Journal of Seismology and Earthquake Engineering
Volume:15 Issue: 3, Fall and Winter 2013

The main scope of this paper is to study an application of evolutionary algorithms, i.e. genetic algorithm and genetic programming (GP), for obtaining Peak Ground Acceleration (PGA) prediction model in the case of Iranian database. The proposed GP model is compared with a set of existing attenuation relationships via several traditional and modern mathematical and statistical methods. A new re-sampling approach is also introduced to assess the stability of the chosen models. The obtained model shows clearly more consistency with the local data in comparison with the other selected models.

In geotechnical problems, uncertainty and inherent variability in soil strength parameters is undeniable. A common and effective way of considering uncertainty isMonte Carlo simulation method. In this method, the primary objective is to obtain the reliability index and probability of failure for the critical slip surface, or in other words, deterministic slip surface. However, the slip surface with the minimum factor of safetymay not be the surface with minimumreliability index or the highest probability of failure. In the present research work, this problemwill be scrutinized on an earth Dam. Furthermore, the global reliability index and probability of failure of the damwill be introduced. Specific gravity, seismic pseudo static coefficient and soil strength parameters including cohesion and friction angle are taken into account as the sources of uncertainty. The results show that the common approach of calculating the reliability index may not be conservative.

In this study, the damping effect of the horizontal baffles inside a liquid storage tank is investigated. For this purpose, a numerical model based on the finite volume method is established.The numerical model is used to evaluate the accuracy of the analytical model developed to estimate the hydrodynamic damping caused by wall bounded baffles. For this purpose, several full-scale baffled tanks with different aspect ratios are numerically analyzed, and the validity of the analytical model is discussed with respect to the numerical results.Next, the reduction in sloshing wave height due to the presence of the baffles is considered for selected tanks subjected to seismic excitations. Finally, a simple procedure to seismically evaluate the reduction of sloshing amplitude due to the presence of baffles is proposed and validated using the time history numerical results.

In this study, the position of the truncation boundary, which is an important issue when modeling the reservoir in the finite element formulation and determining the hydrodynamic pressure on the dam, is investigated. Water in the reservoir is assumed to be compressible, and the Sommerfeld boundary is used to model the far field. The Euler-Lagrange method is used to analyze the reservoir-dam interaction. The serendipity element is used to model the reservoir and dam. The foundation is considered to be rigid. Most researchers assume that the truncation boundary is located three times the total height of the dam away from the dambody, and they use the Sommerfeld boundary at this location. For this purpose, the reservoir was disconnected at different intervals. The Pine Flat Dam is commonly analyzed with a Taft earthquake. The results of the analyses for different positions of the truncation boundary in reservoir models show that the position of the Sommerfeld boundary at five times the height of the damaway fromthe dam body is a proper place for the truncation boundary.Moreover, comparing the results of the present study based on the compressibility of water and those of previous research based on the incompressibility assumption indicates that the maximum hydrodynamic pressure is approximately 153.8% for a Taft earthquake. Therefore, the assumption of water compressibility plays an important role when evaluating the Pine Flat Dam.

The focus of this paper is on the responses of steel moment resisting frames to pulselike near-fault ground motions. Decaying sinusoidal record is a simulated record that possesses specifications of real pulse-like near-fault earthquake. Incremental Dynamic Analysis (IDA) is a newapproach of dynamic analysis that is appropriate for assessment of the structural performance. IDA is performed on steel moment resisting frames with different heights subjected to real and simulated records. PGV is employed as a suitable intensitymeasure for pulse-like near-fault records. Results show that at the low intensity of excitations, decaying sinusoidal record is a suitable alternative for pulse-like near-fault records. Besides, the response to the decaying sinusoidal records shows lower scattering than the responses to the real records especially when PGV is used as the intensity measure. Moreover, the structural responses under decaying sinusoidal records showstrong dependency of this simulated record to the pulse frequency.

In this study, a new structural system for either retrofitting or design of new structures has been presented that provides more resistance for lateral loading conditions in comparison with the conventional systems. This new structural system can be utilized for a wide range of steel or concrete infrastructure systems frombridges to jackets as offshore structures. In this paper, structural performance of the new systemis compared with conventional systemfor bridges. The structural response of piers in long- and medium-span bridges has been studied. A comparative study is carried out through static pushover analysis of four medium-span bridge piers and reveals the new system has a higher load-carrying capacity compared with the conventional system, whilst no significant changes are observed for period-based ductility. A probabilistic analysis of the structural collapse is carried out through incremental dynamic analysis (IDA). The results from IDA analyses show higher seismic safety for the new system compared to the conventional system. Besides, a time history analyses for far-field earthquake ground motions to evaluate structural response of a long span bridge was conducted. The results indicate that the stiffness degradation observed in the conventional system caused more damage than the stiffness degradation observed in the new system.