The earthquake hazard of New York City was revised recently based on the recommendation of a group of seismologists.  The maximum acceleration of bedrock for earthquake of 2500-year return period ranges from 0.2g ~ 0.33g, depending on bedrock stiffness. The input acceleration may amplify considerably if a thick layer of soil exists above the bedrock.  This level of earthquake should be considered in the evaluation and retrofit of New York City’s critical transportation infrastructure, including bridges as well as underground highway and subway tunnels.

Several critical tunnels of New York City, including those across Hudson River, across the East River and the Brooklyn-Battery Tunnel across New York Harbor, are partly embedded in saturated soft soils, some of which are sandy soils and may liquefy under earthquake loading.  The liquefaction of surrounding soil may displace the tunnel and subsequently damage the linings.  The damage could be more severe if part of the tunnel is embedded in non-liquefiable soil or rock while the adjacent portion is in liquefiable soil.

The issues of liquefaction-induced damage to underground structures raised attentions after the 1995 Kobe Earthquake in Japan, following which extensive experimental and numerical studies revealed some underlying mechanisms of liquefaction-induced damage of underground structure.  Some mitigation measures were also identified.  However, in these studies, the three-dimensional effects, including the three-dimensional earthquake input and the embedment of tunnel in both liquefiable and non-liquefiable ground media, were not properly taken into account.  These effects were not considered in prior studies partly due to the restrictions of experimental and numerical capacities.  It is difficult to duplicate 3D effects in reduced-scale or centrifuge models; and 3D liquefaction analysis using numerical method has always been a challenge in geotechnical engineering.  However, the issue is important to the critical underground structures in New York. It is necessary to check whether the existing joints in these tunnels are able to overcome the possible tunnel displacement due to earthquake-induced soil liquefaction, which can only be accomplished by 3D analysis.

In this study, the Finite Element program, TNO-DIANA93 will be used to investigate this problem, targeting typical subaqueous transportation tunnels in New York City.  The program has the capacity to consider 3D soil liquefaction and earthquake input.  The Finite Element procedure will be firstly calibrated against existing centrifuge tests for its capacity in modeling soil liquefaction and its effect on underground structure.  3D numerical models will then be established to investigate the effects of both liquefiable and non-liquefiable ground media on tunnel response under 3D earthquake loading as well as the influences of various parameters such as the depth of soil underneath the tunnel, the burial depth of tunnel, and the depth of water above the tunnel.