Featured Application The conclusions of this article can be used to predict the uplift of tunnels or underground structures induced by soil liquefaction considering vertical earthquake motion.Abstract The uplift of underground structures induced by soil liquefaction can damage underground structure systems. Numerical simulations have shown that uplift is positively correlated with the energy of horizontal input motion. However, the effects of vertical input motion on uplift have not been studied comprehensively in the past. Previous studies on the vertical motion concluded that the effects of vertical motion on uplift depend on the overall characteristics of earthquake motion. These motion characteristics have only been studied separately in previous studies. A comprehensive study to explore the interactions and overall effects of these characteristics on the uplift of underground structures is essential. In this study, the FLAC program with the PM4Sand model was used as a numerical tool to explore the effects of vertical input motion on the uplift of underground structures. The numerical model was calibrated using centrifuge test results, and 48 earthquake motions were selected as input motions to study the effects of the overall characteristics of earthquake motions on the uplift of underground structures. The simulation results show that the frequency content characteristics of horizontal and vertical motion are the major factors affecting the uplift magnitude and the responses of liquefiable soils. However, most simulation cases show that the inclusion of vertical motion causes a 10% difference in the tunnel uplift, compared to cases without vertical motion.

Multi-span reinforced concrete (RC) curved box-girder bridges are commonly designed to facilitate traffic flow at highway interchanges. The Aksemsettin Viaduct (henceforth, A Viaduct for brevity) in Istanbul, Turkey, is an eleven-span interchange bridge with a total length of 596.8 m. Located in a high seismicity zone, the A Viaduct is designed with a curved deck, multiple bearings that have different isolation mechanisms at different bents and directions, ten rectangular columns with unequal heights, and a mix of pile foundations and spread footings. The significant length of the viaduct crossed by eleven spans also makes it susceptible to varying ground motion excitations at different foundations. To evaluate the effects of the degree of modeling detail and analysis complexity on the estimated seismic performance, the present study conducts a comprehensive fragility assessment of the specimen viaduct under various ground motion excitation schemes. First, a three-dimensional finite element model is developed with detailed simulations for the deck, columns, bearings, foundations, and abutment components. To enable different ground motion excitations at each foundation, 57 sets of spatially varying ground motions are simulated by considering the realistic surface topography and soil stratigraphy at the bridge site. Cyclic pushover analyses are performed along multiple loading directions to develop the directiondependent capacity limit state models for hollow rectangular columns. Subsequently, a demand-capacity ratio method is utilized to develop reliable fragility models for bridge columns. Component- and system-level fragilities of the A Viaduct are then assessed under uniform versus multi-support excitations, vertical motions, and ground motions with varying incidence angles. To further capture the seismic damage discrepancies of the same components at different locations, seismic repair cost ratios of the A Viaduct are assessed when subjected to uniform and multi-support excitations. This study highlights the significance of considering multi-support excitations to achieve more realistic seismic fragility and loss estimates for multi-span long curved highway bridges.