"Magnetic Field Architecture" and the HENDO Hoverboard - *ALMOST* a real hoverboard / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

"Magnetic Field Architecture" and the HENDO Hoverboard - *ALMOST* a real hoverboard

first previous 2 3 4 5 next last

Move Thread

LAN_403

BigBad

Thu Oct 23 2014, 12:28AM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

Oh there's no problem.

I find you need about 3mm or more thickness to get decent lift; but pretty much the thicker the better (it's because inductance goes as the square of the thickness, surface resistivity goes inversely); but obviously there's no point in it being thicker than the skin depth at the frequency you excite it at; but you can always slow the spin down; you want one skin depth thickness.

2m x 1m x 3mm ally costs about Â£125 by the look of it.

so it's moderately expensive, but not obscenely so, the board will probably be more expensive unless you're covering a large area.

However, if you make the poles very big, then the required thickness goes down, but you tend to end up with very, very big magnets if you're not careful, turns out thin, wide magnets don't work very well; they need to be about half as tall as they are wide for good lift. edit: you can probably get away with thinner metal if you spin faster to get the skin depth down, and use lots of magnets rather than big magnets

There are some stability issues, if you lean, one edge will sort of dig in and spin you around.

Artlav wrote ...

BigBad wrote ...
Yes, it seems to be a souped up version of this:

Drats.
An i was just warming up to print the second motor to do just that...
Foiled again. :)

DOIT

Incidentally, you really, really need an iron backing plate in your disk, about a fifth the thickness of the magnets. Doing that will do something like double or quadruple the maximum lift. Make sure the iron is well saturated.

Shrad

Thu Oct 23 2014, 07:37AM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

that would be a nice thing to try with HDD motors running iron plates and hard disk magnets...

Uspring

Thu Oct 23 2014, 10:23AM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

Compliments on how you guys figured this.

I'm now wondering about the efficiency, i.e. the lift to drag ratio. I'd expect the eddy currents to be proportional to rotational speed of the rotor and the lift and drag forces also. Power consumption would then be proportional to the square of speed, lift only linearly. That would give slower rotors an advantage as long as these can support the weight.

Another question is about the arrangements of magnets. Is that optimal or could they be placed so that their axes are parallel to the circumference of the rotor? Another idea would be AC driven coils with vertical axes. Might be somewhat more efficient since the eddy currents create a directly opposing fields, i.e. lift only and power consumption is only in the eddy currents (assuming the electronics to be 100% efficient).

dexter

Thu Oct 23 2014, 11:23AM

Registered Member #42796 Joined: Mon Jan 13 2014, 06:34PM
Location:
Posts: 195

Shrad wrote ...

that would be a nice thing to try with HDD motors running iron plates and hard disk magnets...

HDD motors are to week for this as they have very little torque capability

BigBad

Thu Oct 23 2014, 11:28AM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

At a given gap size the lift curve looks like this (this is actually a linear motor curve, which has the same properties):

The blue curve is the drag curve, and needs to be multiplied by the rotation speed and radius and multiplied by a constant to get the power. The red lift curve is fairly flat once you get lift.

Notionally, the constant depends on the resistance of the plate seen by the rotor. The resistance depends inversely on 'how much' copper/aluminium is involved in carrying the current to achieve the lift; i.e. the volume of copper/aluminium under the magnet in which the currents are flowing.

The lift and losses are both proportional to I^2. Obviously you want the resistance to be as low as possible; as the I^2 R losses are the power you need to levitate.

Ideally you would want no resistance (superconducting plate) but that's impractical due to the low temperatures and large areas (but it's certainly possible for small areas).

Failing that, the biggest area of magnet possible because that gives the biggest volume of aluminium/copper.

Now you might think that a big very thin layer of magnet is just the ticket, but when I tested it, it doesn't perform-you get very low levitation forces. You need a layer of magnet that is about half the pole size thick (preferably with an iron backing plate), because that increases the field strength... the pole diameter also determines the lift height because the field of the magnet has a scale height that is the pole diameter.

Interestingly there's no theoretical low limit on how little power you need; levitation doesn't take any power (vertically, force times distance is zero). It's the horizontal drag that's the problem. In practice these kinds of systems usually make excellent room heaters. However large enough area of magnet/thick plates can give low loss because it brings down the I^2 R; you're sharing the force over a large area, and so the resistance is very low.

Uspring

Thu Oct 23 2014, 02:02PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

The lift and losses are both proportional to I^2

Why would lift be proportional to I^2 ? I'd expect the field caused by the eddy current I to be linearly dependent on I and consequently the lift and drag force also. What I can understand is that they seem to level off at higher f, probably due to the skin effect.

Shrad

Thu Oct 23 2014, 02:05PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

dexter wrote ...

Shrad wrote ...

that would be a nice thing to try with HDD motors running iron plates and hard disk magnets...

HDD motors are to week for this as they have very little torque capability

I mean at a lower scale, as when they are rewound they can be pretty nice and could provide great RPM... and you'd easily secure the aluminum body to a frame of some sort...

Of course I don't think a HDD motor would lift a man ;)

BigBad

Thu Oct 23 2014, 02:06PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

Oh wait, my bad. I was thinking of an induction motor; it's I^2 on the primary current there. It's (essentially) because the current creates a proportional current in the secondary, and the two currents, the primary and secondary currents repel each other.

With a permanent magnet system, the ampere current in the permanent magnets is constant, so the lift is proportional to the current (for constant field) in the secondary (i.e. the plate), but the losses go as I^2 R, so you want the lowest resistance you can, but still a high current to give you good lift.

So far as I can tell that changes the detail, but not the broad brush effects.

Sigurthr

Thu Oct 23 2014, 09:58PM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

Btw, in the kickstarter video you can see they're using a RC hobby transmitter to indeed control propulsion on one of their R&D boards.

I wonder what the upper weight capacity on these boards will be. I did some back-of-the-napkin math the other day and came up with somewhere around 4kW to levitate my 100 kilo self. Not real promising. I figure a common hobby 50C 3S 6Ah lipo might get me a minute of "flight time" if the system was incredibly efficient. I am on a diet but the math doesn't look good, haha.

BigBad

Thu Oct 23 2014, 11:20PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

Sounds about right. I looked at this before and reckoned a few horsepower. Basically, we're talking lawnmower engine type levels of power, or, with care, a 13 amp socket could potentially supply the necessary power.

first previous 2 3 4 5 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob