|The observation of the behavior of concrete gravity dams that are exposed to seismic loads is the main factor for the safety evaluation of the dams. In this regard, dam-foundation-reservoir interaction can be seen as one of the most significant concerns for the determination of dam safety. Moreover, hydrodynamic pressures may present because of dam-reservoir-foundation interaction and dynamic loads. As a result of this situation, hydrodynamic forces should be included in the safety evaluation to obtain more accurate seismic responses of dams. The dynamic response of gravity dams can be sufficiently illustrated using two-dimensional finite element analysis in the plane strain condition. This research provides both linear and nonlinear dynamic analyses of a roller-compacted concrete gravity dam considering the soil-structure-fluid interaction by using ANSYS 17.1 software. To evaluate the impact of hydrodynamic pressures on the dynamic response more accurately, different water modeling approaches (Westergaard and Euler) and an empty reservoir condition are used. Furthermore, the effects of foundation flexibility, Poisson’s ratio, and the presence of alluvium on the seismic response are observed. Nonlinear analyses of the dam are performed using the Drucker Prager model. Based on the US Army Corps of Engineers (USACE) criteria, the seismic responses of the dam in terms of acceleration, displacement, and principal stresses are assessed. Additionally, the influence of the various parameters on the modal response of the dam is evaluated regarding natural frequency, the effective damping ratio, and Rayleigh coefficients.
The research reveals that obtained results related to 1st principal stresses from linear transient analyses are significantly higher than the results of nonlinear time-history analyses. On the other hand, linear time history analyses considerably reduce the 3rd principal stresses compared to nonlinear dynamic analyses. However, linear analyses are useful methods to predict the displacements and accelerations on the crest of the dam because the results are substantially similar to the outcomes of nonlinear analyses. Based on the attained seismic responses of the Narli Dam from nonlinear dynamic analyses, maximum and minimum principal stresses never exceed the tensile and compressive strength of concrete. It can be concluded from the analyses that the heel of the dam is the high-stress region in terms of tension while the toe of the dam is more susceptible to compressive stresses. Hydrodynamic pressures significantly influence both the modal and dynamic responses of the dam. When the reservoir is considered in the finite element modeling and compared with the empty reservoir condition, tensile and compressive stresses grow up to 70%. On the other hand, the difference in the outcomes of Euler and Westergaard methods is approximately 10%. The foundation flexibility has a massive influence on the dynamic features of the dam and alters both the modal and seismic response of the dam. Highly flexible foundations suffer from excessive principal stresses. As the rigidity of the rock increases, the tensile and compressive stresses on the dam body decrease up to 50% and 70%, respectively. Furthermore, the Poisson’s ratio of the foundation is the more effective factor that influences the seismic response almost 10% more than the Poisson’s ratio of the concrete which alters the seismic response about 5%. If the thickness of the alluvium increases in the dam site, the dynamic response can change up to 10%.