647 



It is necessary to transform the dome velocity from these arbitrary 

 xmits to feet per second. The linear dimension is obtained from the 

 geometry of the system and/or from a linear scale of some sort appearing 

 on the same print. For a given set of experimental conditions, for 

 example the tank shots, this is a constant factor. The tl~e base is 

 obtained from the timing marks appearing on the edge of the film. 



Knowing the dome velocity, in feet per second, it is easy to cal- 

 culate from Eq. (1) the corresponding peak pressiore. Since, usually, 

 six shots were made for each value of WV3/r, the calculation of ?„, for 

 each value of wV3/R was not made for each shot but rather for the 

 arithmetic mean of the dome velocity for each point. The ordinary mathoda 

 of statistics were applied to each point, the calculations of the standard 

 deviation of a single observation from the mean ( ^ ) and of the stand«ird 

 deviation of the mean of a group of observations ( c^-,) being made in the 

 customary manner. During the course of the 350 or so observations made 

 dxxring this program, the only observations eliminated on any basis other 

 than failure of some experimental unit were three which fell outside of 

 the statistical limits for a chance error. 7) 



The only other parameter in Eq. (l), the calculation of which needs 

 explanation. Is the value of D, the velocity of propagation of the shock 

 wave. This must be obtained for each calculated Pm from a relation such 

 as that given by Arons°^ between Cg, the velocity of sound under the 

 experimental onditions of temperature and salinity, 9) and D for a wide 

 range of Pjn, Since Pm Is unknown at the beginning of the calculation, 

 the correct value of U and Pm must be obtained by successive approximations. 

 This is not an onerous task since each calculation is very simple. 



5. Nonomeasett Pond Results 



Typical photographs taken with the Fastax and Streak cameras are 

 shown in Figs. 5, 6, and 7. The behavior of two well-known explosives, 

 pentollte and tetryl, was investigated. Values of wV3/r were selected 

 such that approxlmatnly equally spaced points would appeeir along the 

 similarity curve plotting wV3/R against Pn, on logarithmic paper. The 

 values were also selected to cover as wide a range of peak pressure as 

 seemed reasonable. The weight of the explosive charges varied from 

 25 grams to 5 lb, the larger charges in general being used for large 

 values of w1/3/r and the smaller charges for small values of wV3/R. 

 However, enough small charges at shallow depths and large charges at 

 deeper depths were Included to show that the dome velocity was dependent 

 only on the peak pressure and not on the charge weight - at least in this 

 range. There is some evidence^O) that Vq might depend somewhat on the 

 charge weight. This evidence is found irtien these results are compared 

 with those from quite large charges at small values of wV3/r. 



Calculations of all peak pressures for both pentollte and pressed 

 tetryl are included in Tables I and II together with the calculated values 

 of ^ and c7« . Similarity curves determined by these results are shown 

 in Figs. 8 and 9. The linear equation representing the similarity curve 

 for pentollte is taken from piezoelectric resultslO). The theoreticsLl 

 curves were calculated by the methods of Kirkwood and Brinkleyli). 



-6- 



