Effect of crest modifications on overtopping discharges over composite seawalls

Effect of crest modifications overtopping discharges over a composite seawall have been investigated in this study. Another purpose of this study is to get information about the effect of impermeable vertical wall which was settled at behind of rouble mound section. Applicability of existing prediction methods for overtopping discharges on rouble mound structures to our set up has been investigated in this study. Overtopping discharges have been experimentally studied based on the new data set of 239 tests in two dimensional wave flume which is equipped with an irregular wave generator. Discharges overtopped over the reference model have been separately collected with three collection tanks with same dimensions to analyze spatial distribution of overtopped discharges. Measured total overtopping discharges were compared to the existing formulas given by EurOtop Manual (2007) and EurOtop II Manual (2016). Suggested prediction method which is given in EurOtop (2007) matched slightly better to the measured wave overtopping rates. Roughness coefficient was found to be γf=0.53 for the reference system presented in this study. A reduction factor for overtopping discharges was investigated by using a storm wall situated at five differrent crest levels. Results showed that overtopping discharge decreases as the crest elevation increases. A reduction in overtopping rate clearly depends on the wave steepness and reductions on overtopping discharges has investigated for storm, transition and swell waves, subsequently. A reduction coefficient has been given depending on wave steepness. After that, parapet was mounted at the storm wall crest for five different crest elevations. Results showed that parapet reduced overtopping discharges. A reduction coeffient for parapet was given. Although there are many studies focused on overtopping discharges on different kind of structures there is still much to consider in terms of spatial distribution behind of the crest. Experimental of the study results showed that spatial distribution of overtopping disharges behind the storm wall is quite random. Spatial distribution of overtopping discharges was also compared to existing prediction formulas given by Juul Jensen (1984), Besley (1999), Lykke Andersen and Burcharth (2006). The results presented in this study predicted relatively better by the formula given by Juul Jensen (1984) and Lykke Andersen and Burcharth (2006). It was shown that spatial distribution of overtopping discharges from this experimental study were considerably underpredicted by the prediction method proposed by Besley (1999). Total overtopping discharges analyzed for the reference system in this study were evaluated by using ANN prediction method by Van Gent (2007). The results showed that the prediction method managed to predict total overtopping discharges, satisfactorily.