In recent years I have done two types of experimentation relating to electromagnetism, (1) interactions of toroidal magnets I have termed "VACE's" (virtual Ampèrian current elements) and (2) the so-called Marinov motor (MM). It now appears that both these bubbles are about to burst. Here, for whatever interest, are the circumstances:
(1) VACE: Hitherto I have worked with Alnico U-magnets paired to roughly approximate the toroidal (doughnut) shape. (The ideal shape would be a thin washer with a small central hole.) I found that the U-magnet toroids interacted with many milligrams of force at separations somewhat greater than their dimensions, with force varying as about the inverse 5th-6th power of separation, and with angular dependence simulating Ampèrian current elements, as I had expected. (This is not quite the same as dipole-dipole angular interaction, though similar.) However, my measurements were only of the component of force parallel to the line between the VACE centers. As Wesley, Graneau, and others have noted, in addition to the forces parallel to the inter-element line there are also torques exerted between VACE's due to their finite lateral dimensions. These were ignored by my measurements and publications on the VACE subject. Because separations are comparable with lateral dimensions, these torques are strong and can have a big effect. R. W. Gray, at his web site http://www.rwgrayprojects.com/energy/VACE/vace01.html describes a very ingenious "paradox" whereby a perpetually turning motor can be built from VACE's if my Ampèrian law of angular interaction fully describes their interactions. To make sure, I built one of the motors he describes, using Edmund stock no. 38107 Alnico U-magnets. Sure enough, the simulated Ampèrian forces act to start rotation, but this is immediately arrested by strong counter-forces due to the neglected torques mentioned above. How can these torques be directly verified? One way is with an analytic balance: Put a test VACE on the balance pan, as in my reported measurements, and then put the force-exerter VACE not above it, as I had it, but to the side, level with the test VACE. That way the (horizontal) inter-element line is perpendicular to the (vertical) line of measured force component action. If the two elements were infinitesimal, as Ampère assumed, Newton's third law would predict zero force. But because of the finite extent of the actual VACE's their various parts have strongly separation-dependent interactions, with resulting non-zero force. I have not done such measurements, but have done a much simpler equivalent: I made a horizontal 4-inch diameter Lucite disk supported on a central ball bearing race, so that it rotated very easily. Near the disk edge I placed one or more of the U-magnet VACE's, so oriented that each simulated a current element pointing azimuthally, say, clockwise (simulating Gray's motor). Opposite one of these, I brought up by hand another VACE pointing radially. Disk rotation through a short distance to a new position of equilibrium was observed. This simulated the balance experiment just discussed, and showed that a net non-zero force is exerted perpendicular to the inter-element line. This is not a violation of Newton's third law, because it results from the finite lateral extent of the VACE's, coupled to strong interactions of their closest portions. Conclusion: Nature has contrived a more complicated interaction of real VACE's, including torques not considered in my previous analyses, of just such a character as to refute Gray's paradox of perpetual motion. His VACE motor does not work, and I hope nobody else will waste time and money to verify this.
A worse disappointment ensued: The whole VACE idea may prove to be an artifact of flux leakage from U-magnet pairs. This conclusion is suggested by the following facts: With the U-magnets the forces observed at separations comparable to their size are of the order of hundreds of milligrams. But I have just received from Charles M. Andrews of Arnold Engineering Co. the very generous gift of several Alnico 5 (?) ring magnets about an inch in diameter circularly magnetized. These he had in stock ... I thought no such existed, but I was wrong. I find that there is very little interaction between these magnets unless they are brought into physical contact. At separation distances comparable with their size the interaction forces are of the order of 1-10 milligrams at most. Of course, there is no perfection of circular magnetization ... the magnetizing current might be a bit off-center. So, it is quite plausible that again these small residual forces are due to flux leakage, and they just happen (more or less accidentally, so to speak) to resemble the angular dependence of the Ampere inter-current-element force law, as far as actions along the inter-element line are concerned. Thus the magnetic interaction physics of VACE's may be wholly conventional - the Lorentz force law - after all.
(2) MM: So far only two investigators have successfully built a "direct MM," that is, a motor in which the central magnetic toroid is fixed in the lab and the surrounding ring rotates continuously. J. D. Kooistra has built and repeatedly demonstrated a working model, and Cyril Smith in England has reported actual torque measurements in rough agreement with a theory based on an unconventional force law F = -qdA/dt, which has recommended itself (to me on theoretical grounds and also to J. P. Wesley) as indicating physical significance of the vector potential A. The conventional Lorentz force law predicts no ring rotation. Many other people, including myself, have built "inverse MM's" in which the ring is stationary and the inner torus rotates up to 90 degrees to a new equilibrium position (so that with current commutation the rotation becomes continuous). But that version appears possibly explicable by the Lorentz force law, so it does not count (except as a new motor type of possible practical interest). Now comes the bubble-bursting part: Kholmetskii in Minsk and Valverde in Argentina have independently tried to build direct MM's and have failed. I know little of the Minsk work, but have the following data from Valverde: A large, toroidal permanent magnet of 10000 gauss (1 Tesla) was surrounded by a conducting ring carrying currents up to 100 amps DC. The electrical contacts allowing rotation were, I believe, mercury. But no rotation occurred. This is very life-threatening to the whole idea, because 100 amps should have moved mountains. At the moment the MM is under a cloud. The topic is a puzzle.
T. E. Phipps, Jr.
908 South Busey Avenue
Urbana, IL 61801
24 July, 1999