According to Kamphuis (1975) , the movable-bed phase of the model 

 study is subjected to four relaxed basic scaling criteria: (a) The 

 particle Reynolds niomber, (b) the densimetric Froude niomber, (c) the 

 relative density, and (d) the relative length-scale relating water 

 motion to sediment size. Ideally, all of these basic scaling criteria 

 must be satisfied simultaneously, but this is impossible in practice. 

 As more of these criteria are ignored, the model will perform succes- 

 sively less like the prototype, and scale effects (nonsimilarity be- 

 tween model and prototype) increase. Only a lightweight material can 

 be used to keep the model and prototype particle Reynolds numbers iden- 

 tical. Any deviation from unity is rather small (in all cases) and is 

 not considered to limit the model seriously. Similarity of the densi- 

 metric Froude number is considered to be the most important of the four 

 modeling criteria. If the model densimetric Froude number is less than 

 some critical value and the prototype number is greater than this crit- 

 ical value, the model is useless. The model and prototype densimetric 

 Froude numbers should be equal, or incorrect scaling will result in 

 considerable distortion of the sediment motion parameters with the 

 exaggerated time scales for sediment motion, and the model will take 

 longer to move the material than it theoretically should. This means 

 that sediment motion will start later in the model (in shallower water), 

 but in the area where material moves freely, the nonsimilarity of the 

 densimetric Froude numbers will manifest itself in adjustment of the 

 time scale for sediment motion. The time scale also varies with depth, 

 and moreover, if initial motion and depositional patterns are important, 

 it is necessary to model the densimetric Froude number correctly. 



The nonsimilarity of the model and prototype ratios of sediment par- 

 ticle density to water density affects the process in two distinct ways. 

 The acceleration of the particle is changed and the particle becomes rela- 

 tively too heavy when no longer submerged. For a lightweight material, 

 the individual particles are relatively heavier in the surf zone than if 

 sand were used. Therefore, the beach material has a tendency to pile up 

 immediately past the surf zone, and the particles will remain in this 

 location because they become relatively heavier when not submerged. This 

 results in a highly distorted version of sediment transport in the surf 

 zone. It is very difficult to duplicate prototype conditions in the 

 littoral zone using lightweight materials. 



Coastal movable-bed models suffer from various scale effects when the 

 particle sizes are not scaled down geometrically. Since this is the case 

 for most models, the prediction of bed morphology time scales is virtu- 

 ally inf)ossible. Thus, verification using historical survey data remains 

 a necessary step. Because of the variety of scale effects, coastal 

 movable-bed modeling continues to be as much an art as an exact science. 



b. Prototype Data Requirements . Perhaps the most important aspect 

 of the design phase of a movable-bed model study is to assure the ade- 

 quacy of the prototype data. The model is constructed to conform to pro- 

 totype surveys; adjustment of the model to accurately reproduce prototype 

 hydraulics or sedimentation patterns is based on prototype measurements. 



476 



