Immersed tunnels subjected to a sunken ship load

Abstract

A sunken ship load needs more and more to be taken into account in the design of an immersed tunnel. Such a load is often governing the design. The designer gets often a prescribed value (in kN/m2) from the client for a sunken ship load. The range in prescribed values show however a big scatter. There are also a lot of tunnels over which a lot of ships are passing, without having a sunken ship load taken into account in the design. This gives the need to get more insight in the behaviour of a sunken ship to predict a representative sunken ship load. Another important question is, if a ship sinks on an immersed tunnel, whether the tunnel collapses or not. If the tunnel collapses, the users are not safe anymore. The insight in the behaviour of a sunken ship is gained through a literature study. From this literature study it resulted that a sunken ship load depends on a lot of parameters. It depends on the mass density of the carried cargo, the size of the ship, the depth of the waterway, and the way how a ship sinks. To determine a sunken ship load, all events in which a ship hits the tunnel are set out in a fault tree. All those events are evaluated with respect to the probability of occurrence an the magnitude of load. The magnitude of load is determined with the aid of a developed model. In that model the different parameters on which a sunken ship load depends are taken into account. The event in which a ship sinks with its bow on the tunnel deck appears to be the most important one. Sunken ship loads ranging from 50 – 300kN/m2 are found then. The response of an immersed tunnel depends on the type of tunnel. Concrete tunnels appear to be the most important type of tunnel. Concrete tunnels on a typical Dutch subsoil are evaluated with respect to a large sunken ship load of 300kN/m2. For the cross sectional analysis, an globally increase in U.C. of 1.3 for shear and a decrease in U.C. of 0.7 for bending moment was found. For the longitudinal analysis, monolithic tunnels obtain shear forces in the joints which are three times bigger than the strength. The bending moments cause through cracking of the tunnel structure. The deformation of the water seals remain below the requirements. The shear forces in the joints for a segmented tunnel lie in the same order of magnitude as the strength. The bending moments remain small and cause no through cracking of the tunnel structure. The deformations of the water seals can become twice as high as allowed. The Wijkertunnel is used as a case study to evaluate the response of a specific tunnel when subjected to a sunken ship load. The Wijkertunnel is a segmented tunnel with an immersed part of 575m consisting out of 6 elements. Each element contains four segments. The sunken ship loads applied on the tunnel are derived from the shipping characteristics of the ships passing that tunnel. The tunnel is evaluated for a small ship (a general cargo carrier) and a big ship (an iron ore bulk carrier). For the general cargo carrier only the event of the (bow of the) ship sinking on the tunnel deck is evaluated. For the iron ore bulk carrier the same event is evaluated, and also the event if the ship sinks just next to the tunnel. If the tunnel sinks just next to the tunnel, a load is induced on the tunnel wall. From the cross sectional analysis it resulted that one tunnel tube fails under the load from the general cargo carrier and the load from the iron ore bulk carrier when sinking on the deck. Shear failure is the governing failure mechanism. From the longitudinal analysis it followed that the shear kay fails in the wall if the tunnel is loaded from aside. Failure of one tube of the cross section implies that the structure is no longer structural safe. The users of the tunnel are in great danger then. Failure of the shear key in the wall causes only big leakage problems. The tunnel remains structural safe then. Users of the tunnel are still able to leave the tunnel safely. The risk of a sunken ship load leading to failure of the Wijkertunnel is lower than the risk of a BLEVE. The risk of a BLEVE is often accepted and not covered in the design of immersed tunnels. It is therefore concluded that the risk of a sunken ship load also can be accepted.