Geology Reference
In-Depth Information
calculating the initial risk occurrence measure,
no new simulations are needed, and the results
in Table 4 are obtained at a small additional
computational cost (requiring only approxima-
tion of the densities through Kernel estimation
and calculation of the relative entropy integral).
These results demonstrate some significant vari-
ability, indicating that there is a big dependence
of the sensitivity analysis and the importance of
the various risk factors, on the exact performance
quantification chosen. Especially interesting is
the fact that for the bridge with the dampers, the
epicentral distance and the pulse amplitude are
the only important risk factors for all different
quantifications.
descriptions for the seismic hazard, exploiting an
end-to-end simulation-based approach.
CONCLUSION
The design of supplemental dampers for seismic
risk reduction of isolated multi-span bridges was
the focus of this Chapter. The basis of the sug-
gested approach is a probabilistic framework that
explicitly addresses all sources of uncertainty,
related either to future excitations or to the struc-
tural configuration, by appropriate probability
models. In this setting, seismic risk is expressed
by a multidimensional integral, corresponding to
the expected value of the risk occurrence measure
over the space of the uncertain model parameters.
Through appropriate definition of the risk occur-
rence measure this approach facilitates diverse
risk quantifications. Stochastic simulation was
suggested for evaluation of the multidimensional
integral describing risk and an efficient algorithm
was discussed for performing the associated
design optimization and selecting the optimal
parameters for the damper implementation. An
efficient sampling-based probabilistic importance
analysis was also presented, based on information
entropy principles, for investigating the influence
of each of the model parameters on the overall
seismic risk.
Due to fact that the framework is based on
stochastic simulation, consideration of complex
nonlinear models for the bridge system and the
excitation was feasible. The adopted bridge model
explicitly addressed nonlinear characteristics of
the isolators and the dampers, the dynamic behav-
ior of the abutments and the effect of pounding
between the neighboring spans to each other as well
to the abutments. A realistic probabilistic model for
future near-fault excitations was also considered.
An illustrative example was presented that con-
sidered the design of nonlinear viscous dampers
for protection of a two-span bridge. The fragility
of the bridge system related to seismic pounding
FUTURE RESEARCH DIRECTIONS
The design of structural systems for seismic risk
mitigation requires explicit consideration of the
uncertainties for the characteristics of future ex-
citations as well as for the system properties. The
constant advances in computer and computational
science, especially the widespread implementa-
tion of distributed computing, have created a new
era for computer simulation and it is currently
acknowledged that Simulation-Based Engineer-
ing Science constitutes a critical new paradigm
for uncertainty quantification and propagation
and is providing great new potentials for detailed
modeling and solution of problems that were until
recently considered computationally intractable.
Future research efforts need to exploit these char-
acteristics and focus on modeling approaches for
quantification or seismic risk and computational
frameworks for estimation of this risk that will
explicitly address all important characteristics of
the built systems and its environment, with no need
to establish any type of approximations for com-
putational simplicity. This will only be established
through development of high fidelity numerical
models for the dynamic behavior of structural
systems and adoption of complex probabilistic
Search WWH ::




Custom Search