PRINCIPLES OF INDIRECT MEASUREMENT 



381 



relative intensities of the light reflected from the sub- 

 marine model antl from the sphere, the scale factor of 

 the submarine model, and the expression for the 

 target strength of a sphere [equation (10) in Chap- 

 ter 19]. 



The technique of these optical measurements was 

 not simple. Light from a motion picture projector 

 bulb passed through a polarizing element rotated by a 

 sj'nchronous motor and was focused on the submarine 

 model. As a result, the plane of polarization of the 

 incident light rotated at a high speed. Upon reflection 

 from the model, the light passed through a second 

 polarizing element and fell on a photoelectric cell; 

 this second polarizing element was stationary, but 

 adjustable. In effect, the two polarizing elements 

 modulated the intensity of the light incident on the 

 cell and made possible the use of a-c instead of d-c 

 amplifying and measuring equipment. Moreover, the 

 use of modulated polarized light greatly reduced the 

 error caused by light scattered from the walls and 

 other objects in the room in addition to the desired 

 reflected light. 



At the same time, the photoelectric cell was also 

 exposed to light from a neon lamp which was supplied 

 with current from both a battery and a step-down 

 transformer. As a result, the neon light contained a 

 small a-c component whose intensity was directly 

 proportional to the alternating current through the 

 lamp, which was measured on a vacuum tube volt- 

 meter. Since the light reflected from the model was 

 adjustable in phase, by use of the second polarizing 

 element, and since the light from the neon lamp was 

 adjustable in magnitude, one was balanced against 

 the other, thus canceling out the a-c component of the 

 light reaching the photocell. When the a-c output 

 from the photocell vanished, this condition of balance 

 was obtained, and the voltmeter reading of the a-c 

 lamp current was then proportional to the intensity 

 of the model-reflected light. The use of this null 

 method made it unnecessary to rely on a calibration 

 of the photoelectric cell. 



To compute the target strengths, spheres from 1 to 

 12J/^ in. in radius were substituted for the submarine 

 models, and a similar procedure was followed. Photo- 

 graphs were also made at different aspect angles and 

 are illustrated in Figures 2 through 5. 



22.1.3 



Mountain Lakes 



At Mountain Lakes, [New Jersey, a model of 

 HMS/M Graph, similar to the model used at 



UCDWR and at MIT, was suspended in water in 

 the path of sound from a supersonic transmitter.' The 

 model, built to a 1 :60 scale, was constructed of copper 

 0.5 mm thick, plated with nickel 0.025 mm thick as a 

 protection against corrosion. The model was sus- 

 pended approximately 2}/^ ft below the surface of the 

 lake by wii-es at distances between 1 and 17 ft from 

 the tran.sducers, corresponding to full-scale target 

 ranges between 20 and 340 yd. 



Pulses were not used in the indirect measurements 

 at any of the laboratories. At Mountain Lakes, con- 

 tinuous sound was transmitted by a quartz crystal 

 projector, and the echo was received by a separate 

 similar unit which served as a hydrophone. The model 

 .scale was 1 : 60. Since the importance of nonspecular 

 reflection depends on the ratio of the wavelength to 

 the dimensions of the target, it was necessary to 

 scale the wavelength similarl}'. Consequently, an 

 actual echo-ranging frequency of 24 kc, which is 

 standard for most Navy gear, would require a 

 frequency of 1,440 kc in tests with a 1:60 scale 

 model. However, since the response of the trans- 

 ducers was somewhat higher at higher frequencies, 

 a frequency of 1,565 kc was used most of the time; 

 the corresponding actual echo-ranging frequency 

 was 26 kc. 



A beat-frequency oscillator, with a fixed frequency 

 of 15 mc, provided signals between 50 and 3,600 kc, 

 which were amplified and sent through coaxial trans- 

 mission lines to the projector. The received echo was 

 amplified by a preamplifier in the hydrophone hous- 

 ing, demodulated by the detector circuit and recorded 

 on a continuous strip of paper as the submarine was 

 slowly rotated about a vertical axis. The known cali- 

 brations of the transducer and receiver were used, 

 together with an assumed inverse square transmission 

 loss to determine the target strength by using equa- 

 tion (6) of Chapter 19. Under the controlled condi- 

 tions possible at a reference station on a lake, the 

 calibration is less difficult than it is for gear mounted 

 on a ship at sea; thus the calibration error in these 

 tests was probably small. Also, at such close ranges, 

 temperature gradients and surface reflections are 

 negligible. At a frequency of 1,565 kc, the attenua- 

 tion coefficient predicted from Figure 7 in Chapter 21 

 is about 0.6 db per yd. At ranges of only a few feet, 

 this attenuation is negligible and the transmission 

 loss may safely be assumed to obey the inverse 

 square law. At ranges as great as 17 ft, however, this 

 assumption may lead to target strengths which are 

 about 6 db too low. 



