544 



cases in which two £ind three gaviges were used, the gauges were spaced 

 at 90° intervals. In all of these cases, then, the shock-wave would 

 tend to move the whole ring, and the damage could be compared to 

 damages obtained in the usual case in which foxir gauges were mounted 

 at 90° intervals about the ring so that the latter had no tendency to 

 move. No important differences were found. 



In order to obtain evidence as to the effect of the ring itself 

 on damage, two types of shots were fired and compared with the usual 

 shots. In the first type, sheet iron was wired to the ring. The 

 sheet iron was 6-1/2 in. high and extended ell arovmd the outside of 

 the ring. In the second type, the gauges were mounted on rubber shock 

 mounts about 3/8 in. thick between the gauges and the ring. This 

 experiment was to test the effect of the rigidity of the mounting on 

 damage. Neither the sheet iron nor the shock mounts were fotmd to have 

 any significant effect on damage, so it would appear that the reflections 

 from the ring and the rigidity of the ring make no appreciable contri- 

 butions to damage. 



Two shots were also fired using Lot 3 steel diaphragms, 3«78 lb. 

 TNT charges, and an inside wooden frame (Figure 11) in addition to the 

 ring. These shots were compared with two previous shots in which the 

 ring was not iised. The average ratio of the damage obtained with the 

 wooden frame and ring to the damage obtained with the wooden frame 

 alone was fovind to be I.07. Of eight such ratios, however, the spread 

 was considerable (0.95 to 1.2lt-), due perhaps to the variability in the 

 wood. This indicates that the major portion of the "frame effect" is 

 due to the presence of the wood smd not the break-up of the frame or 

 the movement of the gauges. 



3. Results Using Wooden Frames 



A considerable ntmiber of tests were conducted with damage gauges 

 before the importance of the effect of wooden frames on damage was 

 discovered. These experiments are described in the following sub- 

 divisions. It should be kept in mind that, in general, a new frame 

 was used for every shot and that, therefore, an important variable 

 was not controlled except insofar as the same kind of lumber was 

 used imtil it was used up. It is probably imsafe to compare results 

 of shots which were separated by a considerable time intervsil. Shot 

 numbers indicate the order in which the tests were conducted. 



(a) Determination of time of damage. A simple method was devised 

 for determining the approximate time during which deformation of the 

 diaphragms in UEIRL damaige gauges occurred. The method consisted of 

 firing shots in which the gauges and charge were suspended in a plane 

 parallel to the water surface and successively decreasing the depth of 

 the rig until the gauges began to record markedly less damage. The 

 decrease in damsige was caused by the reflection of a rarefaction (ten- 

 sion) wave from the water-air interface which cut off the latter part 

 of the pressure wave. At the critical depth -vrtiere the deformation first 



