Underground structures, tunnels, subways, metro stations and parking lots, are crucial components of the build environment and transportation networks. Considering their importance for life save and economy, appropriate seismic design is of prior significance. Their seismic performance during past earthquakes is generally better than aboveground structures. However several cases of severe damage to total collapse have been reported in the literature, with that of the Daikai metro station in Kobe during the Hyogoken-Nambu earthquake (1995) being one of the most characteristic. These recent damages revealed some important weaknesses in the current seismic design practices. The aim of this chapter is not to make another general presentation of the methods used for the seismic design of underground structures, but rather to discuss and highlight the most important needs for an improved seismic performance and design. In that respect it is important to consider that the specific geometric and conceptual features of underground structures make their seismic behavior and performance very distinct from the behavior of aboveground structures, as they are subjected to strong seismic ground deformations and distortions, rather than inertial loads. Several methods are available, from simplified analytical elastic solutions, to sophisticated and in principal more accurate, full dynamic numerical models. Most of them have noticeable weaknesses on the description of the physical phenomenon, the design assumptions and principles and the evaluation of the parameters they need. The chapter presents a short but comprehensive review of the available design methods, denoting the crucial issues and the problems that an engineer could face during the seismic analysis. The main issues discussed herein cover the following topics: (i) force based design against displacement based design, (ii) deformation modes of rectangular underground structures under seismic excitation, (iii) seismic earth pressures on underground structures, (iv) seismic shear stresses distribution on the perimeter of the structure, (v) appropriateness of the presently used impedance functions to model the inertial and the kinematic soil-structure interaction effects, (vi) design seismic input motion, accounting of the incoherence effects and the spatial variation of the motion and (vii) effect of the build environment (i.e. city-effects) on the seismic response of underground structures. The discussion is based on detailed numerical analysis of specific cases and recent experimental results in centrifuge tests. Other important issues like the design of submerged tunnels to liquefaction risk, or the complexity to evaluate the response of the joints of submerged tunnels are also shortly addressed. Finally we present the most recent developments on the evaluation of adequate fragility curves for shallow tunnels.

Performance and Seismic Design of Underground Structures

Seismic behaviour of tunnels: From experiments to analysis

SUMMARY
In order to investigate the seismic failure characteristics of a structure on the liquefiable ground, a series of shaking table tests were conducted based on a plaster model of a three-story and three-span subway station. The dynamic responses of the structure and ground soil under main shock and aftershock ground motions were studied. The sand boils and waterspouts phenomena, ground surface cracks, and earthquake-induced ground surface settlements were observed in the testing. For the structure, the upward movement, local damage and member cracking were obtained. Under the main shock, there appeared longer liquefaction duration for the ground soil while the pore pressure dissipated slowly. The acceleration amplification effect of the liquefied soil was weakened, and the soil showed a remarkable shock absorption and concentration effect with low frequency component of ground motion. However, under the aftershock, the dissipation of pore pressure in the ground soil became obvious. The peak acceleration of the structure reduced with the buried depth. Dynamic soil pressure on the side wall was smaller in the middle and larger at both ends. The interior column of the model structure was the weakest member. The peak strain and damage degree for both sides of the interior column exhibited an ‘S’ type distribution along the height. Moreover, the seismic response of both ground soil and subway station structure exhibited a remarkable spatial effect. The seismic damage development process and failure mechanism of the structure illustrated in this study can provide references for the engineers and researcher. Copyright © 2013 John Wiley & Sons, Ltd.

Shaking table test on the seismic failure characteristics of a subway station structure on liquefiable ground

A 40-m long test was conducted to estimate the seismic performance of the tunnel of the Hong Kong–Zhuhai–Macau Bridge project under non-uniform earthquake loadings. The test used twelve connected model boxes, a synthetic model soil and a scaled model tunnel. The model boxes included four “active” boxes, which were fixed to four isolated shaking tables that worked as excitation sources, and eight “inactive” boxes, which were passively excited by the active boxes through connections. The soil at the tunnel location was simulated with a mixture of sawdust and sand with mass proportion 1:2.5, that yielded dynamic properties analogous to those of the in situ soil. The tunnel model, at a 1/60 scale, was composed of 98 sections, each with dimensions 600 × 375 × 170 mm, that were made of aluminum, which best approximated the target response of the actual tunnel, given the size of the model, geometry and scaled engineering properties required. In the tests, a non-uniform seismic excitation was provided by imposing the seismic wave to the active model boxes with a time lag equal to the time that would take a seismic wave to travel from one active box to the next along the axis of the tunnel. The test started with the assembly of the tunnel sections and installation of transducers on the tunnel at critical locations. After placement of the soil and the tunnel, the seismic loadings were applied through the active boxes. The test results showed that the acceleration response of the tunnel was larger than that of the surrounding soil. It was also found that the deformation of the tunnel joints under non-uniform excitation was larger than under uniform excitation, to the extent that it could jeopardize the safety of the tunnel had it been designed solely using the uniform excitation. The results of the experiments clearly showed that the effects of non-uniform seismic excitation should be considered for the seismic design of long tunnels.

Multi-point shaking table test of a long tunnel subjected to non-uniform seismic loadings

Response characteristics of rectangular tunnels in soft soil subjected to transversal ground shaking

The Daikai Station, a cut and cover structure in the subway system in Kobe, collapsed during the Hyogoken-Nambu earthquake of January 17, 1995 in Japan. The Daikai Station is the first well-documented underground structure not crossing an active fault that has completely collapsed during an earthquake without liquefaction of the surrounding soil. What makes this case even more interesting is that tunnel sections adjacent to the station, with similar structural characteristics and analogous soil conditions, did not collapse. Dynamic numerical analyses have been conducted to investigate the load transfer mechanisms between the underground structure and the surrounding soil and to identify the causes for different behavior of similar sections of the station subjected to the same seismic loading. A hysteretic nonlinear soil model has been used for the analysis. The model captures well the soil's shear modulus degradation and the increase of damping with strain. The results from the analyses show that, for a given earthquake, there are two key factors that determine the response of an underground structure: the relative stiffness between the structure and the degraded surrounding ground, and the frictional characteristics of the interface. A stiff structure has small deformations; because the adjacent soil movement is restricted by the structure, the shear modulus degradation of the soil is limited which contributes to reduce further deformation of the soil and thus decreases the displacement demand on the structure. A strong interface is capable of transmitting larger shear to the structure but in turn increases the confinement of the soil surrounding the structure which limits the soil's shear modulus degradation. The model predicts larger deformations in the section that collapsed because this section had a smaller stiffness, and thus triggered drifts in critical structural elements which were larger than at other sections of the station which remained stable.

Load Transfer Mechanisms between Underground Structure and Surrounding Ground: Evaluation of the Failure of the Daikai Station

An iterative procedure is proposed to estimate seismic-induced distortions of cut-and-cover rectangular structures. The procedure is based on an existing analytical solution for deep rectangular st...

A practical iterative procedure to estimate seismic-induced deformations of shallow rectangular structures

A key requirement in the evaluation of sliding stability of new and existing concrete gravity dams is the prediction of the distribution of pore pressure and shear strength in foundation joints and...

Hydromechanical analysis of flow behavior in concrete gravity dam foundations

The deformations of underground structures during seismic events are largely determined by the relative stiffness between the structure and the surrounding ground. A new analytical solution has been developed to estimate the deformations of buried structures. Soils exhibit a marked non-linear hysteretic behavior with cycles of loading. Such behavior can be incorporated into the analysis by an iterative scheme where the soils’ shear stiffness is reduced with the structure’s deformations. Initial verification of the method is done by comparing its predictions with those from a dynamic numerical analysis of the failure of the Daikai station in Kobe during the 1995 earthquake.

G. Fernández

Papers

Load Transfer Mechanisms between Underground Structure and Surrounding Ground: Evaluation of the Failure of the Daikai Station

A practical iterative procedure to estimate seismic-induced deformations of shallow rectangular structures

Hydromechanical analysis of flow behavior in concrete gravity dam foundations

A Practical Method for Assessment of Seismic-Induced Deformations of Underground Structures