has been done with riprap, which consists of stone having a wide mass distribution 

 (Thompson and Shuttler 1976; Davies et al. 1994). Riprap may not exhibit this 

 secondary armoring because the small stones in the mass distribution are continuously 

 moving. Deterioration of a riprap layer is examined in Chapter 7. 



The sawtooth nature of the damage curves was caused by partial self- 

 healing. If a stone moved out of its location and started downslope, the next profile 

 would show a hole and relatively increased damage. But during the following wave 

 burst, the hole would fill in through adjustment of surrounding stone so that the next 

 profile would show a partially healed structure. This healing process can be clearly seen 

 in Figure 5.10 just after the water level was raised. Over time, the structure became 

 more and more loosened, to the point that it could no longer heal itself because the 

 surrounding stones were also displaced. 



Series B' and C were similar to Series A' in that the wave and water level 

 sequences were similar; but the durations and ordering of the storms were different for 

 B' and C. The results of Series B' and C are tabulated in Appendix B. Mean damage 

 versus number of waves for the storm sequences of Series B' and C are shown in 

 Figures 5.1 1 and 5.12. Also shown are the variability as a standard deviation above and 

 below the mean. For these series, the 32 profiles were used to compute each data point, 

 corresponding to the four repeats with eight profiles each. The characteristics of armor 

 deterioration for Series B' were similar to Series A' because the order of the storms was 

 the same for both series. But for Series C, Waves 4-6 at the high-water level increased 



98 



