SEISMIC OPTIMIZATION OF CONCRETE GRAVITY DAM WITH RUBBER DAMPER USING PROBABILISTIC ANALYSIS

In designing engineering systems, achieving the desired goals with low cost is the ultimate aim. Therefore, minimizing cost and optimizing engineering designs have been considered in the engineering sciences for a long time. The optimization process is important in building concrete dams due to the large volume of concrete used in the dam body. Much of the dynamic force acting on the dam caused by the hydrodynamic force is generated through the interaction between the dam and reservoir, which increases the amount of seismic responses of the dam. Therefore, hydrodynamic dampers with different materials were used
for the upstream of the dam body to reduce the effect of the dam and reservoir interaction on seismic response. The use of hydrodynamic dampers, in addition to reducing the hydrodynamic pressure against the dam, the economic aspect of the design is also less expensive than other designs. In this paper, in order to increase the efficiency of hydrodynamic dampers, the optimized dimensions through Monte Carlo probability analysis as a modern method of optimizing structures, were studied. ANSYS software that is based on finite element method is used for modeling and analysis. Flexibility of foundation is considered in modeling and Sommerfeld boundary condition is used for far field reservoir boundary. The water is taken as an inviscid, compressible and irrotational fluid with small displacements. According to governing equations and geometrical shape of dam, the problem is considered two-dimensional, and seismic analysis is performed by using Newmark time integration scheme. The pine flat dams in California are selected as a case study, and EL Centro, San Fernando, and North Ridge earthquakes is applied for the model. According to the obtained results of the probabilistic analysis model, which shows the sensitivity of the dam seismic responses to the rubber damper dimensions, a suitable range of the rubber damper dimensions to the safe and optimum design has been proposed.