DESIGN FACTORS INFLUENCING RELIAIULITY OF RELAYS 989 



sliding or rockiiio; on it and the tondoncy to lock is thus largely avoided. 

 The likelihood of locking is further decreased because the restoring force 

 is greater and is applied closer to the contact, and because of the impact 

 of the card on the spring when the relay releases. With the contact closed 

 there is clearance lietween the contact spring and the bottom of the slot 

 in the cai-d, and thus when the card hits the spring in opening, it is al- 

 ready mo\-ing and has acquired kinetic energy. This energy is availa])l(! 

 on impact to oN'ercome any locks which may have occurred. 



Kecent studies have shown that erosion of electrical contacts on clo- 

 sure is due almost entirely to an arc occurring, in most cases, before the 

 contacts touch. ^ When there is no arc there is no erosion. It has also 

 been obser\cd that the occurrence of an arc between the approaching 

 contacts is influencetl by operation in the presence of various organic 

 vapors; for example, benzene derivatives.'^ The effect of such operation 

 is to permit arcing on lower currents than is the case with clean contacts 

 and results in increased erosion rates. This is true for noble metal con- 

 tacts, and when so exposed they are said to have become "activated". 

 A metal surface which has been activated by organic vapor remains 

 active indefinitely if there is no arcing at the surfaces. With continued 

 operation and accompanying arcing, the activating material is burned 

 away, and the surface returns to the inactive condition, provided no 

 contaminating vapor is present. 



Some materials used in relays may give off organic vapors which can 

 aggravate the arcing at the contacts. A series of experiments has been 

 made by placing various materials under test in a small glass enclosure 

 and proceeding to find if, and at what elevated temperature, vapor from 

 the materials will give rise to arcing on "make", with contacts that are 

 operating within the enclosure. The materials tested varied widely in 

 their effects upon arcing at the relay contacts. In the solid organic group, 

 they ranged from polystyrene, which produced arcing at room tempera- 

 ture, to teflon, which did not cause arcing until heated abo^'e 200°C. 



The precise correlation bet^^•een the results of these tests and the 

 changes in erosion rates, which occur when these materials are used in 

 the relay construction, has not yet been established. However, they may 

 be used as an aid in the choice of such materials. Cases have come to our 

 attention both in the laboratory and the field where the erosion rates of 

 relay contacts operating in confined chambers were many fold those 

 which occurred when the relays wei-e operated in the open. This was at 

 least partially ascribed to the presence of contaminating materials known 

 to be present. 



Another type of failure that is quite generally experienced in relay 



