584 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1953 



cient to relax all the stress that can be relaxed. On remeasuring the 

 resistance, it was found that there was no change within the experi- 

 mental error of 1 per cent, which corresponds to a resistance of 3 X 10~^ 

 ohms. A similar result is found by studying the corrosion of surfaces of 

 copper and nickel silver in solderless wrapped connections when they are 

 fully relaxed and subjected to a corrosive atmosphere as discussed 

 in the paper by Van Horn. 



The stripping force of the aluminum-aluminum connection subjected 

 to a temperature of 200°C for 12 hours has actually been found to in- 

 crease by a factor of about 2 which suggests that the two parts have 

 diffused into each other and formed a permanent connection. This has 

 been confirmed by the increase in force required to unwrap the wire. 

 Since the activation energy of self diffusion is about the same as the 

 activation energy for stress relaxation, then as the hoop stress is relaxed 

 at high temperatures, solid state diffusion takes place and a diffusion 

 joint is formed in aluminum. The same process, both for stress relaxa- 

 tion and diffusion, should take place at a much lower rate at lower tem- 

 peratures and as the hoop stress is relaxed, a diffusion connection between 

 the two parts is formed so that no decrease in the conductivity of the 

 connection occurs and an actual increase in the strength of the connec- 

 tion results. The same process should result between any two materials 

 provided the energy of seK diffusion from one into the other is less or 

 equal to the activation energy of stress relaxation for the weakest 

 component. 



It is planned to tin plate both terminals and wires for all of 

 the wi'apped solderless connections used in the telephone system. In 

 order to find how the two processes of stress relaxation and self diffusion, 

 which progress as a function of time and temperature, affect the two 

 fundamental properties electrical conductivity and mechanical strength 

 of the connections, a large number of connections were wound up and 

 subjected to temperatures of 200°C for times corresponding to 0.9, 0.8 

 etc., of the initial stress as determined from Fig. 21. The electrical re- 

 sistances of the connections were determined before and after the treat- 

 ment and the stripping force was also measured. Within the experimental 

 error the resistances of the connections remained the same while the 

 average stripping force for twenty connections for each point are shown 

 by Fig. 23. No significant change in the stripping force occurred out to 

 values of stress relaxation less than 0.2 times the initial hoop stress. 

 These experiments show that as the hoop stress is relaxed by time and 

 temperature, self diffusion occurs between the two parts of the connec- 



