SJ^6 MOTION OF THE GAS SPHERE 



doubt of the discrepancy, as predictions regarding migration of the 

 bubble and its effect on bubble pressure pulses are likewise not well con- 

 firmed (see part (c) and section 9.6). 



B. Photographic studies of bubble migration. The most complete 

 set of data on motions of explosion bubbles as affected by free and rigid 

 surfaces is undoubtedly contained in the records obtained by Lt. D. C. 

 Campbell at the Taylor Model Basin (17). In these experiments, most 

 of the bubbles were formed by the explosion of Du Pont Number 8 det- 

 onator caps, although a few results are given for larger mine detonators 

 and for J^ ounce tetryl charges. The bubbles were photographed using 

 the high-speed stroboscopic motion picture camera technique developed 

 by Edgerton, most of the exposures being made at 1,500 frames/sec. 

 In Campbell's report a large number of sequence records are shown 

 taken under the following conditions : 



(1) Charge at various depths, surfaces remote. 



(2) Charge fired at several distances from a rigid vertical concrete 



wall. 



(3) Charge fired at several depths beneath a steel sheathed wooden 



bottom boat. 



(4) Charge fired at several distances above a concrete block. 



(5) Charge fired midway between two steel plates with several sepa- 



rations. 



The charges fired under condition (1) served essentially as control 

 shots for the later series, in which surfaces were present, and show the 

 characteristic radius-time curves already discussed. In some cases as 

 many as 17 oscillations of steadily decreasing amplitude were observed, 

 but after two or three cycles the changes in amplitude become so small 

 that detection of a cycle is very difficult. The charges placed near the 

 vertical w^all show very strikingl}^ the attractive effect appearing at the 

 minimum radius. Displacement-time curves obtained from the films 

 are reproduced in Fig. 8.23 and the rapid motion at the time of the mini- 

 mum is clearly evident, the maximum velocity of the bubble center 

 toward the wall being of the order of 200 ft. /sec. for a charge fired 9 

 inches from the wall. At close distances, this velocity decreases, and 

 the bubble becomes flattened on the side toward the wall, being very 

 nearly a hemisphere when initiated by a contact explosion. Much the 

 same phenomena were observed for charges fired near the concrete bot- 

 tom. 



Charges fired close to the free water surface showed the expected 

 repulsion. A downward motion of part of the gas volume occurred 

 even for a charge fired 3 inches under the surface, although the bubble 

 lost part of itself through the surface during the first expansion. This 

 free water repulsion was also observed for a charge fired 12 inches below 

 the Avooden boat, but a charge fired 6 inches beneath it acquired an up- 



