[Ninth Symposium on Eatiquake Engheeing ( 9SEE - 90), Roorkee, Deceber 14-16 1990, Vol Seismic Design of Frame Staging for Elevated Water Tanks Jain, Sudhir K. Sameer, U.S.” Tassistant Professor, Department of Civil Engineering, indian Institute of Technology, Kanpur-208016 ‘Lecturer, Department of Applied Mechanics, Karnataka Regional Engineering College, Surathkal Z SYNOPSIS 1.5. code provigiona for aseiamic aeaign of elevated water tanka have been reviewed. It la seen that, due to absence of a suitable value of performance fuctor for tanks, the code provides for rather low seismic design force for these atructurea. Simple expreaelone are given which allow calculation of ataging stiffness, and hence the time period, while incorporating beam flexibility. 4. INTRODUCTION In India the selamlc deaign criteria is provided by 15:1893-1984 (40) which gives mininum loading atandarde and 18:4326-1976 (11) which containa deaign and detailing requirementa for conatruction of bulldinga. 18:11682-1985 (12) gives the criteria for design of RCC staging of auch eteuctures. In thie paper provisions pertaining ta aaatamic deaien af elevated vater tovers are reviewed and aeveral suggestions given for making these more rational. The deaign seiemic force for the water tank depends on ite flexibility and hence on the time period. Many engineere In the country tend to evaluate the time period by taking the column atiffneas aa 12Et/L3 which i# baged on the asaumption of bracing beama being infinitely cigid. In practice these beams are quite flexible and therefore the above assumption very such overestimates the etaging stiffness. An approximate procedure is presented for calculation of staging stiffness, and hence time period, giving due conalderation to the beam flexibility 4-1132, REVIEW OF INDIAN CODAL PROVISIONS 2.1 Lateral Design Force 18:1893 requires that the design lateral force shall be taken as F = a,U; where a, BIE,S,/a; @ = coefficient depending upon the 9011-foundation syste B= aelemic zone factor; and S,/q = average acceleration coefficient obtained from acceleration performance factor, K, which the code uses for computing base shear for buildings, 1a absent. Thia implies that ; I= importance facto: apectra given in the code. Heri K 1a 1.0 for the elevated vater tovere which ia aleo the valug/ used for bulldings vith ductile ao the elevated tank type structures have lover energy absorbing capacity nt realeting frames. Thia ia unreasonable ae and poor ductility as compared to those in ductile’ moment resisting frame bulldinga. The all over the world prescribe a performance factor which Ia 2.5 to 4.0 times higher for elevated water tanks than that for the ductile buildings. Thus while the lateral design force prescribed for buildings in Indian codes is of about the same order that in the other codes for zonea of comparable selemicity; the deeign lateral force provided by Indian code for elevated tanke ia far below that by seismic codes in other countries. Hence it ie nec: lamic cod: ary to Introduce a auitable value of performance factor, say 3.0, for elevated tanke in 1S:1893. ree of Freedom Idealieation 2.2 Single D. 18:1892 auggeate a eingle degre: tanks vhich is reasonable only for cloged tanke completely full of water. For tanka with a fre of freedom idealisation of elevated water aucface tyo-mase idealigation (7) le preferred. This view hae been supported by eeveral investigatora (1, 2, 3, 6, 9, 15, 17) based on experimental and computational observations. Calewlationa for a fey tanks Indicate that the alngle degree of freedom representation overestimates the lateral design force; the difference in value depends on the geometrical properties of the tank and relative atiffness of the staging. Hence, there le a good « of freedom idealiaation in IS:1893 alec. for incorporating two degre: 2.3 Staging Stiffness and Time Period In both eingle- and two-degree of freedom idealisations atiffnes: the staging needs to be obtained for calculating the time period. It has been emphasised that a design criteria must aleo include the procedure for time period calculation (8). Thie is because there can be large of aig ivariation in time period calculated by different designers based on different assumptiona and/or different procedures. The Handbook SP: 22-1982 (18) takes the column etiffneae = 12RI/L? which ssaumes horizontal bracings to be infinitely rigid. However the bracing beams are the stiffness clearly not rigid and thie assumption overeatim substantially. It ia auggeated that IS:1893 muat apecify that in time period calculation bracing girders are not to be treated as rigid. 2.4 Hydrodynamic Pressure on Tank Valle 18:1893-1984 giv: formulae for determining thie presaure letelbution due to impulsive preasure (and not that due fo convective pressure) which are based on Ref (7) Hovever, the code must give complete details of the equivalent mechanical analog model proposed by Housner in the came publication. Further, clause 5.2.7.1 of the code atatea that the convective presaurea can be ignored in comparigon with can be a the impulsive pressures. Hovever, the convective preseur dominant factor for certain proportions of the tank and the structure. Moreover, I8:11682-1985 clearly states that “whereever required the affect of aurge due to vave formation of the water may be conaidered.” Hence 1$:1893 must provide complete detaila of Housner‘s model and omit the clause 5.2.7.1. Tt Ja known that the a substantially greater than tho le e€fecta on flexible tanka are Induced in aimilarly excited rigid tanks (S, 20). Simple procedures for evaluating hydrodynamic force in flexible tanks have been developed (5). It le di about larger forcea Induced in flexible tanka and to provide procedure ireable to caution the designer for their analysia in the IS code or in SP:22. 2.5 Ductility Requirements Ae ite title itaelf auggeata, 15:4326-1976 gives deaign requirementa for bulldinge only. This makes many deeign engineere think that ductile detailing 1a not required for elevated tanka even in zonee IV and V. Hence the acope of IS: 4326-1976 aut be enlarged to alae include etructures other than buildings. A figure in 1$:11682-1985 tates that "where deaign selamic coefficient ie 0.05 or more reference to clausea 7.2 to 7.4 of 18:4326-1976 ahall be made to cater for ductility requirement” Thia ia in right spirit. Hovever, the term “design aelaaic soefflelent™ in ita present form la very confusing. IS:1893 defines the 4115horizontal selemic coefficient, a , in tvo ways which mean tvo different things. The future editiona of these codes must be very clear and specific on ductile detailing requirements for tanke 2.6 Design for Torsion It has been recognised that it is impossible to prevent torsional response in elevated water tovere (16, 19). The elevated towers conatructed are seldom truly symmetric due to presence of atairca: Further, during Intenge shaking, with fallure of one or two bracinge, the it mandatory to atructure will go into the toreion mode. SEAOC-1980 mak. design elevated water towers for shear etresa developed due to horizontal toraion resulting from an accidental eccentricity equal to 5% of the largest lateral dimension. It is very much desirable for 1S:1893 to alac require deaign for accidental torsion 3. STAGING STIFFNESS Well known protal thed (4) has been auitably developed to account for the bracing flexibility and the three dimenaional behaviour of the ateucture (13). The point of inflection is assumed to occur at the aidepan of beama and columns. The compatibility requires thet lateral deflection la aame in all the columna of a panel. Hence, the columna share lateral force in proportion to their lateral atiffnesa. Conaidering a column between two bracing levels (Fla. 1), the deflection In column, A,» due to shear, V, can be calculated. This gives the lateral stiffness of one column ae (13) 12et v.18 .te See a me Sree Be Ee FT winn 78, 7 EE | i . . ee ze tec tote a oe BD Tbe 7 ye See ye te ets a isc co : 7 where Ey, E, = modulus of elasticity of beam and column, respectively; I, I, = moment of Inertia of beam and coluan, reapectively; A = helant 4116of panel under considerations h,, hy = height of the panel just above and the panel jJuat below the panel under conaideration, reapectively; WM, = ting the column under congideration at the bracing number of beans level; L = apan of bracing girder; and a = angle the bracing girder makes vith the direction of lateral force. Moat tank etagings have identical bracing girdera and have equal panel heights. Moreover, the top end of column in topmost panel and bottom end of column in bottommost panel is fixed against rotation. Using these congiderations, column atifinesa can be obtained as (13) 128 1 Ek, cle ba . Intermediate pancleS (4) home gs fee ake |S i ® [ te ne 1281, Bkpe ° A —— (top and bottom panels) (5) h ba c vnere Ek, .* Ek,,> Ekyy: k, = E,l,/hs and the summation ie from 1 to N, To evaluate the stiffness of a panel, etiffneas of Individual columns may be calculated from Eq. (4) or Eq. (5) and thia may be summed for all the columna of that panel. | A direct expreasion for the determination of panel atiffneas, without calculating individual column stiffness, has been obtained by further approximatione (13). For tank stagings with uniform panel height, identical columna, identical bracing girders, and agauming that all the columns are located on the periphery of one circle panel atiffness Lo obtained By ole 128,10 ae Kees ge [Pe] intermediate pane = , te Tt bh BI 12E 1 LN, bb x * Soe (te id dott 1a) a panet = SE Le Ey] to and bottom panel a aa _ BKAMPLE Consider the water tank in Example (6) of Handbook (SP:22) ignoring ite diagonal bracings. The tank has 4 panels of height 4 m each (total aayheight 16 m) with 8 columns Incated on a circle of radius 4.5 m. Column gine ia 520 mm diameter. Beam aize ia asauued to be 200 ¥ 500 mm. Both in tank 5 columna and beams are aaaumed to be of M20 concrete. The m full condition le 9 ¥ 10° kg and in tank empty condition it ie 3 X 10 ka. o af Now, E, = B, = 5700 720 = 25,500 MPa; N, = 8; I, = 2.08 X 10° mm T, = 3.59% 109 a"; h = 4,000 mm; and L = 3,534 mm. The moment of Inertia has been calculated asauming gross uncracked section ignoring ateel. Substituting th: Into Eqs. (6) and (7), the panel atiffness ia obtained as 33,900 N/mm for the tvo Intermediate panela and 54,400 N/am for the tvo end panela. Thia givea the ataging atiffnesa aa 10,400 N/mm. On the other hand, If one were to treat the bracing girdere as rigid, panel atiffnesa is obtained as 137,000 N/mm and the staging atiffnesa aa 34,200 N/mm. Thia ataging etiffneea ie 3.3 timea what it @hould be. The time period, for tank {411 condition, 1a 1.85 aee evalu considering beama flexible, and 1.02 sec otherwise. It is thua clear that the beam flexibility could not be ignored. Let thie tank be located in selamic zone V (F, = 0.40). Let @ = 1.0 and I = 1.5. For 5 % damping, S,/ @ = 0.11 for T = 1.02 sec and S, /e 0.06 for T = 1.85 sec. If staging ia deeigned treating the beans ae rigid and with no performance factor, a = (7 IF, S, /g =0.066 and the deaign lateral load V = a, = 594 KN. However, introducing performance factor x 3.0 and accounting for the beam flexibility, a, =K@ I F, 8, /e 0.108 and the deaign selamic force V = 972 kN. Thue the incre in design force for thia tank is only about 64% Thia Increase will be reduced somevhat 1f tvo-degree of freedom ideallaation 1a adopted. In the above example, beam length (L) and the column height (h) have been taken on the basia of centre to centre distances. Thia Ignores the additional stiftn Joint respectively. provided by rigid member zones at beam-coluan - To incorporate thia, one may take clear apan and clear height, 5. SUMMARY AND CONCLUSIONS The Indian code provisions for elemic design of elevated tanke have been revieved. The code must include an appropriate value of Performance factor, say 3.0, for calculation of seismic design force for water tanka. An earthquake design criteria ia incomplete unless cl 411gspecifications are included on how to calculate time period. The code is defficlent on thia. The Handbook (SP:22) aolves an example problem treating the bracing beams ae infinitely rigid which ia unrealistic. The code should alao give more detaile of the Houaner'a mechanical analog model for hydrodynamic forces. Proviaion for eloshing of liquid ehould be Included. The requirements regarding ductile detailing of tanka are vague in the present form of codea and need to be clearly apecif{led. A method flexibility and nt type analysie hae been for calculating the staging stiffn including bei without having to resort to finite ele: Presented. The method is based on the well knoun portal method which hae been suitably developed to incorporate the beam flexibility and the three dimenaional behaviour of the staging. REFERENCES 1. Boyce, U.H., 1973, Vibration Tests on A Simple Water Tover, Proc. SUCEE, Rome, Italy, Vol.1, pp. 220-225. 2. Cloud, U.K. 1963, Period Measurements on Structures in Chile, Bull Selam Soc Am, Vol.53, pp. 439-480. 3. Gracia, SM., 1969, Earthquake R. of Cylindrical Tanke, Proc. 4UCEE, Santiago, Chile, pp. 169-182. 4. Green B.N., 1978, Earthquake Resistant Design and Conetruction, Van Noateand Reinhold Company, New York. 5. Haroun, M.A. and Housner, G.U., 1981, Selemic Design of Liquid Storage Tanks, J. of Technical Council, ASCE, Vol.107, pp. 191-207. 6. Haroun, M.A. and Ellaithy, H.M., 1985 “"Seismically Induced Fluid J. of Technical Topica in Civil Enge, ASCE. 7. Howsner, G.W., 1963, The Dynamic Behaviour of Water Tanka, Hull of Seiam Soc of Am, Vol. 53, pp. 381-387. @. Howaner, G.W. and Jennings, P.C., 1982, Earthquake Design Criteria, Rarthquake Engineering Research Inatitute, Berkeley, California. 9. Tfrim, M. and Bratu, C., 1969, The Effect of Selamic Action on the Dynamic Behaviour of Elevated Water Tanks, Proc 4UCEB, Santiago, Chile, Vol. 4, pp. 127-142. : 10, IS: 1893-1984, Indian Standard Criteria for Earthquake Reeletant Deaign of Steucturea, Bureau of Indian Standards, Nev Delhi. 41. IS: 4326-1976, Indian Standard Code of Practice for Farthquake Resistant Design and Construction of Buildings, BIS, New DeJhi. 12. IS: 11682-1985, Indian Standard Criteria for Dealgn of RCC Staging ponge Analysis and Selemic Design Forces on Elevated Tank: 411913. 44a. 15. 16. a7. 18. 19. 20 for Qverhead Water Tanks, Rufeau of Indian Standards, Ne@ Delhi- Sameer, S.U., 1990, A Study on Seismic Analysis and Design of Elevated Water Tank Stagings, M. Tech Theale, IIT Kanpur SEAOC-1980, Recommended Lateral Force Requirements and Commentary, Structural Engineers Association of California, San Francisco. Shepherd, R., 1972, Tvo Mass Representation of @ Water Tower Structure, J. of Sound & Vib., 23(3), 391-396. Shepherd, RB. 1973, The Selamic Responae of Blevated Water Tanke Supported on Crossed Braced Tovers, Proc. SUCEE, Rome, Italy, Vol. 1. Sonobe, ¥. and Nishikava, T., 1969, Study on the Earthquake Proof Design of Elevated Water Tanks, Proc. 4UCEE, Santiago, Chile, Vol.4. SP; 22-1982, Explanatory Handbook on Codes for Earthquake Engineering 18: 1893-1975 & IS: 4326-1976, BIS New Delhi Steinbrugge, K.V. and Moran, D.F., 1954, An Rnaineering Study of the South Californian Earthquake of July 21, 1952 and its Aftershocks, Bull Selem Soc Am, Vol. 44, Appendix R, pp. 436-453 Veletaoe, A.S., 1973, Selamic Effects in Flexible Tanke, Proc. SUCER, Rome, Italy, Vol.1, po. 1-24 quid Storage fpr =O Bs an ee toes ae « Se er Bete na Be = Ops Byz ees Bee Point of inflection eink of intiection Fig. (1) Deflection Betueen Successive Joints 4-120