The points in the force records corresponding to the wave crests 

 were chosen as the origin (and end) of the digitized records. These 

 points were determined by averaging the midpoints of three or four hori- 

 zontal lines drawn through the crests of the wave record at several 

 elevations above the Stillwater level (SWL) . These midpoints were 

 approximately identical except for some of the larger, longer waves in 

 which the peak of the wave crest did not exactly coincide with the mid- 

 point of the zero crossings of the wave crest. A sample data record is 

 given in Figure 23. 



The three-dimensional experimental data were handled differently; 

 the data were recorded on magnetic tape with an electronic digital data 

 acquisition system. This instrument sampled alternatingly from the two 

 channels (wave and force) at a rate of 100 samples per second, result- 

 ing in 50 samples per second from each channel. 



Tlie origin at the wave crest and the wave period were determined 

 from the digitized wave records, rather than directly from the strip- 

 chart records. Since positive readings of the wave profile corres- 

 ponded to the crest and negative readings corresponded to the trough, 

 the point of origin of the wave crest was determined by taking the mid- 

 point of the positive readings between zero crossings on the wave 

 profile. The crest was thus defined as the data point closest to the 

 midpoint of the zero crossings. The wave period was determined from 

 the number of readings between two successive crests, since there was 

 a time interval of 1/50 second between each reading. Thus, the wave 

 period was determined to the nearest 0.02 second. 



The origin of the force record was taken as the force reading cor- 

 responding to the defined origin at the center of the wave crest surface 

 profile. In reality, there was a small timelag of 1/100 second between 

 the wave profile readings and the corresponding force readings. This 

 small timelag was ignored in the analysis, since it was felt that the 

 accuracy of the defined origin at the wave crest was only good to the 

 nearest 1/50 second, the time interval between successive readings of 

 the wave record. 



Only one wave cycle was used for analysis of the electronically 

 digitized data. Since the data were on magnetic tape, it was impossible 

 to determine that two successive waves had exactly the same period and 

 height until after the calculations were completed on the computer. 

 Thus, if the waves had slightly different periods, the time phase cor- 

 relation of the corresponding force readings would be slightly in error 

 when taken over two wave cycles. In addition, since the accuracy, 

 resolution, and rapid sampling rate of the electronic digitizer allowed 

 more readings per wave cycle than the manual digitizing method, a suf- 

 ficiently large number of force readings could be obtained in one wave 

 cycle. 



56 



