Babcock-Paul: Mathematician review (Step 2)

Reviewer role: PhD mathematician (derivations, notation, units, hidden assumptions). Source: US 8,860,273 B2, "External Field Interaction Motor" (Babcock, Babcock, Smith; Flyback Energy, Inc.).

Source gap (repeated from Step 1): the PDF has no text layer. I rendered pages 10-17 at 200 dpi, OCR'd them with tesseract and read the description and all 20 claims [read: page image p.10-17, OCR]. Page 1 was checked as an image. The seven drawing sheets (pp. 2-9) were not studied, so the circuit of FIG. 12 is judged from its verbal description only. The patent contains NO measurements, efficiency figures, torque curves or equations of any kind [read]. There is therefore nothing quantitative to check for internal consistency, only qualitative physical statements. Anything about Flyback Energy beyond the patent is not used in any verdict.

Web verification: no verdict below depends on a medium- or low-confidence [memory] fact. The physics used (Maxwell stress, virtual work, Faraday/Lenz, E = 1/2 L I^2) is textbook, tagged [memory] high confidence. I did not run web checks and say so plainly rather than inventing URLs.

Corrections to Step 1

1. claims.md says the claim "Stopping current ... greatly reduces counter-electromotive force". The patent says "substantially reduced" and, in the claims, only that current is disabled "to reduce" CEMF (claims 10, 11) or that the overlapped coil is held "substantially zero-current" (claim 14) [read]. "Greatly" is Step 1's word, not the patent's. Minor.
2. claims.md claim 3 says flyback energy raises efficiency. The patent says captured reverse EMF "can be used to supplement the input power source to increase motor efficiency or ... to run external devices" and that it "may be used as a secondary power source" [read p.11-12]. So there are two uses, one of which (running external devices) is just a regenerative output, not an efficiency gain of the motor. Step 1 did not mention the second use.
3. facts.json key claim 4 ("substantially improved acceleration and thrust/torque at high speed compared with conventional motors") is in the patent only as "some embodiments may achieve" [read p.3]. It is a hedged advantage statement, not a claim in the 20 claims. Step 1 listed it among key claims but claims.md correctly folded it into claim 2.
4. Step 1 missed three statements that matter mathematically and are worth flagging: (i) "A motor armature that produces little or no counter-motive force can be driven to a given amount [of] torque and speed with much less electrical power" [read p.12]; (ii) the flyback circuit "produces a reduction in input impedance" [read p.10]; (iii) "The increased area that the fields interact in compensates for loss of force" [read p.11]. Each is assessed below.
5. Step 1's other facts (inventors, dates, 20 claims, no over-unity claim, no test data) agree with what I read. The patent states no output above input. Confirmed.

Background mathematics (so the checks below are self-contained)

• Force from stored magnetic energy (virtual work): F = -dW/dx at constant flux (or +dW/dx at constant current), where W is magnetic field energy and x is displacement. The patent's sentence that the armature is attracted "when the field lines interact to cancel, or weaken, the energy stored in the magnetic field" [read p.11] is this principle in words. It is correct.
• Maxwell stress: magnetic pressure on a surface is about B^2/(2 mu0), with mu0 = 4 pi x 10^-7 T m/A. Example: B = 0.5 T gives 0.25 / (2 x 1.2566e-6) = about 9.9e4 Pa. Over 10 cm^2 (1e-3 m^2) that is about 99 N, an upper bound for perfect coupling. This is the scale a permanent-magnet motor works at; nothing exotic.
• Faraday and Lenz: a conductor (coil) linking flux Phi that changes in time has emf e = -N dPhi/dt, opposing the change. In a motor this is the back-emf (CEMF).
• Energy balance of an ideal motor: electrical power converted to mechanical, ei, equals Fv (force times velocity), with e the back-emf and i the current. Input power = F*v + I^2 R losses + d(W_stored)/dt.
• Energy in an inductor: E = 1/2 L I^2.

Claim 1: Thrust from armature and stator fields interacting in arc-shaped regions outside the air gap

a. Mathematically consistent? Yes, qualitatively. A force between two magnetic field distributions is given by integrating the Maxwell stress (or the force density J x B, or for a permanent magnet F = grad(m . B)) over a region. Nothing requires that region to be a narrow gap. A force between a permanent-magnet array and a solenoid's fringe field outside its ends is ordinary magnetostatics. The patent's attraction/repulsion table (FIG. 9: opposite radial components attract, same-sign repel) is a correct qualitative rule [read p.9]. Weak point: the patent never computes the force, so "the increased area ... compensates for loss of force" [read p.11] is an assertion. The fringe field of a solenoid decays quickly away from its ends, and a solenoid with a ferrous core presents little radial flux outside its ends compared with a gap geometry. Whether the net force is competitive is a number that must come from a field calculation (finite-element) that is not given. Hidden assumption: the armature flux is strong enough in the interaction region at the same moment the stator flux is. b. Consistent with experiment? Yes. It is a variant of the segmented-stator, permanent-magnet-rotor family (stepper, brushless DC, switched-reluctance and linear synchronous motors) whose thrust is well measured. No known experiment contradicts it. Whether this particular geometry is better than a conventional one is not shown. c. Tested? The patent reports no test [read]. I found no independent test in the folder. Not verified by me on the web. No replication is on record that I can cite. d. If true, what changes? Nothing in mainstream physics. It would be an engineering geometry choice, not new physics.

Verdict: Plausible, untested (for the existence of thrust: effectively Established as ordinary magnetostatics; for "competitive with conventional geometries": Never properly tested, because no neutral comparison is on record and the patent offers none).

Claim 2: Stopping current in a stator coil while the armature overlaps it greatly reduces back-emf and so allows high thrust at high speed

a. Mathematically consistent? Split. - Part 1, true: if a coil carries no current it cannot dissipate or deliver power, and the emf e = -N dPhi/dt it sees is not driving any current in a closed circuit. In the open-circuit state the coil still has an induced voltage across its terminals (Faraday still holds); what is eliminated is current and so the opposing force on the armature and the loss. The patent says as much only loosely: "it cannot induce counter-motive force because those windings may be in open condition and electrical current cannot flow from induction" [read p.12]. This is correct as a statement about current, imprecise as a statement about emf (the induced voltage still exists, it just does no work). Step 1's "reduces CEMF" should be read as "removes the opposing current and force", not "removes the voltage". - Part 2, a genuine benefit: at high speed a conventional motor is voltage-limited because back-emf grows in proportion to speed and cancels the supply. A coil that is off when the magnets are over it is not fighting that voltage, and is energised only when its flux is changing little. This is the same idea as field-weakening and as "commutation timing" in brushless drives. So higher speed at the same supply voltage is plausible. - Part 3, not correct as written: "A motor armature that produces little or no counter-motive force can be driven to a given amount [of] torque and speed with much less electrical power" [read p.12]. Take an ideal motor: converted power ei = Fv. For a required thrust F and speed v, mechanical power Fv must be supplied by the electrical side regardless of how the back-emf is distributed. Reducing e at a given Fv means i must rise (or, if the force is made by a different coil, that coil carries the conversion). Zero current in the coil under the armature means that coil contributes no force, so other coils (those not overlapped) must produce all of it, and they have their own back-emf. The reduction is therefore of voltage headroom and of I^2 R and core losses, not of the Fv term. "Much less electrical power" can only mean less loss. The patent gives no loss budget. Worked check: F = 50 N, v = 20 m/s gives Fv = 1000 W mechanical. With an ideal 100 percent efficient machine, input is 1000 W no matter how the back-emf is arranged. If the real machine today is 90 percent efficient, the input is 1111 W; the most this idea can ever save is the 111 W of loss, about 10 percent of input, and only if it eliminated all of it. The "much less" phrase cannot exceed that. (This calculation uses assumed illustrative numbers, not patent data.) - Hidden assumption: the coil actually needs no current while overlapped because the armature magnets supply the field; the overlapped coil then contributes zero force, which is exactly the "loss of force" the patent mentions and tries to offset by area (see claim 1). b. Consistent with experiment? Back-emf proportional to speed and voltage-limited top speed are measured, universal facts of electric machines. Open-circuit (unenergised) windings carrying no current produce no drag torque from induced current: also standard. Commutation advance and sectional switching are used in commercial brushless and linear machines. No experiment contradicts the claim. c. Tested? Not by the patent [read]. Phase-timed segmented drives exist commercially, but I have no source tying them to this specific geometry. No replication is claimed. d. If true, what changes? Nothing. At most a new motor topology.

Verdict: split. Weak form (switching off overlapped coils removes drag current, raises voltage-limited top speed): Established as ordinary motor-drive physics, Plausible, untested for this specific geometry. Strong form (given torque and speed with "much less electrical power"): Refuted by calculation if read as less than F*v; if read as less than loss, Never properly tested and unquantified (not dismissed without testing; simply never tested in the document).

Claim 3: Inductive flyback energy captured in a capacitor and reused as a secondary power source, raising efficiency

a. Mathematically consistent? Qualitatively yes, with two statements that are wrong or loose. - Capture: when coil current I is interrupted, the stored energy E = 1/2 L I^2 must go somewhere; with a diode path into a capacitor it charges the capacitor (E_C = 1/2 C V^2). Described as "at a voltage above the source input voltage" [read p.10], which is consistent (energy is transferred from low-voltage, high-current to high-voltage, low-current state; in the patent's words "collapses the Reverse EMF from a high-voltage low-current state to a low-voltage high-current state", p.12, an apparent reversal of directions; either direction is possible depending on circuit, but the text is internally loose). - Worked example: L = 10 mH, I = 5 A gives E = 0.5 x 0.010 x 25 = 0.125 J per switch-off. At 1 kHz per coil that is up to 125 W cycling through the capacitor. But this 0.125 J was put into the coil by the source in the first place. Recovering it returns at most 0.125 J; net energy saved per cycle is at most the recovered energy, and in practice is recovered energy times capture efficiency (say 80 percent gives 0.1 J). It cannot exceed what was stored, so the gain over a plain freewheeling diode is bounded by the energy that a freewheeling diode would have dissipated in the coil resistance and switch. This is a standard snubber/energy-recovery idea; the patent itself cites He (APEC 2000) on soft-switching snubber energy recovery [read, per Step 1]. - "This produces a reduction in input impedance by the stator segment which makes it possible to drive the stator segment with high frequency Direct Current input pulses" [read p.10]. As stated this is not correct. Inductive reactance X = 2 pi f L is a property of the inductor and rises with frequency. A switching circuit does not change L. What it can do is allow a quicker current rise by applying a higher voltage (V = L dI/dt) and clean turn-off. "Impedance to high frequencies may be reduced or eliminated" [read p.12] is likewise unsupported by any equation. This is imprecise language, not a hidden over-unity claim. - Hidden assumption: captured energy is reused at the same cycle's efficiency; if the capacitor is charged above the source voltage, discharge back into the coil first (path A, A' described p.10), so the net draw from the source is reduced by the recovered amount. That is consistent bookkeeping. b. Consistent with experiment? Yes. Regenerative flyback recovery, snubber energy recovery and switched-reluctance drive energy return are standard, measured techniques. No experiment contradicts it. The magnitude of the efficiency gain is not given. c. Tested? Not in the patent [read]. The related patents it incorporates (US 7,602,157, "Supply Architecture for Inductive Loads"; US 2009/0073726) are cited by number only [read p.10-11]; I did not read them. No independent test in the folder. d. If true, what changes? Nothing. The claim is bounded by conservation of energy: gain at most equals the energy that would have been wasted. Nothing in the patent claims output exceeding input [read]. If a reader or marketing material claims "free" or "over-unity" from the recovered flyback, that would be refuted by the energy balance above, but that claim is not made in the patent.

Verdict: Plausible, untested as to size of gain; Established that the technique recovers some energy. Any reading of "secondary power source" as a net energy source beyond the input: Refuted by calculation (energy recovered cannot exceed energy stored, 1/2 L I^2, which the source supplied).

Claim 4: Lighter, cheaper motor (non-metallic housing, little shielding, modular coils) with high thrust-to-weight ratio

a. Mathematically consistent? There are no equations, only a design argument: the field is claimed weak outside the active segment ("relatively low stray flux", p.9), so shielding is not needed, and a non-conductive housing avoids eddy-current drag [read p.11]. The eddy-current point is correct: a moving magnet near a conductor induces eddy currents that drag. The stray-flux claim is qualitative and not computed. "High thrust/weight ratio" is a ratio of two numbers the patent does not give. Tension with claim 1: the armature field is said to be strong in the external regions where forces act, while stray flux is said to be low; both can hold only if the strong region is confined to the interaction region, which a field plot would need to show. Not demonstrated. b. Consistent with experiment? Non-metallic housings and ironless/air-core designs are used in practice (ironless linear motors, coreless motors) and trade lower weight and cogging for lower force density; this is a known tradeoff, so a claim of simultaneously high thrust and low weight is plausible but not a free lunch. Not verified by me on the web. c. Tested? No data in the patent [read]. No replication. d. If true, what changes? Nothing in physics; a performance claim.

Verdict: Never properly tested (a testable engineering claim: measure thrust per mass against a conventional machine; the patent has no neutral test, and nothing indicates it was dismissed without testing, it was simply never reported).

Overall assessment

• The mathematics is not wrong in any way that matters for the central idea, but it is almost entirely absent. The patent is a verbal engineering disclosure. Mathematical issues are confined to three loose statements: "much less electrical power" for given torque and speed (violates the F*v budget if taken literally), "reduction in input impedance" (inductance is not changed by switching), and "increased area ... compensates for loss of force" (unquantified).
• Nothing is over-unity and nothing needs new physics. A grant is not validation of operation: the USPTO does not test for function [memory, high confidence; standard patent-office practice].
• The one calculation that would settle claims 1 and 4 is a finite-element magnetostatic calculation of force versus armature position and of stray field for the described geometry, followed by a dynamometer measurement of input power, output power and mass versus a conventional brushless motor of equal rating. The one calculation that settles claim 3 is the bookkeeping 1/2 L I^2 per switch-off against the measured input and output power.

Verdict table