545 



appears to be effected, the i>ath difference between direct and reflected 

 waves, divided by the velocity of the shock-wave (ca. the velocity of 

 sound in water), gives the time delay in the arrivsuL of the reflected 

 wave, and thus (if we disregard deformation of the diaphragm caused by 

 its own inertia) the time required for normal deformation. Comparison 

 of the deformation time with the pressure time c\irve of the shock-wave 

 will show what part of the shock-wave caused the damage. It will be 

 noted that no accoiint has been taken of the contribution of bubble pulse 

 waves to the deformation, but other experiments (Section V, k,c) have 

 shown that in norma] use of the gauge the contributions of bubble pulses 

 were negligible. 



The results of these tests are listed in Table VIII, and seme of the 

 results are shown graphicsLLly in Figures 15 and l6. The deformation 

 times range frOTi 2l»0 to 36O microseconds. 



(b) Double shots. According to Kirkwood, curved diaphragms with 

 the convex side toward the charge should be deformed more than those 

 exposed with the concave side toward the charge. To determine the 

 relative capacities of convex and concave diaphragms to withstand de- 

 formation, the following experiments were carried out. The first ex- 

 periment consisted of reversing a damaged diaphragm in the gauge and 

 firing a second charge identical with the first. Ihat is, for the 

 second shot the dent of the damaged diaphragm was bulged out toward 

 the charge. The resiilt of the second shot was a deformation in the 

 opposite direction in which the maximum depression (measured from the 

 original undeformed plane surface as reference point) was greater by 

 25 to hCf. The dent from the second shot had an asymnetriceLL contour, 

 having the greatest depression two-thirds of the way toward the side of 

 the bulge nesfi" the bottom of the gauge. 



In the second experiment, diaphragms were exposed to the explosions 

 of two Similar charges in succession without removing the diaphragms fr(»n 

 the gaiiges. It was fotind that the damage was Increased by the second shot 

 but not to the same extent as on the reversed diaphragms discussed above. 

 Table IX lists the data. 



(c) Study of errors. A series of experiments were conducted in 

 which we investigated the various sources of errors in our use of the 

 UERL damage gauges. This was done by making various deliberate changes 

 of the sort that might occtir accidentally in the experimental set-up and 

 then determining the extent that the damage was affected by this change. 

 Charges of approximately k lbs. loose tetryl (density about 1 gm/cm3) 

 were used so that no initiator other than a No. 8 DuPont cap would be 

 necessary. Charge-to-ga\ige distemce was h8 in. for these tests, unless 

 otherwise specified. 



(i) Due to variation of charge-to-gauge distance . When the gauges 

 were displaced with respect to the charge, the change in damage was about 

 3^ per inch of displacement. Hence, if the charge should shift 1 in. 

 toward one gauge, that gauge would have 6^ more damage than the one 

 opposite it. The damage averaged over the two opposite gauges should 



