Heat Dissipation at the Electrodes of a Short Electric Arc 



By L. H. GERMER 



Platinum contacts are brought together 60 times a second, discharging on each 

 closure a condenser of 0.01 mf capacity charged to 40 volts. The heat flowing 

 along each electrode is calculated from a temperature diiTerence measured by 

 thermocouples, and from this is determined the energy dissipated at each con- 

 tact. If there is no arc on closure, the energy is the same on the two contacts, 

 and is small. If there is an arc between the contacts before they touch, about 

 58 per cent of its energy is dissipated upon the anode and about 42 per cent 

 upon the cathode. The distribution is the same in an arc between clean "inac- 

 tive" electrodes and in the entirely different kind of arc occurring between car- 

 bonized "active" electrodes. This information may be significant in developing 

 an understanding of closure arcs which are the sole cause of the erosion of elec- 

 trical contacts on closure. 



THIS paper is an account of direct calorimetric measurements of the 

 energy dissipated at positive and negative electrodes when they are 

 brought together to discharge a condenser. The experiments are called for 

 by the fact that the erosion at the closure of electrical contacts is due to 

 arcing/ and understanding how the energy of a closure arc is distributed 

 between the electrodes is likely to help in developing a comprehensive 

 theory of this arc which in turn may aid in the control of contact erosion. 

 The experimental method is an adaptation to the present problem of a 

 procedure^ used earlier in which crossed wires are separated and brought 

 together 60 times per second by means of a magnetic loudspeaker unit, 

 each closure discharging a condenser which is recharged after the wires 

 have been separated. For the present experiments the two wires are made of 

 platinum and are rather heavy, and the flow of heat in each of them is 

 measured by a pair of thermocouples. There is a known length of wire 

 between the two thermocouples of each pair which are connected in series 

 to oppose each other, so that a galvanometer in either circuit will give a 

 deflection proportional to the difference in temperature across the wire.'* 

 The flow of heat along each wire is calculated from this temperature dif- 

 ference and the thermal conductivity and dimensions of the wire. After 

 making some corrections this gives the amount of energy dissipated upon 

 the electrode at each discharge of the condenser. 



The two platinum test wires have diameters of 0.0635 cm and each is 

 about 2.2 cm long from its end to the point where it is clamped in a very 



iL. H. Germer, //. App. Phys. 22, 955 (1951). 



2 J. J. Lander and L. H. Germer, //. App. Phys. 19, 910 (1948), pp. 918-919. 



3 This is the experimental arrangement used by J. J. Lander in measuring heat flow in 

 his determinations of Thomson coefficients. Phys. Rev. 74, 479 (1948), Fig. 3. 



933 



