Abstract

     This paper proposes an innovative seismic design approach for shallow rectangular cut-and-cover underground subway or railway stations. The traditional approach is to design rigid frame-like structures by connecting rigidly the main horizontal and vertical structural elements (side walls, top, bottom and intermediate slabs, and central columns); on the contrary, the proposed strategy consists of joining them by means of hinged and sliding connections, in order to obtain structures whose lateral stiffness is almost zero. The objective of this approach is to be able to adapt to the transverse racking motion imposed by the seismic ground motion without significantly increasing the internal forces in the structural members. The aforementioned flexibility of the joints is achieved by interposing rubber bearings between the connected structural elements. As a case study, an existing 2-story 3-bay subway station located in Southwest China is redesigned with the proposed technology; its seismic performance is numerically investigated by performing nonlinear dynamic analyses for a number of horizontal transverse input ground motions (accelerograms) representing the site seismicity. Such inputs are scaled to fit PGAs ranging from 0.1 to 0.6 g. As expected, the results of the time-history analyses reveal that the seismic damage to the structural members is significantly alleviated in the sliding-hinged alternative solution. This conclusion can be understood as a preliminary confirmation of the satisfactory seismic performance of the proposed technology.

Introduction

     The nowadays rapid development of buried structures, such as tunnels, underground parking lots, subway and railway stations, etc. provides alternatives to the insufficient ground space. Taking subway stations as an example, according to the International Association of Public Transport [1], there were 11,084 stations spreading over 13,903 km of subway operating lines at the end of 2017 in the world, and in the next few years another 1700 km were to be built. However, recent and past investigation reveals that earthquakes can damage underground stations [[2], [3], [4], [5], [6], [7], [8]]; therefore, further research is required on their seismic design.

     The traditional seismic design of underground stations is based on providing sufficient stiffness and strength to absorb the strains imposed by the seismic ground motion without excessive damage. This approach is basically correct and leads to safe constructions, but omits that earthquakes are indirect actions (i.e. imposed displacements): designing stiffer structures generates an increase of the soil-structure interaction force, thus providing little benefit (if any). Another smarter strategy is to try to reduce that force, either by lowering the lateral stiffness of the soil (locally) or the station. The first objective is achieved with seismic isolation [9,10], and the second with structural flexibilization measures [9,10]. Both approaches are briefed in the next two paragraphs, respectively.

Conclusions

     This paper proposes a new alternative seismic design solution for rectangular cut-and-cover underground stations; it consists of replacing the traditional rigid connections between the structural elements with hinged and sliding joints, as to obtain near-zero lateral stiffness. The required flexibility of the joints is achieved through the use of rubber bearings. The ensuing structures are expected to be able to accommodate the imposed seismic racking without relevant strains in the structure. As a case study, an existing 2-story 3-bay subway station located in Southwest China is redesigned with the proposed technology; its seismic performance is investigated numerically by time-history analyses for a set of input seismic accelerograms. These inputs are scaled to fit PGAs ranging from 0.1 to 0.6 g.

      The main output of this study is that the damage to the structural elements is considerably lessened in the sliding-hinged alternative solution; this trend is particularly intense for not only central columns but also structural connections. On the other hand, the acceleration and drift do not alter much; therefore, no higher damage to the acceleration and drift-sensitive nonstructural components is to be expected in the proposed alternative solution. These conclusions hold for all the considered levels of PGA. The remarks obtained can be read as a preliminary validation of the proposed seismic design approach.