SOLDERLESS WRAPPED CONNECTIONS — PART III 593 



lished should have a wide apphcation in many areas. In telephone prac- 

 tice they provide a reasonable latitude for variations in the process of 

 making the connection, including tolerances in the parts, and at the 

 same time guarantee a good product. 



Most of the product tests which have been made so far apply specifi- 

 cally to connections which use Standard No. 24 tinned solid copper 

 wire such as is used in 95 per cent of the telephone switching plant, and 

 terminals having a rectangular cross-section punched from sheet nickel 

 silver, brass or bronze and whose dimensions are one-sixteenth inch wide 

 by the thickness of the stock (0.013" to 0.062''). These terminals are 

 typical of those which are or could be used on relays, switches, resistors, 

 capacitors, terminal strips, etc. Studies with #20 and #22 wire have 

 been made with results similar to those with # 24 wire. 



A similar investigation is under way on connections to terminals 

 which are made from round silicon copper and nickel silver wire such 

 as are used on the wire spring relay, and where the wire terminal had 

 been prepared for connection by various treatments such as flattening, 

 coining, serrating, annealing, etc. 



AGING OF WRAPPED CONNECTIONS 



Assuming that a connection can be wrapped with sufficient mechan- 

 ical strength to withstand handling, vibration, etc., there appear to be 

 two factors which might cause the connection to fail after a period of 

 time. These factors are (1) relaxation of the internal stresses in the 

 metals, and (2) corrosion of the metal surfaces. As Mr. Mason points 

 out in his paper, it now appears that solid state diffusion of metal across 

 the boundary between the wire and the terminal improves the con- 

 nection as much or more than relaxation degrades it. Tests have been 

 designed to relax the connections in a short time to the same degree 

 that will occur in the normal forty-year life which is required. These 

 tests will be described later. 



Several investigators have studied the rate of surface corrosion in 

 indoor atmospheres of metals such as are used in these connections. 



Studies of corrosion* where oxidation is the primary factor indicate 

 for example that the corrosion in zinc varies linearly with time and on 

 copper it varies as the square root of the time exposed. The corrosion 

 products are ZnO and CU2O. The corrosion rate for brasses fall between 



* British Non-ferrous Metals Research Association, Investigation on the 

 Atmospheric Corrosion of Non-ferrous Metals, First and Second Experimental 

 Reports to the Atmospheric Corrosion Resistance Committee, May, 1924, and 

 May, 1927, W. H. J. Vernon. 



