499 
-3- 
These formulae apply to T.N.T. Amatol, or Guncotton in the region where the maximum pressure does 
not exceed two tons per square inch, provided the charge and gauge are both more than 10 feet from the 
surface or bottom of the water and are sufficiently removed from other bodies or alr spaces. Relation 
(3) above shows that about one quarter of the chemical energy of a charge of these explosives is converted 
into mechanical energy in the form of a pressure pulse in water. The values of (1), (2) and (3) above 
are little affected by the methods of filling the charge or, except in extreme cases, by the method of 
initiation or, except as stated above, by the depth of the charge. 
Reflexion of the pressure train by the surface and bottom of the sea or by the hul} 
of a ship 
Records have been obtained or reflected pressure-trains in water. Observations of surface 
reflexions indicate that sea water can probably support a momentary tension not greater than 200 Ibs. per 
square inch. Tensions of 100 bs. per square inch and 170 Ibs. per square inch have been observed for 
water in bulk and at a water-steel surface respectively. Records of bottom reflexions show that the 
ratio of the maximum pressures of the reflected and primary pressure waves is about 1:2 for a bottom 
composed of mud and sand. As a result of this phenomenon, the maximum’ pressure developed near a mine 
Vaid on the bottom of the sea will be greater than that developed by a similar mine moored in mid-water by 
about 50%. A record was obtained during the "Gargon" trials which showed a sudden drop in the pressure 
at the gauge, due to the arrival of the reflected pressure-train from the hull, 
Additional observations. 
The experimental methods used in obtaining the above data were used to determine the velocity 
with which the disturbance due to an explosion is propagated in sea water. No certain difference between 
the velocity of these high pressure waves and that of ordinary sound waves was detected. Preliminary 
experiments were also made with a method for determining experimentally the rate of growth of the” bubble 
of hot gases produced by an underwater explosion 
BEARING OF THIS WORK ON FUTURE EXPERIMENTS. 
The experimental methods elaborated for the work described in this report will be used, with 
others, in the series of model experiments which is to be carried out on the damaging capacityof pressure— 
waves due to explosions. The work which has been done has given us reasonably reliable means of predicting 
the form of the pressure-time curve under different conditions, and of calculating its more important 
Characteristics. The next step is to link this information with "damage done" to different types of 
structure. One difficulty in working out estimates of damage is to arrive at a satisfactory numerical 
assessment of damage done, but already the theoretical investigation of this question has given encouraging 
results, and the model experiments, with possibly one cr more large trials, shculd give information which, 
when taken into conjunction with our knowledge of the characteristics of pressure-time curves, will be of 
great value in determining what explosions can safely be withstood by the hul} of a vessel. 
Introduction and object of investigation. 
For a considerable number of years efforts have been made by numerous physicists and engineers to 
obtain information, on the form, magnitude and destructive effect of the pressure wave produced by an 
underwater explosion but the results achieved are often very conflicting and throw litt¥e or no Vight on 
the problem, 0n the other hand certain experiments, notably by H. WwW, Hildiar, and Or. G. W. Walker have 
cleared up many obscure points and have placed the whole problem on a more scientific basis, Hilliar's 
work, using the ‘Copper crusher-gauge’ method, has yielded data not only regarding the maximum pressure 
developed in the explosive wave but also as t» the form of the pressure-time relation.. Walker's ‘spray 
method' has given a means of determining maximum pressure, but throws no light on the question of the 
variation of pressure within the pulse. The ‘copper diaphragm method', originally proposed by walker Is 
a promising form of ‘damage indicator’, but the interpretation of results is difficult at this stage. 
None of the above methods, however, gives a direct indication or record of the pressure-time relation 
in the explosive pulse. Hiltiarts method gives indirectly an indication of the rate of fall of pressure 
TrOM ooses 
