Generation of Near-Fault Response Spectrum of a Large Dam, Using the Spectral-Element Method and a Hybrid Simulation Technique

Abstract eng:
Upper Gotvand surface hydro-plant in the southwest of Iran is scheduled to have four 250 M.W. power generating units in first phase and will be extended to eight units. The plant is located as ~2 Km from the “Lahbari” known active reverse fault, hence considering near fault aspects in design of such important structure is necessary. Although after recent events e.g. the 2003 Bam, Iran, the 2003 L’Aquila, Italy, and 2011 Christchurch, New Zealand earthquakes, we have been provided with some additional records in the near-fault region but fifty years of strong-motion records worldwide is not sufficient to cover the whole range of site and propagation path conditions, rupture processes and geometric relationships between source and site that are possible from earthquakes in the near source regions. Lack of sufficient instrumental data imposes the use of simulation/numerical methods to build ground-motion time histories for deterministic scenarios. Here, we attempt to assess broad band ground motion for the Gotvand hydroelectric plant. Notably the focus is placed on the Lahbari fault for its potential of causing large earthquakes of Mmax of about 7. Low frequencies (0.1-1.0 Hz) are dealt with a deterministic numerical approach and the spectral-element method [1] is used within a regional 1D velocity model. The spectral-element method is similar to a finite element method. It is based on a primal variational formulation of the equations of motion which allows to naturally taking into account both interface and free boundary surface conditions, allowing a good resolution of evanescent interface and surface waves [1, 2]. For high frequencies (1.0-20.0 Hz) a stochastic approach is used. Calculation for each frequency band is performed separately and the total ground motion is obtained by summing up the outputs of the two methods in the time domain. Appropriate source and regional parameters are selected for modeling stochastic high frequency component and proper values for long period pulse are defined.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket