ARCING OF ELIOCTIUCAL CONTACTS IN TELIOI'IIONIO SWITCHING CIRCUITS 549 



From a oroup of transients similar to Fij^. 9, obtained at different con- 

 ditions, the plot in Fig. 11 was made. It indicates that in general, the 

 ratio (ra,)d2/n/(F„,).s/„( starts decreasing at about //,/, = 2 X 10'' cms/sec 

 and at 2 X 10 the arc initiation voltage is only 50 per cent of the corres- 

 ponding static value. As shown in the iiguiv a deionizing velocity of lO" 

 cms/sec is just about two orders of mtigiiitude too high to account for 

 this phenomcMion. It should be added, howi^ver, that while all the ions 

 have cleai'ed the gap, it has been proposed" that the life time of an ion 

 on a surface (ihn can be long enough to enhance the initiation of the 

 next arc. If this mechanism is accepted, our data would indicate that 

 the life time of the ions was only of the order of 10"'^ second. 



b. Residual Atonis 



After an arc, the contact gap contains some metal atoms e\'aporated 

 from the electrodes by the arc. These atoms will clear the gap by travel- 

 ling to and condensing on the electrodes and by lateral diffusion. A crude 

 approximation is given here of the time of recollection of the atoms on 

 the electrodes based on their initial momentum. 



One may visualize the arc spot on an electrode to have a temperature 

 distribution extending from submelting temperatures to a range of 

 boiling temperatures, corresponding to the arc pressures. The lowest 

 temperature is probably the normal l)oiling temperature of the contact 

 metal. At the termination of the arc, the metal atoms produced at the 

 lowest l)oiling temperature are the slowest and last to recondense on the 



1.0 

 0.9 



UcH = S/tcH IN CM PER SEC 



Fig. 11 — Apparent relation l)et\veon arc initiation voltage and velocity of 

 charging. E = 50 volts, L = 0.010 henry, A' = 40 ohms and C as indicated for Pd 

 contacts in atmospheric air. 



