concluded that Froude scaling of the hydrodynamics is necessary so that wave steepness is not distorted, 

 the fall speed parameter should be constant between prototype and model, and wave heights should be as 

 large as possible. These are precisely the criteria suggested by Dean (1985). 



42. Kriebel, Dally, and Dean (1986a) adopted the criteria given by Dean (1985) to examine profile 

 erosion and accretion characteristics in a 2-D movable-bed model. (Greater detail is given in Kriebel, Dally, 

 and Dean 1986b.) They pointed out that the scaling criteria are not universally accepted; however, 

 preservation of the fall speed parameter had been used successfully by others. They supported undistorted 

 models by noting that in undistorted models 



. . . ambiguous definitions of length and time scales are eliminated, unrealistic augmentation of 

 gravity forces are avoided, and interpretation of all physical quantities is clarified. 



43. Kriebel, Dally, and Dean (1986a) examined the validity of the proposed scaling guidance by 

 attempting to reproduce the profile development observed by Saville (1957) in a prototype-size wave tank 

 using uniform waves. The scale model bed material was quartz sand with a mean diameter of 0.15 mm. 

 Application of the fall speed parameter criterion with undistorted hydrodynamic scaling gave a length scale 

 of 9.6, corresponding to a prototype grain size diameter of 0.4-mm quartz sand. Reproduction of the 

 selected erosive condition from SaviUe's prototype-scale tests showed good overall profile development in 

 the bar and trough and off'shore geometry, lending credibility to the scaling guidance for the energetic 

 erosive condition. Similar attempts to reproduce the selected accretive test case were not successful. 

 Kriebel, Dally, and Dean (1986a) noted that the wave generator in their experiments was not sufficient to 

 reproduce the scaled longer period waves, and reflections in the flume caused reflection bars that seemed to 

 "lock up" sediment. Their detailed report (Kriebel, Dally, and Dean 1986b) indicates noticeable cross-tank 

 profile variation due to the longer swell-type wave conditions. Trial tests with irregular wave trains were 

 reported to reduce wave reflection, along with bottom ripples, and they suggested this should produce 

 better results in accretive model tests. Erosive tests aimed at investigating the eff'ect of the initial model 

 profile found that the inner surf zone and beach face areas were not affected by different initial profiles, but 

 the offshore region in the vicinity of the bar was different. This was attributed to different wave shoaling 

 and breaking characteristics. They concluded that the scaling criteria performed well for the erosive 

 conditions, but that realistic initial profiles must be used in physical modeling due to its effect on the 

 incident wave characteristics. They do not recommend using movable-bed models to simulate beach 

 recovery under regular (monochromatic) wave conditions. 



44. Dette and Uliczka (1986) compared beach profile development observed in a prototype-scale wave 

 tank with similar tests conducted at 1:10 scale. The objective of the comparison was to test validity of 

 various scaling relationships and to examine the effects of the initial profile. The model used the same sand 

 (grain size 0.33 mm) as was used in the prototype. Best comparisons between model and prototype were 

 found when the model profiles were scaled to prototype using the distorted model guidance given by 



18 



