58 



TESTING TECHNIQUE 



constant is at least the length of the line divided by 

 the velocity of sound. Unless deep water is available, 

 this time may be greater than the time required for 

 an interfering reflected pulse to appear. 



While the outline given above should be of aid in 

 fixing the geometry for a test, it must be remembered 

 that each test has its own particular factors involved 

 and thus no general rule can be made. Experience 

 and judgment are of primary value in making a 

 proper choice. It is nearly always helpful to repeat 

 tests with a changed geometry, particularly when it 

 is suspected that reflections or proximity effects are 

 causing trouble. One then has an internal check, on 

 the validity of the corrections that may have been 

 applied, as well as a means of recognizing the pres- 

 ence of interfering reflections or other extraneous 

 effects. 



ss THE ESTABLISHMENT OF 



SOUND FIELDS 



551 Absolute and Relative Calibrations 



The calibration of underwater sound devices re- 

 quires reliable knowledge of the magnitude of sound 

 fields in water. Two techniques for establishing these 

 magnitudes may be distinguished. One involves a 

 direct absolute measurement of the field intensity; 

 the other uses previously calibrated instruments eith- 

 er to establish a known sound field or to measure one 

 which is present. Among the available methods of 

 the first type are the Rayleigh disk method, the radia- 

 tion pressure method, the reciprocity method, and 

 various motional impedance methods. The second 

 technique requires calibrated instruments, whose 

 calibrations have been obtained either by some of 

 the absolute methods mentioned above, or by such 

 methods as computation from their design or cali- 

 bration in air. In the routine calibration and testing 

 of underwater sound devices, the method of relative 

 calibration with a known standard is far more con- 

 venient than a direct absolute calibration, which re- 

 quires considerable care and is more time-consuming. 

 It has the disadvantage that it is based on the stability 

 of the reference standard, but numbers of these 

 standards have been constructed whose calibration 

 remains sufficiently constant for the accuracy usually 

 desired in relative calibrations. The most satisfactory 

 absolute method in water is the reciprocity method, 

 and it is now exclusively used by USRL for the abso- 



lute calibration of standards. In the following sec- 

 tions the various methods of obtaining absolute 

 calibrations are discussed. 



5.5.2 



Absolute Calibration from 

 Design of Standard 



In principle, if the design of a transducer is com- 

 pletely specified, one can theoretically calibrate the 

 device over its entire frequency range by solving the 

 equations of acoustics, mechanics, and electromag- 

 netism involved in its operation. Actually, the equa- 

 tions are too complicated to allow a practical solution 

 unless some assumptions are made. In spite of these 

 approximations, it may often be possible to obtain a 

 reasonably valid theoretical response characteristic 

 for the device over a considerable range of frequen- 

 cies. This method has been applied with consider- 

 able success to certain standards in use at USRL, in 

 particular to the 3A Rochelle salt crystal hydrophone 

 and to the 1A pressure-gradient type hydrophone 

 designed by the Bell Telephone Laboratories. 924 The 

 principal difficulties in the method are the following: 



1. It is not feasible to take into account all pos- 

 sible modes of vibration of the device and its housing, 

 but only the desired mode and perhaps a few closely- 

 coupled ones. However, some of the neglected modes 

 are excited in operation, particularly near their 

 resonant frequencies, and may introduce "break-ups" 

 in the response which will not be included in the 

 computed calibration. 



2. At all frequencies, but particularly at those hav- 

 ing wave lengths of the order of magnitude of the 

 dimensions of the device and higher, diffraction 

 around the instrument plays an important role. This 

 diffraction effect is very difficult to compute, and 

 computations have been carried out only in some 

 highly idealized cases. Thus, it is difficult to include 

 in the theoretical calibration precisely the effect of 

 diffraction. 



3. Some of the constants of the mechanical ele- 

 ments involved in the construction of an instrument 

 are not easily measurable. In particular, mechanical 

 resistance as a function of frequency as well as of the 

 effective mass and stiffness of various elements may 

 be difficult to obtain over the desired frequency 

 range. 



4. The method can be applied only to the rela- 

 tively few instruments designed with this method of 

 calibration in view, and then only by skilled person- 



