the maximum count that can be stored is 255; therefore, the sample time 

 must be on the order of 25.5 milliseconds C^aximum count divided by maxi- 

 mum frequency output from voltage to frequency converter) . The wave 

 buoys use an 8-bit register with a 32.5-millisecond sample time 

 while the wave staffs use a 16-bit register with a 250-millisecond sam- 

 ple time. 



The error due to gain instability and nonlinearity of the voltage 

 to frequency converter is of such low magnitude that it can be neglected 

 and the overall accuracy of the recording is essentially the same as 

 given for the wave staff unit by itself (i.e., between +1 and +3 percent 

 depending on the range of operation on wave staff) . 



2. Spar Buoy Design . 



Spar buoys were used at two of the sites because of their advantage 

 in handling and transport and because they minimized the placement diffi- 

 culties due to navigational hazards, water depth, and tidal conditions. 

 The spar buoys were made of two PVC pipes coupled together near the cen- 

 ter of the buoy. The lower section is a 15 foot by 6 inch pipe filled 

 with styrofoam. The top section is 12 feet by 3 inches wherein the upper 

 8 feet is wound with a resistance wire which measures wave elevation. 

 The wave staff electronics are mounted inside the top section, above the 

 waterline, with the remainder being filled with a wood core to add stiff- 

 ness. The buoys also have a 2.5-foot diameter damping plate mounted 

 on the bottom and are anchored using a dual point mooring system with 

 the anchor lines attached at the center of drag on the buoy to prevent 

 it from being pulled underwater in strong currents. One of these buoys 

 was tested in the Puget Sound just north of Seattle. Its performance 

 exceeded expectations both in terms of minimized response to the waves 

 and accuracy of wave height measurement. Figure K-1 gives a sample of 

 the output from the buoy's wave staff in saltwater for a plus and minus 

 1 foot excitation of the buoy in heave. This was accomplished by push- 

 ing the buoy up and down by hand. Some distortion results from this 

 approach which shows up in the output of the accelerometer mounted at 

 the center of the response of the buoy in heave and roll in calm water. 

 The natural periods for heave and roll taken from these plots are approxi- 

 mately 18 and 14 seconds, respectively. These are well out of the range 

 of the 3-to-4-second wave periods expected at the site. Visual obser- 

 vations of the buoy in waves in excess of 1.5 feet indicated no ob- 

 servable heave or roll motion, but some yaw about the anchor line caused 

 by the current and wind. This motion resulted in less than a 1 foot 

 variation from the buoy's horizontal position in calm water and appeared 

 to have periods in excess of 30 to 60 seconds. For comparative mea- 

 surements, the buoy was located about 30 feet from an existing four-gage 

 array of 1-inch diameter Oceanographic Services, Inc. resistance wire 

 wave staffs. A comparison of simultaneous output from the two wave 

 staffs (buoy mounted and stationary) is shown in Figure K-4. The auto- 

 spectras computed from data obtained from one of the stationary wave 

 staffs and from the spar buoy, in a 25-miles per hour storm with 



220 



