were done on six different beam widths, 2, 3, 4, 5, 6 and 7 modules at a 

 constant water depth of 240 cm. Results were presented graphically in 

 the form of transmission coefficients versus frequency for each beam 

 width (McGregor 1978, Figure 7). All tests were run with the same 

 incident wave spectrum with a peak frequency of 0.67 Hz. Values of C^ 

 versus Ln/B have been obtained by measurement from McGregor's 

 Figure 7. Results agree very well with those of Harms and are shown in 

 Figure 10. Unfortunately, there is no data for Lp/B < 0.99. 



As seen in Figure 10, the results of the present field study 

 agree quite closely with earlier model test results. For practical 

 purposes, the most important part of Figure 10 is for values of C^ 

 less than 0.5. In this range the Harms curve shows the best agreement 

 with the field results. For values of C^ larger than 0.4, the Harms 

 and McGregor results underpredict C-f 



The field results indicate a levelling-off or residual value 

 of C* of approximately 0.15, while the model results tend towards zero 

 for small values of L/B. This may be attributable to viscous scale 

 effects causing more attenuation in the model than in prototype. 

 It may also be due to reflected and diffracted wave energy contributions 

 in the field. Unfortunately the prototype-scale CERC model tests do not 

 include data for values of L/B less than 0.8. 



84 



