MEASUREMENT OF PRESSURES 179 



cables subjected to stresses in packing glands through which they are 

 led to external circuits from the pressure cylinder. Difficulty with 

 these signals can be avoided by using rigid leads through bushings in the 

 pressure chamber, but if this is done, the crystal elements of a gauge 

 must be calibrated prior to their assembly in the hnal gauge plus con- 

 necting cable, and the possibility enters of changes resulting from such 

 mechanical operations. 



A second source of error is the pyroelectric effect, or generation of 

 electrical charge in a piezoelectric material as a result of temperature 

 changes of the crystal. For tourmaline the magnitude of the effect is 

 such that a temperature increase of 1° C. gives the same signal as a 

 negative pressure of 200 lb. /in.-. Compression of either a gauge crystal 

 or its surrounding medium always leads to such temperature changes, 

 which can be calculated from the thermodynamic properties of the sub- 

 stance (specific heat, compressibility). Such temperature changes 

 must occur both in the calibration and use of the gauge. Transient 

 underwater pressure waves are in nearly all cases of such short durations 

 that the crystal is heated adiabatically, and if the calibration conditions 

 are also adiabatic there will be no error from the pyroelectric charge. 

 If, however, the time of measurement is long enough, there may be an 

 appreciable heat exchange with the external medium. In general, suit- 

 able pressure fluids exhibit much greater temperature rises than tour- 

 maline. The apparent piezoelectric charge therefore decreases with 

 time, owing to the negative sign of the pyroelectric effect for tourma- 

 line, as has been experimentally observed (9). The presence of this 

 error has the same effect as electrical leakage and sets a further condi- 

 tion on the maximum time of measurement. 



If the sources of error described are realized and their effect can be 

 made small, the quasi-static method furnishes a fairly satisfactory 

 means for routine calibration which is widely used with good success. 

 Both manual and diaphragm pressure release are practicable, the former 

 being more convenient if many readings are wanted and the time neces- 

 sary is not too long. Pressures up to about 5,000 Ib./in.^ are readily 

 obtained and measured with simple equipment and the electrical meas- 

 urement presents no formidable difficulties. The method falls short of 

 the ideal in that only relatively slowly changing pressures are involved 

 which are not transient pressures in the sense of a travelling pressure 

 wave. 



B. Steady-state calibrations. Several methods are available which 

 permit a better approximation to dynamic conditions than static or 

 quasi-state calibrations. One of these, on which the Stanolind Oil and 

 Gas Co. Research Laboratory has done preliminary tests (104), utilizes 

 a reciprocating pump which produces approximately sinusoidal pres- 

 sure variations up to 30 Ib./in.^ at a frequency of 30 cps. This speed is 



