EARTHQUAKE SAFETY EVALUATION OF ATATURK DAM

More documents

Recommendations

Info

given. For this purpose, the sliding movement was assumed to occur in the direction of the tangent at the highest point of the sliding surface. Table 3: Permanent displacements of critical sliding block on upstream slope (Notation: ayield: yield acceleration; dh: horizontal displacement; dv: vertical displacement) Peak angles of friction Residual angles of friction Considering only Considering only Considering horizontal and horizontal acceleration horizontal acceleration vertical accelerations ayield (g) Max./ Mean/ Min. dh (m) dv (m) ayield (g) Max./ Mean/ Min. (a) With average dynamic material properties dh (m) dv (m) ayield (g) Max./ Mean/ Min. Max. 0.41 0.26 Max. 1.03 0.64 Max. 2.34 1.46 0.239 Mean 0.32 0.20 0.185 Mean 0.91 0.57 0.139 Mean 2.03 1.27 Min. 0.19 0.12 Min. 0.79 0.50 Min. 1.66 1.04 (b) Considering possible upper and lower bounds of dynamic material properties Max. 1.28 0.80 Max. 2.28 1.42 Max. 3.72 2.32 0.239 Mean 0.47 0.29 0.185 Mean 1.07 0.67 0.139 Mean 2.17 1.36 Min. 0.03 0.02 Min. 0.18 0.11 Min. 0.98 0.61 dh (m) CONSEQUENCES <strong>OF</strong> <strong>EARTHQUAKE</strong>-INDUCED SLIDING DISPLACEMENTS The most important consequences of the sliding movements induced by the earthquake loading are discussed below. Reduction of available freeboard During a major earthquake, the crest region can be expected to undergo sliding movements towards both the upstream and downstream sides along a number of criss-crossing failure surfaces. The sliding movements are likely to be of intermittent nature. Consequently, the displacements towards the upstream and downstream sides would occur at different times and, thus, can be assumed to have a cumulative effect on the vertical displacement of the crest region. Hence, the total drop of the level of the core top was assumed to be equal to the sum of the vertical displacements occurring during the sliding displacements towards the upstream and downstream sides. Based on the results for the average properties, the total vertical drop of the core top would be about 1.0 m during the MCE due to the downward vertical displacements of 0.8 m and 0.2 m produced by sliding towards the upstream and downstream sides, respectively. However, if the dynamic shear moduli approach the upper bound values (case E), the total lowering of the level of the core top could become as high as 2.3 m as a result of vertical displacements of about 1.5 m and 0.8 m during the sliding movements caused by the MCE towards the upstream and downstream sides, respectively. Besides the permanent sliding displacements resulting from the dynamic slope instabilities, the earthquake ground shaking also produces a general settlement due to the vibration-induced densification of the embankment materials. The maximum seismic settlement of the top of the core due to this effect was estimated to be of the order of 0.5 m. Considering the effects of both the movements of potential sliding masses and the general seismic settlement, it was estimated that the lowering of the elevation of the core top could approach the order of 3 m during the MCE. Therefore, the freeboard with respect to the core top must be sufficient at all times to accommodate such a reduction without endangering the safety of the dam. dv (m)
Danger of leakage through core Relatively wide filter (transition) zones have been provided on both sides of the core. Except for the reinstated crest region, the upstream filter has an almost uniform width of about 9 m (measured horizontally), and the downstream filter has a width varying from about 8 m at the top to more than 15 m at the base. The filter widths are gradually decreased in the reinstated crest region. As a result, the widths of the upstream and downstream filters reduce to about 6 m and 8 m, respectively, at elevation 545 m a.s.l. The wide upstream and downstream filters form a very effective first line of defence against concentrated earthquake-induced leakage through the dam. Moreover, the core is made of highly plastic clay (mainly CH), a good core material in which it is unlikely that a great deal of erosion would occur even in the worst conditions of cracking and leakage. After the sliding movement caused by the MCE, the filter width at the location of a sliding surface would still be at least about 6 m. This should be sufficient to ensure that uncontrolled leakage does not develop through the dam in the critical period following an earthquake. SUMMARY AND CONCLUSIONS The main points and the conclusions of this study are as follows: 1. The present study deals with the effects of the earthquake ground shaking only. 2. The dynamic analysis and the slope stability calculations were performed using 2-dimensional (2-D) models representing the central cross-section of the dam; 3-D effects were not considered. 3. As dynamic test results were not available, the dynamic material characteristics for the earthquake analysis were selected on the basis of the geotechnical literature. In view of uncertainties concerning the material properties and the random nature of the ground motion, the possible scatter of the results was investigated by analyzing a number of cases. 4. On the basis of the currently available information about the seismic hazard at the dam site, the peak ground accelerations (PGA’s) of the horizontal and vertical components of the maximum credible earthquake (MCE) were taken as 0.50 g and 0.33 g, respectively. The input motion in the dynamic analysis consisted of synthetic accelerograms compatible with the elastic response spectrum of the Turkish earthquake code. 5. The horizontal crest accelerations calculated for the upper and lower bounds of the damping ratios deviate from those calculated for the average damping ratios by about 15%. For obvious reasons, the earthquake response becomes smaller when the damping ratios increase. 6. The peak value of the horizontal component of the absolute crest acceleration calculated in the various cases under the MCE ground excitation is in the range of 0.42 g to 0.93 g and that of the vertical component in the range of 0.26 g to 0.49 g. 7. From the viewpoint of the earthquake safety of the dam, sliding displacements in the crest region are of main interest. A large number of potential sliding surfaces involving the crest area were examined in this study. 8. As the excess pore pressures generated by the earthquake shaking in the upstream shell and the clay core are likely to be quite small, they have not been considered in this study. 9. The sliding blocks on the upstream slope have significantly lower yield accelerations than those on the downstream slope, whereby it is assumed that the angle of friction of the submerged upstream shell is equal to that of the downstream shell. 10. The yield accelerations of the upstream sliding blocks are in the range of 0.20 g to 0.26 g for the peak friction angles and in the range of 0.16 g to 0.20 g for the residual friction angles, when only the horizontal earthquake accelerations are considered. If the vertical accelerations are also conservatively applied in the unfavourable direction, the horizontal yield accelerations of the upstream sliding blocks could become as low as 0.12 g to 0.16 g. 11. Based on the results obtained using the average dynamic material properties, the total vertical drop of the core top during the MCE is expected to be about 1.0 m due to downward displacements of 0.8 m and 0.2 m produced by sliding towards the upstream and downstream sides, respectively. In
Page 1 and 2: 1 st National Symposium and Exposit
Page 3 and 4: DESIGN EARTHQUAKE GROUND MOTION The
Page 5 and 6: Fig. 1: Dependence of dynamic shear
Page 7: Table 2: Material properties for sl

EARTHQUAKE SAFETY EVALUATION OF ATATURK DAM

Create successful ePaper yourself

Delete template?

Save as template?