Abstract
In this study, seven input variables are used to estimate wave forces on the crown wall, and explanatory variables are ranked using neural network techniques. 274 small-scale 2D tests, including both wave overtopping and pressure on crown-wall measurements, were used to calibrate the wave force predictors. Wave overtopping (log Q) was the most relevant variable to estimate horizontal wave forces and overturning moments, while the relative foundation level (Fc/L0p) was the most relevant variable to estimate wave up-lift forces. The new wave force estimators showed prediction errors slightly higher than the formulas given in the literature, but using fewer parameters and explanatory variables. The range of application of the new formulas is 1.67 < Rc/(γfHm0) < 6.55, 1.39 < ξ0p < 7.77, 0.36 < γfRu0.1%/Rc < 1.41, 0.00 < (Rc-Ac)/Ch < 0.59, 2.64 < Lm/Gc < 6.54, 0.00 < Fc/L0p < 0.03 and −6.00 < logQ < −2.78. Compared to pressure on crown walls, the mean wave overtopping rate is relatively easy to measure in small-scale tests and prototypes. The new estimators of wave forces on the crown wall can be used to indirectly calculate forces on models when only overtopping rates are measured. If wave overtopping is one order of magnitude higher, the wave forces and overturning moments on the crown wall increase between 11% and 60%, considerably reducing the crown wall stability.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call