Ship impact

A collision of a ship in a revetment can damage the revetment, but for large ships and certain sailing speeds, penetration can be so deep that there is also a risk that the structures behind the revetment will be damaged.

In the past model tests have been performed (Brinck-Kjaer et al, 1983 and Denver, 1983) to investigate the damage that can be expected. Scale model tests using Froude scaling can simulate the ship movements and the wave action. However, in such tests the stresses in the revetment are different from the stresses in the prototype. To model the soil mechanical properties of the revetment properly, a comparable stress situation in model and prototype is necessary. Therefore impact tests have been performed in a geotechnical centrifuge. In such a centrifuge the acceleration is increased with the scale (a 1:200 model is tested at 200 times the acceleration of gravity) resulting in similar stress situations in model and prototype.

In total, a number of 12 tests have been performed. Variations between tests reside in the angle of collision, the types of revetments, the design of the foundation behind the seawall. From the tests, the penetration appears to be roughly equal to the square of the velocity. The penetration for the angled tests was measured along the line of penetration. This means that when 1 cm of penetration is measured in a 30 degrees angled test, the penetration perpendicular to the revetment is only 0.87 cm. However, even when corrected for this difference it appeared that the penetration is quite large for the angled tests. It was concluded that a different deformation pattern (compared with the perpendicular tests) deformation leads to relatively high penetration. 

The ship has to penetrate some distance into the revetment before the revetment will develop a resistance to penetration. Only then there is a high reaction force and only then large pressures can be expected on the foundation. The maximum pressure was found at the position where the bow penetrates the maximum distance into the revetment. No clear plastic deformation (failure zones displacements of more than 1mm in the model) in the sand was observed.

Conclusion
It has been possible to simulate the consequences of ship impact on a foundation in a geotechnical centrifuge. Penetration of the ship into the revetment and loading on the foundation could be investigated. Penetration of the ship into the revetment increases with the square of the impact velocity. An impact out of the center of the foundation leads to a rapid reduction of the loading on the foundation. An angled impact leads to hardly any reduction of the loading. No recordable permanent movement of the foundations due to impact was registered.