee design methods for slabs considering as torsion-free gillages. H e has covered the yiek lines which are curved in plan also.
2.4 Direcl design method

this method of design based on elastic stress fields. The steel is wood" determined by using elastic stress distribution at ultimate loads in conjunction with the yield criterion. Recently Hago and ~ h a t t " reported test results designed by the above method and it was found that the slabs behaved satisfactorily both at working load and at ultimate loads. To design by this method, computer facilities are required and also they haxe assumed uncracked stiffness of slab throughout the analysis.

3. Studies on rectangularlskew slabs

3.1 Memhrane uction in restrained slahs

The above mentioned methods have been fornlulated without considering the effect o f change of geometry o i the slab. The changes in the geometry of the slab and boundary restraints have considerable influence on the load-carrying capacity of the slab. These effects are called membrane actions and there exist two types of membrane action. wr. ( I ) compressive and (2) tensile membrvne actions. This is reviewed briefly in tht following sections.

3.1 Conlpressive membrane action

When a slab is surrounded by and is continuous with either stiff heams or additlonai slabs. compressive membrane action develops within the depth of the slab. This phenomenon can he explained as follows. After cracking load, thcre occurs a substantial shift in the position oE neutral axis in the cross-section. This effects in the outward movenlent of the edges. This movement is prevented by the restraints. Hence compressive forces develop in the slab. This compressive force acts above the mid-depth 1 at the centre of the slab and heluw at the edges of the slab and hence causes arching 0 h m actton. It increases the moment of resistance of the slab cross-section which in turn significainly increases the ultimate flexural resistance of the slab. In 1939. ~ v o z d e v ' ~ conducted tests on restrained slabs and showed that the compressive membrane forces were due to the different level of neutrnl axis in the positive and negative yield lines. Much work has been done on membrane action by Soviet researchers. Their cofltribut% are not available in English. In 1955, Ockleston* conducted load tests on a three-stow reinforced concrete building at Johannesberg and recorded the collapse three or four times of Johansen's yield-line load. powell'" tested nine h d dsabout ~ ma~l-lcale restrained, isotropic rectangular slabs with different steel ratio. H e noticed ghat for a slab with 0.25% steel ratio, the ultimate load was 8.2 times higher than



johsnsen.sload. HIStest resuits showed that highcr cnhancernent could be achieved with lower stee] ratio. ~rotchie'" ~hristiansen'". Park'". 4acohsenS'. Kohcrtss', Brotchie and Hoilevsi, Hopkins and kark", Desayi and K u l k a r n i have prcsentcd methods of including compressive mcahrcinc actlon for reslraincd recrangular siahs based deformationtheory. Also Desayi and Kulk;rrni have recently critically reviewed the Some investigators have used the flow thcory iotroduced by Sawczuck in ,964. ~ ~ w c z u c k ~ ~' n , n : ~ " ,~ o r l e y " ,Hung and N m y i o have reported methods of for restrained rectangular slabs using flow ihcary. Braestrupf'" has reviewed the treatments of donie effect of reinforced concrete slabs up to 197s. He made a clear distinction between the tluw theory and deformatiam theory. He developed solutionsfor a restrained slab strip by flow theory as we!! a \ deformation theory anti silo~cd that for a given membrane force, thc dctlection predicted by deformation theory w a s twice that predicted by flow theory. Aiso the load-carrying capacity dropped below the flexural collapse load of unrestrained rl;~h,if deformallon theory was used. H e pintedout that lhesc predictions are due to thc assuniplion inade in deformation theory, ;e, concrete ic assumed to assert tts ulti~natc compressive stress during unloading. H e concluded that deformation theory led lo unrealistic predictions. However. ~ h a t t o ~ a d h ~ points a ~ " o u t that rigid platic approach applied to concrete itself is far from reality and hence comparison of thc two theories is n o t appnlpriatc. Also tlow theory should no1 he taken as a suitable method, as it predicts maximum detlcxion of a slab when total superimposed load ceases to exist. He pointed out that the rigid plastic analysis through either approach is not useful for slahs of practical interest, as the one-bay stripa may tail beforc thc concrete compressive stress block at the critical sec'ion reaches its elastic limit. This leads tu (he possibility of developing deformation theory ahereas flow theory cannot be developed. While the above cited investigations are on restrained rectangular slabs, some work has been dOnc on the strength and behavioui of slab-beam systems. P a ~ k " ~ Hayes , and Taylor"', Datta and ~arnesh'"'.", Desayi and Killkarni" have pmposcd mcthoda for estimating ultimate load of rectangular slab-beam system. Tong and Batchelor"' reported test results of square two-way bridge slabs subjected to concentrated loads. They have proposcd a method of estimating the failure load which includes compressive membrane action. They concluded that if steel of low Percentage is used, punching failure could be eliminated and flexural failure mode results in. They recommended a minimum pcrccntagc of 0.25% for two-way bridge slabs to avoidinstant failure. The study of compressive membrane action has hccn made on other shapes also. Desayi and Prahhakilra6* presented a method of analysis for restrained skew slabs including compressive membrane action. The next section deals with tensile membrane action occurring in restrained slabs.


In restrained slabs. the load-carrying capacity is enhanccd by compressive niernbrane action till the ascending !(&-deflection curve reaches a peak point. While in the descending portion of the load-deflection curve, the membrane force changes from comPre~sion to tension. The load-deflection curve descends to a secondary resistance



point at which the membrane action in the central region of the slab changes to tensile, Beyond this, the tensile membrane action spreads throughout the slab till the fractureof steel. park3'. KeenanhY,Black7', ~ e r z o g "have presented procedures for calculatingthe incipient collapse deflection capacity. Iqbal and ~ e r e c h o ~have ' reviewed their contributions in detail.


Membrane action in simply supported slabs

In simply supported slabs. membrane forces develop at large deflections. With large deflections a1 midspan, the central region of the supported edges tend t o move inwards but are restrained from domg so by adjacent outer regions. This creates the central area of the tensile membrane stresses within the slab together with a surrounding ring of compression. This effect enhances the load-carrying capacity of the slab. Park5", Taylor er ~ 1 ' Hayes7'. ~. Kemp7', Sawczuck and Wlnnicki7', Morleysx, Desayi and Kulkarni", proposed methods of analysis for rectangular simply supported reinforced concrerte slabs indiiding membrane actlon. l'rabhakara7%xtended the work of Desayi and Kulkarni" t o simply supported skew slabs. The above studies were concerned with the ultimate strength of reinforced concrete rcctangulariskew slabs. But these should be complimented by investigation on ~ ~ ~ i c e a b i limit l i t y states namely deflection and crackwidth. In the following sections the literature a~ailahle on the studies on deflections and crackwidths of reinforced concrete slabs is reviewed.

5. Deflections of two-way reinforced concrete slabs Deflection is one of the serviceability limit states t o be satisfied in the limit-state designof structures. ThereEore the estimation of deflection under working loads is often necessary in the design of concrete slabs. But an accurate estimation of deflection is complicated hecause of the influence of factors like cracking of concrete, creep and shrinkage effects itime-dependent effects) and nonlinear properties of the material. A s a convenient way of emtrolling deflections, some codes of practice specify limiting span to depth ratio. This is conservative because the effect of all factors affecting deflections is included in a single parameter namely depth. Clarke et a17%ave compared the code clauses of 25 countries related to deflections of concrete memhers. In 1974, ACI Committee 435'' reviewed the methods like classical, cross-beam analogy, analogous grid work method. static ratio, wide beam and equivalent frame methods for the determination of deflections o i two-wy slabs. The ACI manual of concrete practicex1 recommends the prwedure developed by Branson based on beam test results for the computation of slab d~fle&ons. This requires further examination due to difference in the structural Wanchoo " hehaviour of the beam and the slab. Scanlon and Murray8', Bell and ~ l m s ~ ~ incorporated May". Yukio et d", Cope and Vasudeva ~ a o " " , ~ h o n o h r i c h ~ have the influence of cracking in developing finite element analysis t o predict the



;oa&deflection curve of slabs. Bazant el u!"%asc provided a summary of the work done on the appltcation of finite clement method to i-cinlorccd concrctc structures up to 1978. ~ h method j ~ cannot bc readily used in design otfices as it requires good computer proposed ' a trilinear !oad-deflection plot foi two-way slabs lron, f,rLllltleb. . . ' ' . . ~~erzog" uncracLed stage up to failure of the slab. His theoretical load-deflection plot consisted of three zones. Zone I is between zero and cracking load, Zone 2 represents the plot from crackingto yield-linc load and Zone 3 corresponds to plastic deformatron beyond \ield.line load. Shukla and n/littal"fl reported semi-empirical formulae for determining ;he &flections of reclangular reinlorced co:lcrete slabs based on their test results of nine rilnplysupported squaw slabs rested under uniformly distributed load. BransonY1 gave dctalls of [he available methods of computing deflections of two-way slabs. Desayi and ~ u l k a r n i " . ~ ~proposed ~" methods for predicting load-deflection cut-ves for restrained, pitially restrained rind simply supported rectangular reinforced concrete slabs in the form of piecewise straight lines. Thzy accounted for cracking of concrete and yielding of stecl by suitably modifying the flexural rigidity of the .;lab and changcs that occur in the support conditions due to possible yielding after the format~onof yield lines. The nothors"."' have conlputed shol-t-time deflections of reinforced concrete rectangular slabs using a deteriorating momcnt of incrtia function. Dcsapi and Prabhakara"' proposed a method for predlctinfi load-deflection curves of restrained skew slabs. The studies cited a h o w concern only with the cst~mation of instantaneous deflection . ~ f reinforced concrete I-ectangulariskew slabs. But the serviceability of slabs depends to a large degree on their long-term deflection performance. Hcnce methods are needed to estimate the long-term deflection of slabs. ACI Code'' recommends the use of a multiplier constanl to short-time deflections to determine long-term deflections. But Taylor'sy4 and Heiman's" investigations showed that ACI Codc mcthod did not predict the long-term dcflectlons satisfactorily. Rangan"" proposed a method for the computatlon of instantaneous and lorig-term deflections of flat plates and slabs. His results were within 24% of the lncasured valuec of threc existing floors. Rangan and McMullen '7 presented a rational approach to control both total and incr;mental long-term deflections uf flat slabs and plates by limitation o f span-depth ratio. They developed the equations based on a limiting long term total dcflections of L11240 where L I is longer centre-to-centre suan. They concluded that the use of minimum thickness as per ACI -318-71 and AS 1480-1974'%ill result in long-term deflections greater than L11240. IS -1%-1978" specified formulae for estimation of short-term, shrinkage and creep deflectlam. The authors"" examined these formulae with the use of available 22 rectangular simply supported reinforced concrete slabs (short-term tests only). It was foulld that the load factors obtained by limiting the total deilcction to span1250 were barge. In some cases this load factor was found to be as high as 5.93 ag-inst the recommended load factor of 1. 5 for the limit state of collapse. Hence it was concluded that IS 456-1978'" Codc clauses probably require furthcr examination in respect of the computatiori of deflection of slabs including those due to shrinkage and creep. Also more test data on long-term deflections of two-way slabs are lequired. Gilbert"" extended Rangan's approach for beams to two-way slabs. But thls requlres finite element simulation. Tam and Scanlan102 proposed a method for computing immediate deflections



including shrinkage effects. They have compared with the field measurements. Graham and scanlan"" have presented a method for estimating the deflection due to shrinkage restraint, construction loads. They proposed multiplier constants for estimating the two-way slab deflections. Limit state of crack~ng is another serviceability criterion to be satisfied in the limit-state design of structures. The studies pertaining to the predictiona( maximum crackwidth in rectangularlskew slabs are presented in the next section.
6. Prediction of crackwidth

In reinforced concrete slabs. the crack appears as soon as the tensile stress due to the bending moment exceeds the tensile strength of concrete. This may impair appearanceof the slab or increase the risk of corrosion of the reinforcement. In order to avoid rhis, the crackwidth should be kept below the acceptable limit. ACI Code, CP 110"" and CEB-FIP re corn mend at ion^'^^' have specified the maximum allowable crackwidths for structural members under different exposure conditions. A crackwidth study on one-way s!abs with welded wire fabric by Srinivasa Rao and S ~ b r a h m a n y a m " ' ~ indicates that a steel stress of 2500 kg/cm2 can be allowed without violating the strength and crackwidth limitations for one-way slabs. In developing their mathematical model, they have included the force exerted by transverse bars in iongitudinal direction. Nawy and his "8 have done a major study on cracking of two-way rectangular siabs. The associates~!~7proposed equations were based on the fracture hypothesis on the stress concentration at the intersection of orthogonal bars. The limitations of the proposed equations are in choosing the appropriate multiplying constant for the particular type of loading and specified boundary conditions. and the experimental results being those of rectangular .;lab reinforced with welded wire fabric. Desayi and Kulkarni"" extended the method propsed by Desayi"' for the determination of maximum crackwidth of partially prestressed concrete beams to two-way rectangular reinforced concrete slabs. Two-way action was included by considering the bearing exerted by transverse bars. A uniform distribution of bond stress over the length of the bar and tensile stress in the stretched concrete area was assumed in their analysis. This was further extended to skew and rectangular slabs by Deszyi and Prabhakara'". It was found from their investigation that the assumption of parabolic distribution for bond stress and uniform distribution of karing stress resulted in a better agreement with the test data of 348 crackwidth measurements on rectangular and skew slabs. In recent times. considerable work has been in progress to study the cracking behaviour of concrete using fracture mechanics conceprs. These have yet to find application in the design of slabs. 7. S t u d i e s on circular slabs
7.1 Membrana action

in restrainedlsimplj supported slabs

WOO^'" analysed the membrane action in isotropic circular slabs using the basic equations of ::lrg:-defiection plate theory. He assumed the material to b e rigid plastic



andcomputedload-deflection plots for simply supported and restrained isotropic circular slabs. D~~ to the above assumption, the maximum load-carrying capacity was obtained at zero deflection for restrained circular slabs. Also, the theoretical load-deflection plots gartfromJohansen's load at zero deflection for simply supported circnlar slabs. As such, his theoreticalcurve does not predict the elasto-plastic behaviour of simply supported and restraiced circular slabs tested under uniformly distributed load. However his is useful to estimate ultimate strength of slabs, i i a suitable value of deflection at ,,,rimate lozd is assumed. Desayi and Kulkarnii1%xtended Wood's approach to oflhotropicrestraiI12dcircular slabs. Two case5 of ortbotropy. namely, (1) M I> M , and p) MI i M2 were considered in their analysis. where M , and M,are the plastic moment capacitiesin circumferential and radial directions respectively. While the above inyestigations of the section of the slab in circumferential and radial directions were based on deformation theory. some investigators have applied flow theory to circular sisbs. ~ a n a $ gave ~ rigid plastic flow theory soiution not only for restrained circular slabs but also to square slabs and slab strips. The comparison between theoretical and eyerimental deflection curve was not reported. ~ o r l e y 'used ~ the rlou, theory for predicting the load-deflzction plots for rigid plastic slabs with and without edge restraints. The essential difference in the above two approaches is that, Janas used the dissipation method whereas Morley used equilibrium method. Calladine"' applied flow !heory to circular slabs and showed that the upper-bound calculations on the strength of circular plates and slabs turn out to be simpler using a three-dimensional theory instead o f two-dimensional slab theory. H e illustrated his method on metal plates, sandwich plates and reinforced concrete slabs, simply supported or restrained, under point or uniformly distributed loading. The theoretical load-deflection plots for restrained circular slabs under point loading and simply supported circular slabs under uniformly distributed loading were given,in his paper. But theoretical treatment on restrained reinforced concrete slabs under uniform distributed loading was not available in it. Morley"* tested circular slabs with ring beams and reported a method of analysis at the Euromech Conference in 1974. Moondra and SharmaUb reported the behaviour of simply supported circular slabs without and with square, rectangular and triangular openings. They compared the experimental resultc of seven slabs with the Johansen's load and observed that the ultimate load to be 1.37-1.70 times the yield lines load. They concluded that the ultimate load capacity is independent of the shape of the opening if the area of the opening of different shapes is kept constant. A I - ~ a s s a n i "used ~ the deformation theory for the ascending part of the load-deflection curve and flow theory for the descending portion of the curve. H e treated elasto-plastic strips using the above procedure. Braestrup and M o r ~ e y "proposed ~ a modified rigid plastic theory for circular slabs with ring beams. They assumed that the membrane action starts at an initial elastic deflection. This deflection was assumed empirically as 0.03 times the thickness of slab. An alternative procedure also was suggested by them to obtain the deflection at which membrane forces started developing. From their load-deflection plots, it is observed that the empirical elastic deflection corresponds to that at the Johansen's load of Simply Supported circular slabs. Hence they assumed a linear load-deflection plot Up to Johansen's load of corresponding simply supported circular slab. Their experimental resultsalso showed considerable linear hehaviour whereas available literature on orher



shapes of slabs showed non-linear behaviour up t o Johansen's load. Also the assumption of the same initial deflection irrespective of size of the ring beam is also not clear, H~~~~~~ the effect of ring beam on membrane action was considered in their analyslsby adding the inplane slab flexibility and ring beam flexibility in the form of a boundary spring. The agreement between the theoretical and experimental curves is good. But complete details of the experimental work were not presented by them, gave some steps to obtain the initial value of deflection according tothe theory presented in the paper. Recently, the authors have developed methods of estimating the ultimate load of simply supported and restrained, isotropic and orthotropic circular slabs sub~ected t o distributed loading taking into account the membrane action, load-deflection behavia"! and estimation of widths of cracks that form in these ~labs""-'~".

8. Study on slabs with different layouts of reinforcement

~ e d ~ e y 'conducted ~ ~ " ~ experiments on simply supported square slabs with different arrangements of reinforcement. He kept the total amount of reinforcement to be the same for all slabs. H e noticed a reduction of 15% of collapse loads for square slabs with diagonal reinforcement compared with uniform reinforcement parallel to the edges. Taylor et a17%ested ten two-way simply supported square slabs to investigate the effect of arrangement of the reinforcement. The slabs were designed to have same flexuralresistance under uniformly distributed load. They concluded that slabs with variable rein. forcement were stiffer than uniformly reinforced slabs up to Johansen's load. However. the use of variable reinforcement did not enhance the load-carrying capacity. They observed that the use of variable reinforcement reduces crackwidth in the central portion of the slabs. bur increases the crack width in the corner regions. Reddy et allz6 tested 12 simply supported square reinforced concrete slabs under cental concentrated load to study the effect of direction of reinforcement, effect of banded reinforcement and the effect of circular and spiral reinforcement layout in square slabs. The investigation showed that the best arrangemen: of reinforcement for strength and stiffness wasby having the bars parallel to the sides at a uniform spacing. Ultimate loads of all slabs reported were higher than Johansen's load. Clark12' studied the effect of arrangement of relnfomment for reinforced concrete skew bridge slabs. H e suggested that saving in steel could be achieved by placing the steel parallel and ~ e r ~ e n d i c u l to a r the abutments in the central region and changing to the arrangement wh'ich'has steel parallel to the free edge near the edges.

Slabs with different layouts of reinforcement based on optimisation theory have been Sharpe and ~ l ~ d e Thakkar ' ~ ~ , and Sridhar &=tigated by R~nany'~-'~',~elchers~~~, ~ a o ' " . The principal canclusions were: load enhancement in optimised slabs was not much more than conventional slabs; at a particular load, optimised slabs exhibited smaller defiecti~ns when compared to conventional slabs; although the reinforcement 1 u n h e was saved in optimised slabs, the optimal layout was quite complicated. While :



muchtheoretical advancement has been done in optimisation of slabs, reported test resultson the same appear to be very few. M~spratt''~.'" has reported tests on opti&lv designed slabs He suggested that Monte Carlo technique could be used for of slabs as the properties of the slab are random variables.

9. Conclusions

1. Many investigations are come aCiOSS on the determination of ultimate strength of rectangularslabs. Studies on the ultimal? strength of skew and circular slabs are comparatively few. While the studies recognize the enhancement in the strength beyond Johansen's load, attempts to include this enhancement in design of slabs are yet to be made.
1. Studies on the determination of deflections and crackwidth in slabs are comparatively

less. More experimental work on these aspects including the influence of creep under sustained loading is required in the examination of serviceability limit states of slabs.


Theory of plarer nnd shells. Second edn , McGraw-HIII Inc , New York. N.Y., 1959 Om en elementaer beregntngmetode af krydsarmerede plader, Ingen~zmren. 1921, 30, 507-515. Beregnmg af krydsarmerede jernbetonpladers Bygnmgsslariske Meddr, 1931, 3, 1-18. brudmoment,


Bruchrnomente der kreuzwe~se bewehrten platten, Mem. Asso Inlernales Ponts Charpenres, 1932. 1, 277-295. Nagle pladetormler, Bygnrngsstairske Meddr, 1932, 4 , 77-84 Brrrdl~nzeteoner,Jul. Gjellcrups Forlag. Copenhagen, 1943, p. 191. ( Yield 11ne theory, translated by Cement and Concrete Associat~on, London. 1962, p 181.) P!udeformler, Polytrkntsk Foremng. Copenhagen. 1946, p. 186. 2nd edn, Polyteknisk Foremng, Copenhagen, 19.19, p. 186,3rd edn. Polytcknxsk Forlag, 1968, p. 241. (Yteld-/mefurrnulae /or shbs. transiated by Cement and Concrete Assoclarion. London, 1972, p. 106.) Ywld irne theory, 1963, Cement and Concrete Assoc~ation,London. line theory lo culculatiom of flexural The upplicarion of rhp slrengrh of dabs and flnr slob floors, Com~teEuropeen du Beton. 1962, Bull. No. 35 Coefficients for the des~gnof slabs by yield-line theory. Concrete, 1969, 3, 171-172.



i b e r f m r n t \ for d~wgnlng rietx oi ocm-and-slab panels hy y,eld.l,6. thcuiy. Comrnr. I'Jbil. 3, 486 onalyse of rcmurccd amccete \l;~h\. A m Poiylrrhmm Scondtila wm. Civil Engmcenng ano B u ~ l d m gConsiruction Scrirj, Copenhagen. D m n d t h . 1964. C.20 Jamnv~ktstcoil tor A r m e i i d e hctuiigpl.firtor. 171-181 Thc rheory ol the \trip method :or drhign Engi ( L o n d ). I'IhS. 41, 385-307

Belong. 1956, 11,

\lab\, I'ioc I m e


The strip metnod of \:ah i h g n ~ n q or w iowri-wund ~ o l u t ~ o n i ~ %log. ( b m r . Rr., , 1975. 27, 23-20
Lltrmale loau d r x g n ot slmldy .S<J.. IcI6Y, 4, l45-15!1
W ~ ~ I ~ I T 4:ib-her?n1 C~I

\plem>. Bids

Ulllmdtc load d c r g n ot \keu h h i hy he 5irlp method. Bias Sn 1970, 5, 117-121

Dengn of two-way reri&;ced

concrete rectangular dabs h? Hillerhwg's arrip method. Iridiun Concr. . I . 1970. 4 , 158-166 Ultmdte luad tcsts o l s b b i dsvgned by the srrlp metliod, Proc h a Civ. Engr ( L o r d ), IYhX, 11, 313-331

Design o f skcu,siabi hy thc slrrp method. J S m c t Engiig , 1973.1,


Rehwiour of Iixed square siaba designzd by rhe arrip method. 1 Strucr Cigng. 1978. 6 , 36-43



Desrgn ot rectanguldr rerntorced concrete slabs supported on a!l the s~dcs with a short disconunuous edgc. Bldg Scr.. 1970, 5 , 181-185 B p t ~ m ~ s esjnp d method for the dealgn of freely supported slabs, Indian Concr. 1.. 1971, 45, 390-394. Moment coeffs~enlsfor the design of rectangular reinforced concrete slabs with fixed or contrnuous edges, J. lnstn Engr (India) 1972, 52, 224-277. Moment coefi~cientsfor the design of recranguiar slabs with short discontmuous edges, Indrnn Concr. 3 , 1972, 46, 466-468

Sirrmlernnoden for piflllrirpu peluie. I iizkd plarror m rn, ut gwen av Svenska R~ksbyggen,Stockholm, 1959 (Srr~p method for slabs on colurnns, L-shupedplares, eic.. translated by F. A Blakey, Commonwealth Scientific and Industr~alResearch Organisation, Melbourne, 1964)
L m t des~gnof reinforced concrete slabs, J. Engng Mech. ASCE, 1964, 5, 321-342.

A strip mcthod of aLb design with concentrated loads on supports, Srrucr. Engr, 1971, 49, 543-538. 36. MAI.L!CK, S. K. Gupra, A. P.

The shcar distribution method for the design of reinforced concrete slabs, I. Srrucr. Engng (Roorkee), 1971, 3, 103-112.

A generalised strip deflection method.of relnicrced concrete slab des~gn, Proc. Insin Civ Engr (Lond-), Part 2, 1978, 65, 163-174.
Computation of thc strip denexion method, Proc lnsrn Civ. Engr ( L o n d ) Part 2, 1980, 69, 499-509 Des~gn tables for slabs using strip deflection mcthod, Indian Concr. I., 1983, 57, 205-219. The advanced strip method-a Res., 1982, 34, 175-181. simple design tool, Mag. Concr.

Birnomenr equiilbrium of tinrte segments of Hillerborg plate, Mag. Concr. Res., 1979, 31, 142-150. Bimoment design of some rectangular slabs, Mag. Concr. Res., 1983, 35, 3-18. The reinforcement In slabs in accordance with predetermined field of moments, Concrete, 1969, 2, 69-76. HAGO,A. H.


Tests on RC slabs des~gnedby direct design procedure, 3. Am. C o n o . Inst., 1986, 83, 916-924. Obosnovanie norm procktirovaniya, zhetobetonnykh konstmktsii, Sronrelnayor Promyshlenmost, 1939, 17. Load tests on a three storey reinforced concrete building in Johannesberg, Struct. Engr, 1955, 33, 304-322.

The uitimate srrengih of concretepanelr subjected lo unifomly disiribnred loads, Ph.D. Thesis, Cambridge University, 1956.


A re!ined theory for reinforced concrete slabs, I. lnrm Engr, 1963, 35, 292. The effect of membrane stresses on the ultimate strength of an interlor panel In a reinforced concrete dab, Strun. Engr, 1 9 6 3 , ~ 261-265 Uitimate strength of rectangular concrete slabs under short-temuni. form loading with edges restrained against lateral movement, pror, Inrm Civ. Engr (Lond.) 1964, 28, 125-150. Membrane flexural modes of restrained slabs, J. Strucr. Div, ASCE , 1967, 93, 85-111. Load carrying capacity of slab strips restrained against longitud~n~l expansion, Concrere, 1969, 3, 369-378. Membrane action m slabs, crackmg, deflection and ultlmate loadof concrete slab systems, ACI, SP-30, 1971, 345-377. Tests on a reinforced concrete slab and beam designed with allow. ance for membrane actlon, cracking, deflection and ultimate loadof concrete slab systems, ACI, SP-30, 1971, 223-250. Load deflection behwiour of restramed reinforced concrete slabs,J. Struct. D w , ASCE, 1977, 103, 405-419. Membrane action in flexure of rectangular pljtes with restrained edges, Flexural rnechan~cs of reinforced concrete, ACIIASCEJP-12, Detroit, 1965, 347-358. Large piastic deformations of reinforced concrete slabs, Int. J. ~ b h h Slrucr., 1968, 4 , 61-74. Yield line theory for remforced concrete slabs at moderately large deflexions, Mag. Concr. Res., 1967, 19, 211-222. Lirmr strength and serviceability factors in uniformly loaded, isotroplcaily reinforced two-way slabs, SP-30, 30-1-323. Dome effect in RC slabs: Rigid-plastic analysis, J. Struct. Div., ASCE, 1980, 106, 1237-1253. Discussion on ref. 60, J . Srrucr. Div., ASCE, 1981, 107, 1017. The lateral stiffness and strength required to ensure membrane action at the ultimate load of a reinforced concrete slab.and-beam floor, Mag. Concr. Res., 1965, 17, 29-38. Some tests on reinforced concrete beam-slab panels, Mag. C m . Res., 1969, 67, 113-120.


DA~A T. , K. b 5 . DATTA. T . K. *ND R a r w , C. K.

66. DESAYI, P. AND Kummi. A. B.


C. K.


Ultimate strength of reinforced concrete slah-beam systems-A new approach, Indian Concr. J . , 1973, 47, 301-308. Some experimental studies on a reinforced concrete slab-beam system, Mag. Concr.' Res., 1975, 27, 111-120 Loaddeflection behaviour of reinforced concrete slab-beam paneb, Indian Conm. J., 1979, 53, 277-283.

67, TONG, P.



Compressive membrane enhancement in two-way bridge slabs, Paper SP30-12, ACI SP-30, Detroit, 1971, 271-286.
Load-deflection behaviour of restrained reinforced concrete skew slabs, I. Struct. Div., ASCE, 1981, 107, 873-888.

Strength and behaviour of restrained ranforced concrete slabs under static nnd dynamrc loadmgs, Tech. Rep. R621, U S . Naval Civil Engng Lab . Port Huencme. Californ~a. 1969
Ultimate strength study of nvo-way concrete slabs, 1. S m c r Dzv , ASCE, 1975, 101, 311-324 The membrane action of RC slabs after tests, Beton Stahlbetonbau, (Berlin), 1976, 71, 270-275.

Design criterion for defiction capacity of conventionally reinforced concrete slabs, Phme-I, Stme of an report, Construction Technology Laboratories, Portland Cement Assoclat~on,Illinois, 1979, 139
74. HAYES, B.

H. A N 0

Effect of the arrangement of reinforcement on the behaviour of r a n forced concrete slabs, Mag. Concr. Res., 1966, 18, 85-94. Allowing for membrane actlon in the plasnc analysis of rectangular reinforced concrete slabs, Mag. Concr. Res., 1968, 20, 205-211. Yield of a square reinforced concrete slab on simple supports allow. mg for membrane forces, Struct. Engr, 1967, 45, 235-240 Plastic behaviour of simply supported plates at moderately large deflexions, Inr. J. Solidr Strucr., 1965, 1, 97-111. Load-deflection bebaviour of simply supported rectangular remlorced concrete slahs. Proc. Itrr Aw Bridge Srnm. E q r . 1. 1978, p 11178, 1-16.

Strength and deformanon of reinforced concrete skew slabs, Ph.D. Thesis, Indian Institute of Science, Bangalore, India, 1979, 15-160.
Deflection-Problems and treatment in vanous countries, ACI, SP-43, 1974, 129-178.

80. ACI Committee 435

Deflection of two-way reinbrced concrete floor systems, State of the a n , Deflection of concr. slruct, ACI, SP-43, 1974. 318-328.
Bdilding code rcqulrerncnts o i reinforced concrete, ACI Standard. ACIJIb', 1977.

An analysis to deternune the effects of cracking in reinforced concrete slabs. Pwc. Conf finrte ei~mentmethods rip. engng, McCiN University, 1972, 841
Nonlinear analysis of ieinforced concrete slabs, Mag. Concr. Res., 1972, 24, 63-70. Cracking analysis of reinforced concrete plates, L Srruct. Div., ASCE, 1975, 101, 201-215. Nonlinear anaiysir of reinforced concrete slabs by finite element method, Tech. R e p , Osaka Universl!~, 1976, 26, 595-604.



Nonitnear finite eiemenr analysls of concrete slab structures, pra Instn Ctv Engr (Lond.), Part 2, 1977, 63, 159-180. Behawour of relnforced concrete structures predicted by finiteele ment method, Computers Struct., 1977, 7, 365-376. Finite element andiysls of re~nforcedconcrete structures, ~~~~~l~ Presenri Speciaiiry Course on reinforced concrete construction, Polytechnic, Milan, Italy, 1978. Deflection of two-way slabs from uncracked stare up to rupture,~j~ Bavtechnik (Berlin), 1977, 9, 313-317.


90. SHUKLA. S. N. M ~ A L M K.


Short term deflection in two-way retnforced concrete slabs after nac. kmg, J . ACI., 1976. 73, 416-419. Deflection of two-way reinforced concrete slab systems, Defom. aon gf concrete structures, McGraw-HIU, 1977, 279-342.

92. DESAYL, P. AUD MwI'. K. U

93 DESRII, P. AND Mumv. K. U. 94. T n x o n , P. J.

Short tlme deflectton of rectangular slmply supponed relnforced concrete slabs, Advances m concrete slab technoiogy, R. K. Dhr and J 6 . L. Munday (eds), Peigamon Press, Oxford, 1980, 110-117. Short time deflexion of rectangular, restramed, reinforced concrrte slabs, Proc. lnsrn Crv. Engr (Lond.), Part 2, 1979, 529-536. The tninal and long-term deflection of ranforced concrete flat platen ana siabs, Constr. Rev. (Sydney), 1970, 43, 68-74.
A comparison of measured and calculated deflections of flexural members in four retnforced concrete buddlngs, ACI SP-43-21. Dej7mtion of coflcrele srrucrurei, 1974, 515-525.

Long term deflection of flat plates and slabs, J. ACI, Proc.. 1976,4, 205-223.
A rattonal approach to control of slab deflections, J. ACI, Proc 1978, 75, 256-262.

Concrere Structures Code, AS-1480.74, Standards Association of Australia, Sydney, 1974. Indian Standards code of pracrrce for piam and reinforced concrete (Third revision), IS 456-1978, Indran Standards Institution, New Deihi, 1978. Discussion on deflection of RC rectangular simply suppotted slabs and the recommendations in IS 456-1978, Institution of Engineen (India), Bangalore, 13-16, October 1980. Deflectios control of siabs using aliowable span to depth ratios, 1. ACL Proc., 1985, 82, 67-72. Deflection of two-way slabs subjected to restrained volume change and transverse loads, 1. ACI, Proc , 1986, 83, 737-744. bDg-time multipliers ior estimating two-way slab deflections, 1. ACl, Pmc., 1986, 83, 899-907. Code of pracnce for the rrrucrural me of convele, Parr I, D&n. materiolc and workmanship, British Standards Institution. CPlIo. 1972.



Infernational recommendations for ?he design of concrete siructures, CEB-FIP, 1970.


107. NAWY, E. G .

Crack~ngand uitrmate strength of one way slabs re~nforcedwirh welded wire fabric, Indian Concr. J 1971, 45, 115-121, 139,

Crackwidth control in weided fabnc reinforced centraliy loaded twoway slabs, Causes, mecharrim and control of crackrng m concrete, Anlatican Concrete Institute, Detroit, SP-20, 1'168, 211-235. Crackwidth control in reinforced concrete two-way slabs, J Srrucr. Dzv., ASCE, 1970, 96, 701-721.

109. NAWY,E. 0. BMR, K. W


Further siudies on flexural crack controi in nructural slab systems, Cracking, deflectron and uitimare load of concrete slab system, ACI, D e t m t 1971, SP-30, 1-42 Determination of maxlmum crackwdth in two-way reinforced concrete slabs, Proc. Insrn Civ. Engr (Lond.),Part 2 , 1976,61,343-349. Discussion on the paper Part 2, 1977, 53, 227-231.
A method for determ~ning the spacfng and width of cracks in partially prestressed concrete beams, Proc. lnstn Cfv. Engr (Lond.),


Part 2, 1975, 59, 411-428. De~erminarionof maximum crachw~dth 1n remforced concrete skew and rectangular slabs, Proc. Insm Civ. Engr (Lond.), 1979, 67, 1077-1090. Effect of membrane action on the plasric collapse load of circular orthotropic slabs wlth fixea edges, Int J. Mech. Sci., 1978, 20, 97-108. Simple ideas in the iarge-deflection plastic theory of plates and slabs, Engineering plartzciry, Heyman and Leckie (eds), Cambndge University Press, Cambridge, England. 1968, 93-127. Some experiments on circular concrete dabs with lateral restraints, presented at the 1974 Euromech conf large drflecrion plasticiry, Warsaw, Poland. Yield line analysis of circular slabs with openings, J. Instn Engr (India), 1980, 60, 265-270. Behnviow of axially restmined slabs, Ph.D. Thesis, University College, London, 1978. Dome effect in KC slabs, elastic-plastic analysis, J. Snucl. DIV., ASCE, 1980, 6 , 1255-1262. Short t ~ m e deflexions of simply supported reinforced concrete circular slabs, Proc, lnsrn Cw. Engr (Lond.), Part 2 , 1981, 71, 323-339. some tests on reinforced conclete circular slabs with ring beams, Torsteel Res. Foundation, 1981, 1, 26-33.

~~~i~~~ crackwidth in R C circular slabs, Inr. J. Struct. (Roorkee), 1985, 5 , 127-148.


Membrane analysis and load-deflection behawour of a a t i o p ~ aim. ~ ply supported re~nforced concrete circular slabs. J Indian lnsr. sa,, 1988, 68, 109-128. Membrane analysis of restrained RC nrcular slabs, Int. I. slruet (Roorkre), 1987, 7, 1-15.

Srrength and deformnrion of remforced concrete circulor slabs, p h , ~ , Thesis, Indian Institute of Science, Bangalore, 1982.
Private communication to R. H.Wood (ref. 14) on work carned out on behalf of the Home Office. Expenmental behaviour of RC slabs with various arrangemen- of reinforcement, J . Imm Engr (Indcal, 1972, 53, 75-78

Theprovuion of reinforcement m simply supported skew bridgeslabs m accordance with elastrc moments, Tech. rep. No. 42450. Cement and Concrete Associat~on,London, 1970.
The behaviour of optimal reinforced concrete slabs, Ciu. Engng Tmm., Imfn Engr ( A w t . ) , 1967, CE9, 283-294.
129. ROZVANY, G. I. N. C m n a ~ r D. , E. 130. Rozvmu, G. I. N ADIDM, S. R.

Slabs with variable reinforcement, J. Srrucr. Div., ASCE, 1971,97, 1521-1530. Absolute minimum volume of reinforcement in slabs, 1. Srruc~D m , ASCE, 1972, 98, 1217-1222. Optimal slabs and grillages of constrained geometry, I. Engng Mech Divn, ASCE, 1975, 101, 755-771. Optimal design of variable thickness reinforced plates, Civ. Engng Trans., Imtn Engr, ( A m f . ) , 1973, 15, 99-102. Rational design of reinforced concrete slabs, Civ. Engng Trans. Instn Engr (Ausf), 1967, CE9, 209-216. Minimum reinforcement in clamped square slabs, J. Stnrcr. Div., ASCE, 1970, 96, 751-756. Destructive tests on a rationally designed slab, Mag. Concr. Res., 1970, ZZ, 25-36. Behanour of simply supported slabs, J . Strvct. Div., ASCE, 1969, 95, 2703-2721.


131. Rozvmu. G . I . N.. GANGAD~XA~ C. AH AND .

WU, R.
132. MELCHERS, R. E.


M. C. AND 134. THAKLLAR, SR~~HA bR e . 1. K.

135. MUSPRATT, M. A.

i 3 6 . M u m r r , M. A.