Risk Assessment for the Seismic protection Of industRial facilities
WP2. Simulation and synthesis of seismic ground motions
This WP has two major directions: (a) developing a methodology to synthesize of realistic strong motion time histories using the stochastic modelling approach, and (b) applying the methodology to generate ground motions for targeted sites where major industrial facilities exist in Greece.
Regarding the first direction, the state-of-the-art techniques should be extended so as to provide the capability to generate correlated time histories, reflecting the proximity (~100m) of recording sites, and also to incorporate basin-edge-generated surface waves. As it has already been pointed out, the simulation techniques that have been developed so far do not account for the long period surface waves (usually T~3sec and longer) that are generated in a sedimentary basin from the conversion of body waves at the edges of the basin, if the seismic source is located outside the basin (e.g. Chi-Chi, Taiwan), or from channelling of seismic energy in the waveguide of the sediments in the form of surface waves, if the source is located in the basin (e.g. Imperial Valley 1979). Such waves affect long-period structures, such as liquid storage tanks which are typically located in the flat expanses provided by such basins. For this purpose we will expand the capability of our simulation techniques to incorporate surface waves in the synthetic motions for sites where the conditions warrant their presence. The technique that will be used is based on the physics of surface wave propagation and incorporates the dispersion characteristics of the sedimentary deposit (i.e. group and phase velocities). Incorporation of such waves, with the appropriate arrival time, in the synthetic motions will render the simulated time-series non-stationary with respect to their frequency content as well as to their intensity. In the past, earthquake engineers have proposed simulation techniques of non-stationary processes (i.e Grigoriu et al. 1988). We will explore how our physics-based technique relates to the above mentioned phenomenological techniques.
A second task will be the simulation of spatially correlated ground motions. To accomplish this we will rely on the spectral-representation-based simulation algorithm proposed by Deodatis (1996) to generate sample functions of a non-stationary, multi-variate stochastic process with evolutionary power, according to its prescribed non-stationary cross-spectral density matrix. An effort will be made to infuse as much physics into the selection of the various parameters and functions as the current state of knowledge allows.
For the synthesis of ground motions for sites where major industrial facilities exist in Greece (i.e petrochemical industrial complex at Aspropyrgos), all available strong motion records will be collected and used as a basis to establish the scaling law of the source spectra for earthquake sources. It is noted that the source spectrum describes the frequency content of the radiation emitted by the earthquake source/fault, before it has been affected by propagation path and recording site effects, while the scaling law of the source spectrum describes how the source spectrum scales/changes with the earthquake size. The methodology that we will follow has been established by Ηalldorsson & Papageorgiou (2005). Using the information collected from all available strong motion records, we will calibrate the Specific Barrier Model (SBM) which, subsequently, will be used for synthesizing/predicting ground motion at selected sites in Greece for selected potential earthquake sources. Furthermore, we will gather information regarding attenuation characteristics of the region (e.g. quality factor Q(f)). Finally, we will review the available information regarding site effects at the selected sites where ground motion is synthesized.
The research team of the University of Patras (UPa) will lead the efforts of this WP. The Team Leader and Team Members of UPa have long and ongoing research experience on the aforementioned topics which guarantees the successful outcome of this WP (Papageorgiou & Aki, 1983a,b, Papageorgiou, 2003, Ηalldorsson & Papageorgiou 2005, Mavroeidis & Papageorgiou 2003, Mavroeidis et al. (2004), Halldorsson et al. 2009).
Key intermediate goals: Sample time histories and corresponding response spectra to be used in the seismic risk process after the disaggregation stage of a typical Probabilistic Seismic Hazard Analysis (PSHA).