Journal of Waterway, Port, Coastal, and Ocean Engineering

The roles of frequency dispersion, nonlinearity, and laminar viscosity in the evolution of long waves over distances of many wavelengths in constant water depth are investigated with numerical solutions of the Boussinesq equations. Pronounced frequency doubling and trebling is predicted, and the initial evolution to a wave shape with a pitched‐forward front face and peaky crests is followed by development of a steep rear face and a nearly symmetric crest/trough profile. While reducing overall energy levels, laminar viscosity acts to prolong cycling of third moments and to inhibit the onset of disordered evolution characteristic of nonlinear, inviscid systems. Preliminary laboratory results show some qualitative similarities to the numerical simulations. However, these laboratory experiments were not suitable for detailed model‐data comparisons because dissipation in the flume could not be accounted for with either laminar or quadratic damping models. More carefully controlled experiments are required to assess the importance of viscosity (and the accuracy of the Boussinesq model) in the evolution of nonlinear waves over distances of many wavelengths.

The linear theory for waves impinging obliquely on a vertical, porous breakwater with an impermeable crown is examined. The wave‐induced flow within the breakwater is modeled by the 1972 theory of Sollitt and Cross. Using eigen‐function expansions for the velocity potentials seaward of the breakwater, beneath the crown, and behind the structure, the reflection and transmission coefficients are calculated to determine the sheltering provided by the breakwater. The results agree well with other methods, including a plane wave analysis, obtained by neglecting the evanescent wave modes. The distribution of the wave‐induced pressure along the base of the crown is also calculated, showing that the linear pressure gradient assumption usually applied for uplift forces on the breakwater crown is inaccurate.

Measurements of impact pressures and forces together with simultaneous video records of the wave motion at and in front of caisson breakwaters were performed in a wave flume. Based on the results of this model study, criteria are developed for wave‐breaking and breaker‐type classification in the presence of vertical structures. As a result, four main breaker types are identified and discussed. The corresponding impact loads (pressure, pressure distribution, and total forces) are described. The results obtained suggest that the observed breaker shapes can now be identified by the recorded force histories, as well as by the impact pressure histories and distributions, without any use of the corresponding video pictures. The effect of the caisson front geometry on the impact forces is also briefly discussed. It is concluded that the shape of both the colliding breaker and structure front must be accounted for in the future prediction of impact loads.

The influence of tank walls on the mean drift loads on an array of bottom‐mounted surface‐piercing vertical cylinders of arbitrary uniform cross section in a narrow wave tank is investigated theoretically. The problem is rendered two‐dimensional by specifying analytically the z‐dependency of the linearized velocity potential. Application of Green's second identity, utilizing an appropriate two‐dimensional Green's function, yields an integral equation for the velocity potential on the cylinder surfaces. Discretizing this equation leads to a matrix system for the potential at the nodal points on the cylinders. The mean drift forces and moments on the array members are obtained via a direct pressure integration. A forward difference approximation is used to calculate the fluid velocity components on the cylinders from the corresponding potentials. Numerical results are presented that illustrate the effect of the tank walls on the mean drift loads for various array configurations. It is found that the mean drift loads experienced by the array members in a narrow tank are quite different from those experienced by the same array members in the open sea.

This paper investigates the effect of spoiler orientations on wave‐induced scour around a submarine pipeline. Experiments conducted under pure wave action show that the spoiler can increase both the scour beneath (tunnel scour) and downstream of (lee‐wake scour) the pipeline. The increase in tunnel scour depth is due to the higher blockage in the flow. This high flow blockage causes the deflection of a strong flow downwards through the tunnel beneath the pipe. The lee‐wake scour depth is affected by the increase in strength of the reattached flow as flow separation occurs at the edge of the spoiler. At some spoiler orientations, the formation of a secondary hole is observed at the edge of the spoiler. Secondary scour holes form when the spoiler is located near the sediment bed. The local disturbance of flow at the edge of the spoiler is the cause of the formation of these secondary scour holes.

Irregular water waves were generated in a flat‐bottomed, two‐dimensional wave tank with a partially reflecting spending beach. The horizontal and vertical velocity of the fluid particles was measured with a laser Doppler velocimeter at middepth and near bottom in the water column. The results show that a linear theory equation predicts the root‐mean‐square (rms) water particle velocity to within 10% given the measured incident wave height spectrum, reflection coefficients, and water depth. An extension of the Gaussian probability distribution based on the Gram‐Charlier series was applied to the velocity time series. This extended distribution did not fit the measured data better than the ordinary Gaussian distribution. The Beta‐Rayleigh probability distribution function was applied to peak velocities. A slight improvement is observed over the ordinary Rayleigh distribution. The parameters of the Beta‐Rayleigh distribution are empirically correlated with parameters of the incident wave spectrum.

Previous studies by De Zoysa and Chucheepsakul on a three‐dimensional static analysis of marine cables yielded contradictory results for an example problem. This work addresses this discrepancy and furnishes the correct expressions for the set of differential equations governing the cable problem. The so‐called shooting‐optimization method is proposed as a solution technique. De Zoysa's cable problem was resolved, and presented herein are the correct variations of the tension components at the bottom end of the cable with respect to different angles between the ocean current direction and the projection of the ship‐to‐buggy line onto the horizontal plane.