SUORT-RANOE PROPAGATION IN DEEP WATER 



201 



4900 

 



SOUND VELOCITY 

 IN FEET PER SECOND 



4920 4940 



£0 

 5 60 



120 



RANGE IN YARDS 

 800 1000 



INCLINATION 

 OF RAYS AT 

 HYDROPHONE 



Figure 6. Sample velocity-depth curve and ray diagram for the day on which the data of Figures 5, 7, 8, 9, 12, 14, 15, 

 and 16B were taken. 



shadow zone the intensity, defined as the square of 

 the pressure at the first positive peak, decreases at a 

 rate of 35 or 40 db per kyd, down to a level which is 

 about 30 db below the value which would be obtained 

 by extrapolating the pressures obtained at short 

 ranges according to the inverse square law. Less com- 

 plete intensity data obtained on other days give com- 

 parable values for the decrease in intensity. This de- 

 crease is of the same order as that which would be ex- 

 pected for diffracted sound in an ideal medium in 

 which the velocity of sound varies with depth in the 

 manner shown in Figure 6." However, as has been 

 noted in Section 5.4, a very similar decrease is ob- 

 served for the case of 24-kc sinusoidal sound ;'^ for 

 this case, however, the rapid decrease of intensity 

 with increasing range ceases after the intensity has 

 fallen to about 40 db below the inverse square 

 extrapolation, and beyond this point the decrease in 

 intensity seems once again to be described by an 

 inverse square law. For this and other reasons the 

 supersonic signal received at a considerable distance 

 inside the shadow zone is believed to arrive there by 

 some sort of scattering process, rather than by dif- 

 fraction. It does not seem likely, however, that 

 scattering contributes appreciably to the observed 

 intensities of the explosive pulses plotted in Figure 5 

 or in Figures 3 and 4. For the disturbance produced 

 at the hydrophone by scattered sound is a superposi- 

 tion of the d.sturbances produced by various scatter- 

 ing centers, and since these have different times of 

 travel, the number of scattering centers which can 

 contribute to the disturbance at the hydrophone at a 



given instant increases with the length of the pulse. 

 The explosive pulse is so short that one would expect 

 the scattered intensity to be lower by 30 db, at the 

 very least, than that from a 100-msec pulse of sinu- 

 soidal sound having the same initial amplitude and a 

 frequency of the same order as that which pre- 

 dominates in the explosive pulse in the shadow zone. 

 It is thus hard to see how the scattered intensity 

 could be comparable with the shadow zone inten- 

 sities observed in Figures 3, 4, and 5. Thus we are 

 forced to consider diffraction as the mechanism by 

 which an explosive pulse penetrates the shadow zone, 

 at least in cases such as Figure 5 and the afternoon 

 shots of Figure 4, where a true shadow zone is pro- 

 duced by downward refraction. For a split-beam pat- 

 tern like Figure 3, the existence of a shadow zone in 

 the ray diagram is due to the fact that the assumed 

 velocity-depth curve has a discontinuity in slope; 

 since the true variation of velocity with depth is un- 

 doubtedly represented by a smooth curve, it is better 

 in this case to speak of a zone of low intensity, rather 

 than of a shadow zone, and the argument just given 

 for the occurrence of diffraction is less compelling. 



The diffraction hypothesis receives support from 

 a study of th^ shapes of pulses received in or near the 

 shadow zone. Figure 7 shows the oscillograms for 

 some of the shots plotted in Figure 5. It will be seen 

 that as the shadow boundary is approached, the 

 direct and surface-reflected pulses merge, and that 

 within the shadow zone the pulse is oscillatory. The 

 time of rise to the first maximum begins to increase 

 suddenly at about the position of the shadow bound- 



