Abstract :
The energy liberated when an arc is drawn in an oil circuitbreaker has been considered from several points of view in two earlier papers.* In the present paper the manner of dissipation of this energy is dealt with in considerable detail for various currents and for 5 500 volts (R.M.S.). The energy is considered under the following heads:?(1) energy dissipated at the contact surfaces, (2) that radiated from the arc, (3) that required to heat, vaporize and break up the oil, (4) that used in raising the gas to the arc temperature, and (5) that used in dissociating the hydrogen present. The energy under heads (1) and (2) has been found approximately by experiment. That under head (3) has been calculated. From the energy remaining to be accounted for, the three terms involved in (4) and (5) have been calculated and a value has been obtained for the temperature of the arc in the calculation. The temperature so calculated, i.e. about 3500? K., is not in accord with ideas hitherto prevalent as to the temperature of such arcs [e.g. the 1 750? (? C. ?) given by Dr. Kesselring]. It is, however, the only estimate available to the author´s knowledge with any claims to reliability, it does not contradict any available definite evidence, and indeed it receives support from what is available. This temperature is practically independent of the current in the arc over the range of current investigated, viz. 270 to 2 000 amperes. Taking the mean value found for the temperature of the region in which the phenomena described occur?i.e. 3 000? K. ?we can now calculate all the separate losses for any test from a knowledge of the volume and constitution of the gases liberated, together with the arc duration, arc length and mean current, and make it possible for the first time to set out a´ balance sheet of all the energy involved. This is carried out for a representative or ?sample? test in the open-tank circuit-breaker in Section (6), where it will be seen that the balance is accurate - to within about 5 per cent, thus checking that the constants involved are mutually consistent. The energy calculated for about 80 individual tests in the opentank switch from the mean values of the constants involved, including the arc temperature, agrees with the observed values with a maximum deviation of about ?40 per cent, the mean percentage deviations being only 12 and 13 per cent, respectively, for the two assumptions made as to the state of dissociation of the gases at the arc. This must be regarded as highly satisfactory, considering the wide range of variations generally found in pursuing quantitative measurements on the arc. It is well known that under identical external conditions wide variations of performance are observed, even in repeat´tests, due ito fortuitous) variations in1 the.form and activity of the arc itself, of the moving oil, etc., and in each particular case these must affect most or all of the quantities which appear on the balance sheet. The effect of change of conditions on the energy distribution has been considered for (a) open-top tank and copper contacts, (b) open-top tank and aluminium contacts, (c) closed tank and copper contacts. A discussion and criticism of work on this subject by other investigators are also given.
