In the LWT tests the following parameters were varied systematically: wave 

 height, embankment slope, riprap weight, and wave period. Wave heights varied 

 from 0.43 to 1.83 meters; wave periods ranged from 2.8 to 11.3 seconds; the 

 embankment slopes tested were 1 on 2.5, 1 on 3.5, and 1 on 5; and the median 

 riprap weight varied from 12.2 to 54.4 kilograms. 



2. Small-Scale Tests . 



The small-scale tests were run in one of CERC's small wave tanks, 0.46 

 meter wide by 0.91 meter deep by 45.7 meters long, which was used to repli- 

 cate the LWT at a 1:10 Froude scale. The width of the small tank is one-tenth 

 of the LWT, and the distance from the toe of the embankment to the mean position 

 of the blade was made one-tenth of the distance in the LWT. Figure 1 shows a 

 profile view of the tank used in the small-scale tests. 



In the small-scale tests the wave height and wave period were varied, but 

 the embankment slope and median riprap weight were fixed. The embankment 

 slope for the small-scale tests was 1 on 3.5 which was one of the slopes 

 tested in the LWT. The median riprap weight was fixed at 0.034 kilogram which 

 replicates the stockpile of riprap with a median riprap weight of 34 kilograms 

 used in the LWT when scaled down using the Froude scale. 



The riprap armor layer in the small-scale tests was the same material and 

 from the same quarry as that used in the prototype tests, diorite with a 

 specific gravity of 2.71. The gradation of the model armor ranged in weight 

 from four times to one-eighth the median weight of 0.034 kilogram, the same 

 gradation as that used in the prototype (Fig. 2). A gradation analysis was run 

 both before and after testing to determine if the gradation of the armor unit 

 changed over time. As shown in Figure 2, the two gradations appear to be about 

 the same. 



The filter layer consisted of small gravel with a 3- to 8-millimeter diam- 

 eter and was constructed to approximately one-tenth prototype scale; model 

 and prototype filter layer gradations are shown in Figure 3. 



The core material was compacted sand with a median diameter of 0.2 milli- 

 meter. No attempt was made to replicate the core material at a 1:10 Froude 

 scale which is effectively impermeable in both model and prototype. 



Waves were run in bursts of short duration with an interval of about 1 

 minute between bursts to allow the wave energy in the tank to dampen out. The 

 duration of the wave burst was set equal to 



Wd = ^ (.1) 



where W, is the wave burst duration, S the distance from wave blade to toe 

 of embankment, and Cg the group velocity of the waves. The number of wave 

 bursts run at a particular wave height in the model, which was equal to the 

 number of wave bursts in the prototype, was normally enough to ensure that the 

 riprap profile was at equilibrium for the given wave height and period. The 

 minimum number of waves run at a particular wave height ranged from 340 for 

 the longest period waves to 1,050 for the shortest period waves. After the 

 required number of waves had been generated, the condition of riprap surface 



