Statistical Evaluation of the Seismic Service Ability of a Bullet Train Along the Dead Sea Fault

Abstract eng:
The bullet train from Be’er-Sheva to Eilat in Israel is planned to be built in great vicinity to the Dead Sea Fault (DSF) and therefore susceptible to different seismic risks. Bullet trains are especially vulnerable to differential deformations and one of the engineering tasks associated with the train design is to evaluate the expected serviceability level with time. The paper presents statistical analyses conducted for the prediction of the anticipated deformation of the railway caused by multiple and single earthquake events as function of the exposure period. The analyses incorporate both a probabilistic seismic hazard analysis (PSHA) and statistical disaggregation to select controlling events, using local seismicity data. The highlight of the analyses is its comprehensive approach to include four main features, from viewpoints of geotechnical engineering and earthquake engineering. Firstly, the soil constitutive model considers the effects of nonlinearity and plastic compaction due to cyclic loading. Plastic deformations during earthquakes are widely known phenomena and thus important to be taken into account for accurate deformation prediction. Secondly, the soil properties are distributed heterogeneously so that differential deformations can be evaluated, which are likely to occur in natural sediments. Thirdly, using an attenuation model, a potential cumulative absolute velocity (CAV) is evaluated at different return period based on Israeli local seismic zones and levels. Using statistical disaggregation, well representative earthquakes are selected and used for geomechanical analyses in order to predict deformation patterns. Forth, based on the results of 81 cases investigated, the fragility curves for the serviceability of the train are plotted. The resulting curves can be used as guideline to build the railway track for given exposure period of the structure and deformation tolerance.

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