because the revetment integrity is maintained. Both iip^o ^"^ ^D=t 

 are given in terms of the significant wave height at the toe of the embank- 

 ment. Test conditions and results are presented in Table 1. 



1 . Discussion . 



The tests show that riprap stability is greatly improved when a stone 

 overlay is used. It might be suspected that when the waves become large 

 enough to move the overlay stone, open areas would be created in the over- 

 lay through which the underlying riprap could be removed by the wave action. 

 Small open areas did appear between the overlay stone when the wave height 

 reached the zero-damage level and gradually enlarged with increasing wave 

 height. These open areas started near the Stillwater level and slowly 

 migrated upslope, sometimes extending to the upper limit of the active 

 wave action. The open areas developed through the general shifting around 

 of the overlay stones, which tended to pack more tightly just below the 

 Stillwater level, rather than by the actual removal of overlay stones. 

 Even as the open areas enlarged, the exposed riprap underlayer had little 

 tendency to be removed and remained sheltered from the wave action by the 

 overlay stones until the waves approached the failure wave height. Near 

 the failure wave height, riprap was removed from the open areas, undermin- 

 ing the stability of the adjacent overlay stones which shifted around and 

 further enlarge the area; at times, the overlay stones were also removed. 

 The riprap and overlay stones once removed from the open area were depos- 

 ited by the wave in the zone just below the Stillwater level. 



In test S-7, where the wave generator was run continuously rather than 

 in 20-second bursts, the prolonged attack of high waves failed to remove 

 the riprap through the overlay stones. There was no observable displace- 

 ment of overlay stones at a wave height of 0.363 foot CH-l centimeters), 

 the wave height increment just below the estimated zero-damage wave height. 

 These observations were supported by time-lapse movies of the riprap 

 motion. Table 2 shows the following three comparisons of the stone over- 

 lay stability, using the average weight stability number, N| : 



(a) The stability of 100- to 67-percent overlay coverage; 



(b) the stability of a rounded to angular stone overlay; 

 and 



(c) the stability of a quarrystone overlay for small-scale 

 and prototype tests. 



Since the number of tests involved in the comparisons (Table 2) were 

 small, the results are regarded only as trends. The last comparison indi- 

 cates surprisingly large-scale effects, but the differences in stability 

 may be partly due to differences in the gradation and shape of the over- 

 lay. The gradation of the stone overlay used in the prototype tests 

 (Table 3) is considered wider than the stone overlay used in the small- 

 scale tests, although the gradation of the overlays in the small-scale 

 tests were not documented. The quarrystone overlay used in the prototype 



17 



