452 



34 



First evidence of the separation of the data into two groups ap- 

 pears in the graph of t „" in Figure 26, which consists of two branches. The 

 U-branch is practically rectilinear, extending from 0.30 millisecond at 7 

 Inches to 0.3U millisecond at a charge distance of t8 inches. The L-branch 

 is also linear with charge distance and extends from 0.2U millisecond at 12 

 Inches to 0.275 millisecond at 32 inches charge distance. The two groups 

 are separated still farther in the «i -graphs. Figure 27. The U-branch is 

 rectilinear and extends from 0.^9 millisecond at 7 inches to 0.6 millisecond 

 at l8 inches. The L-branch exhibits a slight tendency to depart from linear- 

 ity. It extends from 0.435 millisecond at a charge distance of 12 inches to 

 about 0.56 millisecond at 32 inches, with considerable scatter in the data. 



The fact that t^, frequently referred to as the swing time of the 

 diaphragms, is not constant, but seems to increase slowly with charge dis- 

 tance in both the U and the L groups, is probably a combined result of many 

 separate effects. The major part of this variation, however, may very likely 

 be ascribed to the increase in plastic stress due to increased strain rates 

 at shorter charge distances. The result of this increase in stress would be 

 to increase the velocity of the bending wave, thereby cutting down its time 

 of transit from edge to center. 



At time (, the diaphragm appears to be almost conical in shape. The 

 departure from a cone may be attributed, in part at least, to the strain-rate 



w 

 a 



2.0 



.1.6 



1.2 



0.8 



o 0.4 



12 16 20 24 



d , Chorge Distonce in inches 



28 



32 



Figure 28 - Central Deflection Zc^ at the Time of Arrival 

 of the Bending Wave from the Edge 



Note that, at a charge distance of 6 inches, the deflection is large because the center has received 

 the blow caused by closing up of cavitation before the bending wave arrives. 



