Communications - Tesla and the Induction Motor

To the Editor:

I read with interest Ronald Kline’s article in the April 1987 issue of Technology and Culture and found his account of the parallel development of the induction motor and its “engineering science” well written and informative, for the most part. However, I noticed certain technical misconceptions that appear to have led to inaccurate historical conclusions. Specifically, Kline misrepresents Tesla’s role in the development of the induction motor and its theory, and this misrepresentation appears to be based on an incorrect analysis of the motors described in Tesla’s U.S. patent no. 381,968 (referred to hereafter as ’968).

Contrary to Kline’s assertion on page 290 (and repeated on page 299) that ’968 was Tesla’s “first patent for the induction motor,” the motors disclosed in that patent are not induction motors at all. Induction motors depend for their operation on the interaction between an electromagnetic field produced by currents fed to the motor’s “primary” coils from an outside source and an electromagnetic field produced by currents induced in “secondary” coils or conductors by the rotation (“slip”) of the primary field relative thereto. The motors in ’968 clearly do not depend for their operation on such induced currents; they have no operative secondary coils or conductors. In fact, the elements of these motors which do interact with the generator-fed coils are designed to minimize induced currents, since these would merely dissipate power in the form of heat. Moreover, there is no slip between their generator-fed electromagnetic fields and any interacting motor element. The motors are, in fact, synchronous motors - motors whose rotors revolve at the same speed as their revolving electromagnetic fields.

Specifically, the motors of figures 9, 13, and 15 are either “hysteresis” or “reluctance” types, depending on which of the described, alternate shapes is chosen for their rotors. In either case, they depend for their operation on the condition that the rotating electromagnetic fields of their stator coils induce magnetic poles (not electric currents) in their ferromagnetic rotors. The motor of figure 10 is simply an inside-out version of the motors of figures 9 and 15. Here the coils rotate while the interacting ferromagnetic material remains stationary. Finally, the motor of figure 17, though of neither the hysteresis nor the reluctance type, is also a synchronous motor. It has sets of coils on both the rotor and stator, with both sets being energized directly from the same generator source. Here again, no induced currents contribute to the operation of the motor.

A more significant error, in consideration of Kline’s overall argument, is his assertion that the motor of figure 10 “worked exactly like the one Ferraris built” (p. 290). The Ferraris motor illustrated in figure 3 of Kline’s article is neither a hysteresis motor nor a reluctance motor. In fact, it is not even a synchronous motor. The difference between the Ferraris motor and the Tesla motor can best be seen in Tesla’s motor of figure 13 of ’968. Both motors have salient poles (the difference in their number is due merely to Tesla’s choice of three-phase energization as opposed to Ferraris’s two-phase), but Ferraris’s rotor is made of copper, a highly conductive, nonmagnetic material, whereas Tesla’s is made of iron or steel, a highly magnetic but less conductive material made even less conductive by being laminated. Thus, the rotating electromagnetic field of Ferraris’s motor cannot possibly induce ferromagnetic poles in his rotor as it does in Tesla’s; but it does readily induce an electric current that, in turn, establishes the rotor’s electromagnetic field - which it cannot do in Tesla’s. The strength of the induced current is proportional to the slip between the field and the rotor and would, of course, disappear entirely if the rotor and the field were forced to rotate at the same speed. Hence, Tesla’s ’968 patent is not “evidence” that “Tesla thought that Arago’s disk worked on the principle of magnetic attraction rather than eddy currents.”

The Ferraris motor is, in fact, a true induction motor and much more nearly resembles the two-phase induction motor shown in figure 3 of the second Tesla patent referred to by Kline, U.S. patent no. 382,279. But Tesla mounts his closed-circuited copper conductors on an iron core, while there appears to be no iron in Ferraris’s rotor. Because of the high permeability of iron relative to air, Tesla correctly states that his core greatly adds to the magnetic effect produced by his induced currents. Thus, here again, Tesla’s patent fails to show that this improvement over Ferraris “resulted, ironically, from his misunderstanding of Arago’s rotations.”

Kline’s point regarding Tesla seems to be that he did not contribute much to “engineering science” because he misunderstood the science involved. Even if the first part of that premise is correct (the article does not explicitly address this issue), the second part is certainly not supported by Kline’s argument. It does demonstrate, however, the importance to the history of technology as history of technology (rather than as, say, cultural or economic history) of getting the technology right.

KENDALL J. DOOD

Mr. Dood is a patent classifier at the U.S. Patent and Trademark Office in Crystal City, Virginia.

To the Editor:

I was most interested in Ronald Kline’s fine, thoroughly researched article. It makes a significant contribution by identifying the migration of thought from the concept of producing an electrically revolving field and combining this knowledge with that of Arago’s rotations, which together led to the invention of the induction motor. An important distinction is also made in identifying Tesla’s “second” induction motor patent (382,279, actually his third patent utilizing the rotating magnetic field principle and covering induced-current actions in windings) as “the standard form,” the conventional, customarily understood induction motor in universal use today. However, the recorded application date for the first Tesla revolving field patent (381,968, covering eddy-current induction actions) preceded the third by only seven weeks. It would seem that this is much too short a time, lacking other supportive information, to infer a developmental progression from the eddy-current actions in the first patent and the induced-current actions in windings in the third patent. That is, in all probability such conceptual evolution must have occurred over a more extended period before the application dates for both patents.

I think there is some confusion in Kline’s article with respect to the distinctly different environments for the rotating magnetic fields described and demonstrated by Ferraris and Tesla (in his third induction motor patent cited above, including those subsequently based on it) and the conceived utility of those effects. The extension of eddy-current induction to linear conductors (i.e., induction machines having closed-circuit windings) seems to me a strained explanation. I find that eddy-current effects are almost exclusively reserved for the treatment of continuous, extended, or bulk conductors as opposed to linear conductors.

Let me introduce at this point a fundamental definition for any magnetic motor operation: Torque on the rotor results by virtue of the interaction of magnetic fields, both attractive and repulsive. between the rotor and stator. Kline’s remark, “Tesla did recognize the value of induced currents in his second [rather, third] patent [382,279], but he again interpreted their benefit in terms of magnetic attraction” (pp. 290-91), is therefore not clear. Inasmuch as rotative effort results only from field forces of attraction (and repulsion), I am at a loss how to interpret it. It seems to imply that Tesla’s understanding of the mechanism accounting for rotational torque was incorrect.

The most critical consequences of the confusion I have referred to appear on page 289, last two lines: “This two-phase motor worked on the same principles as the one Ferraris built,” and on page 291, “Tesla said that he added these windings to increase the magnetization of the secondary’s iron core rather than to provide a better path for eddy currents. Hence, Tesla’s major improvement over Ferraris - the use of closed-circuited windings - resulted, ironically, from his misunderstanding of Arago’s rotations.” I do not find Kline’s observation, “Tesla said that he added these windings [with ends connected together]...” (my emphasis), to be supported by any of the three references cited in note 17, either explicitly or by inference.

It seems quite probable, however, that Tesla had earlier recognized the two different induction motor actions - the eddy-current induction in unwound disk/drum rotors and stators, and the induced-current actions in windings - merely by virtue of the near-simultaneous patent applications. Kline’s observation that “Tesla’s motor was more powerful and efficient because its secondary windings channeled the eddy currents into paths where they interacted more strongly with the revolving magnetic field” (pp. 289-90) is not supported by the reference (n. 14). The method of induction or its characteristic is not described in the reference, a conclusion drawn in Kline’s thesis.

If one chooses to use eddy currents as an explanation for the induced current in an induction motor, then engineering design from that perspective would dictate isolating the high-permeability/lower-conductivity magnetic core from the closed, higher-conductivity/low-permeability conductor surrounding it. This was what Tesla accomplished that was new by his third induction motor patent. His was not merely a “more powerful” Ferraris device, it was a specifically different arrangement.

In considering Tesla’s first patent (381,968), Kline says that “he described a ... permanent-magnet secondary” (p. 290). Nowhere in the subject patent is a “permanent magnet” secondary (stator) described. This member is identified in the patent as either an “annulus, R,” an “iron shell, R',” or a “field-magnet...ring, R.” If these stators were indeed permanent magnets, the steel (or soft iron) would be in saturation, and any induced-eddy currents would give rise to inappreciable and, for all practical purposes, ineffective Δϕ.

The term “eddy current” is rarely used today outside of an association with such manifestations as the production of heat (deliberate) or core losses (unavoidable), Foucault damping forces, “skin effect,” and nonmagnetic transducers. Eddy currents were, in early literature, also called “parasitical currents,” being viewed as producing no useful effects. I believe eddy currents are quite analogous to the phenomenon of parasitic oscillations in tank circuits (undesired, self-sustaining oscillations at frequencies other than the operating frequency). Eddy currents are, of course, induced not only in the magnetic rotor and stator of motors but also in the massive copper bars of large machines, resulting in a localized heating effect in these conductors rather than the principal inductive current of the secondary circuit suggested in Kline’s article.

Kline remarks that “The century-old priority dispute about who invented the [induction] motor is unresolved and will likely remain so.” I thought that every drop of blood had been spilled in that contest and in analysis of it decades ago! At the Muzej Nikole Tesle in Belgrade there are exhibit panels showing drawings of two- and three-phase AC motors, one drawing dated March 10, 1884. These early drawings antedate any schematic by Ferraris and refute the assertion by Giovanni Silva of unsupported backdating of Tesla’s invention appearing in “Galileo Ferraris, the Rotating Magnetic Field and the Asynchronous Motor,” L’Elettrotecnica 34 (September 1947). Other documentation in the Tesla museum could undoubtedly reflect interestingly on these conceptual evolution and priority issues.

Leland I. Anderson

Mr. Anderson is currently a systems analyst with the U.S. Department of the Interior, Bureau of Land Management, Denver.