Anda di halaman 1dari 19
No.22-7 A proposed draft for IS: 1893 provisions on seismic design of buildings — Part I*: Commentary and examples ‘Sudhir K. Jain ‘The provisions for seismic design of buildings contained fy 1S:1893-1984 need to be revised im view of many efciencies that are currently being fl. Pat I of this paper dlacused a proposed draft on provisions for seismic esi of buildings for inclusion In the nest edition of the code. This paper provides a detailed commentary to ‘explain the proposed eadal provisions. ased on the detailed review! of 15:1993.1989 provi: Jonson seismic design of buildings, a revised drat forthe same as en presented ia act Tf this pape In order "plain these provisions and wo give the itet Bend some ofthe clases, ths paper provides detailed commentary. Inne following sections, clase number ea in Pat 1 of the paper. For instance, clause C341 of his pape ew tains discussion about clase 34.0 of Pat Only those ‘lmses of Part T which rege discussion ar into in the commentary. Figures and tbles of Pat Hare given numbers staring with C. Thus, for example “Tale # refers tothe Table 4 of Part To hs paper, while "Table C8" erst Table Cl of Par Tf the pape. COMMENTARY Symbols (22) The 1984 edton ofthe cn considers aie ton im seismic sk in dierent pts ofthe cour tr "tsi rz eof” () ine seismic eon meta and though “sesmic zone fator” CF) nthe fe sponse specu method. Tht i ely mo eed fr dfn ‘0 diferent parameters forthe same purpose in ft is Simply five times. Hence, in the new provisions, a in- septate zoe factor (2) haben dein. * Ato fon, Deparment of Cel Eons fa ste of Tohnaloy, Kanpur 208 046, India. Symbol “A's been assigned to represent the de sipn horizons acclertion spectrum ative a afer con sidering all the relevant factors such as the imprtance factor (zone aco 2) response reduction fact, nt sel profi fctor (3). Tiss the spectrum 0 be finaly wed for design of u pacar type of buldng at hit sit, inspect of the atalyss procedure used (sai oF dynam) Ground Motion (C31.1. ‘The Nonridge earthquake of Sanwary 17.194 in southern California has early shor the valeablityof prestressed horizontal memters to verte! component of ground ino ton. To check the structure for vertical component of mo tion, it may be sulcent to consider the structure, execpt forthe lage span seuctares, 8 igi for vera ibeations and to sijec ito zeo-petad vical aseeratons, with oreo factor (ie, the eine ene 80521). Assumptions (C32): (€22(0 The elastic modulus for materials such as conerte and maomy is ificl 1 speci. Ts value varies with sess level, loading conditions (sac vers dynamic), ‘UAL O# STRUCTURAL ENGINEERING VOL.22 NO 29ULY 1985 » material strength, age of the material, etc. Hence, there tends to be a very large vasiation in the value of elastic ‘modulus specified by different codes even for a specific rade of concrete under static condition. For instance, ACE318% recommends # as 4700 YF (MPs), while IS : 456-1978° suggests $700 Vig (MPa) or about 6370". Yiy = characteristic cube strength, c= characteristic eylinder strength = 0.8 fa) ie E given by the IS code is about 14 times the value given by the ACI fede for the same grade of concrete. Further, actual concrete strength in a structure is usually more than the specified 28-day strength and it also increases with time, Thete are further difficulties with choosing the value of modulus of elasticity for concrete of seismic analysis. The value given in the codes, such as ACI-318° and 1S:456° is ‘often the secant modulus: its value is prescribed with a view to obtain a conservative estimate of deflections, ic. lower stifess. On the ober hand, the dynamic modulus of ‘concrete refers to almost pure clastic effects and is equal to the intial tangent modulus and is appreciably higher than the secant modulus. When a strcture fs new and subjected to low amplitude of ground motion, the dysamic modulus ‘of clastcity may be applicable, However, long time expo sure ofthe stricture t9 wind pressures may overcome the initial stiffness properties, and the modulus of elasticity of cconerete may tend to be close to the secant modulus, The value of modulus of elasticity to be used in analysis has ‘vo opposite implications on scismic design, For calculi tion ofthe design seismic fore itis uneonservative to have low value of modulus of elasticity; this leads to high time period and lower design seismic coefficient. However, for the drift riteria (deflection condition) itis unconservative to make a higher estimate ofthe stiffness, Hence, there are no easy answers t0 the question of hat value of modulus of elasticity should be use for sei mic analysis. Considering the enormous ‘variations, this clause allows the designer to use elastic modulus as for static condition. However, a safeguard has been intruced (44.3 and 4.62) against using a very high value of natural period forcalculation, Load Combinations and Increase in Permissible ‘Stresses (C33): €33.1: The design ground motion can occur slong any direction of building. Moreover, the mation has diferent directions at different time instants, The eartiquake ground ‘motion can be thought of in terms of components in the 180. horizontal and one vertical directions. For buildings with lateral force resisting elements oriented along two principal icetions, itis usually sufieient to design the bulking for the earthquake force acting in «- and )- directions sepa rately; i, not for fores acting in both the dicections simul taneously (Fig. C1(a)). During earthquake shaking, when the resultant ground motion isin a direction her than x andy, the motion can be resolved into the wand y- oa ae = “ on, toon, te, © TFIGCI\@) EARTHOUAKE LOAD CONDITION FoR DUSIGN ‘OF HUILDINGS WHT LATERAL LOAD RESISTING SYSTEMS ORIENTED ALONG TWO PRINCIPAL DIRECTIONS, (0) TO. ENSURE No ELEMENT 1s UNDER DE SIGNED, EARTHOUAKE FORCE SHOULD BE ‘CONSIDERED IV ALL POSSIBLE DIRECTIONS I WHICH THE ELEMENTS ARE ORIENTED: (ALTERNATIVE TO CONDITION (by THE BUILDING MAY BE DESIGNED CONSIDERING FULL DESIGN LOAD IN ONE DIRECTION AND 4% DESIGN LOAD INTHE OTHER DIRECHON, [ACTING SIMULTANEOUSLY AND VICEVERSA components, as elements in the two principal directions ‘which are normally ble to withstand, except forthe earner columns for which this may be unconservative, However, when the lateral force resisting elements re not oriented along the x- and y= divection, design based ‘on earthquake force in x- and y-direction, separately, leads {o underdesign of the elements. In such » ease, one should design the structure for earthquake force ating along all possible directions in which the seismic load reiting ele ‘ments are oriented (Fig. C1(b). One way to get around the Aiticuty of having to consider too many possible earthquake ietions iso design the seucture for (Fig. C1(6) 1. Full design force in the xdieetion (EL) acting Simultaneously with 40% of the design foree in the pdirection (EL,); ie, FL, +04 EL,), and " JOURNAL OF STRUCTURAL ENGINEERING VOI.22 NO 2 IDLY 1995 Full design force in the y-direction (EL,) acting simultaneously with 40% of the design force inthe seditetion (EL); ie. (04 EL, + EL) ‘This combination ensures that elements oriented in any direction will have sufficient lateral stength. Iti also a ‘200d practice to design the comer columns of otherwise ‘orthogonal sytem as per these combinations. Design Spectrum (C3. C342: The present code” provides diferent design spectra i sein the seismic coefficient and the response spectrum ‘methods. The drat provisions provide fora common design spectrum which is applicable irrespective of whether the Assign force is calculated bythe static or dyamic procedure Several important changes have been introduced in the new design spectrum 44 The performance factor (K) in the earlier version, has been replaced by a response reduction factor (R). The soil-foundation factor (B) has been replaced by a soi- profile factor (5), and the basic horizontal coefficient (4) and seismic zone factor (F,) have been replaced by the zone factor Z). The terms representing the importance of structure (0) and the strate flexibility effect (C) are the same. b. In the earlier version, the code directly specified the design seismic fore; this was often misunderstood as the maximum expected force onthe structure. In Tine with the world-wide trend in this regard, the code ‘now tris to distinguish the two. The terms (21 CS) represent the spectrum corresponding to the maxi ‘mum expected earthquake force, if the structure isto respond elastically, andthe design force is arived at by dividing this force by R. The term R gives a cleat indication of the level of overstrength and ductility that a structure i expected to have The term Z now represents the realistic values, as fraction of ccoleration due to gravity, ofthe expected peak ground acceleration in diferent seismic zones. For instance, the code specifies zone IV for areas Which are likely to sustain shaking of intensity VIII on the Modified Mezcal sale. The value of Z (© 0.30) for zone IV gives the value of peak ground acceleration as 030g which may be reasonably expected in shaking intensity VIL 4. Adoption of realistic values of peak ground accelera- tion asthe seismic zone factor has als rationalized the relative values of design force for different sis: mic zones. As the intensity of shaking goes up one level on the MM scale (say from VI to VII), the peak ground acceleration almost doubles. Tn eatlier code this was not duly reflected since the seismic force in ferent zones varied in the ratio 1:24:58 ‘notes change inrouced i that he soi foundtion factor (has eon replaced hy the si profi factor Factor B, depending on the type of soil and the type of | Toundation, was intended to inereae the design force for systems tht are more wlneabe to ier setlomonts. However, in rn eahquake stato, ‘uildings dono ser hich earn induced in eri force on account of sera t erat setlement Alo the problem offre stle- Imnt cannot be adresse by inreasing the desian senmic force onthe bung: instead thas 10 be askressed hy a proper choice of the foundation, On the other and, revo obsned inthe past eth avakesclearyshow that the average sceeeaon {pecan tends to be ificat for sien wh fie So profiles (Fig C2), The sc profile factor ()con- Sie ths variation, 8 Su ae aa 3 G.€2 EFFECT OF SOI. PROFILE ON SHAPE OF [RESPONSE SPECTRUM, JOURNAL OF STRUCTURAL ENGINEERING VOL.22 NO 2JULY 1995 ‘The product of terms (C) and (S), shown in Fig2 of the draft code, represents the shape of design spec- trum with peak ground acceleration scaled to the value of 1.0 This shape is same as the average shape of acceleration response spectrum, except inthe pe- riod range O- 0.1 sec (Fig.C3). In this ange, the plot of Fig.2 is ata constant valve as against the response spectrum which varies from I.0 at ero period to the ‘maximum value (equal 1 2.0) at period of around 0.1 sec. The shape of design spectrum is mosified in this range in view ofthe fact that ductility does not help in reducing the maximum force on stiff stu ‘ures ith fundamental period in the range 0 10 O11 sec?" However, itis acceptable if one were to use the shape of response spectrum in this range, for ‘modes other than the fundamental mode (Fg.C’) ‘The basic philosophy of earthquake-esistnt design is that a stricture should not collapse during a severe earthquake, although it may undergo some structural as well us nonstructural damage. Hence, a building i ‘usually designed for a much less force than what would be required it the building were to remain eas- tic during a severe earthquake shaking, Fig. C4 shows s