An experimental investigation of tsunami forces on coastal structures

This thesis explores the complex fluid interactions between inundating long-period waves and buildings. Several aspects of these are found to be poorly described in the literature, particularly the flow conditions around, and forces applied to, emergent structures. Numerical models capable of reproducing the propagation phase of tsunami waves reliably are insufficiently detailed to capture the complex behaviour near structures. They also have no means of directly determining the force on a building, are agnostic to its presence and often lack reliable validation data. By critically reviewing the literature this thesis shows that experimental modelling to date has been confined to short-wave tests, commonly using paddle type wavemakers, which are not comparable to tsunami waves at scale. These experiments form the basis of many current design guidelines. The loading imposed by long-period wave inundation on buildings is characterised through experimental modelling and development of a simple numerical model capable of predicting the force on a building. Large-scale testing with a novel pneumatic long-wave generator and complementary smaller-scale steady-flow experiments are carried out. The experiments are unique and examine onshore long-wave inundation loading on emergent bodies at wave periods never before tested in a laboratory. Significant differences between the loading regime of long- and short-period waves are observed in the experiments, demonstrating that short-period waves are insuf- ficient models for tsunami inundation. The experiments to date also represent the only known database of long-period wave tests at scale which are representative of actual tsunami periods, so can provide an excellent validation resource for numerical models. The numerical model presented within this thesis is novel because of its inclusion of flow blocking and drag to estimate forces on buildings. It is firstly applied to steady experiments which are used to parameterise the drag inclusion, secondly to externally published unsteady bore/surge experiments and finally to a set of largescale unsteady long-period wave experiments. Results of this very simple model show that it performs very well in all three situations examined in this thesis. Excellent estimations of measured full-body force on a blockage, as well as the flow conditions around it are achieved.