Discussion on “limits of injurious sparking in direct-current commutation”

Mr. Reid's paper gives us a theory to account for the phenomenon which I think is the best we have so far although to my mind it does not explain all of the facts that we observe. The theory that the blackening of the commutator bars is due to volatilization of the copper is confirmed by the following fact: when the sparking has not been too severe there is a very delicate layer of copper deposited on the under side of the brush which layer I have observed may be completely removed from one brush and deposited upon a brush of opopsite polarity if, for instance in the case of a generator, the polarity of the generator be reversed. The fact that this layer of copper changes its brush when the current changes its direction, implies that the action is electrolytic, and we know could be electrolytic only if the copper were volatilized. I cannot account for sufficient energy in the contact of the brush with the copper bar to produce this volatilization in the manner Mr. Reid points out. I have run a carbon brush at enormous densities on plain brass and copper rings and have not had such deposits. I think the deposit is the result of some change in the brush after the machine has started running. It is an old maxim with me that all new brushes are good brushes. Whenever new brushes are put upon machines and freshly sandpapered, they work well. Blackening of the commutator, which occurs after the machine has been running some time, I have always attributed to excessive action commencing at the edge of the brush and causing an eating back of its contact surface. The depth affected may be only 0.01 of an inch, but the eaten or burned area travels back under the brush until it gets one-eighth or one-quarter inch back and then the forces that cause it balance with other forces and it stops. Evidence of this is seen in many brushes which, taken out of the machine, show a perfectly polished surface at the entering edge and a rather, dull surface at the leaving edge, these two surfaces being sharply separated from each other by exactly straight lines parallel with the mica. These peculiar markings are also found on brushes, sometimes in several degrees. There will be a bright surface, then a dull surface, then a surface of different texture. I therefore think that these surfaces may indicate that the brush is not making such good contact at one place as at another place. I do not mean it is out of contact, but the pressure of contact may be very much lighter at some than at other areas. Such lighter contact might account for sufficient energy to volatilize the copper at that point. The reason why I think this is so is that experiments have recently been made with brushes of greatly increased contact resistance. A brush of this kind has been devised by Mr. F. W. Young, and when these brushes have been used on machines which previously performed as Mr. Reid has described, the phenomenon he observed either in the neutral or the reverse field these strike one as eminently the only factors of importance. On page 318 Mr. Reid also brings in the effect of the distorting and weakening of the reversal flux by the armature ampere turns. In this he appears to bring in a needless complication; in other words, sparking is the result of too high a voltage between two consecutive bars. This voltage can be caused by only one thing; that is, the coil cutting magnetic lines. These magnetic lines constitute the field. This field is a resultant of two fields, one due to the field winding, and the other to the armature winding. The sparking of the machine at any load may be considered in two parts. In the first place the sparking due to the external field which he has already taken into consideration under the title of the reversal electromotive force and in the second place the sparking due to the field set up by the armature. In other words, this latter is due to the voltage set up in the armature coil by the armature coil cutting the lines which have been set up by the current flowing through the armature coils. This, however, is what is meant by self-induction. That is, the quantities inserted in the fundamental equation, pages 299 and 319, are entirely sufficient. Stated differently, the stability factor or ratio between armature ampere turns and field turns for the air-gap is not a factor of first importance. What must be done in designing machines is either to keep down what is called the self-induction or reactive voltage, or to provide a reversing field sufficient to render it innocuous.