CHOICE OF TESTING GEOMETRY 



57 



not, however, be smaller than several times the larg- 

 est dimension of either instrument. Surface screens 

 and bottom screens may be helpful in reducing re- 

 flections. If the instruments have a relatively flat 

 response, that is, low Q or time constant, pulses may 

 be used to advantage with a possible increase of test- 

 ing distance. 



3. If one of the instruments is a pressure-gradient 

 device and the other nondirectional, and if no spher- 

 ical wave correction is to be applied, the testing dis- 

 tance should be greater than half a wave length. II 

 a spherical wave correction is to be applied lo the 

 results, it should not be greater than 5 db if it is pos- 

 sible to avoid it. This means the testing distance 

 should exceed y 10 of a wave length. As a rule reflec- 

 tions are not severe until the testing distance is 

 greater than the depth, and surface screens may in- 

 crease this distance. This means that reasonably good 

 calibrations can be obtained down to frequencies 

 where the wave length is about ten times the depth. 



4. If two pressure-gradient instruments are used, 

 the testing distance should not be less than a wave 

 length, since closer to a transmitter of this type the 

 pressure gradient is no longer radial. If it is necessary 

 to work closer, a correction can be computed and ap- 

 plied to the results. 



5. When a directional transducer of the piston 

 type and a nondirectional instrument are used, the 

 testing distance selected should be such that errors 

 due to proximity effects are about equal to errors due 

 to reflection interference. The possible variations in 

 this case are many and depend on the size of t he- 

 transducer and the frequency, but by an examina- 

 tion of Figures 4 and 14 one usually can reach a 

 reasonable compromise. If possible, the reflected in- 

 tensity given in Figure 4 should be 10 db or more 

 down. To avoid spherical wave corrections, the test- 

 ing distance should be larger than D 2 /X. If it is nec- 

 essary to use shorter distances, a spherical wave cor- 

 rection from Figure 14 may be applied, but any 

 correction greater than 5 db must be considered un- 

 reliable. The pulse technique may be of value in such 

 tests. 



6. The same considerations given in the preceding 

 paragraph apply to a pressure-gradient instrument 

 facing a directional transducer of the piston type, ex- 

 cept that a different spherical wave correction must 

 be applied. The fact that the pressure-gradient in- 

 strument has directionality itself reduces reflection 

 interference difficulties. The reflected intensity is be- 



low the direct intensity in decibels by the sum of the 

 values obtained for each instrument from Figure 4. 

 7. When two piston-type transducers face each 

 other, reflection interference is not usually a source 

 of trouble because of the directivity of the trans- 

 ducers, except at very great separations. On the other 

 hand, proximity effects become of greater signifi- 

 cance, and the separation should be great enough so 

 that: 



a. Spherical wave effects are small. This requires 

 that any testing distance (/ satisfy the conditions 



cl> D S+± 



+ 



6Dj I)., 



and 



d > ID, 

 d > 2D.„ 



(50) 



where D 1 and I)., are the diameters of the two 

 pistons. This also keeps the pressure due to one 

 transducer uniform over the face of the other in 

 spite of the directivity of the instruments, 

 b. No standing wave pattern is formed between 

 the faces of the two instruments. This generally 

 is taken care of if the criteria in the preceding 

 paragraph are met. 



When measuring directivity patterns, the last two 

 conditions of equation (50) should be changed to 

 d > 10D, and d > \()D-, to avoid the effect of vary- 

 ing pressure due to the inverse-square law. 



8. Essentially the same considerations apply to line 

 instruments as apply to piston-type transducers pro- 

 vided one substitutes the length of the line for the 

 diameter of the piston. If the line is suspended ver- 

 tically, there are ordinarily no great difficulties in 

 obtaining a calibration except possibly at low fre- 

 quencies. Spherical wave corrections are useful when 

 a testing distance sufficiently great to eliminate prox- 

 imity effects cannot be used. If a line is suspended 

 horizontally, surface reflection usually causes great 

 difficulty and makes it almost impossible to obtain 

 good directivity patterns except with great testing 

 depths. The inverse-square-law effect also makes it 

 difficult to obtain good directivity patterns except at 

 great distances, which are consequent on great depth. 

 The pulsing technique may be of aid. but it must be 

 remembered that, with the line mounted end-on with 

 respect to the source or receiver, the acoustic time 



