1087 
The charges were fired in three strings, one charge of each weight 
being fired in each Seiys) Charge-to-gauge distances were varied for 
each string to obtain wh, /R values over as wide a range as practicable. 
IV. RESULTS OF TESTS 
The data obtained are presented in Tables A-I, A-II, and A-III of 
the Appendix. The results are presented as the mean and percentage de- 
viation from the mean of the readings of pairs of equivalent gauges, 
and the + or - sign refers to the deviation of the first gauge listed. 
Where no percentage follows the table entry, the value is obtained from 
a single gauge. 
aS Piezoeleotrio Gauge Results 
(a) Peak pressure:--The log of the peak pressure plotted as 
a function of log Wl 3/R for spherical TNT charges is shown in Figs 2. 
The best straight line through the experimental points may be represent- 
ed by the following equation: 
p= 2.16 x 104 (W1/3/p)1+13 1p/in.? 
The standard deviation of the experimental points from this line is 8.9. 
Values from Kirkwood's theory, which were calculated from Ref. 4 and are 
represented by the dotted line, are in better than 6% agreement with the 
above equation for values of W1/3/R < 0.2, and differ by only 13% at the 
upper limit of the experimental curvee There is a slight tendency for 
the experimental curve to deviate upward at the highest pressure values 
obtained, but the theoretical line shows a much greater curvature than 
the experimental. 
(b) Time constants:--Sinoe the first portion of the presse 
time curve fits reasonably closely an exponential of the form P=P.97% 8 
out to t < @, the time constant 9 (i.e., the time required for the curve 
to fall to a pressure value = Pmax/ ®) of the initial portion of the shock 
wave may be determined. Curves of experimental values of oft 3 plotted 
against w1/3/R for TNT, and the line from Kirkwood's theoretical calcula- 
tions for TNT are shown in Fige 3. Though the theoretical line drops off 
sharply the best line through the experimental points is only very slightly 
curved downward at the high values of wl? 9 7 in each case, and might almost 
equally well be a straight line represented by the following equation: 
@ = 0.060 W1/3 (w1/3/p)-+18 
(c) Impulse:--Impulse values were obtained by integration of 
the pressure time curve to 6-7 times the nominal time constant which, for 
each ,particular value of Wl/3/R, was taken from the line representing 
ofiid/ versus w/3/R shown in Fige 3. The use of these faired time con- 
stants prevents the low precision of the measured time constants from 
introducing an unjustified scatter in the momentum and energy. Integra- 
tion to a fixed multiple of the time constant, for each W/ 7R, gives 
values which are more nearly proportional to the total momentum or total 
ae 
