energy in the incident wave spectra at lower frequencies. This possi- 

 bility can be backed up in part by data from two similar projects under- 

 taken in the past 2 years on a styrofoam-filled concrete- type break- 

 water (Christensen and Richey, 1974). The first is located in relative- 

 ly sheltered water while the second breakwater is located near the open 

 ocean where swell becomes an influence and results in a much broader 

 spread of spectral energy over the frequency band in question. The first 

 breakwater showed similar response curves to Friday Harbor while the 

 second tended to approach one near the lower frequencies' and stay there 

 (see App. H., Figs. H-10 and H-11). 



All of the anchor cable data showed a very dominant amount of ener- 

 gy at lower frequencies. Appendix I shows the results of the low-fre- 

 quency analysis for three of the anchor cable force signals for record 

 FH 7-8. The autospectra for the force gages show several large peaks in 

 this lower frequency band (App. I, Figs. I-l, 1-2, and 1-3). The exact 

 location of these peaks varies for different records, but in all records 

 analyzed, the dominant amount of energy in the force spectra was con- 

 tained in this lower frequency band of approximately 0.015 to 0.05 hertz. 

 In most cases, however, a relatively dominant peak appeared in the 56- 

 to 63-second-period range. The anchor forces measured were all quite low; 

 the largest range was only 628 pounds. The cables are spaced at 50-foot 

 intervals . 



The phase and coherency spectra for three of the force gages for re- 

 cord FH 7-8 are given in Appendix I (Figs. 1-4 through 1-7). They show 

 a strong linear relationship between the gages on the same side of the 

 breakwater and for the opposing gages. The forces in the two anchor 

 lines on the same side were in phase; the two opposing were 180° out 

 of phase. This indicates that the sway or roll motions are dominant in 

 this frequency range. The accelerations at these lower frequencies were 

 too small to be recorded and could not be used to help confirm which 

 motion was involved. However, in the overall frequency range (0 to 1.0 

 hertz) the variances for sway and heave were two orders of magnitude 

 greater than roll, which indicates that sway would have to be the domi- 

 nant motion involved here. 



The analysis of the complete frequency range for the force data for 

 FH 7-8 is shown in Appendix J (Figs. J-2 through J-8) . The data were 

 high-pass filtered (f^ = 0.1) for these spectra. A comparison of the 

 variances computed for the high- and low-frequency sections of the force 

 spectra showed that over 90 percent of the energy in all cases analyzed 

 was contained in the lower frequencies. A summary of the force data, 

 without any filtering, for all the data collected in this experiment is 

 given in a table. 



The autospectra for the force gages (FH 7-8) are relatively spread 

 out (see App. J, Figs. J-2, J-3, and J-4) , with the outside force gages 

 showing a greater response to the lower frequency incident wave energy 

 than the inside gages. However, the outside and the opposing gages show 

 relatively high coherency, with the outside gages being in phase and the 



60 



