3. 
1025 
Because of distortion occurring in reproduction, horizontal and vertical scales are 
provided in the photographs in the form of black or white lines drawn on the prints. The 
distances between the inside edges of the lines corresponds in the actual experiments to 
7 in. in the 10 and 50 microsecond series pictures and the miscellaneous photographs, 
to 13 in. in the 100 microsec. series and to 26 in. in the 200 microsec. series. 
The pieces of some of the longer cylindrical charges which had been broken in 
handling were rejoined after fusion at the points of fracture. It is interesting to 
note that the remelted parts are visible in the photographs. In many of the photographs, 
twine, pieces of detonator wire, and part of the dqtonator itself are visible in addition 
to the charge, bubble, and shock wave. 
The focal length of the camera lens was determined at two distances by photographing 
a grid underwater so that calculations based on the lens equations could be made. In so 
doing, it was found that the lens used gave some "pincushion" distortion of rectangular 
objects, but this distortion is not sufficient to change the relative axial lengths by 
an amount greater than 0.6 in. 
Figure 105 shows the relation between bubble radius and shock wave radius for 1/2 
lb. spherical charges of cast pentolite, Since the time intervals for these shots can 
be estimated from the positions of the shock waves, an approximate time scale is also 
included. 
Figure 106 indicates the rates st which the shock waves from the asymmetric 
cylindrical charges approach a spherical shape. 
ime-constant mea ments of shock wave tical di on. 
Several photographic methods have been developed which allow calculation of the 
peak pressure of a shock wave, and for one of these methods the calculations have been 
extended to make possible an evaluation of the time constant of the shock wave. These 
methods make use of the fact that light rays passing through a shock wave at favorable 
angles are considerably distarted due to the increase of refractive index in the region 
of high pressure. A detailed theoretical discussion of these methods is to be found in 
Appendix II. 
(a)  Spheric hock wave; c e not on the optical axis; charge in grid plane. -- 
The arrangement with which most of the work has been done involves placing a 
transparent lucite grid marked off with lines 1/4 in. apart in front of the 
camera, and the shock-wave-producing charge off to one side and in the same 
plane with the grid. (See Figure 133, Appendix II.) A flash charge is placed 
behind the grid and is timed by means of primacord to go off when the shock wave 
from the main charge is crossing the grid. The results of both peak pressure and 
time-constant determinations from four shots using this method agree eseentially 
with UERL piezoelectric results for similar conditions. 
(4) Peak pressure. Two of these four shots were made with 250 gm tetryl as 
the main charge, and the photographs were taken when the shock wave radius 
was about 15 in, One of these photographs is reproduced in Figure 107. The 
theory (Appendix II.1) relates the amount of apparent displacement of an 
individual intersection of grid lines with the average refractive index of 
that section of the shock wave through which the corresponding light ray must 
pass. This average refractive index is then converted to the corresponding 
pressure, Paye This calculation is made for several grid intersections at 
different rettii, Plotting (on semi-log paper) log p,, against the corres- 
ponding distances of the mid-points of the refracted Prght rays behind the 
shock front, ({RI - r,,), a straight line is obtained to within the pre- 
cision of measurement. Figure 108 is such a plot for a shock wave 14.9 in. 
from a 250 gm tetryl charge, while Figs. 109 and 110 are the results of 
measurements on two different prints of another shock wave 14.4 in. from 
250 gm tetryl. The average deviations from the straight lines drawn are 
approximately 7%, 6%, and 3%, respectively. The peak pressure is assumed 
to be the extrapolation of this line to zero distance behind the front. Any 
errors introduced by assumption that the calculated Pay's correspond to the 
43 15518 
