DEPENDENCE ON FREQUENCY 



409 



standard for most Navy sonar gear, although some 

 measurements have been made at frequencies of 12, 

 18, 45, and 00 kc. The indirect measurements used 

 scale models; at Mountain Lakes, with a 1:60 scale 

 model of the Graph, a frequency of 1,565 kc was em- 

 ployed to simulate an echo-ranging frequency of 

 26 kc. At MIT visible light was used, and the cor- 

 responding full-scale frequency was very much 

 higher than for any of the other measurements. 



23.6.1 Theory 



The target strength of a submarine depends on 

 frequency according to equation (36) in Chapter 20, 

 especially if nonspecular reflection contributes ap- 

 preciably to the target strength. Specular reflection 

 depends only slightly on frequency, as described in 

 Section 20.4. Beam echoes result largely from specu- 

 lar reflection, as pointed out in Section 23.8.1. Thus 

 the variation of target strength with frequency at 

 beam aspects may be expected to be slight. 



At off-beam aspects, however, specular reflection 

 is much less important and nonspecular reflection 

 may become appreciable; this effect is discussed in 

 Section 23.8.2. At these aspects, the whole sub- 

 marine appears to scatter sound. If reflections from 

 the superstructure or exterior protuberances on the 

 submarine, such as rails, guns, and periscopes, con- 

 tribute appreciably to the target strength, the target 

 strength may depend on frequency as long as the 

 dimensions of these scatterers are of the same order 

 of magnitude as the wavelength. Section 20.5 de- 

 scribes the origins of nonspecular reflection. 



23.6.2 



Direct Measurements 



No reliable direct measurements substantiate this 

 expected dependence of target strength on frequency. 

 Figure 22 shows target strength as a function of as- 

 pect angle plotted for frequencies of 24 and 60 kc for 

 a fleet-type submarine at San Diego at signal lengths 

 of 10, 30, and 100 msec. Each point represents the 

 average of all observations in a 30-degree sector 

 centered at the point indicated. 



A clear-cut dependence of target strength on fre- 

 quency is apparent in this illustration, as well as in 

 Figures 20 and 21. Figure 21 gives quite reasonable 

 values for the target strength at 60 kc, but Figure 20 

 shows values about 10 db lower, for a frequency of 

 24 kc; thus the dependence on frequency is still evi- 

 dent. It is unlikely, however, that the increase in tar- 

 get strength with frequency would be not only so 



great but also so nearly uniform at all aspect angles, 

 as Figure 22 shows. Furthermore, the measurements 

 cannot be relied on for two reasons : the transmission 

 loss was not known but estimated, and the calibra- 

 tion of the 60-kc gear was less reliable than the 24-kc 

 gear calibration. The estimated attenuation co- 

 efficient of 20 db per kyd used for these computations 

 is larger than that measured elsewhere and is per- 

 haps excessive, since attenuation coefficients of only 

 10 db per kyd at 60 kc were measured at Fort 

 Lauderdale. However, correcting the high San Diego 

 target strengths at 60 kc by reducing the attenuation 

 coefficient from 20 to 10 db still results in values 

 greater than those obtained elsewhere. The remain- 

 ing discrepancy may perhaps be attributed to faulty 

 calibration of the gear, described in Section 21.4, 

 since this discrepancy is systematic and apparently 

 independent of aspect angle. 



In addition, the high target strengths measured at 

 60 kc at San Diego do not seem substantiated by 

 target strength measurements at that frequency at 

 Fort Lauderdale. These measurements gave a target 

 strength of 25 db for a fleet- type submarine at beam 

 aspect at 60 kc, assiuning an attenuation coefficient 

 of 12 db per kyd. An assumption of an attenuation 

 coefficient of 20 db per kyd would raise this to only 

 33 db, compared with the maximum value of 44 db 

 recorded at San Diego for the target strength of a 

 fleet-type submarine. Furthermore, wake echoes 

 measured with the same equipment at San Diego 

 and described in Section 33.4.2 were found to be 

 much higher at 60 kc than at 24 kc, contrary to 

 theoretical expectations. These results support the 

 suggestion that calibration errors are responsible for 

 the high values obtained at San Diego. However, in 

 view of the many uncertainties in this subject, the 

 possibility that submarine target strengths are sys- 

 tematically some 10 db higher at 60 kc than at 24 kc 

 cannot be entirely ruled out, even though there is 

 little if any theoretical expectation of such a varia- 

 tion. 



No difference was apparent between the measure- 

 ments at 12 kc and 24 kc made by Woods Hole ob- 

 servers; "> both target strength-aspect curves were 

 very similar. However, the target strengths reported 

 are so much larger than all other measurements else- 

 where that calibration errors were probably present. 

 Therefore, since the calibration at 12 kc was quite 

 different from that at 24 kc, and since all the values 

 seem very uncertain, the lack of any frequency de- 

 pendence cannot be considered significant. 



