axial generator with lamination core

[mbouwer]

For less weight it would be much nicer to make a carbon-epoxy version of the blade.

[MattM]

How do you bake the carbon if you are using foam?

[JW]

When you work with carbon fiber mat you use epoxy to seal it. Its real similar to working with fiberglass but lighter.

[mbouwer]

Driven yaw bearing, main shaft with magnet rings, blade suspension and adjustment.

This test set-up again gives me better understanding of how to make the various components.

And can reproduce in case a windmillfriend wants to build along with me.

[mbouwer]

Now I would like to make an 150 mm improved version of my axial.
Designing a generator is not my area of ​​expertise, so I will definitely need some help.

[mbouwer]

Just finished the mold for stacking and gluing laminations for the 27 coil cores (thickness 20mm)

[mbouwer]

Today a friend gave me this Sparta bicycle motor and the intention is to turn it into a windmill generator.

[SparWeb]

Excellent!
That could be a big step forward for your projects.

[mbouwer]

As a start I'm going to try to shape a usable main axis.

[Bruce S]

NICE!!
With 24 coils , how shall you be configuring it?

Cheers
Bruce S

[mbouwer]

Thanks for your optimism, but I don't think I will be able to make a suitable windmill generator out of this for the time being.

[mbouwer]

The 27 glued packages for the axial.
I expect more from this.

[mbouwer]

A discussion about a generator for a small windmill will eventually always bring us the same answer:

"Look around at all those examples with their specially designed direct drive generators.
Always realized through cooperation of different disciplines".

[Adriaan Kragten]

Today a friend gave me this Sparta bicycle motor and the intention is to turn it into a windmill generator.

(Attachment Link)

If you count the number of magnets you get 20, so there are ten north poles and ten south poles. If you count the number of coils you get twentyfour. So the ratio is 24/20 = 6/5. These different numbers are chosen to prevent a high peak in the cogging torque but I doubt if this is effective enough to use this motor as a generator for a wind turbine. The angle in between the magnet poles of the armature is 360/20 = 18°. The angle in between the stator poles is 360/24 = 15°. So the difference is 3°. This means that you will get 360/3 = 120 preference positions per revolution. This isn't a high number and therefore the preference positions will be rather strong. These rather strong preference positions won't be a problem for a bicycle motor but it may give your wind turbine a rather high starting wind speed.

You must know very well what you are doing if you want to use this motor as a generator and if you use the original winding or if you make a new one. The winding must be a 3-phase winding. This means that there are 8 coils of phase U, 8 coils of phase V and 8 coils of phase W. I assume that the coil sequence is: U1, U2, V1, V2, W1, W2, U3, U4, V3, V4, W3, W4, U5, U6, V5, V6, W5, W6, U7, U8, V7, V8, W7, W8 right hand. So every phase has four coil bundles of each two coils. I assume that the sequence of the magnets is: N1, S1, N2, S2, N3, S3, N4, S4, N5, S5, N6, S6, N7, S7, N8, S8 right hand.

Next you should make a picture of the armature and the stator and in this picture you should take coil U1 opposite to magnet N1. If coil U1 is opposite to magnet N1, coil U2 is about opposite to magnet S1 (with 3° shift). This means that the winding direction of the two coils of one coil bundle must be opposite to make that the voltage generated in coil U1 is strengthened by the voltage generated in coil U2. Assume that coil U1 is wound right hand. So coil U2 must be wound left hand.

In the picture you can see that if magnet N1 is opposite to coil U1, magnet N6 is opposite to coil U5 and that magnet S6 is about opposite to coil U6. So coil U5 has to be wound right hand and coils U6 has to be wound left hand. So the winding direction of two opposite coil bundles is the same!

In the picture you can see that if magnet N1 is opposite to coil U1, magnet S3 is opposite to coil U3 and that magnet N4 is about opposite to coil U4. So now a south pole is opposite to the left coil of a coil bundle. This means that coil U3 must be wound left hand and that coil U4 must be wound right hand. The same counts for the opposite coils U7 and U8. So the winding direction of these four coils of phase U is just opposite as the winding direction of the first four coils of phase U! So it is very important to make no mistake in the winding direction of the eight coils of one phase.

The eight coils of phase V and the eight coils of phase W have the same pattern for the winding direction. It is easy to prove that there is a phase angle of 120° in between the coils of phase U, V and W. However, the coil patterns have to be positioned such with respect to each other that there is an angle of 12° in between a coil of phase V and a north pole and an angle of 24° in between a coil of phase W and a north pole. This is the case for coil V3 and pole N4 and for coil W5 and pole N7. So these two coils must be wound right hand just as coil U1. The mechanical angle in between coils U1, V3 and W5 is 120°. Assume right hand is R and left hand is L. So the sequence of the coil number and the winding direction is: U1R, U2L, V1L, V2R, W1R, W2L, U3L, U4R, V3R, V4L, W3L, W4R, U5R, U6L, V5L, V6R, W5R, W6L, U7L, U8R, V7R, V8L, W7L, W8R.

The armature has the same magnetic position with respect to the stator if it has rotated 36°. So a rotational angle beta = 36° corresponds to a phase angle alpha = 360°. So a rotational angle beta = 1° corresponds to a phase angle alpha = 10°. So a rotational angle beta = 3° corresponds to a phase angle alpha = 30°. This means that there is a phase angle of 30° in between the voltages generated in coil U1 and coil U2. Adding two identical sinusoidal voltages which are out of phase results in a new sinusoidal voltage which has a maximum if the magnet N1 has rotated 1.5° right hand from the drawn position in the figure. Beta = 1.5° corresponds to alpha = 15°. This results in a summerised maximum voltage of 2 * sin 75° * Vmax = 1.932 * Vmax. If both voltages would be in phase, the maximum voltage would be 2 * Vmax if Vmax is the maximum voltage of one coil. The fact that the voltages are 30° out of phase, results therefore in a reduction of the voltage by a factor 1.932 / 2 = 0.966 which is certainly acceptable. 

If a new winding is laid, the number of turns per coil and the wire thickness can be found by try and error for a certain wind turbine rotor. However, I see no good reason why you should rewind the stator if the winding direction of the coils is as I predicted.
 

[joestue]

the only reason to rewind that core would have been to change the voltage. already its going to be pretty optimal for the intended purpose.

you could add more copper if there is room.

sometimes it is beneficial to wind every other tooth of the core, but in this case, winding every tooth has better thermal conduction from the windings to the core, and more surface area for cooling.

those did look like bread loaf magnets so the cogging torque was probably rather small.

[Adriaan Kragten]

the only reason to rewind that core would have been to change the voltage. already its going to be pretty optimal for the intended purpose.

I would never rewind the stator as with the standard winding, you have three different voltages depending on how the coil bundles of one phase are connected. The two coils of one coil bundle must always be connected in series because in parallel connection of coils which are out of phase with each other results in an internal short-circuit current! If all four coil bundles of one phase are connected in series, you have the highest voltage and the lowest current at a certain rpm. If two coil bundles, connected in series, are connected in parallel to two other coil bundles, connected in series, the voltage halves and the current doubles.  If four coil bundles, connected in series, are connected in parallel to four other coil bundles, connected in series, the voltage halves and the current doubles again. Another advantage of the standard winding is that it is wound by a machine and that therefore all wires are lying very close to each other. You will never get such a nice winding if you wind the coils by hand.

[mbouwer]

It seems to me that there are not many windmill friends here who want to build a small windmill themselves and that is why it was so nice that a friend came up with the idea to try to turn that Sparta motor into a windmill generator.

But in my opinion it is much better to continue building an axial.

[mbouwer]

Given my setups, I believe I'm now at the point where building the components for a small windmill is doable.
The mechanical components are light, inexpensive and scalable.

The next step: the electrical design of the axial generator,
is definitely not my area of ​​expertise.

That is why I would like to work together with a student ( profile assignment )

[Adriaan Kragten]

Today a friend gave me this Sparta bicycle motor and the intention is to turn it into a windmill generator.

(Attachment Link)

I have tried if I could find more information about this motor and I found the Dutch YouTube films: "Sparta Ion GD Design" and "Sparta ion E Bike hubmotor, display, accu hack, controller swap". I think that this motor can be used as generator of a small wind turbine if the inside electronics are removed or simply no longer used. Only the three big phase wires are used and connected to a 3-phase rectifier for battery charging. As I have explained in my earlier post, this generator will have 120 preference positions and so the rotor of the wind turbine must have a rather high starting torque coefficient to get a sufficiently low starting wind speed. I have designed such a rotor for the 28-pole, 1-phase Nexus hub dynamo. This VIRYA-1.04 windmill is described in a manual which can be found at the bottom of the list with KD-reports on my website. I think that this rotor can be scaled up to a diameter of maximal 2 m for this much bigger Sparta motor. The VIRYA-1.04 head should also be scaled up to get a proper working safety system.

Sparta is a Dutch bicycle brand and it might be that such motors can only be found in The Netherlands but it might be that one can find a similar front or back wheel motor of other brands in other parts of the world. If one has found such a motor, one should first measure it with the original winding for a 12 V battery load to find the Pmech-n curve because with this curve it can be checked if the generator matches with the optimum cubic line of the chosen wind turbine rotor. One can vary the rotor diameter and the design tip speed ratio to get optimal matching. The formula for the optimum cubic line is given as formula 8.1 of my public report KD 35.

[mbouwer]

Don't you also find it much more fascinating to think about a self-built axial generator?

[Adriaan Kragten]

Don't you also find it much more fascinating to think about a self-built axial generator?

I have designed and built several PM-generators but not because it is more fascinating than buying one. Most PM-generators which you can buy have disadvantages which make that use in a wind turbine is difficult or even impossible.
One of the disadavantages of most PM-generators which have iron in the coils, is that they have a rather large peak in the cogging torque resulting in a high starting wind speed if a rotor with a low starting torque coefficient is used. It is possible to design a PM-generator with iron in the coils and with a low peak in the cogging torque if there is only a difference of two in between the number of armature and stator poles (see public report KD 580).
Another disadvantage of most PM-generators which you can buy is that a too high rotational speed is needed if the generator is used for 24 V battery charging. If you make the generator yourself, you can chose the winding and so you can change the voltage at a certain rpm such that the matching with a certain wind turbine rotor is optimal. But building of only one generator will be much more expensive than buying a generator which is mass produced.
Another disadvantage of most PM-generators which you can buy is that mostly no measured Pmech-n and Pel-n curves are given. Sometimes the characteristics for a resistance load are given but these characteristics are completely different as for a battery load. So if you buy such a generator, you have to measure the wanted characteristics yourself and if the result is not what you need for the wanted rotor of the wind turbine, the whole purchase is a waste of money.

The Sparta wheel motor which we discussed earlier, has 20 armature poles and 24 coils resulting in 120 preference postions per revolution. If they would have used 22 poles in stead of 20 poles, the difference in pole angle would have been 1.3636° resulting in 264 preference positions per revolution. Such a motor would be much better for a wind turbine generator. If they would have used 26 poles in stead of 20 poles, the difference in pole angle would have been 1.1538° resulting in 312 preference positions per revolution. Such a motor would even be better for a wind turbine generator than a 22-pole motor. The winding would be different. Each phase must now have two bundles of four coils.

You can design an axial flux or a radial flux PM-generator. Each type has certain advantages and certain disadvantages. Although most people on this forum use axial flux generators with no iron in the coils, there are some very good arguments to chose for a radial flux generator based on the housing and winding of an asynchronous 3-phase motor.

[mbouwer]

My idea is that problems and questions raised here have often long since been solved by the professional windmill builders. All we have to do is look around us to try to implement their solutions.

[mbouwer]

Adriaan Kragten:
[Most PM-generators which you can buy have disadvantages which make that use in a wind turbine is difficult or even impossible.]

Based on what Adriaan writes I would like to put forward some examples.
Much more fun than continuing to search is building a generator yourself.
A good example is of course the booklet by Hugh Piggott.

[mbouwer]

example 2
Adriaan writes:"One of the disadavantages of most PM-generators which have iron in the coils...."

Isn't it a big advantage for diy builders to have iron in the coils because you can then use much lighter magnets?

[mbouwer]

example 3
...... disadvantage with iron in the coils, is that they have a rather large peak in the cogging torque resulting in a high starting wind speed ....

Another reason to diligently look for ways to apply active blade adjustment.
From the starting position you already have a large torque even with little wind.

[MattM]

Here I thought the iron was simply to hold a complete circuit to prolong the life of the magnets.  I've read the magnets separated by a distance - without a magnetic circuit completed - slowly lose their magnetism.

[mbouwer]

@MattM,
If so, we have another reason to put an iron core in the coils.

example 4
[ Building of only one generator will be much more expensive than buying a generator which is mass produced  ]

With creativity building your own is less expensive.
But also it has often been said on this forum that suitable generators for small windmills are not for sale.

We need a design that can also be built in practice.

[kitestrings]

MattM,

What is "slowly"?  Is this something that happens over years, decades...  Do you have a source that describes this in more detail? Thx, ~ks

[Adriaan Kragten]

example 2
Adriaan writes:"One of the disadavantages of most PM-generators which have iron in the coils...."

Isn't it a big advantage for diy builders to have iron in the coils because you can then use much lighter magnets?

(Attachment Link)

Yes, the advantage of having iron in the coils is that the air gap is much smaller and therefore you can use thinner and so cheaper magnets and still have a large flux density in the coil. But for use of magnets in a motor, a certain peak on the cogging torque isn't a problem and therefore designers of most PM-motors prevent this cogging only up to a rather high limit. But for a modern fast running rotor of a wind turbine, the starting torque coefficient is only about 10 % of the optimum torque coefficient and this results in a high starting wind speed if the peak on the cogging torque is too high. For a PM-generator with iron in the coils, one therefore has to pay more attention to the reduction of the peak on the cogging torque than for a PM-motor. But with the present high prices of neodymium magnets it becomes more and more important to reduce the magnet costs and then a well desiged PM-generator with iron in the coils is the cheapest option for a certain maximum torque level. I have compared PM-generators with and without iron in the coils in chapter 7 of my public report KD 341.

Neodymium magnets don't loose their remanence if they are not short-circuited by an iron loop. They can loose their magnetism if the temperature becomes too high. The temperature at which this starts to happen depends on the magnet quality. This critical temperature will normally not be reached in a PM-generator of a small wind turbine, even not if the generator is short-circuited. I have measured PM-generators for short-circuit for which you have the largest counter acting magnetic field and they didn't become weaker.

[mbouwer]

@Adriaan Kragten,
So it's about making the best possible combination of magnet, lamellar package and coil.

example 5
[ .....So if you buy a generator, you have to measure the characteristics yourself..... ]

Building your own generator involves testing.

[MattM]

MattM,

What is "slowly"?  Is this something that happens over years, decades...  Do you have a source that describes this in more detail? Thx, ~ks

Sadly there is no formula I could find.  Magnets in a perfect state should lose 1% of their strength naturally every century is the only specific statistic I could find.  Heat, kinetic collisions, and being exposed to opposing (or dissimilar) magnetic fields are the ways magnets quickly lose charge, but even then the data tends to be specific to a magnet's material properties.  The last one would be like putting a neo against a ferrous-ceramic magnet, the latter will almost instantly begin to lose its field.

[Adriaan Kragten]

@MattM,

But also it has often been said on this forum that suitable generators for small windmills are not for sale.

(Attachment Link)

That isn't true. The Chinese company Hefei Top Grand supplies a large range of well designed axial flux PM-generators with no iron in the coils which can very well be used in a small wind turbine. I have designed several VIRYA wind turbines which make use of such a generator (see list with KD-reports on my website). I have even bought a small one and measured it (see report KD 595). But the characteristic of these generators is only given for a fixed resistance as load and this characteristic differs strongly from the characteristic of a battery load or of a load which is gained if the windmill is grid connected by an inverter. In all my KD-reports in which generators of Hefei Top Grand are used, I give a method how to derive the wanted characteristic for a battery load or for grid connection from the measured characteristic for a resistance load. I have advised Hefei Top Grand to measure their generators differently because no one is dissipating the generated power in a resistor but they don't respond to this advise. But many other Chinese suppliers of PM-generators give no measured characteristics at all. So measurements for a resistance load at least give some information.

[kitestrings]

I think the only other consideration with pre-manufactured unit is what the characteristics and quality of bearings are.  The ones I've seen are compact and nicely packaged, but I'm not sure how robust the bearings and castings/hubs that support them - probably fine if it is a geared, or chain-driven application where the loading is radial.  Most of the home-built axial's employ heavier, tapered roller bearings which can be sized to meet the rotor loading.

I suppose it could be coupled to the rotor, which is in-turn separately supported, but that adds components, cost and complexity; always trade-offs.