Progress achieved over the last few decades in computer technology and in numerical techniques has enabled engineers to address complicated, properly posed, boundary value problems associated with the seismic response of long-span bridges with relatively small requirements in calculation time and equipment. Such problems may involve a variety of structural elements, complex geometries and loading conditions, as well as soil-structure interaction effects. However, this progress has helped in solving only one part of the engineer's work, the implementation of efficient numerical algorithms and the study of realistic structural models. The other equally important part is the generation of reliable synthetic ground motions (incorporating near-source effects) for realistic simulations of the seismic response of long-span bridges. Near-fault strong ground motions are characterized by long-period velocity pulses, as well as by displacement time histories exhibiting an impulsive character and/or significant permanent displacements associated with the tectonic deformation that the site experienced during the earthquake. The waveforms of these near-source time series depend on the fault type (e.g., strike-slip, reverse, etc), the direction of ground motion component with respect to the strike direction of the causative fault (i.e., strike-normal, strike-parallel), as well as on the type of the rupture (i.e., dislocation-like versus crack-like rupture). These near-field ground motions can be detrimental for long-period structures such as long-span bridges, high-rise buildings, base isolated buildings or bridges, and should be systematically considered and studied in the seismic hazard characterization of flexible structures. Until recently however, the importance of the long-period ground motion components for the seismic response of long-span bridges was underestimated. The gradually increasing number of near-fault ground motion seismograms recorded by broadband digital strong motion instruments has recently enabled seismologists to understand and analyze the character of the near-source ground motions, and engineers to reevaluate and reconsider the design practices of long-span bridges. Despite the progress that has been accomplished, the recorded near-source strong ground motions should be complemented by analytical and numerical techniques that generate reliable synthetic ground motions appropriate for the engineering design of long-span bridges. In this direction, a modeling approach that combines (depending on the simulated frequency range) both deterministic and stochastic in nature methodologies can be employed. Alternatively, simple and reliable analytical models that adequately describe the nature of the impulsive near-fault motions both qualitatively and quantitatively may be used. Such mathematical models should be able to analytically represent empirical observations that are based on available near-field records. Furthermore, the input parameters of these models should have a clear physical meaning and be related to basic physical parameters of the fault rupture. In this study, we discuss the main characteristics of the near-source ground motions, as well as their importance for the seismic response of long-span bridges. In addition, we present a simple mathematical expression for the representation of near-fault ground motions that fulfills the requirements and serves the purposes addressed above.
展开▼