the electrical terminal leads and the insufficent bonding 

 of the compound to the adjoining wall of the end cap cavity. 

 Water under high pressure (2,600 psig) entered the housing 

 through the cable armor wires, ruptured the bond between the 

 sealing compound and the end-cap cavity wallf and extruded 

 the compound into and around the phenolic plate mentioned 

 earlier. 



Since no evidence of failure of the mechanical cable 

 lock in the housing was observed (it had been subjected to 

 a tensile load, under water, of approximately 7,000 pounds), 

 it was decided that a test should be run to compare it with 

 a poured epoxy fitting of the type used successfully by the 

 David Taylor Model Basin. A poured fitting of this type 

 was designed in the form of a clevis for attachment in a ten- 

 sile testing machine, and a comparative test to destruction 

 was run. Failure occurred at a tensile load of 17^300 pounds, 

 The rated breaking strength of the cable was 18,000 pounds. 

 The failure occiirred just inside the mechanical lock, approxi- 

 mately 5/8 inch before the first bend of the wires into the 

 spool. Three wires were broken; examination revealed a 

 "necking down" indicative of tensile failiare in the wires. 



After this housing failed to maintain a watertight in- 

 strument compartment, a "test" housing was designed speci- 

 fically to test the sealing characteristics of ccrunercially 

 available watertight electrical connectors. 



1.21 



