April 2004 * The Indian Concrete Journal 11

Discussion Forum

Seismic performance of conventional multi-storey building with open ground floors for

vehicular parking

DDDDDiscussion iscussion iscussion iscussion iscussion FFFFForumorumorumorumorum

The discusser would like to congratulatethe authors Mr Ravi and VasantKanitkar for throwing light on a verypractical topic, which is very important tothe design of multi-storey buildings of ourcountry1. The draft code of IS 1893, whichwas under circulation for more than 2 to 3years, did not contain clause 7.10. Thisclause was introduced in the final code onlyafter witnessing the failure of several multi-storey buildings during the 2001 Bhujearthquake. To my knowledge the factor of2.5 specified in clause 7.10.3(a) and thefactor of 1.5 specified in clause 7.10.3(b)of the code were chosen arbitrarily withoutany scientific investigations. Hence thispaper, which compares the base shearsbased on this approach given by the codewith the results of an inelastic analysis,assumes greater importance.

Still, we have not fully understood thebehaviour of masonry infilled concreteframes subjected to earthquake forces. Thefailure of such buildings with soft storeyat the ground floor level may be due to thefollowing reasons/practices adopted inIndia.

(i) Due to the stiffening effect ofmasonry in the top floors, more than

90 percent of the deflections occurin the ground floor columns (sheartype deformation, as rightlypointed out by the authors), whichresult in the formation of plastichinges at the base and top portionof the ground floor columns andsubsequent failure.

(ii) Normally, the length of groundfloor columns will be larger thanthe other floor columns due to theraised ground level. Many builderseither do not provide plinth beamsor design and detail the plinth beamproperly. (Incidentally, the ACIcode suggests that grade beamsdesigned to act as horizontal tiesbetween pile caps or footings shallhave a depth equal to or greaterthan the span / 20, but need not begreater than 450 mm. It alsosuggests that closed ties shall beprovided at a spacing not to exceedthe lesser of one half the smallestorthogonal cross sectionaldimension or 300 mm) 2. In mostof the cases the ground floorcolumns have to be designed aslong columns but often they aredesigned as short columns only.

(iii) The amount of lateral ties in thecolumns is usually inadequate indiameter and spacing and moreoften not provided with 135o hooksand not detailed near the joints forearthquake forces.

(iv) In many cases, the beam-columnjoint reinforcements are notprovided, even if specified by thedesigners, due to the difficulty ofproviding them2. Also as indicatedby the authors, while providingconcrete for the beams, slabs andcolumn tops (especially when thedepth of beams are different)together, weak joints are formed atlocations of maximum seismicdemands.

(v) Many designers do not carry outthree-dimensional analysis of theframes and design even the cornercolumns for uni-axial bending only.

(vi) Even though ready-mixed concretemay be used for slabs and beams,more often site mixed (withoutproper quality control) concrete isused for column, due to the smallquantity of concrete involved. On

The Indian Concrete Journal * April 200412

Discussion Forum

many occasions, concrete is notvibrated properly, resulting inhoney combing of concrete. Inseveral sites, there is no control overthe water-cement ratio (many donot even use designed mixes),which has profound effect on thestrength and durability of concrete.

(vii)Most of the architects do not haveknowledge about earthquake-resistant design and hence oftenplan the buildings, which areunsymmetrical or have non-uniform mass and stiffnessdistribution either in plan or inelevation. Due to their planning andinsistence, many design engineersprovide floating columns andconcealed beams. Floating columnsand concealed beams should notbe allowed, since they will weakenany building. Especially the floatingcolumns at the tip of cantileverbeams as indicated in Fig 2 shouldbe avoided at any cost.

Unless these practices are corrected(especially in numerous small projectshandled by small companies) buildings willcontinue to have failures, especially duringearthquakes.

The authors suggest that for multi-storey buildings with stilts, one can adoptR=2.5 with earthquake detailing for theframes and omit the scale factor of 2.5specified in the code for moments andshears. By adopting this approach, thebending moments, axial forces and shearforces will be increased uniformly in all thefloors. But as per the behaviour illustratedin Figs 1 and 6 of the paper, the columns inthe stilt floor are affected greatly and failin the event of an earthquake. Also, byadopting R=2.5, the cost of the structurewill increase considerably, since it will affectthe design of columns and beams in all thefloors.

Several researchers have conductedanalytical and experimental investigationson infilled frames under static and lateralloads 4,5,6 and suggested that one can adopta reduced bending moment of wl2/50(which is also suggested in IS:2911 for thedesign of grade beams supporting

masonry) for the design of beams due tothe arching action of masonry. But in normalpractice, only the load due to masonry isconsidered, and do not consider thecomposite action is ignored. Can weconsider this action and reduce the bendingmoments in beams supporting the masonryinfills? What will be the effect of openings(windows and doors)?

As suggested by the authors, it ispractically difficult to isolate the behaviourof masonry from the frames, unless theprecast facades are directly attached, asdone in many developed countries. Butsince masonry is brittle, the infill will failafter a few cycles. Hence the frame willbehave as an infilled frame for a few cyclesonly and after that there will not be anyinfill effect. Have they modelled thebehaviour accordingly? Of course, the bestpractice will be to have reinforced infill, andinclude its behaviour in the analysis anddesign so that one can achieve safety andeconomy.

It is believed that in tall multi-storeybuildings (beyond 15-20 floors), the windload will govern the design more than theearthquake load. In these cases also do thestilt floor columns need to be strengthenedconsidering 2.5 times the earthquake shear?

It will be interesting if the authors madea comparison of the provisions of codes ofother countries with this particularprovision in the Indian code (clause 7.10 ofIS:1893).

Dr N. SubramanianChief Executive,

Computer Design Consultants,763,Quince Orchard Boulevard

Apt#22,Gaithersburg,MD 20878,

USA

References1. KANITKAR, RAVI AND KANITKAR, VASANT Seismic

performance of conventional multi-storeybuilding with open ground floors for vehicularparking, The Indian Concrete Journal, February2004, Vol 78, No 2, pp. 9-104.

2. ______Building code requirements for structuralconcrete, ACI 318 M-02, American ConcreteInstitute, Farmington Hills, Michigan, USA,2002.

3. SUBRAMANIAN, N. and PRAKASH RAO, D.S. Seismicdesign of joints in RC structures, A review, TheIndian Concrete Journal, February 2003, Vol 77,No 2, pp. 883-892.

4. GOVINDAN,P. and SANTHAKUMAR, A.R. Compositeaction of reinforced concrete beams with plainmasonry infills, The Indian Concrete Journal,August 1985, Vol 59, No 8, pp. 204-208, 222.

5. GOVINDAN,P., LAKSHMIPATHY, M. andSANTHAKUMAR, A.R. Behaviour of infilled beamsunder cyclic loading, The Indian Concrete Journal,May 1987, Vol 61, No 3, pp. 124-128.

6. ACHINTYA, DAYARATNAM, P. and JAIN.S.K.,Behaviour of brick infilled RC frame under lateralload, The Indian Concrete Journal, Vol 65, No 9,September 1991, pp. 453-457.

The authors reply:The authors wish to express their appre-ciation to the discusser for writing an illu-minating discussion on our paper. The in-clusion of the factors in section 7.10.3 inIS 1893 : 2002, although arbitrary to someextent and obviously based on a last-minute decision, was definitely a step inthe right direction towards addressing thecollapses seen in the Bhuj earthquake. Weconcur with the discusser that at this stagewe have not fully understood the behav-iour of masonry infilled concrete framessubjected to seismic forces.

Since writing the paper, we have beenactively researching building codes fromthe USA and Europe to better understandtheir approach to such structures. We hopeto broaden this survey to codes applicablein areas such as Taiwan and New Zealand.Our initial research indicates an on-goingdiscussion regarding masonry infilledconcrete frames, at least within theEuropean region. We hope to explore thisissue further and we fully intend to write acompanion paper within the near future.

The discusser has raised the issue ofcolumn failures within the stilt floor, evenwith the use of a lower R factor. Our intentwas to use the lower R factor in conjunctionwith explicit modelling of the effect of theinfill. Simplistically, this could be done byusing wall elements or even cross-bracingto model the rigidity of the infill. Althoughvery rudimentary (since infill strength isnot always easily quantifiable), such anapproach would be applicable within theelastic range of frame and infillperformance and could be used effectivelyto guide the design of stilt buildings. Suchan analysis would reduce inter-storeydisplacements at the upper floors andconcentrate a majority of the displacementand rotation demands within the stilt floor

April 2004 * The Indian Concrete Journal 13

Discussion Forum

columns. Thus, the seismic forcedistribution would not be uniform over theheight of the building. We believe that suchan analysis will not lead to over-designingof the upper floor members, since the inter-storey displacements would be very smallat these upper floors (an inherent effect ofinfill). The use of the lower R factor isintended to address the poor ductility,redundancy and energy dissipationcharacteristics of the stilt type buildings(as shown in Figs 6 and 7 of the paper), inkeeping with the spirit of the code.

In our non-linear analysis effort weexplicitly modelled the infill within theconcrete frames. However, we did not seeany significant failure within the masonryprior to inelastic action within the frame(concentrated at the stilt level). If thestrength of the frames is increased, we maysee the brittle infill show some signs offailure. We have not performed suchparametric analyses yet. The main issuewith such parametric studies would be togauge the strength and workmanship of

the infill. It is anticipated that infill strengthwill have a significant effect on theperformance of infilled concrete frames. Inthe Indian context, the variation in theconstruction of the infill (type of brick andmortar, quality of construction, etc) wouldmake it rather difficult to quantify infillstrength and energy dissipation properties.If all infill conformed to some standard(such as hollow concrete block withdeterminable properties), this processwould be greatly simplified. Still, we hopeto include such parametric studies in ourfuture work.

As far as wind load is concerned, theapproach has to be dual. Wind loads areresisted by structures without resorting toinelastic action (yielding, plastic hinges,etc). Codal seismic forces account for suchinelastic effects inherently (use of the Rfactor). Thus, multi-storey buildings haveto be designed for both wind and seismiceffects, with the appropriate governingdemands dictating the design at a memberlevel, rather than the global building level.

As mentioned earlier, we hope tocompile a comparison of various nationaland international code provisions for suchtypes of buildings. We definitely intend tosubmit such work for publication in theICJ.

The discusser has also pointed outseveral detrimental construction practicescommonly employed in India, which ifremedied, would go a long way towardsmaking concrete frame buildings morereliable under earthquake forces. We concurwith his views and hope for morediscussion on the subject.

Lastly, we thank the discusser for hisinsightful comments on the subject at hand.

Mr Ravi Kanitkar280 Ravenswood Avenue,

Suite 4, Menlo Park,CA 94025,

USA

Mr Vasant Kanitkar21/4, Bund Garden Road

Pune 411 001