Characterizing wave-induced scour processes in a transition zone: A continuum model for equilibrium scour depth at cylindrical piles

The dynamics of wave-induced scour at the base of vertical piles are not well defined for some realistic combinations of wave-induced flow and pile diameter. Scour development at marine structures has a strong dependence on the Keulegan-Carpenter (KC) number, which is a ratio of flow to structure length scale. Observations and descriptions of scour development in the range 1 < KC < 6 are mostly absent from existing literature. The scour process is governed by steady streaming in the wave boundary layer for KC < O(1). A pressure induced downflow, combined with incipient flow separation around the pile, produces weak three-dimensional vorticity at the base of the pile for KC ∼ 6. Lee-wake vortices develop downstream of the structure for KC > 6, and shedding of the vortices commences at higher values of KC. While there is little to no vorticity for KC < 6 at a cylindrical pile, there must be a continuum of the sediment transport initiated by steady streaming of fluid. Analysis of wave-induced flow and pile characteristics representative of field conditions reveal that mobilization of sediment, and hence scour, should occur in the range 1 < KC < 6 even though it is not evident in some laboratory experiments. A comprehensive evaluation of published field and laboratory data regarding equilibrium scour depth at the base of cylindrical piles reveals a marked discontinuity in scour development over the range 1 < KC < 6 that may be the result of laboratory scaling. As such, existing design equations for estimating wave-induced scour may likely underestimate the magnitude in the range 0.1 < KC < 10. A new empirical model for estimating wave-induced scour at the base of cylindrical piles is derived using all available laboratory and field data in such a way as to minimize error over a broad range of flow and pile characteristics. By modifying the existing Sumer's equation, the new equation was created for entire range of KC values. This new equation reduces error by 5% compare with Sumer's equation.