Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Making domed/half sphere closed loops for EDS.

  1. Jun 21, 2007 #1
    How can I wind magnetic wire around a half sphere so that I end up with closed loops?

    How can I wind magnetic wire around a half sphere so that I end up with individual closed loops?

    There may be a better way to accomplish this so any suggestions would help a great deal.

    The way I would attack this is... starting with my half sphere that has a pin installed at the apex, wind magnetic wire around it spinning the half sphere until I have the desired wraps. Once this is done remove the pin so that I can insert a scroll saw blade equal to the pins' diameter. Following a straight line cut through the coiled wire exiting out to the outer perimeter. Insert/solder a copper shim into the new slot effectively closing each wire creating a local-bus.

    Well that my best shot, what do you think?
    Last edited: Jun 21, 2007
  2. jcsd
  3. Jun 29, 2007 #2


    User Avatar

    Staff: Mentor

    I'm not tracking the geometry that you are asking about. Could you maybe upload a sketch?
  4. Jun 30, 2007 #3
    I posed this question in General engineering but got very little response. So, if no one minds I'll try it here.

    Imaging a regular hollow ball, approximately 2.5" in diameter cut in half making a dome or half sphere. Drill a hole at its' apex so that, at a later process you will be able to insert a scroll saw blade. For now insert a pin so that you are able to wind a single 30 gage wire around it in a continues manner until you have wrapped the entire surface of the ball with your magnetic wire. Continue the winding until you have reached the apex angain and the outer perimeter. Repeat this process until the desired winds are reached. You now have a half sphere/dome of magnetic wire, several layers thick. Take the pin out and replace it with a scroll saw blade. Cut the coiled magnetic wire from the apex to the outer edge until the blade is completely through. At this point we should have individual wires that are not joined to each other. Now, to create the closed loops I would think that soldering copper shim material in the gap, that was created by the scroll saw blade, would join each wire making a closed loop of each individual wire. The shim, I understand, will make the local bus that joins each loop.

    I'll make a drawing as soon as I can and upload it.

    I guess, what I'm asking is. Is this process going to give me the individual closed loops I'm looking for? Or, does each individual wire need to be connected as separate units?


    P.S. EDS is the acronym for ElectroDynamic Suspention
  5. Jul 1, 2007 #4


    User Avatar

    Well it sounds like it'd result in a lot of loops and a
    bus that electrically connects all the loops along one line of 'longitude' from pole to equator.

    You'd have to be very careful to choose wire that would
    both tolerate being cut like that and being soldered like
    that, or you could end up with a LOT of open circuits.

    I still really don't fully understand the engineering
    application principles of what you're trying to do, but I'll
    just add that I *assume* it's important to you for all the
    loops to be able to conduct current *only* along rings of
    latitude and *not* along lines of longitude (except for the
    single bus running along the latitude). Were that NOT
    the case, it seems that you might as well just use a sheet
    of foil and not individual rings; the foil would obviously
    act as an 'infinite' number of parallel rings of latitude,
    but would support eddy current conduction in the
    longitude direction as well.
  6. Jul 1, 2007 #5
    A polishing process of all the open ends would have to be adopted, prior to the soldering process. This is still prototyping/R+D, so there are going to be some hurdles to over come.

    The intent is to create an external magnetic field in the closed loop coils that was induced by the PM array which is a result of human input for EDS.

    If my understanding is correct... magnetic field lines need to interact at a perpendicular angle to the wire in order to be able to be induced by the field itself? The rotor design and the wire will in this manner always be perpendicular to the induced magnetic field as the rotor rotates.

  7. Jul 1, 2007 #6


    User Avatar

    Well if the changing magnetic flux is parallel to the wire,
    no EMF will be produced in the wire. If the angle
    between the changing flux and wire is somewhere
    between parallel and right angles, you'll get an increasing
    EMF along the path of the wire as the angle approaches
    a right angle where it'll be maximum.

    Wrapping foil or variously cut 'fans' of it over a hemisphere
    isn't hard; there will either be some wrinkles or seams, but
    that's usually of no consequence.

    Of course it's not hard to beat metal sheet into
    a hemisphere or mold molten metal into one.

    The real key issue electrically is whether it's essential that
    eddy currents NOT flow in the direction of the equator
    to the pole.

    If the flux were always incident normally across
    the plane of the equator of the circle all the EMF
    would be around lines of latitude anyway, so it'd be
    immaterial if there was equator-to-pole conductivity.
    However since you have complex rotations going on,
    etc. I don't really know what eddy current loops are desired
    or possible or not in the hemispheres.

    One thing that's sometimes used to solder insulated
    wires is that some kinds of insulation are not much at
    all capable of resisting the heat of soldering temperatures,
    so merely dipping the ends in solder *might* cause the
    insulation to shrink / burn away and the solder to wick
    up and attach to the bare copper wire ends. It could
    be engineered to work well, but I wouldn't rely on it
    unless I'd carefully controlled the wire, insulation, and
    process temperatures and timing.
  8. Jul 1, 2007 #7
    The rotor is so designed so that the flux will always be perpendicular to the direction of the wire, so the EMF along the path of the wire will be at its' maximum, pending the speed of rotation.

    I don't see the rotors rotation as complex, its' spin in an equatorial plane is fairly slow and in a north to south plane it rotates between 1600 - 2000 RPM. I understand, to limit the eddy currents to a minimum is desirable? Aren't eddy currents basically the resistance in the current flow within he medium the current is interacting?
  9. Jul 2, 2007 #8


    User Avatar

    I wasn't aware of the mechanics of your rotation, but
    I recalled that it seemed to be a two-axis gimbal type setup;
    not outrageously complicated mechanically, I realize.
    I should have said that the field's variance in
    3D geometry given the Halbach array coupled with the
    mechanical motions seemed complex enough that I
    haven't an idea what the desired EMF / eddy current
    interactions would be or should be for your desired effect
    given all the mechanical, geometrical, magnetic,
    and electro-kinetic circumstances.

    "eddy current" is a generic term for a current that
    is induced in a passive conductor as a result of the
    EMF from a varying magnetic field interacting with that

    Certainly sometimes eddy currents are beneficial
    (I worked on building a $250,000 piece of test equipment
    that used them for non-destructive testing of metal
    film resistivity / uniformity / flaws), and in other cases
    they're detrimental.

    When the eddy current is a parasitic effect that reduces
    the inductance of a coil that is supposed to be
    shielded from the environment, or which saps power
    from something like a motor, it's seen as a detriment and
    unintended factor.

    When it's used to measure something like in a
    primitive metal detector, or when it's used for intentional
    heating like in an induction furnace, or in other applications
    when it contributes to the desired result, it's a favorable

    One could just call it "induced current" and leave it at
    that, assuming one is clear about what it's affecting.

    In your case since you seem to desire rings like lines
    of latitude, eddy currents flowing in the hemisphere
    along those circles are apparently beneficial or critical
    to the effect you seek from designing the coils
    in that orientation, and arranging the incident flux to
    couple to them in some cases.

    It's just unclear to me what detriment or benefit
    the same kinds of currents, if it's even geometrically
    possible for them to exist, would have were they to
    be able to flow along lines of longitude on the

    I'm reminded of the technology used to make
    some low frequency transformers and motors called
    laminated cores. If a coil is wound axially
    around a bar of metal, AC flowing in the coil will induce
    EMFs that would cause the current to flow in circles
    inside the bar -- just like the winding of the coil, except
    the current would be in the opposite circular direction
    since it's an opposing eddy current in an (let's say)
    non-magnetic material.

    For motors / transformers such anti-circulating eddy
    currents are loss factors, so to eliminate them, they
    make the bar the coil is wound around from many
    quite thin insulated plates of metal oriented so as to
    insulate and break up (across the insulated boundaries
    from one stacked sheet to the next) eddy currents from
    being able to flow in circles around the path of the
    windings. Since the metal sheets have about the same
    magnetic properties as a solid sheet would, they suit
    the use as a core material, but without large
    eddy current flow there isn't so much power wasted.

    So the only real difference between using insulated
    stacked loops (which are really just like the laminated
    sheets I've described, in a way) and a solid
    hemisphere of foil or metal sheet is simply that the
    parallel loops conduct along the circles of latitude they're
    axially wound around and not at right angles to those,
    except along the single bus-bar that does run along
    a semi-circle of longitude.

    Since you mention EDS, I'd guess the idea is to have
    the eddy currents in the hemispherical cap's loops
    cause mutual repulsion against the
    varying flux from the composite field of the rotor's magnet
    array, thus braking the angular momentum component
    of the rotor that causes changing flux through the
    cap's loops and causing an opposing torque on the
    loops which have eddy currents interacting with the
    rotor's field.

    That would create a stess force between the
    cap loop(s) which are eddy current regions,
    and the sources of changing flux in the rotor.

    Of course the geometry of the gimbal arrangement,
    rotor angular momentum vectors, and flux/eddy
    current field interactions would dictate mechanically
    how that braking force and torque would all effect
    the system as a whole.
  10. Jul 2, 2007 #9


    User Avatar

    Staff: Mentor

    Robin, Multiple posts are not allowed here on the PF, so I merged your General Engineering thread into this one. In the future, please PM me or another Mentor if you want a thread moved.
  11. Jul 2, 2007 #10
    From all the replies I've received thus far, it looks like I need to research NDT (Non-Destructive Testing). NDT, as I understand it, makes use of the magnetic flux field in its' testing procedure. My gut tells me that for the most part NDT uses AC for testing but uses an external flux field which is similar to my purpose. How NDT deals with eddy currents is something I will need to look at as well.

    Anyone that has references/links to so that I can jump write in would be very much appreciated.

    Thanks to all your input, everyone.
  12. Jul 2, 2007 #11
    Thanks, I was hoping that the two could be merged. I suspect that I would like to merge other posts, that are directly related to the original topic (there are no others at this time). I am not familiar or know how to "PM" any of the mentors. Can you be of assistance in clarification/procedure(s)?

    Is it possible to continue the first post "How can I create an external magnetic field" and continue it with "How can I wrap wire to form a closed loop"?

    My apologies for the infraction.
  13. Jul 2, 2007 #12


    User Avatar

    Staff: Mentor

    PM is personal message. You can send one to a user by left-clicking on their underlined user name, and selecting "Send a PM" from the pulldown boxes. For the Mentors and Homework Helpers, you also can send us a PM via the "Staff" list which is a selection at the top of each PF page.

    On merging the "How can I create an external magnetic field" or continuing it or whatever -- I have to be honest and say that I really did not understand that thread or what you were trying to do. Well, I don't understand this thread either, but luckily xez seems to be able to help. I think the main problem is my trying to visualize what you are asking for in each case. I know you are trying hard to explain it, but maybe some new sketches would help out. If there is an overall thing that you are trying to accomplish, then if you could explain that at a high level, perhaps we would be able to suggest a practical way to accomplish it.
  14. Jul 2, 2007 #13


    User Avatar

    NDT would be an interesting thing to study, though I
    think it's technologically tangential to your purposes.
    Indeed NDT mostly uses AC in the several kHz to several
    GHz range for making crack / thickness measurements.

    I believe that the two-dimensional system that approximates part of what Robin07 is trying to accomplish
    would be something like a traditional DC motor:

    1) a broad circular ring of a stator made of numerous
    individual solenoidal coils standing successively
    in-line around the stator ring's circumference laid out
    with one of their poles facing radially inward, and the
    coils' other poles being radially outward.

    2) a rotor wheel spinning inside the stator ring
    having a single permanent magnet north pole in
    close proximity to the stator ring, and its south pole
    somewhere 'far away' from the stator ring e.g. near the
    center of rotor rotation.

    3) The 'north' pole piece of the rotor will rotate
    in the proximity of each of the stator's coils as the rotor rotates.

    4) In the proposed mechanism each of the solenoidal
    coils along the stator e.g. one at 12o'clock, one at
    1o'clock, ...... are short-circuited, thus creating eddy
    current regions along the stator ring.

    5) As the rotor's pole piece approaches
    each stator coil the induced eddy currents from the
    increasing flux through the coil will cause repulsion
    between the rotor pole piece and the proximate stator coil.
    Of course conversely as the rotor pole passes and
    recedes from each stator coil, the decreasing flux
    through the coil will cause attraction between the stator
    coil and the rotor pole. It's not clear to me if this latter
    effect is being fully considered or is not detrimental to
    the sought effect(s).

    6) I think the basic idea has something to do with
    externally mechanically driving the rotation between the
    rotor and stator by spinning the stator or something like
    that and letting the rotor free-wheel inside on its bearing
    reacting to its inertia / angular momentum as well
    as the eddy current coupled rotor-stator forces.

    7) I believe that the desired result has something
    to do with converting rotor angular momentum and
    kinetic energy into moments of force / torque and
    resultant momentum affecting the stator.
    Possibly the idea is conversion of rotor angular momentum
    to 'linear' momentum of the stator-rotor system, or
    possibly it's concerned with applying some kinds of
    torque between the stator and its external supports,
    I'm not quite sure how much of which.

    8) I'm not quite sure of how many stator coils there would
    be in the best 2D analogy to Robin07's 3D conception;
    perhaps there would be only one stator coil at a discrete
    position e.g. 12 o'clock along the stator ring, or perhaps
    there'd be a half-circle of them in line between 12o'clock
    and 6o'clock. In the 3D version there's talk of
    hemispheres and concentric coils et. al. and I'm still
    confused by the intended 3D multiple-coils per distinct
    hemisphere geometry.

    9) Anyway I think that any essential physics that
    affect the 3D concept could be reduced to an
    analogous '2D' rotor + pole piece + stator +
    N * solenoidal eddy current coil system.
    Comprehending the various eddy-current rotor-stator
    torques, and the effects of rotor momentum,
    stator momentum, resultant gyroscopic effects
    (at least wrt. any 3rd-axis accellerations)
    and forces between the stator and an external
    framework would seem to be all that's necessary
    to analyze the system, and I believe that doing so in
    '2D' would be enlightening to the OP.

    10) Though there are various unclear (to me) aspects
    of the posited questions / conceptions, it's certainly an
    interesting electro-mechanical system to consider.
    Sort of a like an gyroscope motor hybrid, really, afaik.
  15. Jul 3, 2007 #14
    Well xez, you've hit the nail on the head for the most part in your previous post. Consider the rotor to be conductive on only one half of the entire spindle/axis itself, which rides in the conductive circular guide as it spins. This allows for intermittent contact (On-Off), switching as it were. As the rotor turns it completes the connection between the upper hemisphere and the lower one. This break in contact occurs when the pole pieces recede from the stator coil. Now in 2D this is all too straight forward but in 3D another switching must be adopted. Consider, just for the moment that we have no individual stator coils, just wire. The wire is wound so that the induced field is opposite to the Haulback pole orientation (right hand rule) this would be typical in both hemispheres. The induced field pole orientation will then be so designed to be opposing and as we know now that induced magnetic flux will all-ways give you an equal and opposite force. Since the spindle switches on and off the breaking due to eddy currents is also minimized. So, what you think?

    I have found that constructing a multitude of small gage wire coils is not in the best interest re: the manufacturing process. Well, that is to say, if it should ever come to that that is. The individual coils configuration can work and very well I might add. But each coil, in a 3D configuration must now be synchronized to switch on and off, a difficult task, I would think. Perhaps not.

    Oh, In your previous post you mention "AFAIK" and "enlightening to the OP", what is afaik and OP
    Last edited: Jul 3, 2007
  16. Jul 4, 2007 #15


    User Avatar

    Robin07, this is starting to make more sense to me in
    certain aspects now that you've elucidated a bit concerning
    the electro-mechanical aspects of the system.

    Previously I was a bit confounded not entirely so much
    by what you'd described, but by what you hadn't mentioned.

    I didn't recall that you had mentioned any kind of
    mechanical or electrical switching relating to stator currents,
    and since you apparently didn't want the field coil current
    circulation to reverse, (i.e. when the rotor flux was
    waning in the field coils) I thought that either I or you
    was perhaps missing something rather essential, or that
    I was just badly misunderstanding some
    geometric / engineering aspect of your conception.

    I'm now of the impression that you're essentially
    talking about a DC motor operating in the inverse
    ('DC' generator) sense in that you're seeking to generate
    strictly clockwise circulation (let's say) of eddy currents
    in stator 'coils' mechanically in front of the
    rotational path of a permanent magnet rotor's pole piece.
    Thus there is mechanical drive (mutual repulsion)
    from the mutual approach of the stator eddy coil
    and rotor pole piece. Given a relativistic
    point of view one could say the rotor moves and drives
    the stator 'around', or the 'stator' shell moves relative
    to a 'fixed' rotor and drives the inner rotor around.
    In reality it's a mutual effect given Newton's law,
    and either one can be viewed as in motion relative to the
    other and the physics are still the same.

    We're more used to seeing motors with inner
    rotors and fixed casings rather than semi-free casings
    being affected by the dynamics of inner / kinetic
    rotor systems perhaps with the exception of toy gyroscopes.

    Electro-mechanical switching analogous to a DC motor's
    terminates the stator coil current when the rotor pole is
    a bit before T.D.C (top dead center -- like a car's piston
    when the spark fires in the combustion cycle)
    alignment with the (eddy) stator coil

    Of course in a DC permanent magnet motor (or in an AC motor,
    for that matter), it's a rotating magnetic field which rotates
    in sync with and just positionally ahead of the mechanical rotor
    that causes the rotor motion, but relative to the field
    the rotor doesn't move at all, and all that's going on are
    a couple of 'static' magnets attracting each other, the fact
    that they're both rotating relative to some unrelated
    shell is immaterial.

    So in the case of interest to you, it's a 'DC' motor
    (generator) where the rotor drives the 'rotation' of the
    stator at the expense of rotor angular momentum w.r.t. the stator.

    I don't see anything physically amiss about those general
    engineering conceptions, and a half-efficient half-wave-AC
    generator (disconnected during the rotor recession w.r.t.
    the stator pole) certainly does the same thing.

    Indeed as you may be aware common exercise bicycles
    use eddy current based braking to generate
    variable levels of mechanical resistance to the
    human power driven rotor, not wholly unlike what you've
    sometimes referred to.

    w.r.t. coils and a bus bar, yes, I agree, doing the coils
    of multitudes of thin wires isn't likely in your best interest
    w.r.t. manufacturing. If you wanted a few score of
    coils you could just use PCB traces on flexible substrates,
    or a metal shell laminated to an insulating shell and C.N.C
    mill insulating trenches in the formerly solid metal layer,
    or use a laminated layer stack of 'rings' as a laminated
    toroidal core might, metal deposition and resist patterning,
    et. al.

    Of course that's assuming you need numerous discrete coils
    which may not be the case at all as far as I see it. . .

    Of course you'd need a couple of bus bars, not just
    one tangential one to provide a couple of points to
    switch the stator loop current, so it'd be more of an 'C' or
    OMEGA rather than a 'Q' with a single perpendicular bus bar
    which couldn't be used to interrupt circumferential
    loop current.

    My apologies for the confusing acronym use and
    rhetorical asides in my previous post(s), in part I was
    also responding in the thread and context of berkeman's
    post and explaining to him and other viewers what my
    interpretations were with respect to what I thought
    you were trying to conceptually express, since he'd
    expressed some confusion about things you'd expressed
    that I thought I could elucidate.

    AFAIK = "As Far As I Know",
    and "OP" is "Original Poster", yourself, in this case, though
    just a common referential acronym to use in forum posts
    when multiple parties are discussing aspects of a thread
    that someone has started when discussion diverges to
    include topical commentary between 3rd party

    With respect to the difficulty of the task of switching
    coil circuits with respect to the kinetic and geometric
    dynamics of a rotor in 3d, it's certainly not difficult
    at all given the scales of speed and size you're talking
    about. Such switching is not particularly different
    that what you'd do in a stepper motor, and there's
    no reason you can't use a 'position encoder' and
    'state machine' (e.g. control logic circuit) to get
    and process the information you need about when to switch,
    just as in a brush-less DC motor controller.

    Tell me, have you ever heard of Maxwell's Demon? That's
    probably the ultimate abstraction of one aspect
    of what you're trying to do, and at least in these
    macroscopic speed/size scales there isn't a problem with respect
    to getting the needed information concerning the available
    (kinetic/magnetic) energy source......

    The principles of reversibility and relativity are also
    helpful in analyzing these sorts of problems.

    W.r.t. the physics, concepts, and some conceivable applications,
    this all encompasses subject areas I've been familiar with
    and interested in for several years.

    Your terminology / expression confused me for a while making
    me think that perhaps you were talking about different kinds
    of apparatus.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook