Seismic Assessment of Concrete Gravity Dams in Terms of Energy Using Capacity Estimation for Near-Fault Ground Motions: Case Study of Pine Flat Dam

Seismic assessment of concrete gravity dams is mainly conducted using capacity estimation of limit states or determination of damage indexes. The main purpose of this study is to estimate the capacity of limit states and damage levels of concrete gravity dams based on energy. Therefore, by selecting Pine Flat gravity dam as a case study, incremental dynamic analysis has been performed using hunt & fill algorithm on the dam-reservoir-rigid foundation system, under near-fault records with forward directivity effects. In this research, Arias intensity, which represents accelerogram cumulative energy has been considered as intensity measure and three parameters of dissipated energy due to fracture (DFE), dissipated energy due to fracture in crest motion into down-stream direction (DFE D/S) and dissipated energy due to fracture in crest motion into up-stream direction (DFE U/S) have been selected as damage measures. In the field of scaling records, previous studies that have used the IDA method to analyse concrete gravity dams have used the stepping algorithm. This algorithm has two main weaknesses. One is inefficiency, because its quality depends very much on the choice of IM step. Another is the implicit coupling of demand and capacity estimation, as the demand and the capacity resolutions are effectively the same in this method and equal to the step size. In the present study, the advanced hunt & fill algorithm was used to scale the records to eliminate the above weaknesses. In this research, using IDA curves and performing a series of additional statistical analyses, limit states, damage levels and global dynamic capacity of the structure were determined based on energy-based parameters. By investigation of the incremental dynamic analysis curves related to DFE D/S, DFE U/S and DFE, the following results were obtained:The study of all IDA curves showed that DFE-IA curves generally consist of two branches. The first branch consists of two parts. The initial part is linear (with a constant slope) that is observed at the beginning of all curves and represents the elastic part of the diagram. The second part of this branch is segment of the inelastic region of the curve. In this part, there are slight changes in the slope of the curve. In general, in the first branch, the structure exhibits hardness behavior. The second branch is the continuation of the inelastic region of the curve. In general, in this branch the slope of the curve is significantly reduced. Relatively severe softening is observed in this branch.In all records up to the intensity level that DFE in U/S direction is equal to zero, the dominant energy response of the dam is the response due to motions into D/S direction and base level cracking. While after the mentioned intensity level, at each intensity level, the overall response of the structure is equal to the sum of the responses in motion into D/S and U/S directions. The results showed that the representation of structural behavior based on the overall energy response seems more comprehensive and accurate.In the case of dissipated energy due to fracture in crest motion into down-stream direction (DFE D/S), points (0.315 KJ, 0.005 m/s), (5.378 KJ, 0.105 m/s), (16.944 KJ, 0.417 m/s) and (19.731 KJ, 0.538 m/s) were extracted corresponding to the BLci, Yielding, NZci and CP limit states, respectively.In the case of dissipated energy due to fracture in crest motion into up-stream direction (DFE U/S), points (0.157 KJ, 0.417 m/s) and (10.454 KJ, 0.538 m/s) were extracted corresponding to the NZci and CP limit states, respectively.In the case of dissipated energy due to fracture (DFE), points (0.315 KJ, 0.005 m/s), (5.378 KJ, 0.105 m/s), (17.26 KJ, 0.417 m/s) and (32.047 KJ, 0.538 m/s) were extracted corresponding to the BLci, Yielding, NZci and CP limit states, respectively.