Talk:Halbach array

untitled section
There must be some misunderstanding or fineprint missing. The configuration is symmetrical and therefore there is no reason why the fields should cancel on one side but not on the other side. --Pjacobi 19:24, 4 October 2005 (UTC)
 * Now I get it. The article is in need of a diagram showing the field lines, like . Also the simplest, because most symmetric configuration, would be a cyclinder magnet inside a toroidal magnet. I guess I have to make a picture to make me clear. --Pjacobi 19:39, 4 October 2005 (UTC)

Arg. I still don't get it. I see the illustrations which look believeable, but the diagram still appears symmetrical. If so, how can it produce an assymetrical field? How can I predict whether the top is canceled or the bottom? The article does not explain this and it is drivng me crazy. --Dan (sorry, no account) —Preceding unsigned comment added by 24.16.144.75 (talk) 01:01, 1 March 2010 (UTC)

is there any information on how much the field is increased on the augmented side?


 * It is not symmetric at all. Have a look at this:

NN NS SS SN NN NS SS SN NN NS SS NS NN SN SS NS NN SN SS NS


 * In the first line, the north and south poles are always 4 units wide. In the second line, they are one and two units wide, alternating. The lower part looks uniform, (that is unmagnetized,) from a much shorter distance than the upper part. Just as a lot of tiny black dots, seen from a distance, look grey.


 * Or, more formally : If the director of the magnetig field gets tilted in one direction by going along the long edge of the array, the poles (or density of field lines) are always further apart on one side(low field), and nearer to each other on the other side (high field). --Maxus96 (talk) 20:56, 3 March 2010 (UTC)

Please provide legend for green/red images
Nice pictures that display info, but they need a legend or explanation. Green is polarity A, red B, and strong side emits polarity A. — Preceding unsigned comment added by 208.80.117.214 (talk) 19:15, 16 June 2023 (UTC)

more easy to understand
Can we make this easyer for somone with less technical know how to understand? Alan2here 20:55, 24 November 2006 (UTC)
 * knowhow
 * The magnetic field spins and drifts; in other words, it tumbels like a wheel or baton. The strength grows where these directions match, on one side, and shrinks where they thwart, on the other. -lysdexia 07:32, 31 March 2007 (UTC)
 * What I asked seems to have happened while I was not watching which is good Alan2here 20:22, 24 July 2007 (UTC)

Am I allowed, or is it even possible to post a picture here? There is still much confusion about the array, I have an extremely simple diagram that came with my Halbach array. People seem to think there is a way you can simply line up magnets to make the Halbach array, this is not correct, nor even possible (think trying to push N "pole" to N "pole")the array is only complete once the last magnet is forced into it's "holder". By holder I mean an aluminum block milled out to hold the magnets, they are held in place by, in this case, grub screws. (Some other Halbach arrays are assembled using an adhesive.) If any of the magnets are not held securely they fly out of their "box" due to the magnetic force of the magnets next to them. —Preceding unsigned comment added by 75.17.201.107 (talk) 04:18, 7 September 2010 (UTC)

something amiss
The field on the strong side of the array should not be twifold as strong because the magnets are discrete, and each has some greattest intensity already. They would onely be stronger if the magnets are squished to half their volume. -lysdexia 07:26, 31 March 2007 (UTC)This is incorrect, while you are correct the array is not stronger by two-fold, it is indeed stronger than if they were just lined up N to S.

The other type of Halbach Array
This is

NSNSNSNS

or

NSNS SNSN NSNS SNSN

Where a N represents a magnet with it's north pole facing up and a S represents a magnet with it's south pole facing up. These are often (possibly incorrectly) called hallbach arrays Alan2here 20:24, 24 July 2007 (UTC)


 * They don't cancel the field though, so I don't think they count.- (User) Wolfkeeper (Talk) 18:31, 18 October 2009 (UTC)

Merge with Halbach cylinder?
I'm not sure if there's a formal definition of Halbach array. It seems to me though that the term 'Halbach array' could apply to three dimensional layouts of magnets as well, so it might be a good idea to combine Halbach array and Halbach cylinder together into one article with redirects. Comments?- (User) WolfKeeper (Talk) 16:02, 28 August 2008 (UTC)


 * I see what you mean but I think these should be re-split. The Halbach Cylinder and Halbach Array are related but distinct structures both used for very different engineering applications. (Also, the cylindrical structure is two dimensional, like the array; in three dimensions both suffer from magnetic end effects.) --Hiltonj (talk) 05:41, 10 November 2008 (UTC)


 * Well... I don't see that there's any good reason to split right now, and the article is not big. The section topics meet the definition we use for Halbach array, they're obviously intimately related and even the wiggler is 3D really.- (User) Wolfkeeper (Talk) 18:36, 25 November 2008 (UTC)

half-life of Halbach arrays?
Isn't it true that all Halbach arrays must eventually fail (demagnetize)? If so, how long do they last? beefman (talk) 21:03, 20 April 2009 (UTC)


 * umm. It depends what type of magnets you use and how much you bash them about and whether they get hot etc. Supermagnet based arrays are very robust and may not degrade appreciably (it's difficult to be sure, they haven't been around very long).- (User) Wolfkeeper (Talk) 00:01, 9 August 2009 (UTC)


 * Every macroscopic dipole must fail in the end, because their macroscopic orientation is just not stable above 0 K. The old trouble with entropy. But never mind, as long as you stay at room temperature, they will last practically forever, at least modern Nd - Materials.
 * Wether a halbach array should demagnetize either faster or slower than on ordinary dipole, i don´t know. Any ideas, you physicists? And, btw., is that old story about magnets staying alive longer if you stick them to a large piece of iron true? --Maxus96 (talk) 13:34, 10 August 2009 (UTC)
 * Been thinking about this a bit longer: I guess that the demagnetization is due to the fact that magnetic field lines repel each other, and thus the field outside one end of the magnet is not parallel to the field inside (except for the center of the pole). If that´s true, then a halbach array should demagnetize much slower, as it is ideally constructed in a way that the field lines are not being bent at the metal/air border. Anybody knows a source for this? --Maxus96 (talk) 17:57, 18 October 2009 (UTC)


 * They won't demagnetise themselves otherwise they would be demagnetised as soon as you put them together. To degauss a hard magnetic material you need to apply the reverse field to make the magnetic material itself to give no field (see remanence). Because Halbach arrays have zero field net at the outside, so it does not reach this strength at any point; there would have to be negative net field on any particular magnet to do that.- (User) Wolfkeeper (Talk) 18:15, 18 October 2009 (UTC)


 * What you say is only true in the short term (years ;-)). As i said above, every macroscopic dipole must fail on a large timescale for thermodynamic reasons. I´m very mistaken if this slow decay isn´t sped up (a little) by pertubing external fields, however small. Of course, it will be hard to even measure that speed up, let alone noticing it in application. The problem is sort of "academic" ;-)). --Maxus96 (talk) 22:05, 28 October 2009 (UTC)


 * (Late to the party, but anyway:) Short answers: No, Halbach arrays usually are stable (if coercivity is high enough, usually the case in Neodymium magnets), and yes, iron can help some magnets live longer (in terms of their magnetization). Demagnetization occurs when (and in those portions where) the opposing H field inside of the magnet exceeds the magnet's coercivity Hcj. A full FEM simulation is required to accurately predict this. Significant opposing internal H field is always generated by the magnet itself (think of 2 elementary magnets side by side, same orientation, opposing each other), depending on it's geometry. Additional field can be introduced by external sources, such as coils or other magnets, as is the case in a Halbach or other arrays. The opposing internal field can also be weakened by suitable external sources, or by flux conducting pieces (e.g. steel or iron parts in a loudspeaker magnet system, conducting the flux towards the air gap). The latter case leads to a higher working point, which comes from a steeper load line (larger permeance coefficient Pc) in the 2nd quadrant of the B-H diagram; this is similar to an electric circuit with internal resistance and external low-resistance load, which leads to a large current (flux) and low voltage at the source terminals (weak internal field). Also, coercivity is temperature-dependent: Through heat it can drop below the present opposing field, leading to partial demagnetization until the self-generated opposing field matches the new coercivity again. Especially with older materials (ferrite, AlNiCo, early Neodymium), coercivity was rather low to begin with; even at room temperature or in arrangements such as Halbach, those magnets operated close to their coercivity limits; that's where the issues with instability arose and where iron helps "staying alive". Some advanced Neodymium magnet grades can withstand >200°C (at Pc=1), and at room temperature are pretty much "indestructible". Langenforth (talk) 07:03, 17 December 2020 (UTC)

Halbach cylinders
... are actually idealised models, aren´t they? If you really want one, you have to build it out of wedges or other kinds of dicrete dipoles. --Maxus96 (talk) 20:37, 8 August 2009 (UTC)


 * So are the cube-style linear ones though. An ideal Halbach array of any description is made of infinitely thin slivers of magnetic material.- (User) Wolfkeeper (Talk) 23:57, 8 August 2009 (UTC)


 * There's no theoretical reason you can't build an ideal one, it's more of a practical issue. They're normally built out of wedges as it's very difficult to impose the correct variation in magnetisation when you fabricate the ferromagnetic material. The practical difference between an eight-wedge design and the ideal one is very small anyway, so there isn't much gain to be had. There are some clever designs, though, that form a perfect magnetic image of a Halbach cylinder in a soft magnetic material. These designs have a continuous magnetisation pattern. If you're interested (and can get access) see "Cylindrical permanent-magnet structures using images in an iron shield", IEEE Trans. Magn., 39, p1983 (2003).--Hiltonj (talk) 06:16, 27 October 2010 (UTC)

Why mirrors?
The radiation in a wiggler is emitted in one direction only, and there is no medium to start lasing anyway, so why do you need a mirror? The page on FEL only states that there is no material to mirror x-rays (tough luck ;-), but no explanation why you should want one in the first case. --Maxus96 (talk) 21:10, 8 September 2009 (UTC)

ERRATA ON THE HALBACH WRIGGLER ARRAY DIAGRAM: Concerning the wriggler array diagram, you will note that the entrance of the electrons (e.g., from an electron gun) so that they cross perpendicular to the initial B-field before curving away due to the Lorentz force interaction is shown, followed by the reverse curving when striking the next perpendicular B-field that is going in the opposite direction, and so on down the line of the Halbach array. The direction of the curving of the charged particle on the diagram is correct for a POSITVE charge (i.e., proton or alpha particle) but not for an electron, where it would curve in the opposite direction. This is because of the right-hand rule that applies for a positve charge traveling with a velocity v perpendicular to a B-field, where it experiences a force F perpendicular to both v and B. The simple formula, which can be taken from any college freshman pysics textbook, is the formula (for the Lorentz force on a positive charge traveling at a velocity v across a B-field) F = qv X B, where q is the positive charge given in units of esu (electrostatic units), v is in meters/second and B is in units of Gauss. For the electron, though, you must put in a minus sign in the formula, so that the Lorentz force is given by F = -qv X B for a negative charge, and then it would meed to follow a left-hand rule. So to correct the wiggler diagram for the electrons moving down the path between the two parallel Halbach arrays in a sinusoidal fashion, you need to have the electrons enter the Halbach on the OPPOSITE side from what is shown on the diagram.

FOI: The right-hand rule is visualized as follows: hold your right hand straight out with the four finger straight out and together and the thunb is extended straight up at right angles to the four fingers. Then imagine an arrow stickng straight out of the right palm at right angles to both fingers and thumb. Now label them as follows: Arrow shows direction of Lorentz force F on the positive charge, the fingers shows the north pole direction of the B-field (fingers are the north pole arrows pointing outward from the knuckles of the right hand), and the thunb is the directon of the velocity of the positive charge that is sitting in the middle of your palm. That is why, if you use a negative chraged particle instead (e.g., an electron), you have to switch hands and use your left hand to show the oposite direction that the Lorentz force will take when using an electron, hence the left-hand rule application when considering the direction of an electron moving perpendicularly to a B-field. [Michael A. Varela, MS in Physics, can be reached at mivarela@sbcglobal.net] —Preceding unsigned comment added by 99.101.188.7 (talk) 23:26, 28 August 2010 (UTC)


 * Thanks guys. Good spot on the electron path. I spent a very short time copying this diagram out of some book as a Halbach array application for the introduction to my thesis. If I'd known it'd cause so much discussion I'd have checked it more thoroughly, especially the electron path! The figure in the book had mirrors so I left them in, I didn't consider the FEL physics. The mirrors have now been removed and the B field should be in the correct direction for the wiggles. --Hiltonj (talk) 05:40, 27 October 2010 (UTC)

Typo in Halbach cylinder section?
The first bullet intem at the end of the "Halbach cylinder" section states that the flux is confined to the centre of the bore (such as k = 3 above, a six pole rotor) but it is obvious from the referred-to picture above to see that k should be 4.

Motivation: it is the number of inward and outward arrows in the picture, that define how many poles the magnetic field has. So K=1 has one that cancels out because it goes in all the directions simultaneously, resulting in no pole, k=2 has two, k=3 has four, and k=4 has six. —Preceding unsigned comment added by Gwrede (talk • contribs) 08:06, 7 January 2010 (UTC)
 * In literature, k is often defined as γ = (1+k)β, γ being the tilt angle of the direction of magnetization at the azimuthal angle β. That makes an outside field for k<0, no field for k=0, uniform inside field for k=1, inside quadrupole for k=2, and so on. I suspect this got mixed up with the definition metioned in the article. --129.13.72.198 (talk) 23:06, 26 May 2010 (UTC)

field perpendicular to array
Does anybody know what the field of a halbach cylinder looks like, perpendicular to the center of the ring? I guess it´s not just like a long solenoid with the vector tilted by 90°? Uniform in the center, zero outside, but that´d be too easy. --Maxus96 (talk) 22:51, 7 April 2010 (UTC)


 * The ones described in the article have exactly the same field pattern all the way along the axial length if that's what you mean.- Wolfkeeper  23:07, 7 April 2010 (UTC)


 * But there are riffs on the theme where the field lines go axial and oscillate sinusoidally along the length, and you can make sort of spiral staircase field shapes and all kinds of others.- Wolfkeeper  23:07, 7 April 2010 (UTC)


 * OK, being a bit more clear ;-) : In a k=2 halbach cylinder, with uniform field inside, how fast does the field decay if i´m leaving the cylinder in the direction of its axis? Quite fast, of course, but is there an analytical equation for it? --Maxus96 (talk) 12:26, 8 April 2010 (UTC)


 * If it's a continuously varying halbach array, the field is zero. If it's got discrete magnets, then it depends on how many magnets you have. For example this shows you the field outside as well as in:

- Wolfkeeper  15:06, 8 April 2010 (UTC)
 * Sorry, my question really was misunderstandable. If you see the cylinder from above, as a circle, the field along the normal of that circle. Out of the plane of paper of that picture. ;-) In an infinitely long cylinder, its constant, and the value is that of the equation mentioned in the article. I wanted to know if there is an analytical equation for the field alog the axis for a finite cylinder, like there is for a helmholtz coil. --Maxus96 (talk) 20:43, 8 April 2010 (UTC)


 * Dunno, the article is largely about infinite arrays, it doesn't seem like it really needs to cover finite arrays right now.- Wolfkeeper  17:55, 12 April 2010 (UTC)

no heading
Please explain how in the world anyone can believe a "lowly refrigerator" magnet or a Brush-less AC motor is using the Halbach array? I have two of these magnets the "flat" and "Cylinder" comparing these to both the refrigerator magnets and the 6 AC Brush-less motors I have sittig on the bench, it is obvious there is no Halbach array in either. If need be I can provide pictures to prove my point. The "Magnetic field viewer" I received with my Halbach array magnets shows exactly the described magnetic field created by the Halbach array when used with my Halbach magnets, while every refrigerator magnet I could find (25 of them)shows a simple north south magnet is used. AC Brush-less motors (something I work on nearly every day)certainly is not your typical magnet set-up but in no way utilizes a Halbach array as I understand them. Again, I'm looking for someone to explain how either of these items is based on the Halbach array. —Preceding unsigned comment added by 75.17.201.107 (talk) 00:32, 7 September 2010 (UTC)
 * Those "refridgerator magnets" (at least the ones on my fridge) only have one magnetic side that sticks to the fridge. I guess this must be some sort of halbach. Advantages are clear: Less magnetic material needed for the same "stickyness", plus credit cards in the back pocket of people leaning against the fridge are not in danger. --Maxus96 (talk) 01:32, 17 May 2016 (UTC)
 * Indeed the similarity is from being one-sided multi-pole. The typical Halbach array is made from rectangular pieces, rotating in 90° steps, but this is only for practical reasons as such bar magnets are easy to make and readily available. It is entirely possible to make a Halbach array from shorter pieces rotating in smaller steps, and this can be decreased further until the rotation is continuous. This would require infinite tiny magnets to be assembled, but fortunately this configuration can also be achieved by continuous material such as the ferrite sheet from which refrigerator magnets are made (ideally isotropic = without homogeneous preferred axis, or with rotating preferred axis). Magnetization is then done by a one-sided coil with straight conductors, in order to generate the pole-stripes in the magnet. Such a continuous Halbach "array" is actually superior to a discrete one, since the flux/field concentration on one side (and away from the other) is now perfect to the microscopic level. Actually, I'm about to update the article to include this, which was actually invented even before Mallinson. Langenforth (talk) 07:20, 17 December 2020 (UTC)

Copyright concern
Halbach array Description IS WORD FOR WORD Taken from other sites. I thought this was not allowed? And no the site I'm specifically speaking about did not take the Wiki description, it existed long before this Wiki was made. I'm only talking about the first section. —Preceding unsigned comment added by 75.17.201.107 (talk) 00:03, 7 September 2010 (UTC)
 * Sounds like a problem that needs to be solved (see WP:COPYVIO)! What specific section are you talking about (I don't see one titled "Description")? Do you mean the introductory paragraph, or all the material before the "Flat Halbach arrays" section begins, or some other paragraph somewhere else? Please let us know the specific source you suspect is the original (the exact URL where you're talking about). DMacks (talk) 05:14, 7 September 2010 (UTC)

Thank you for moving this where it belongs! Including the addition of the title. You are correct I meant the introductory paragraph. I will replace this with the first incident on the web tomorrow. (put in a 18 hour stint at the hospital today gotta hit the rack..) Well an interesting developement in the main site I was referring to. I contacted the "webmaster" of the site in question and asked him/her what the date of orgin on the build of the webpage was and explained briefly why I was asking, without even mentioning any specific website. He/she responded with an apology, admitted to plagerism, not the correct term(?) but is the term he/she used and said their site would remove all offending material within 48 hours... Hmm some people's kids. Okay this brings up another issue, what is the policy on copying word for word a wikipedia article and placing it on one's own site. This is not what happened in this incident as the website did pre-date this wiki by several years, but, it is clear that many sites are using the same information - word-for-word. I give up on tracking who was first etc. in this case for obvious reasons, but I have seen many wiki's copied word for word and placed on other websites. —Preceding unsigned comment added by 75.17.203.222 (talk) 05:06, 22 September 2010 (UTC)
 * As long as they attribute the content to us (and a few other details), then it is completely allowed. That's part of making the world's knowledge freely available. See FAQ/Copyright for details and further links. -- Quiddity (talk) 05:56, 2 October 2010 (UTC)

Intro - second paragraph clean-up
Can someone edit the presumably mischievous 'down under with a shrimp on the barbie' comment in the second paragraph and restore to make sense, thanks! Davey5505 (talk) 19:37, 2 October 2016 (UTC)

JPEG display problem
Noticed 1st three .jpg pictures have display problems. I'm using Ggl Chrome Version 55.0.2883.87 m as of this date, 2017-01-06,

I'm working on other things today. I'll try to return to this for the technical fix.

This article may need other improvements such as adding citations.

I'm turning on article alerts so to be notified of changes.

Cheers, johnswolter — Preceding unsigned comment added by Johnswolter (talk • contribs) 18:13, 6 January 2017 (UTC)

External links modified (January 2018)
Hello fellow Wikipedians,

I have just modified 2 external links on Halbach array. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20110604160523/http://www.uta.edu/physics/main/resources/ug_seminars/papers/HalbachArrays.doc to http://www.uta.edu/physics/main/resources/ug_seminars/papers/HalbachArrays.doc
 * Added archive https://web.archive.org/web/20060328054612/http://www.tipmagazine.com/tip/INPHFA/vol-4/iss-3/p34.pdf to http://www.tipmagazine.com/tip/INPHFA/vol-4/iss-3/p34.pdf
 * Added tag to http://www.reeserail.com/

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 18:05, 22 January 2018 (UTC)

Multiscale Hallbach Expansion
Here is a thingie I made a figure for.