2J^2 SHOCK WAVE MEASUREMENTS 



{a, j(3, 7) for the series listed are summarized in Table 7.3. As has been 

 remarked in section 7.2, impulse and, to a lesser extent, energy flux 

 values depend on the time to which the integration of the pressure-time 

 curve is carried, and this time for the different series is therefore also 

 entered. For comparison with theory, values of the parameters fitted 

 to calculations based on the Kirkwood-Bethe theory (59) are entered in 

 parentheses. 



Important questions from the purely empirical point of view are the 

 accuracy of the experimental values upon which the curves are based 

 and the accuracy with which the curves represent these points, as these 

 determine the confidence with which pressure and derived quantities 

 can be obtained from Eqs. (7.6). The reproducibility of experimental 



Table 7.3. Parameters of shock wave similarity curve for several explosives. 

 Values predicted from the Kirkwood-Bethe theory over the same pressure range 



are given in parentheses. 



values differs for the various series quoted, and specific estimates should 

 be obtained from the original reports, but standard deviations of mean 

 values, where applicable, are of the order of d=2 per cent for peak pres- 

 sure, and ±4 per cent for impulse and energy flux. These estimates of 

 error are based entirely on internal consistency of the data and rather 

 larger errors in absolute value are possible, but enough independent 

 evidence for isolated pressure values exists to make it improbable that 

 these errors are more than a few per cent. The standard deviation of 

 single experimental points from the best straight lines are of the order 

 of 4 per cent for peak pressure, and 8-12 per cent for impulse and 

 energy. It is to be noted that these estimates of error include failure 

 of the straight-line relation to fit the data over the range of measurement 

 as well as scatter of the points, and arc therefore somewhat pessimistic. 

 It is therefore reasonable to conclude that over the experimental range 

 of W^'^/R, the formulas (7.6) give reasonably good values of the para- 

 meters. 



The theoretical estimates listed in Table 7.3 give, on the whole, a 



