A similar relation can be established for the core of rubble-mound sta- 

 bility models to ensure that the pressures in the underlayers, which may 

 affect the stability of the armor units in the protective cover layer, 

 are in similitude. The linear scales of such models usually range from 

 about 1:40 to 1:50. 



In harbor wave action model studies where short -period waves are 

 studied, scale effects in wave reflection from the outer breakwaters are 

 usually not critical. In such studies, waves reflecting from the outer 

 breakwaters can be reduced satisfactorily by wave absorbers around the 

 ocean perimeter of the model and in front of the wave generator. How- 

 ever, when reflected waves from the breakwaters affect the test results 

 adv«rsely, the increase in wave reflection from the breakwaters due to 

 scale effects can be reduced by wire-mesh screens placed on the ocean- 

 side of the structure. The quarrystone in the protective cover layers 

 of the breakwaters would then be sized in accordance with the linear 

 scale of the model. The proper value of the reflection coefficient 

 would be obtained by special two-dimensional tests in a wave flume. In 

 these special tests the model would be as large as possible, preferably 

 with a linear scale between 1:10 and 1:20, depending on the size of the 

 prototype structure, the prototype water depth and wave dimensions, the 

 dimensions of the available wave flume, and the capacity of the wave 

 generator. 



(2) Intermediate- and Long-Period Waves and Distorted-Scale 

 Models . When the problem involves intermediate- and long-period seiche- 

 type waves, harbor models must reproduce a large ocean area because of 

 the long wavelengths involved and because the absorption of reflected 

 waves in the ocean area of the model requires the reproduction of large, 

 additional ocean areas to provide space for the wave absorbers and filters 

 needed to absorb the reflected waves. Exceptionally large absorber- filter 

 areas are required in such models because waves with large periods and 

 small heights are difficult to absorb without considerable reflection 

 from the outer boundaries of the absorber or filter material. The neces- 

 sity of reproducing such large areas in seiche-type models and the ever- 

 present need to conserve funds require a selection of relatively small 

 linear scales. The use of small linear scales results in excessive 

 bottom friction losses in the model, relative to the small losses in the 

 prototype, if geometrically similar models are used (see Figs. 4-3 and 

 4-4); therefore, distorted linear scales are usually adopted to reduce 

 the friction effects. The use of the distorted-scale model, in which 

 the horizontal scale (Lh) is smaller than the vertical scale (Ly) , 

 also allows the use of larger wave heights in the model. This increase 

 in model wave heights provides an easier measurement of waves. In undis- 

 torted linear-scale models, major scale effects are related to bottom 

 friction effects and the energy loss as waves are transmitted through 

 the voids of rubble breakwaters and wave absorbers. In distorted-scale 

 models the effects of bottom friction on the reduction of wave heights 

 with travel distance are reduced in magnitude as a result of the in- 

 creased depths and decreased distances of wave travel, compared with un- 

 distorted models. However, the bottom slopes and the slopes of coastal 



228 



