Large-scale experiments on tsunami inundation and overtopping forces at vertical sea walls

Abstract

Tsunami are very long gravity waves that may cause significant damage to coastal sea walls. The majority of relevant design codes and research papers that describe methods for predicting tsunami loads on coastal walls consider the scenario of transitory force from a bore-led wave. This does not relate to tsunami that do not form bore waves. Bore fronts generally cause short term spikes in force, which may have little effect on the vulnerability of massive structures. Post disaster accounts suggest that most coastal walls show damage that implies failure modes that occur over moderate to long durations. Therefore it is likely that the bore front assumption gives an overly conservative prediction of maximum force, and may not capture the full timescale of tsunami loading. This paper uses a pneumatic tsunami generation facility to determine the force loading on two vertical coastal sea walls during tsunami inundation. Two sea-wall models, 0.15 and 0.25 m high, with crown widths of 0.1 m (7.5 and 12.5 m at a nominal prototype scale of 1:50) are tested. It is shown that bore fronts only occur for short period waves over the bathymetry tested. Bore fronts cause a very short period spike in force, which is followed by a transitory force approximated by the hydrostatic pressure equation. The loading of tsunami length waves of periods ≥ 40 s (280 s prototype at 1:50 scale), which do not break is not greater than 1.2 times the hydrostatic force. Overtopping volume is positively correlated to the time duration of positive upstream head over the crest, rather than its maximum value. Overtopping causes a small increase in the horizontal load due to the addition of a drag and momentum load. The magnitude and time of these effects are small and short-lived in comparison to the hydrostatic load. The results compare well with available equations based on hydrostatic force and the engineer may apply a desired multiplying coefficient of a factor of at least 1.2 to account for any added pressure and momentum, and the factor of safety intended. • Tsunami loads on vertical walls are no greater than 1.2 times the theoretical hydrostatic load. • Only shorter wave periods break and form bores; longer tsunami-length waves do not break. • Tsunami overtopping volume is related to the duration of positive head over the crest. • Tsunami overtopping applies a relatively small (against the hydrostatic load) additional horizontal force.