10. Two types of model tests were conducted during this study: stabil- 

 ity and previous damage tests. Each type followed a prescribed sequence. 



Stability Tests 



11. For a stability test the following test sequence was used: 



a. Rebuild the breakwater from the previously damaged condition. 



b. Survey the breakwater to document its initial condition. 

 £. Calibrate the wave gages. 



d. Select the tape file and signal attenuation setting. 



e. Start the wave generator and run waves. 



f. Collect wave data (several or more times). 

 g_. Stop the wave generator. 



h. Survey the breakwater to document its final condition. 

 The duration of wave action was from 1-1/2 hr for a test using the File 1 

 spectrum to 3-1/2 hr for a File 4 spectrum. Generally, the technicians 

 observing the tests thought that most of the stone movement occurred during 

 the first 10 or 15 min of wave generation, so the final survey is regarded as 

 an equilibrium profile for the structure. In rebuilding the breakwater the 

 technicians rarely touched the stone but merely pushed it around by foot until 

 the shape conformed to the desired initial profile. This procedure was a con- 

 scious effort to avoid overly careful placement of the stone. Outlines of the 

 desired initial profile were fixed to the walls of the testing channel, and a 

 moveable template was used to ensure that the initial profile was reasonably 

 close to the desired profile. Initial configuration of the breakwater for a 

 stability test was a narrow, trapezoidal shape with seaward and landward 

 slopes of IV on 1.5H (Figure 3). Crest widths were three typical stone dimen- 

 sions wide, using the cube root of the volume of the median weight stone W . 

 as the typical dimension d _ . Figure 3 also shows a typical profile after 

 moderately severe wave attack during a stability test. Wave transmission and 

 reflection also were measured during a stability test. 



