Sharma 



The strong correlation between the curves for larger y is very surprising 

 in view of the different origin of the two methods. 



REPLY TO DISCUSSION 



S. D. Sharma 



Professor Ward has presented an instructive diagram (Fig. Dl) showing a 

 comparison of the experimental wave resistance of the Inuid S-201 determined 

 in three different ways. This is the first time such a verification of the mutual 

 consistency of these methods has been attempted. It may appear from the fig- 

 ure that the overall agreement is not quite as good as would be desirable, but it 

 should be remembered that wave resistance is only a fraction of the total re- 

 sistance, and the discrepancies regarded as percentages of the latter are small. 

 This is encouraging in two respects. First, it shows that an approximately in- 

 variant wave resistance can be defined in a real fluid on the basis of these 

 methods. Second, it seems permissible to use, alternatively, any of these three 

 methods depending upon the facilities available. 



Further, it is true that the "X-Y" method can also be used to predict the 

 wave -making resistance of certain linear combinations of the bulb and the main 

 hull under the same assumptions as used by the author. However, the spectral 

 approach is more powerful. For example, given a certain wave profile (longi- 

 tudinal or transverse) of a particular hull form, it is an easy matter to predict 

 by linear superposition the corresponding profile (i.e., longitudinal or trans- 

 verse respectively but not vice versa) of a pair of such hulls moving in tandem 

 or side by side (catamaran) respectively. However, if one takes the trouble of 

 deriving the spectrum by either method, one can predict any profile (longitudi- 

 nal, transverse, or inclined) of any parallel configuration of any number of 

 similar hulls! Of course, in principle, one could also derive the spectrum from 

 the X and Y signals, but it would be a poor way of doing so. 



The additional results of a computer study on the Inuid S-201 supplied by 

 Dr. Eggers in Fig. D2 are welcome. It is evident that if y is not too small, both 

 longitudinal cut methods would asymptotically yield the right value of wave re- 

 sistance as the truncation point -x^ -»oo. Figure D3 has been prepared by taking 

 two cross curves from Dr. Eggers' data and supplementing it with the corre- 

 sponding values corrected for truncation error as explained in the paper. It is 

 obvious that for any given finite length of the cut there is an optimum transverse 

 location (e.g., ca. y= 8 in the present case). This results from the fact that the 

 truncation error increases with increasing y, while the error due to local ef- 

 fects decreases. 



770 



