1. Itdepends on the properties of the dam, the depth of the water, and the absorptiveness of the reservoir bottom materials. The nat ural vibration period of the equivalent SDF system representing the fun- damental mode response of the dam on flexible foundation rock with empty reservoir is (12) Tp RT eee eeeeeeeeeeee Because of the frequency-dependent, added foundation-rock fle arising from dam-foundation rock interaction, the factor Ry > 1. It de- pends on the properties of the dam and foundation rack. ‘The natural vibration period of the equivalent SDF system represent- ing the fundamental mode response of the dam on flexible foundation rock with impounded water is approximately given by (11) RT. (de) ‘The damping ratio of this equivalent SDF system is (11) b-papetete 6 Rey in which & = the damping ratio of the dam on rigid foundation rock with empty reservoir; & = the added damping due to dam-water inter- action and reservoir bottom absorption; and & = the added radiation and. ‘material damping due to dam-foundation rock interaction. Considering, that R, and Ry > 1, Bg, 5 shows that dam-water interaction and dam: foundation rock interaction reduce the effectiveness of structural damp ing. However, usually this reduction is more than compensated by the added damping due to reservoir bottom absorption and due to dam- foundation rock interaction, which leads to an increase in the overall damping of the dam. | . The quantities R,, Ry, &, &, Prly,T.), Land My, which are required to evaluate the equivalent lateral forces, Eq. 1, contain all the modifi- cations of the vibration properties of the equivalent SDF system and of the generalized earthquake force coefficient necessary to account for the effects of dam-water interaction, reservoir bottom absorption, and dam- foundation rock interaction, Even after the considerable simplification necessary to arrive at Eq, 1, its evaluation is still too complicated for practical applications because the aforementioned quantities are com- plicated functions of the hydrodynamic and foundation-rock flexibility ferms (11). Fortunately, as will be seen in a later section, the compu tation of lateral forces can be considerably simplified by recognizing that the cross-sectional geometry of concrete gravity dams does not vary widely. Approximation of Hydrodynamic Pressure—The hydrodynamic pressure function Py(y,7,) and therefore lateral forces are real-valued at the period , for nonabsorptive reservoir bottom materials, ie., a = 1 (3), but complex-valued if the hydrodynamic wave absorption in the al luvium and sediments at the bottom of a reservoir is considered (a < 1), However, the imaginary-valued component of lateral forces is small 1691 J, Strack. Eng. 1987.15 1688-1708,yal igs rer, 812. Copyright ASCE, For personal use. logy, Deki on | Downloaded fom ascelinaryng by Indian Insite of Tes relative to the real-valued component, increasing near the base of the dam (10), where, because of the large stiffness of the dam, it will have litle influence on the stresses in the dam. Consequently, the imaginary- valued component of Pi(y,T,) may be neglected in the evaluation of the lateral forces f(y), and Eq. 1 becomes Ey SPB “ n= Bs @ ©) fone) + spt where ply, T,) = Refpi(y,T,)]. It is only in Eq, 6 that the imaginary-val- ted component of pi(¥, 7) has been dropped, but its more important effect, the contribution to added hydrodynamic damping & in Eq. 5, is stil considered, ‘The generalized mass Ny of the equivalent SDF system, Eq. 22, only depends on the real-valued component of hydrodynamic pressure: ee @® = My + [ Py. Tro(yddy (a) where M, is defined in Eq, 25. However, the generalized earthquake force coefficient L, , Eq. 32, contains both real-valued and imaginary-val- ued components of the hydrodynamic pressure. Again, dropping the imaginary-valued component gives, = 7) iene ry Tdy here Ly is defined in Eg, 3b. ‘STANDARD PROPERTIES FOR FUNDAMENTAL MODE RESPONSE Direct evaluation of Eq,_6 would require a complicated computation of several quantities: p(y, T,) from an infinite series expression; the pe- riod lengthening ratios R, and R, due to dam-water and dam-foundation rock interactions by iterative solution of equations involving frequency- dependent terms; damping ratios & and &, from expressions involving, complicated foundation-rock flexibility and hydrodynamic terms; the in- tegrals in Eq. 7; and the fundamental vibration period and mode shape of the dam (11,12). The required computations would be excessive for purposes of the preliminary design of dams. Recognizing that the cross- section geometry of concrete gravity dams does not vary widely, stan- dard values for the vibration properties of dams and all quantities that depend on them and enter into Eq. 6 are developed in this section. A complete set of standard tables and curves is presented in Ref. 10, avail- able from the University of California at Berkeley. Vibration Properties of Dam.—Computed by the finite element method, the fundamental vibration period, in sec, of a “standard” cross section for nonoverflow monoliths of concrete gravity dams on rigid foundation rock with empty reservoir is (3) H, Tel. ® 4 Staci, 1987113:1688-1708use only al rights reserved \Va82. Copyright ASCE : 2 5 FIG, 1.—Fundamental Period and Mode Shape of Vibration for Concrete Gravity ‘Dams: (a) Standard Period and Mode Shape; (4) Comparison of Standard Values with Properties of Six Dams FIG. 2.—Standard Values for Perlod Lengthening Ratio R, and Added Damping Ratio ¢, due to Hydrodynamics Effects; £, = 3,000,000 pei which H, = the height of the dam, in ft; and E, is the Young's mod- ulus of elasticity of concrete, in psi. The fundamental vibration mode shape dy) of the standard cross section is shown in Fig. 1(@), which is compared in Fig. 1(@) with the mode shape for four idealized cross sec: tions and two actual dams. Because the fundamental vibration periods and mode shapes for these cross sections are similar, it is appropriate to use the standard vibration period and mode shape presented in Fi 1(@) for the preliminary design and safety evaluation of concrete gravity dams. The ordinates of the standard vibration mode shape are tabulated in Ref. 10 for the user's convenience. Modification of Period and Damping: Dam-Water Interaction. —Dam- water interaction and reservoir bottom adsorption modify the natural vibration period (Eq. 42) and the damping ratio (Eq. 5) of the equivalent SDF system representing the fundamental vibration mode response of 1693, 4. Suc. Eng.1987113:1688-17083 ce by lin as Dowaloded from sselibary. the dam, For the standard dam cross section, the period lengthening, ratio R, and added damping §, depend on several parameters, the more significant of which are: Young’s modulus E, of the dam concrete, ratio H/H, of water depth to dam height, and wave reflection coefficient a. This coefficient a is the ratio of the amplitude of the reflected hydro- dynamic pressure wave to the amplitude of a vertically propagating pressure wave incident on the reservoir bottom (6,8,12,14); a = 1 indi cates that pressure waves are completely reflected, and smaller values of a indicate increasingly absorptive materials. ‘The results of many analyses of the “standard” dam cross section, using the procedures developed in Ref. 12 and modified in Ref. 10 for dams with larger elastic modulus £,, are summarized in Ref. 10. The period lengthening ratio R, and added damping & are presented as a function of H/H, for several values of E, and a; the data for one selected value of E, is shown in Fig. 2. Whereas the dependence of R, and f, on. E,, H/H,, and a, and the underlying mechanics of dam-water interac- tion and reservoir bottom absorption are discussed elsewhere in detail (10-12), it is useful to note that R, increases and £, generally, but not always, increases with increasing water depth, absorptiveness of res- ervoir bottom materials, and concrete modulus. The effects of dam-water interaction reservoir bottom absorption may be neglected, and the dam analyzed as if there is no impounded water, if the reservoir depth is not large, H/H, < 0.5; in particular R, = 1 and & = 0. Modification of Period and Damping: Dam-Foundation Rock Inter- action.—-Dam-foundation rock interaction modifies the natural vibration period (Eq. 4b) and added damping ratio (Eq. 5) of the equivalent SDF system representing the fundamental vibration mode response of the dam. For the ‘standard” dam cross section, the period lengthening ratio R, and the added damping & due to dam-foundation rock interaction depend on several parameters, the more significant of which are: moduli ratio E)/E,, where E, and E,~ the Young’s moduli of the dam conerete ‘and foundation rock, respectively; and the constant hysteretic damping. {actor ry for the foundation rock. The period ratio R, is, however, insen- sitive t0 ny The results of many analyses of the “standard” dam cross-section, us- ing the procedures developed in Ref. 12 are summarized in Fig. 3 and tabulated in Ref. 10, The period lengthening ratio R,and added damping, & are presented for many values of E,/E, between 0.2 and 5.0, and 1, 0.01, 0.10, 0.25, and 0.50. Whereas the dependence of R, and & on E)/ E, and ty, and the underlying mechanics of dam-foundation rock inter- action are discussed elsewhere in detail (11,12), itis useful to note that the period ratio R, is essentially independent of ry, but increases as the ‘moduli ratio E;/E, decreases, which for a fixed value of E, implies an increasingly flexible foundation rock. The added damping &; increases with decreasing E)/E, and increasing hysteretic damping factor 7y. The foundation rock may be considered rigid in the simplified analysis if E)/ E, > 4 because then the effects of dam-foundation rock interaction are negligible. Hydrodynamic Pressure.—In order to provide a convenient means for determining ply, T,) in Eqs. 6 and 7, a nondimensional form of this func tion, gp(g)/toH, where j = y/H, and w = the unit weight of water, has 4. Stuct. Bg, 1987 113:1688-17088/12. Copyright ASCE, For personal use only allright eeeved Downloaded fom ascelirary org by Indian Insite of Techrology, Debi on | FIG, 9.—Standard Values for Period Lengthening Ratio f, and Addod Damping Ratlo & due to Dam-Foundation Rock interaction FIG, 4.—Standard Values for Hydrodynamic Pressure Function p() for Full Res- ervokr, La, H/H, 0.75 and 0.50 been computed from the equations presented in Ref. 12 for a = 1.0, 0.90, 0.75, 0.5, 0.25, and 0, and the necessary range of values of Re 0 in which the fundamental vibration period of the impounded water T; = SH/C, where C = the velocity of pressure waves in water. Results are presented in Ref. 10 for several values of a in the form of charts similar to Fig, 4 and also in tables for the user’s convenience. All these data are for full reservoir, H/H, = 1, The function, gp(@)/ioHt for any other value of H/H, is approximately equal to (H/H,) times the function for H/H, = 1 Q). Generalized Mass and Earthquake Force Coefficient.—The general- ized mass M; (Eq. 74) of the equivalent SDF system representing the ‘dam, including hydrodynamic effects, can be conveniently compte from (2) 5 = (RM = (104) Stet. Bag, 1987115:1688-1708,use only al rights reserved ce by lin as Dowaloded fom sselbary. Re Value of A, fora o ® 099 1.242 0.98 0.393 087 0.739 0.96 0.687 0.95 0.585 094 0.599 0.93 0.503 092 0.474 0.90 0.431 8s 0.364 0.80 0.324 070 0.279 =050 0.237 ‘TABLE 2.—Standard Values for Hydrodynamic Force Coefficient 4, inf, = 0.80, 0.75, 0.50, 0.25, and 0 120 | 0.071 om 0.178 0.30 | oar 0361 0.224 in which M, is given by Eq. 25. In order to provide a convenient means to compute the generalized earthquake force coefficient [ , Eq. 7b is ex- pressed as: 1 ‘HY 2a, (BY apne 069 where Fy = wH?/2 = the total hydrostatic force on the dam and A, the integral of the function 2gp(@)/wH1 over the depth of the impounded water, for H/H, = 1. The hydrodynamic force coelficient A, is tabulated in Tables 1 and 2 for a range of values for the period ratio Ry and the wave reflection coefficient ‘Stac Connection Fon HicheR MoDE RESPONSE Because the earthquake response of short vibration period structures, such as concrete gravity dams, is primarily due to the fundamental mode 4. Stet Eag. 1987 113:1688-1708g oi vibration, the response contibutions of the higher vibration modes have, 0 fat, been neglecied jt the simplified analysis procedure pre- sent in the preceding sections, However, the height wine mass ds tribution of concrete gravity dams is such that the effective mass (2) in the fundamental vibration mode is small, e.g., its 35% of the total mass for the standard dam section mentioned earlier. Thus, the contributions of the higher vibration modes to the earthquake forces may not be neg- ligible, and a simple method to consider them is summarized in this section. ‘This simple method utilizes three concepts. Firstly, because the pe- siods of the higher vibration modes of concrete gravity dams are very short, the higher vibration modes respond to earthquake ground motion with litle dynamic amplification in essentially a static manner, leading, to the “static correction” concept (1,13). Secondly, just as in the case of rultistory buildings (18), soilstructure interaction effects may be ne- glected in computing the contributions of the higher vibration modes to the earthquake response of dams. Thirdly, the effects of dam-water in- teraction and water compressibility may be neglected in computing the higher mode responses (10). The maximum earthquake effects assoc ated with the higher vibration modes can then be represented by the equivalent lateral forces (10) 1or=2{aerf— 2 ae] + [sen -2 worvea]}e a “ al" M, J MO" oo In Eq. 11, a, = the maximum ground acceleration; po(y) = a real-valued, frequency-independent function for hydrodynamic pressure on a rigid dam undergoing unit acceleration, with water compressibility neglected, both assumptions being consistent with the “static correction” concept; and B, provides a measure of the portion of p(y) that acts in the fun- ern aaove RESERVOIR sorToM t ob FIG. §.—Standard Values for Hydrodynamic Pressure Function (7) 1697 4. Stas. Eg, 1987 113:1688-1708,8/12. Copyright ASCE, For personal us nly alight eserved tat of Teckology, Delhi oa Downloaded fom ascelinary ong by damental vibration mode. Standard values for p(y) are shown graphi- cally in Fig. 5 and tabulated in Ref. 10 for the user’s convenience. Using, the fundamental mode vibration properties of the standard dam: Fy (HY 8, = 0.052 (# ven (@) where Fy = the total hydrostatic force on the dam. The shape of only the fundamental vibration mode enters into Eq. 11 and the higher mode shapes are not required, thus simplifying the analysis considerably. (2) Response Cowsinarion Dynamic Response.—As shown in the preceding two sections, the maximum effects of earthquake ground motion in the fundamental vi- bration mode of the dam have been represented by equivalent lateral forces f,(y) and those due to all the higher modes by f(y). Static analysis of the dam for these two sets of forces provide the values r, and r for any response quantity, r, e.g,, the shear force or bending moment at any horizontal section, or the shear or bending stresses at any point. Because the maximum responses r; and f. do not occur at the same time during the earthquake, they should be combined to obtain an estimate of the dynamic response r, according to the well-known modal combi- nation rule: square-root-of-the-sum-of-squares (SRSS) of modal maxima leading to w= Vin + Ge ay Because the natural frequencies of lateral vibration ofa concrete dam are well separated, itis not necessry to inhide the conelaton of toda Tespontes in Ej 1.In Ref 1, the SRSS combination rule i shown fo be preferable 1 the stanoiabscutewalves (ABSUM), which provides an overly conservative res ‘the SRSS and ABSUM combination rules are applicable to the com- putaton of any response quantity tats proportional tothe generalized Todal coordinate responses Thus, these combination roles ate gener aly inappropriate to determine the Pinal stresses, However a5 shown in Rel 1 the principal stesss atthe aes ofa dam monolith may be determined by the SROS method i te upateam fave fe nearly Vertical Sn the fect of tall water at the downetream face are sal "otal Response. In onde to obtain the total value of any response uantty 7, the SESS estimate of dynamic response "y should be come Binet with the sae effects ty. The latter may be determined by san: dard analysis procedures fo Compute the ina stresses nw damn por to the earthquake, incaing effect ofthe selfeelght of te dam, hy- Grostate pressures, and temperature changes. In ore to ecogne that the dtecon of lateral earthquake forces revesble,combitations of State and dynamic stresses should allow for te werst case, leading fo the maxim value of toll responce: 2 Ve +O (4) 1608 4. Stet. Bg, 1987 113:1688-1708ny, Delhi on 11/28/12. Copyright ASCE, For personal use ony; al rights reser z ‘This combination of static and dynamic responses is appropriate ifr, Fi, and rg ate oriented similarly. Such is the case for the shearing force ‘or bending moment at any horizontal section, for the shear and bending stresses at any point, but generally not lor principal stresses except un der the restricted conditions previously mentioned. Simpuriep AnaLysis PROCEDURE ‘The maximum effects of an earthquake on a concrete gravity dam are represented by equivalent lateral forces in the simplified analysis pro- cedure. The lateral forces associated with the fundamental vibration mode are computed to include the effects of dam-water interaction, water com- pressibility, reservoir bottom absorption, and dam-foundation rock in- teraction. The response contributions of the higher vibration modes are computed under the assumption that the dynamic amplification of the modes is negligible, the interaction effects between the dam, impounded water, and foundation rock are not significant, and that the effects of water compressibility can be neglected. These approximations provide a practical method for including the most important factors that affect the earthquake response of concrete gravity dams, Selection of System Paremeters.—The simplified analysis procedure requires only a few parameters to describe the dam-water foundation rock system: E,, &, H,, Ej, 1, H, and a, The complete data necessary to implement this proceduse are presented as both figures and tables in Ref. 10, available from the University of California at Berkeley; for brov- ity, only a small portion of the data is shown here in Figs. 2-5, ‘The Young's modulus of elasticity E, for the dam concrete should be based on the design strength of the concrete or suitable test data, if avail- able. The value of E, may be modified to recognize the strain rates rep- resentative of those the concrete may experience during earthquake mo- tions of the dam @). In using the figures and tables mentioned earlier to conservatively include dam-water interaction effects in the compu- tation of earthquake forces (Eq. 6), the E, value should be rounded down to the nearest value for which data are available in Ref. 10: F, = 1,000,000; 2,000,000; 2,500,000; 3,000,000; 3,500,000; 4,000,000; 4,500,000; or 8,000,000 psi. Forced vibration tests on dams indicate that the viscous damping ratio & for concrete dams is in the range of 1-3%. However, for the large motions and high stresses expected in a dam during intense earth- quakes, § = 5% is recommended, The height H, of the dam is measured from the base to the crest. ‘The Young's modulus of elasticity E, and constant hysteretic damping. coetficient 1; of the foundation rock should be determined from a site investigation and appropriate tests. To be conservative, the value of 1 should be rounded down to the nearest value for which data are avail- able: 1 = 0.01, 0.10, 0.25, or 0.50, and the value of E;/E, should be rounded up to the nearest value for which data are available. In the absence of information on damping properties of the foundation rock, a value of nj = 0.10 is recommended. ‘The depth H of the impounded water is measured from the free sur face to the reservoir bottom. It is not necessary for the reservoir bottom and dam base to be at the same elevation. The standard values for unit 1699 4. Struct. Eng. 1987 113:1688-170818/12, Copyright ASCE, For personal se only ll rights reserved Downloaded fom acliray org by Indian Insite of Techrology, Debi on | weight of water and velocity of pressure waves in water are w = 62.4 pet and C = 4,720 fps, respectively. It may be impractical to reliably determine the wave reflection coef- ficient « because the reservoir bottom material may consist of highly variable layers of exposed bedrock, alluvium, silt, and other sediments, and appropriate site investigation techniques have nat been developed, However, to be conservative, the estimated value of « should be rounded. up to the nearest value for which the figures and tables are presented w= 10, 0.90, 0.75, 0.50, 0.25, and 0.00, For proposed new dams or recent dams where sediment deposits are meager, a = 0.90 or 1.0 is recommended and, lacking data, a = 0.75 or 0.90 is recommended for ‘older dams where sediment deposits are substantial. In each case, the larger « value will generally give conservative results, which is appro- priate at the preliminary design stage. Design Earthquake Spectrum.—The horizontal earthquake ground motion is specified by a pscudo-acceleration response spectrum in the simplified analysis procedure. This should be a smooth response spec- trum—without the irregularities inherent in response spectra of individ- ual ground motions—representative of the intensity and frequency char- acteristics of the design earthquakes which should be established after 2 thorough seismological and geological investigation; see Ref. 3 for more deta Conpuraionan SrePs ‘The computation of earthquake response of the dam is organized in three parts Part I.—The earthquake forces and stresses due to the fundamental vibration mode can be determined approximately for purposes of pre- liminary design by the following computational steps: 41. Compute 7), the fundamental vibration period of the dam, in sec, fon rigid foundation rock with an empty reservoir from Eq, 8 in which H, = height of the dam in ft, and E, = design value for Young’s modulus of elasticity of concrete, in psi 2. Compute 7,, the fundamental vibration period of the dam, in sec, including the influence of impounded water from Eq. 4a in which T, was computed in step 1; R, = period ratio determined from Table 2 of Ref 10, which contains a complete set of data such as Fig, 2 presented here, for the design values of E,, the wave reflection coefficient a, and the depth ratio H1/H1,, where Hi = the depth of the impounded water, in f. If H/H4, < 0.5, computation of R, may be avoided by using R, = 1. 3. Compute the period ratio Ry from Eq, 9 in which 7, was computed in step 2, and T{ = 4H/C where C = 4,720 ft/sec 4. Compute fy, the fundamental vibration period of the dam in sec, including the influence of foundation flexibility and of impounded water, from Eq, 4c in which R, was determined in step 2; Ry = period ratio determined from Fig. 3 presented here, or Table 3 of Ref. 10, for the design valte of £,/E,. If E/E, > 4, use Ry ~ 1. 5. Compute the damping ratio & of the dam from Eq, 5 using the period ratios R, and R, determined in steps 2 and 4, respectively; &1 = 1700 4. Stet. Bg, 1987113:1688-1708rights reserved 12. Copyright ASCE. For persona se on Downloaded fom asceliray.org by Indian Insite of Technology, Debio viscous damping ratio for the dam on rigid foundation rock with empty reservoir; & = added damping ratio due to dam-water interaction and reservoir bottom absorption, obtained from Table 2 of Ref. 10, which contains a complete set of data such as Fig. 2 presented here, for the selected values of E,, «, and H/H,; and & = added damping ratio due to dam-foundation rock interaction, obtained from Fig. 3 presented here, or Table 3 of Ref, 10, for the design values of E)/E, and ny. If H/H. < 0.5, use & = 0: if E,/E, > 4, use & = 0; and if the computed value of &<,usease. | 6. Determine spy, 7.) from Table 4 of Ref. 10, which contains a com- plete set of data such as Fig, 4 presented here, corresponding to the Value of Ry computed in step 3—{rounded to one of the two nearest available values, the one giving the larger p(y), the design value of a, and for H/H, = 1; the result is multiplied by (H/H,)*. It H/H, < 0.5, computation of p(y, ,) may be avoided by using ply. I.) = 0. 7. Compute the generalized mass M1, from Eq. Ie: in which R, was computed in step 2, and M; is computed from Eq. 2, in which w,(y) = the weight of the dam /unit height; the fundamental vibration mode shape 4() is given in Fig. 1 presented here, or Table 1 of Ref, 10; and g = 32.2 ft/sec. Evaluation of Eq. 2b may be avoided by obtaining an appro imate value from M; = 0.043 W,/g, where W, = the total weight of the dam monolith, 8. Compute the effective earthquake force coefficient from Eq. 10) in which Ly is computed from Eq. 3b; Fy = wHF/2; and A, is given in Table 1 for the values of R,, and « used in step 6. If H/H, < 0.5, com- ulation of Ly may be avoided by using [; = Ly. Evaluation of Eq. 30 may be avoided by obtaining an approximate value from I = 0.13 WW,/ ¢. (It may be noted that computation of steps 7 and 8 may be avoided by using conservative values: L;/M; = 4 for dams with impounded water, and 1 /M = 3 for dams with empty reservoirs.) 9. Compute fy), the equivalent lateral earthquake forces associated with the fundamental vibration mode from Eq. 6 in which 5,(T,£:) the pscudo-acceleration ordinate of the earthquake design spectrum in ft/sec* at period T; determined in step 4 and damping ratio & deter- ‘mined in step 5; u(y) = weight/unit height or the dam; $(y) = fun- damental vibration mode shape of the dam from Fig. 1(a) presented here, ‘or Table 1 of Ref. 10; M, and L; = generalized mass and earthquake force coefficient determined in steps 7 and 8, respectively; and the hydrody- namic pressue term gp(y,T,) was determined in step 6 10. Determine by static analysis of the dam subjected to equivalent lateral forces f(y), from step 9, applied to the upstream face of the dam, all the response quantities of interest, in particular the stresses through out the dam. Traditional procedures for design calculations may be used wherein the direct and bending stresses across a horizontal section are computed by elementary formulas for stresses in beams. Alternatively, the finite element method may be used for a more accurate static stress analysis Part II.—The earthquake forces and stresses due to the higher vibra- tion modes can be determined approximately for purposes of prelimi nary design by the following computational steps: 1701 4 Staci, 1987113:1688-1708of Technology, Deli on 1/28/12. Copyright ASCE, For persnal use only allright reserved Downloaded Gon asclinery org by India In 11. Compute fy), the lateral forces associated with the higher vibra- tion modes from Eq, 11 in which M, and 1 were determined in steps 7 and 8, respectively; gp.(y) is determined from Fig. 5 presented here, or Table 6 of Ref. 10; 8; is computed from Eq. 12; and ay is the maximum ground acceleration, in ft/sec", of the design earthquake. If H/H, < 0.5, computation of pty) may be avoided by using py(y) = 0 and thus By ~0. 12, Determine by static analysis of the dam subjected to the equivalent lateral forces f.(y), from step 11, applied to the upstream face of the dam, all the response quantities of interest, in particular the stresses throughout the dam. The stress analysis may be carried out by the pro- cedures mentioned in step 10. Part III. —The total earthquake forces and stresses in the dam are de- termined by the following computational step: 13. Compute the total value of any response quantity by Eq. 14, in which r, and 1, are values of the response quantity determined in steps 10 and 11 associated with the fundamental and higher vibration modes, respectively, and is its initial value prior to the earthquake due to various loads, including the sell-weight of the dam, hydrostatic pres- sure, and thermal effects, Use of Metric Units.— Because the standard values for most quantities required in the simplified analysis procedure are presented in nondi- ‘mensional form, implementation of the procedure in metric units is straightforward, The few expressions and data requiring conversion to ‘metric units are noted in Ref. 10. EVALUATION OF SMPLIFIED ANALYSIS PROCEDURE As mentioned earlier (11,12) and in the preceding sections, various approximations were introduced to develop the simplified analysis pro- cedure, and these were individually checked to ensure that they would lead to acceptable results. In order to provide an overall evaluation of the simplified analysis procedure, it is used to determine the earth- quake-induced stresses in Pine Flat Dam, and the resalts are compared with those obtained from a refined response history analysis, rigorously including dam-water-foundation rock interaction and reservoir bottom absorption effects, in which the dam is idealized as a finite element sys- tem. System and Ground Motion.—The system properties for the simpli- fied analysis are taken to be the same as those assumed for the complete response history analysis; the tallest, nonoverflow monolith of the dam is shown in Ref. 3; height of the dam, H, = 400 ft, modulus of elasticity of concrete, E, = 3.25 x 10* psi; unit weight of concrete = 155 pef; damp- ing salio 6, ="5%; modulus of elasticity of foundation rock E; = 3.25 10° psi; constant hysteretic damping coefficient of foundation rock, ty = 0.10; depth of water, HY = 381 ft; and, at the reservoir bottom, the wave reflection coefficient « = 0.5. ‘The ground motion for which the dam is analyzed is the S69E com: ponent of the ground motion recorded at the Taft Lincoln School Tannel 702 4. Stuct. Bg, 1987 113:1688-1708rights reserved. i z a ownlotded from ascelinary.ngby Indian Insite of Technology, Debi on | ‘TABLE 3.—Pine Flat Dam Analysis Gases, Simplified Procedure Parameters, an Fundamental Mode Properties from Simplified and Retined Analysis Procedures ——T TUNOMVENTAL HODE PROPERTIES pias Prone 00 | arg fat im Gratis] Aelaed [Sting] Retest) coma) smet |i] | | & | |ecenaee| ace] proce) poset) ng wt @ [ar |e fo | elo en [on | | st mesa fio fo fo | eau | osy | om | oan [ va 2 fre Jeol” izo}ie Joanlo | a7 | cae | nor | aoe | Sat "Scr csponding oT ano slond pocste during the Kem County, California, earthquake of July 21, 1952. The response spectrum for this ground motion is shown in Ref. 10. Such an irregular spectrum of an individual ground motion is inappropriate in conjunction with the simplified procedure, wherein a smooth design ‘spectrum is recommended, but is used here to provide direct compari- son with the results obtained from the refined analysis procedure. ‘Computation of Earthquake Forces.—The dam is analyzed by the simplified analysis procedure for the four cases listed in Table 3. Imple- mentation of the step-by-step analysis procedure described in the pre- ceding section is summarized next with additional details available in Ref. 10; all computations are performed for a unit width of the monolith: 1. For E, = 8.25 x 10% psi and H, = 400 ft, from Eq. 8, Ty = (1.4400) /V5.25 10 = 0.311 see 2. For E, = 30 x 10° psi (rounded down from 3.25 x 10° psi), a = 0.50 and Hi/H, = 381/400 = 0.95, Table 2 of Ref. 10 gives R = 1.213, so 1, = (1.213)0.311) = 0.377 sec. 3. From Eq, 9, Ti = (4)(381)/4,720 = 0.923 see and Ry = = 0.86, 4. For E/E, = 1, Table 3 of Ref. 10 gives R, = 1.187, so T= (1.187)0.311) = 0.369 see for Case 3, and Ty = (1.187)(0.377) ~ 0.448 sec for Case 4. 5, For Cases 2 and 4, &, = 0.080 from Table 2 of Ref, 10 for E, = 3.0 x 10° psi (rounded down from 3.25 x 10° psi), « = 0.50, and H/H, = 0.95, For Cases 3 and 4, & = 0.058 from Table 3 of Ref. 10 for Fy and 4y = 0.10, With 6, = 0.05, Eq. 5 gives: & = (0.05)/(1.213) +.0.030 [O71 for Case 2; & = (0.05)/(1.187)" + 0.068 = 0.098 for Case 3; and & = (0.08)/((1.213)(1-187))) + 0.030 + 0.068 = 0.123 for Case 4. 6. The values of gp(y) are obtained at 11 equally spaced levels from Table 4 of Ref. 10 for Rx = 0.90 (by rounding Ry ~ 0.86 from step 3) and a = 0,50, and multiplied by (0,0624)(381)(0.95) = 21.5 kips/t. 7. Evaluating Eq, 2b in discrete form gives My = (1/g)(500 kips). From Bq. 102, M; = (1.2137(0/9)(800) = (736 kips)/g. 8, Eq, 3b in discrete form gives L, = (1,390 kips)/g. From Table 2 A, = 0.274 for R. = 0.90 and « = 0.50. Eq, 10) then gives L, = 1,390/g + {(0.0524)(381)*/2g}0.95)*(0.274) = @,510 kips)/g.-Consequentiy, for Cases 323/0.377 1708 4. Struct. Eng. 1987 113:1688-1708rights reserved 12. Copyright ASCE. For persona se on Insite of Technology, Deion 1 Downloaded fom ascelibray org by nd 1 and 3, Li/M, = 1,390/500 = 2.78, and for Cases 2 and 4, f/My 2,510/736 ~ 3.41 9. For each of the four cases, Eq. 6 was evaluated to obtain the equiv- alent lateral forces f(y) at 11 equally spaced intervals along the height ff the dam, including the top and bottom, by substituting values for the quantities computed in the preceding steps, computing the weight of the dam/unit height %,(y) from the monolith dimensions and the unit weight of concrete, by substituting 6{y) from Fig, 1(a) or Table 1 of Ref. 10, and the §,(71,£,) from the response spectrum corresponding to the Ty and & obtained in steps 4 and 5 (Table 3). 10. The static stress analysis of the dam subjected to the equivalent lateral forces fi(y), from step 9, applied to the upstream face of the dam is described in the next subsection, leading to response value ry at a particular location in the dam. 11. For each of the four cases, Eq. 11 was evaluated to obtain the equivalent lateral forces f(y) at 11 equally spaced intervals along the height of the dam, including the top and bottom, by substituting nu- ‘merical values for the quantities computed in the preceding, steps; ob- {taining ga(y) from Table 6 of Ref. 10; using Eq. 12 to compute B, = O1c52(.0834) 381 2g10.95)' = 12.3 Kips), leading to By/M = 212.5/ 500 = 0.425; and substituting a, = 0.18 12, The static stress analysis of the dam subjected to the equivalent lateral forces .(y), from step 11, applied to the upstream face of the dam is described in the next subsection, leading to response value r. at a particular location in the dam, 13. Compute the maximum total value of any response quantity by combining 7 from step 10, r< from step 12, and ry, the intial value prior to the earthquake, according to Eq. 14; this is described further in the next subsection. Computation of Stresses.—The equivalent lateral earthquake forces {i(@) and f(y) representing the maximum effects of the fundamental and higher vibration modes, respectively, were computed in steps 9 and 11 Dividing the dam into ten Blacks of equal height, each of these sets of distributed forces is replaced by statically equivalent concentrated forces at the centroids of the blocks. Considering the dam monolith to be a cantilever beam, the bending stresses are computed at the bottom of each block using elementary formulas for stresses in beams. The normal bending stresses are then transformed to principal stresses (Ref. 4). In this simple stress analysis, the foundation rock is implicitly assumed to be rigid. ‘The principal stresses a, and a at any location in the dam due to the forces f(y) and f.(y), respectively, may be combined using the SRSS combination rule, Eq, 13 (10). Based on this rule, the combined values 0% along with the fundamental mode values ay for the maximum prin- cipal stresses are presented in Table 4 forthe four analysis cases. These stresses occur at the upstream face when the earthquake forces actin the downstream direction, and at the downstream face when the earth- ‘quake forces actin the upstream direction, The maximum principal stresses at both faces of the monolith are shown in Figs. 6 and 7. Comparison with Refined Analysis Procedure. The Pine Flat dam 1704 4. Suuct ng. 1987113:1688-1708rights reserved. 12, Copyright ASCE. For persona wse on Downloaded fom asclirayrg by Indian Institute of Techuology, Debio | TABLE 4.—HHaximum Principal Stresses (in psi) in Pine Flat Dam due to S69E Component of Taft Ground Motion (intial Static Stresses Are Excluded) —_ tear Fon owen Faw coe | “Sa | via | “irony | srss | case | "Moaraur | ones | cat fs 1 [mea Emmy | af ae | aya ae ae aja fem] Gs | ae | | oa | as | a [ewe [er | ee | oa || a | “rom ed pede, Ra ‘monolith was analyzed by the computer program EAGD-84 (5) in which the response history of the dam, idealized as a finite element system, due to the Taft ground motion is computed considering rigorously the effects of dam-water-foundation rock interaction and of reservoir bottom, absorption, The results from Ref. 9 at some intermediate steps of the analysis, in particular the resonant period anal damping ratio of the fun- damental resonant response, are presented in Table 3, and the envelope values of the earthquake-induced maximum principal stresses are pre- sented in Table 4 and in Figs. 6 and 7. It is apparent from Table 3 that the simplified procedure leads to ex- cellent estimates of the resonant vibration period and the damping ratio for the fundamental mode. Because the response of concrete gravity dams is dominated by the fundamental mode, this comparison provides a con- FIG. 6.—Maximum Principal Stresses FIG, 7.—Maximum Principal Stresses (in psi in Pine Flat Dam on Rigid Foun- (in pai) in Pine Flat Oam on Flexible dation Rock due to S89E Component of Foundation Rock due to S69E Compo- alt Ground Motion (Initial Static nent of Taft Ground Motion (Initial Static ‘Stresses Aro Excluded): (a) Case 1— Strosses Are Excluded): (a) Case 3— Empty Resorvoir; (4) Case 2—Full Res- Empty Reservolr; (b) Case 4—Full Res- corvelr vol 1705 4. Struct. Eng, 1987 113:1688.17084 ite of Technology, Deh on 11/2812. Copyright ASCE, For personal we on Downloaded fom sscelbary.org by Indian I firmation that the simplified analysis procedure is able to represent the important effects of dam-water interaction, reservoir bottom absorption, and dam-foundation rock interaction. ‘As shown in Table 4, the stresses obtained by the simplified procedure considering only the fundamental vibration mode response are about the same as those obtained by including higher mode responses with the SRSS combination rule, For the system parameters and excitation considered in this example, the higher mode responses add only 2-10 psi to the stresses, indicating that these,can be neglected, in this case and many other cases. However, in some cases depending on the vi- bration periods and mode shapes of the dam and on the shape of the earthquake response spectrum, the higher mode responses may be more significant and should be included, ‘Considering only the fundamental mode response or obtaining the SRSS ‘combination of fundamental and higher mode responses, the simplified procedure provides estimates of the maximum stress on the upstream face that are sufficiently accurate—in comparison with the “exact” re- sults—to be useful in the preliminary design phase. The accuracy of these stresses depends, in part, on how well the resonant vibration period and damping ratio for the fundamental mode are estimated in the simplified procedure; e.g., in Case 4, the maximum stress at the upstream face is, overestimated by the simplified procedure primarily because it under- estimates the damping ratio by about 2% (Table 3). While the simplified procedure provides excellent estimates of the maximum stress on the Upstream face, at the same time it overestimates by 25-50% the maxi mum stress on the downstream face. This large discrepancy is primarily due to the limitations of elementary beam theory in predicting stresses near sloped faces. Similarly, the beam theory is ineapable of reproducing. the stress concentration in the heel area of dams predicted by the refined. analysis (Figs. 6 and 7), and the stresses in that area are therefore under- estimated, ‘The simplified procedure gives conservative, but reasonable, maxi- ‘mum stresses, with the exception of the heel area of dams with full res- ervoir, The quality of the approximation is satisfactory for the prelimi- nary phase in the design of new dams and in the safety evaluation of existing dams, considering the complicated effects of dam-water-foun- dation rock interactions and reservoir bottom absorption, the number of approximations necessary to develop the simplified analysis procedure, and noting that the results generally err on the conservative side. Conctusions A procedure was presented in 1978 for earthquake resistant design of new conerete gravity dams and for the seismic safety evaluation of ex- isting dams (3). The procedure was based on two performance require- ments: firstly, dams should remain essentially within the elastic range ‘of behavior for the most intense geound shaking expected to occur dur ing the useful life of the structure; and secondly, some cracking, which is limited enough that it does not impair the ability of the dam to contain the impounded water and is economically repairable, may be permitted 706 4. Struct. Eng. 1987 113:1688-1708righ reserved tate of Technology, Delhi on 11/28/12. Copyright ASCE, For persona se only; i if the most intense ground shaking that the seismic environment is ca- pable of producing were to occur. ‘A two-stage procedure was proposed for the analysis phase of elastic design and safety evaluation of dams: (1) A simplified analysis proce: dure in which the response due only to the fundamental vibration mode is estimated directly from the earthquake design spectrum; and (2) a re fined response history analysis procedure for finite element idealizations of the dam monolith. The former was recommended for the preliminary phase of design and safety evaluation of dams and the latter for accu- rately computing the dynamic response and checking the adequacy of the preliminary evaluation (3). ‘At the time (1978) the dam design procedure was presented, both of, these analysis procedures included the effects of dam-water interaction and compressibility of water, but assumed rigid, non-absorptive reser- voir bottom materials and neglected the effects of dam-foundation rock interaction, Recently (1984), the refined response history analysis pro- cedure and computer program have been extended to also include the absorptive effects of reservoir bottom materials and dam-foundation rock interaction effects (5,6). In this paper, the simplified analysis procedure has been extended to include these important effects so that they can now also be considered in the preliminary phase of design and safety evaluation of dams. Also included now in the simplified procedure is a “static correction”” method to consider the response contributions of the higher vibration modes, Thus the design procedure proposed in 1978 @), which is conceptually still valid, should utilize the new refined and simplified analysis procedures. ACKNOWLEDGMENTS ‘This research investigation was supported by Grants CEE-8120308 and (CEE-8401139 from the National Science Foundation to the University of California, Berkeley, for which the writers are grateful. This work was started while Gregory Fenves was a graduate student at Berkeley but ‘was completed subsequent to the submission of his doctoral dissertation in May 1984. Both writers are grateful to Hanchen Tan, a graduate stu- dent at the University of California at Berkeley, for computing the stan- dard data described in this paper. Appenou.—Rerenences 1, Anagnostopoulos, &, A., “Wave and Earthquake Response of Offshore Struc tures: Evaluation of Modal Solutions,” jourtal of te Structural Division, ASCE, Vol. 108, No. 10, Oct., 1982, pp. 2175-2191 2. Chopra, A. K., Dynamics of Stnictures, A Priner, Earthquake Engineering Re: search Institute, Berkeley, Calif, 1981 3. Chopra, A. K,, “Earthquake Resistant Design of Concrete Gravity Dams,”" Journal ofthe Siractural Division, ASCE, Vol. 104, No. 6, Jun., 1978, pp. 9 or. 44. Design of Gravity Dams, U.S. Bureau of Reclamation, U.S. Government Print- ing Office, Denver, Colo., 1976. 5, Fenves, G., and Chopra, A. K,, “EAGD-St: A Computer Program for Earth- quake Analysis of Concrete Gravity Dams,” Report No, UCB/EERC-84/11, Earthquake Engineering Research Center, University of California, Berkeley, Calif 1984, 1707 4. Stet. Eng. 1987 113:1688-1708tof Technology, Deli on 1/28/12. Copyright ASCE, For personal use nly alight served ded from ascelibray org by Indian Insti 10. 2. 1B 4. Fenves, G., and Chopra, A. K., “Earthquake Analysis of Concrete Gravity Dame inckading Reservoir Botiom Absorption and Dam-Water-Foundation Rock Interaction,” Enrthuake Enhnering an Structural Dy, Vol. 12, No. BSepeo, Let, presse Fenves, G. and Chopra, A, K, “Effects of Reservoir Bottom Absorption and am-Water-Foundation Rock interaction on Frequency Response Functions for Concrete Gravity Dame,” Enrthquate Engineering und Strctural Dynamics, Vol. 13, No. 1 jan Feb. 1985, pp. 13-31 Fenves, G. and Chopra, A. K., “Eifece of Reservoir Bottom Absorption on Earthquake Response of Concrete Gravity Dams,” Earthquake Engincering and Strctiral Dyan, Vol. 11, No. 6, Nov-Dec. 1983, pp. 809-823, Fenves, G., and Chopra, A. K., "Reservoir Bottom Absorption Effects in Earthauake Response’ of Concrete Gravity Dame" Journal of Structural Em {incering, ASCE, Vol. 111, No. 3, Mar. 1965, pp. 545-562 Fenves, G., and Chopra, AK, “Simplified Anaiysis for Earthquake Resist fant Design of Concrete Gravity Dame,” Report Nov UICB/EERC 5/10, Bath. quake Engineering Research Center, University of California, Berkeley, Cal Bee. Fenves, G., and Chopra, A. K., “Simplified Earthquake Analysis of Concrete Gravity Dim Combined #jdrodyfamic end Foundation incraion EF fects,” Joural of Engineering Mechrics, ASCE, Vol. 111, pp. 730-756. Fenves, G., and Chopra, A. K., “Simplified Earthquake Analysis of Concrete Gravity Dams: Separate Hydrodynamic and Foundation interaction Elfects,” Journal of Bgnring Mechs, ASCE, Vo. 11, NO 6, Ju 1985, pp 715- 7 Hansteen, ©. B., and Bell, K,, “On the Accuracy of Mode Superposition Analysis in Stctoral Dynamics," Earthquake Engineering end Structural Dy tums, Val. 7, No.5, Sep-Oct, 1979, pp. 405-411 Rosenblueth, "Presion Hidrodindmica en Presa Debi a la Aceleracisn Vertical con Refraccon en el Fonda," Pracedinge, 2nd Congreso Nacionl de IngentertSisnten, Vol. , Sociedad Mexicana de Ingenieria snd Universidad Veracruzana, Veracruz, Mexico, 1968 (in Spanish) Veletos, AS., "Dynamics of Structure Foundation Systems,” Structural and Geotechnical Mécharcs, W. . Hall, Ea, Prentice Hal, Chilton, NJ. 1977. 0. 6, Jun., 1985, 1708 4. Suet Eng. 1987113:1688-1708