Abstract

The number and size of hydroelectric dams have increased greatly across the Canadian landscape since 1910. The concrete gravity dams should perform satisfactorily during a seismic event as in case of failure, the release of impounded reservoir water can cause catastrophic damage in the downstream communities. Traditionally the foundation in a dam is modeled by a sub-structuring approach for the purpose of seismic performance analysis. The main disadvantage of sub-structuring approach is that it cannot be used for solving nonlinear dynamic problems. Therefore, in that case seismic response analysis must be carried out in time domain as it allows inclusion of nonlinear behavior in a system. In this study, different earthquake input mechanisms has been studied considering the following models A) massless foundation , B) free-field earthquake input at dam foundation interface and C) deconvolved earthquake input model. Deconvolution is a mathematical process which allows the adjustment of the amplitude and frequency contents of an earthquake ground motion applied at the base of the foundation to achieve the desired output at the dam-foundation interface. It has been observed that the existing procedures of deconvolution are inadequate for the high frequency earthquake records. A Modified deconvolution procedure has been proposed here for efficient deconvolution of high frequency earthquake records.The above discussed input mechanisms are studied in more detail with two different geometrical models. It has been found that model C is the most rational and accurate one compared to the other models.

In the second part of this study elastic and inelastic analysis of dams with different geometries has been performed to study the existing guidelines for the seismic assessment of dams. The performance of the dam models has been assessed using both the high frequency and low frequency earthquake records scaled at 0.35g. Based on the study it can be concluded that the different numerical models induce slight differences in the results relating to the tensile damage in dam foundation system. However, the results are consistent with each other. Also, the results from the linear elastic analysis provided valuable insight about seismic performance of concrete gravity dams but they fail to account for the existing deterioration in the dam.