508 
-i2- 
the error introduced in using a tourmaline gauge adiabatically (as In an explosion) and calibrating 
the gauge isothermally does not exceed 0.33% - negligible quantity. No experimental verification 
is given. In spite of this reassuring theoretical deduction, however, it is desirable that the 
experimental calibration method should, as far as possible, approach the ideal adiabatic condition. 
Towards this end, the pressure apparatus shown in Figure 10(a) and (b) has been designed. The pressure 
vessel consists of a heavy wrought iron cylinder capable of withstanding a steady hydrostatic pressure of 
the order of % ton per square inch, This cylinder is fitted with a watertight gland, for the cable of 
the p.e, gauge, and a special device for suddenly releasing the pressure on the gauge. This release 
consists essentially of an annealed copper disc (14 inches diameter effective, and .007 Inches thick) 
closing a hole on the lid of the pressure vessel. The copper disc, whilst making a good pressure joint is 
held between the jaws of a circular cutting punch - a slight rotation (30°) of a hand screw causing this 
punch to shear the copper disc completely around its circumference. Pressure is applied to the p.e. gauge 
in the pressure vessel, by means of a hydrostatic piston pump (with non-return valve) hand operated, the 
pressure being indicated on a Bourdon pressure gauge* 0 = 1200 Ibs./sq.in. When all joints are tight the 
pressure vessel holds a steady pressure of 1000 1bs./sq.in. without leakage. 'f now the copper diaphragm 
is sheared by rotation of the cutting arm the pressure in the water filling the vessel is suddenly released 
through an aperature 1 inches diameter. This sudden release develops a charge on the p.e. gauge proport= 
ional to the change of pressure. The records indicate that the time involved in the fall of pressure Is 
a little under 0,01 second. Although this time may be considered large compared with the time of rise of 
pressure, Viz., .0001 second, in an explosion impulse, it must still be regarded as of short duration when 
we are considering adiabatic and isothermal effects. ¥t seems reasonable to suppose therefore that this 
method of calibration is essentially adiabatic, and the piezo-ciectric sensitivity of the gauge thus 
obtained is applicable without correction, to the case of the explosion record, 
In making a calibration record it is important of course to measure the.capacity and insulation of 
the p.e. circuit and to observe the voltege generating the cathode rays. The time axis of the retord is 
made by one of the methods described in Section |i iv. a. The rotating potentiometer method and the slow 
traverse method have both proved satisfactory and much superior to the older a,c. method. Records 
illustrating the three methods are shown on Sheet |. Records e, f, g and h. From such records as these 
the sensitivity of the gauge is determined. The pressure sensitivity of a gauge is usually expressed in 
terms of the pressure required to produce 1 cm. deflection of the cathode rays when the generating voltage 
V is 3000 volts and the capacity of the circuit is 0.01 4 farad**. All observations are therefore reduced 
to these standard conditions, making use of the Information that 8a 2? where p is the pressure in 
cy 
lbs./sq.ine, C Capacity in microfarads and V the generating voltage of the cathode rays. 
The graphs shown in Figure 11 were experimentally determined in this manner and indicate:- 
The deflection of the cathode rays is 
(i} Directly proportional to the pressure applied and 
(ii) Inversely proportional to the capacity of the circuit, 
In both cases the generating voltage V is kept constant, tn (i) the capacity c is kept constant whilst Pp 
is varied and in (ii) P is kept constant whilst c is varied. 
In Section || (b) (iv) above it was pointed out that the voltage — developed in the p.e. gauge 
circuit is proportional to the area A of tourmaline crystals (c= KPA). Gauges of different sizes have 
c 
been constructed and this point also has been verified experimentally = a 7 inches diameter gauge having 
approximately double the sensitivity of one 5 inches diameter. Slight differences between gauges of the 
same size have of course been observed as the crystals mounted in the gauges are irregular and do not 
completely cover the surface of the supporting metal plate. To summarise, = the theoreticali relation 
§a@ PA has been experimentally verified. 
cv 
ThE sevens 
i This gauge is calibrated at intervals and has been found to remain correct, 
ae These values of voltage and capacity were Chosen as standard since they represent average 
values over the whole working period. 
(2 Sane . P 
+ Qe p Be 6 Tv ON eyi 
