653 



The last method of estimating T to be presented here is that based on 

 a comparison of Pu as obtained from Eq. (1) and dome velocity measure- 

 ments and Pji obtained for the same values of wV3/R from piezoelectric 

 measurements. It is well-known that piezoelectric results give a similarity 

 curve which is a straight line when plotted on logarithmic paper. 

 Inspection of Figs. 8 and 9 will show that the values of ?^ calculated 

 from Eq. (1) tend to fall below the piezoelectric straight line and/or the 

 theoretical curves for the smaller values of wV3/r. The straight line 

 relationship must be assumed for the case of the tank series (Fig. 13) 

 because piezoelectric results are not available for such small explosions. 

 This departure from a straight line is undoubtedly due to our lack of 

 knowledge of the proper value of T to put in Eq. (2). If T were known 

 for each set of conditions, then Pj, as calculated from Eq. (2) would 

 probably be more nearly the same as that obtained from other records. This 

 T correction should also be applied to the higher values of P^,, but P^, is 

 so large for high values of wV3/r that T is Insignificant by comparison. 



If the above explanation is correct, one c«in obtain an approximate 

 value of T by correcting the values of P^, obtained by the dome velocity 

 method so as to bring them more nearly in accord with curves from other 

 sources. This estimation is questionable in the case of pentolite (Fig. 8) 

 but the value of 1500 lb/in. 2 is necessary to make the first point match 

 the piezoelectric straight line. This assumption would not place the 

 other points unreasonably high. The similarity curve for pressed tetryl 

 charges is not as well known as is probably necessary for this calcvilation^) 

 taut a comparit^^n with known experimental values shows that a value of about 

 2000 lb/in. 2 is necessary for correspondence at the lower values of wV3/r. 

 Since no similarity curve is known for Engineer's Special Detonators it 

 seems reasonable to assume vsuLues which would correct the cmrve of Fig. 13 

 to the straight line of Fig. 13a. This gives a value of 3300 lb/in. 2 which 

 corresponds to T - 6600 lb/in, 2, the highest value of T obtedned by any of 

 the considerations of this section. The value seems reasonable because 

 it allows all of the results obtained with lowered surface tension to be 

 included in the area between the experimental and corrected curves, Fig. 15a. 

 The logic of this sta cement will be discussed later. 



The various estimations of T are summarized in Table IX. 



TABLE JX 

 SUMM&RT OF ESTIMATED VALUES OF T 



Method of Catlculation T 

 (lb/in,2) 



Critical Depth 1500 (800 - 2000) 



Dome Periphery ^50 (300 - 900) 

 Extrapolation to Vg - 3600 



Similarity Curve Correction 4500 (3000 - «^00) 



-12- 



