550 
3: 
-2- 
ctip and the bar was inserted in the frame so that the lower surface of the hose-pipe rested on the 
lower shelf when the arrangement was vertical. The bar was adjusted centrally relative to the hole 
in the lower plate, so as to give an all-round clearance of 4 inch, and with the bar in this position,. 
the gap between the bar and the hole in the upper shelf was packed with strips of sheet rubber so as 
to prevent lateral motion of the bar. 
The ring switch, used to trigger the sweep of the oscillograph spots, was of the type 
previously described; it was situated a few inches above the lower shelf in the girder frame. The 
sweep circuit was adjusted so that the displacement of the measuring end of the bar was recorded on 
the return trace. 
The bar condenser unit was a Cylindrical unit measuring the longitudinal displacement of fhe 
end of the bar. The lccating tube of the unit was an easy sliding fit on the pressure bar, and the 
unit was suspended from the girder frame by three adjustable weak springs as shown in Figure 1; 
these arrangements, together with the rubber cylinders supporting the insulated cylinder of the unit, 
ensured that the latter remained at rest during the upward motion of the measuring end of the pressure 
bar due to the explosion. The girder frame was covered with a wooden box (not shown in Figure 1) 
which protected the apparatus from damage by water. 
The charge consisted of 1.765 1b. of T.N.T., together with 0.79 oz. C.£. primer, contained in 
a cylindrical charge case of mild steel plate, inch thick. The charge case was suspended from 
lugs attached to the lower end of the girder frame by spun yarn stings, approximately = inch diameter, 
16 
the distance between the lower end of the bar and the upper surface of the charge case being e inches 
“jn air. When the bar was lowered into the water, the spun yarn contracted and a separate experiment 
showed that the contraction under the conditions of this experiment was 1 mm. per foot length of spun 
yarn; the contraction in a length of 15 foot was thus 1.5 cm, = 0.59 inches, so that the true 
distance between the pressure end of the bar and the upper surface of the charge case was 6.04 inches. 
Experimental Results. 
The oscillogram of the p.d. developed across the input terminals of the amplifier in the 
experiment is shown in the lower trace in Figure 2; the upper trace in this figure was produced by 
a timing wave of period 0.0525 milliseconds, The analysis of the oscillogram of Figure 2 is given 
in curve (A) of Figure 3, where the displacement & of the measuring end of the bar is plotted as 
ordinate, and time t, reckoned from the arrival of the stress pulse at this end, is taken as abscissa; 
the values of é and t calculated from the readings of the measuring microscope are indicated by dots 
in Figure 3. 
Assuming that dispersion effects in the bar are negligible and that a uniform pressure P is 
applied over the lower plane surface of the bar, the cylindrical surface being free from stress, 
the value of P corresponding to the point (t, &) in Figure 3 is given by equation 
Pe ace maton Ole Le ate, Maant etex'. aeeos Mere (3.4) 
where ) is the density of the material of the pressure bar c is the velocity of extensional waves of 
infinite wavelength in the bar and& is the slope of the (€, t) curve at the pdint considered. since 
the values of 0 and ¢ were not determined for the particular bar used in this experiment, the value 
of pc (4.10 x 10° gm./sqecm.sec.), found in earlier experiments for the same type of steel, has been 
adopted for calculating P from the (€, t) curve. 
Figure 3 curve (A), shows that, after an initial curved portion, the €, t) curve is sensibly 
linear over the region t = 0,035 milliseconds to t = 0,143 milliseconds. {f this Vifear portion 
is produced backwards to cut the t — axis, the length of the intercept on this axis corresponds to 
about 0.024 milliseconds. tt can be shown that when the length of this intercept is not zero, the 
finite length can be due to three causes:— (1) the finite time taken by the applied pressure to rise 
from zero to a finite value, (2) the distortion of the pressure pulse due to dispersion in the bar, 
(3) the distortion of the p.d. developed across the bar condenser unit, due to imperfections in the 
amplifier seeeee 
