For Series A', lasting 28.5 hr, the general experimental strategy was 

 designed to provide an indication of the long-term deterioration of a structure. The 

 structure was exposed to waves until failure, where the underlayer was visible through a 

 hole in the armor layer at least D^q in diameter. For this series, the experiment began at 

 the lower water depth and progressed through each of the three wave heights listed in 

 Tables 4.3 and 4.4, reaching an apparent equilibrium (visually observed) at each. Then 

 the water level was raised, and each of the three wave conditions at the greater depth 

 was run until an apparent equilibrium was achieved. The structures were exposed to a 

 total of approximately 60,000 waves for Series A'. This is equivalent to 100 to 250 hr of 

 accumulated prototype storms for mean wave periods ranging from 7 to 15 sec. The 

 number of waves was computed by dividing the run length by the mean period of the 

 incident wave time series. 



For Series B' and C, the testing strategy was intended to simulate damage 

 from a sequence of individual storms. So for B' and C, each wave-water level condition 

 was run for approximately 4,200 waves, representing roughlylO-hr prototype storms. 

 For B' and C, each sequence was run twice. And because there were two structures in 

 the flume, this testing provided an original and three repeats to determine statistical 

 variabilities. The total run length for B' and C was 8.5 and 9 hr or approximately 

 18,000 waves. Damage did not progress to failure for Series B' and C. In Series B', the 

 three wave heights at the low water level were run first (Waves 1, 2, 3), followed by the 

 two highest wave heights at the high water level (Waves 5, 6). Series C was similar 



82 



