A large framework was constructed on which the necessary 

 equipment was mounted (fig. 10). The equip:m.ent included 

 two Eastman Type 3 high-speed cameras mounted to 

 photograph the bubbles from two different angles, eight 

 1000-watt lamps, the underwater spark capsule, and pneu- 

 matic source. To accomodate the heavy currents required 

 for the lights, welding cables were used for conductors. 

 All the components were enclosed in watertight housings. 



Spark Capsule 



Since it is impractical to run very heavy currents, such 

 as those generated in the underwater spark, down long cables, 

 the entire spark sound source was built into a discarded air 

 flask from a torpedo (fig. 11). The flask was cut in half and 

 a flange welded at the center. It contained a 3000-watt- 

 second energy storage system and 15, 000-volt dc power 

 supply. Vacuum relays, such as previously described, were 

 used for keying and safety discharge. Figure 12 is an inside 

 view of the capsule. Note that the spark gaps were mounted 

 directly at the bottom of the capsule to keep the lead length 

 to a minimum. A small, multiple conductor cable connected 

 the capsule with the surface and supplied 110-volt, 60-c/s, ac 

 to the unit. A single pair of wires triggered the source from 

 the surface. This arrangement worked well and would be 

 suitable for lowering the equipment from a ship or helicopter. 

 A switch inside the camera would automatically synchronize 

 the firing with the camera. The spark capsule demonstrated 

 that an intense sound source can be made into a small unit. 

 For use aboard a submarine, the capsule could be attached 

 on the deck, outside the hull, and thus conserve space inside. 



Analysis of the Underwater Spark Signals 



In the tests reported here, the underwater spark dis- 

 charge data were obtained by means of three separate 

 methods for later analysis in the laboratory. One was 

 high-speed photography of the bubble formation. Speeds 

 of approximately 2 800 frames per second were used. In 

 addition, the signal, received on a calibrated hydrophone 

 separated 20 meters from the source, was recorded on a 

 Magnecord Model PT6J tape recorder. Tape speeds of 



26 



