axial generator with lamination core

[Adriaan Kragten]

I have written several reports in which a pitch control system is described for a small wind turbine like KD 437, KD 622 and KD 654. It is not simple to derive the mathematical description of the functioning and mechanically it is also not simple. So for small wind turbines, I would advise to turn the rotor out of the wind. An overview of methodes how this can be done is given in KD 485.

But all big wind turbines have a pitch control system as turning a rotor with a diameter of about hundred meter out of the wind or using very big vanes is practically impossible. This has to do with scale laws but also because one wants that the pitch movement can be activated by different signals like the rotational speed, the generated electrical power, the generator and gear box temperature, vibrations and expectation of heavy storms for which the rotor is stopped by turning the blades about 90° into the vane position. Normally this works well but it has as disadvantage that the movement is steered by a computer. If there is an error in the chain of activities which powers the blade pitch motor, then the blades won't move. In this case the rotor can speed up to dangerous high rotational speeds. So this system isn't fail safe and several big accidents have happen. I prefere systems which are directly activated by the rotational speed or the rotor thrust but it appears to be very difficult to realise this for a very big wind turbine. A reason is that turning the blades requires a rather high torque and if the electric system fails, this torque can no longer be supplied.

[mbouwer]

To get a small safe self-build windmill in the garden and use the energy:
All around us professionals show us how to do this.

[Adriaan Kragten]

To get a small safe self-build windmill in the garden and use the energy:
All around us professionals show us how to do this.

(Attachment Link)

If you would accept the same risk as accepted for big wind turbines, there is a simple solution. Assume that you have a 3-bladed rotor with fixed blades and no rotor eccentricity. Assume that the vane is coupled to the head by a worm wheel gear box with a gear ratio of at least 40 and that the vane is driven by an electric motor. For this big gear ratio, the motor can drive the vane but the vane can't drive the motor. The vane can move over 90° and the movement can be steered by any signal. So if you measure the wind speed, the rotor can be turned out of the wind above a certain wind speed. If you measure the rotational speed, the rotor can be turned out of the wind above a certain rotational speed. If you measure the power, the rotor can be turned out of the wind above a certain power. If you want to stop the rotor for whatever reason, you can turn the vane 90°. Then the rotor will be out of the wind for any wind speed and for any wind direction. This is a very safe position for big storms. If you want the rotor 45° out of the wind at stormy days or if the battery is full, you can turn the vane over 45°. This all looks very nice.

However, you need power to drive the motor. If the rotor is 90° out of the wind, it generates no power and so the power must come from a battery or from the grid if you want to turn the rotor in the wind again. If there is something wrong in between the motor and the device which steers the motor, the rotor will not turn out of the wind resulting in a very high rotional speed and a very high thrust at high wind speeds. I would never accept the risk that this would happen but for big wind turbines, this policy is common practice. So don't think that you get an ideal small wind turbine if you copy the safety philosophy of a big one.

[mbouwer]

There are so many windmills around us and almost nothing ever goes wrong.
So why not embrace modern technology?

[Adriaan Kragten]

About 40 years ago there was a group of about 20 people in The Netherlands who had built their own wind turbine. We had a meeting about the progress every month in Utrecht and in the mean time we visited each others wind turbine. Most of those wind turbines had no automatic safety system activated by the rotational speed or the thrust. One farmer build a very nice 3-bladed rotor with a diameter of about 8 m and positive pitch controle activated by a hydraulic cylinder which drives a pin hrough the hollow rotor shaft. The pin was connected to levers at all three blade roots in the same way as it was done by Vestas at that time. He could activate the system when the rotor was running and it worked really nicely. However, it worked only if he gave the order to increase or decrease the blade angle. I said to him that it must work automatically if the wind speed or the rotational speed was too high but he didn't agree. Some months later there was a big thunder storm and he wasn't at home. The rotor thrust became that high that the tower broke at the base. The whole wind turbine felt on his greenhouse and this cost him a lot of money. Almost none of the twenty windturbines except mine were alive about five years later. If you look at Google you can find many movies of big wind turbines which have failed. The law of Murphy claims that if something can fail, once it will fail. Big companies can take precautions to prevent failure and mostly this works. But amateurs can't forsee what happens to their wind turbine in thunder storms so they certainly should not take a safety system which isn't fail safe by itself.

The inclined hinge main vane safety system is technically simple and everybody copies it. But it is very difficult to predict what happens at very high wind speeds under dynamic load. I have built two wind turbines with this system when I was working at the University of Eindhoven. The CWD 2740 water pumping windmill and the 1 m diameter Wesp electricity generating windmill. The CWD 2740 had a vane which locked automatically if it has moved about 100°. This makes the windmill safe at thunder storms. But it must be unlocked if the storm is over. This wasn't difficult as the lattice tower had a height of only 5.5 m. It was even possible to put the vane in the locked position by climbing in the tower and put the right hand at the main vane and the left hand at the side vane (the rotor had no eccentricity because of the pump rod). Then some swings were given at the main vane untill it locked. The functioning of the Wasp has been described earlier. It is not for nothing that I have developed the hinged side vane safety system which is much more stable at high wind speeds. But every system has its advantages and disadvantages and you must have a lot of knowledge to make the correct choise for your situation. I think that pitch control is too difficult for most amateurs. If you don't have the knowledge, you must trust others or learn from your own mistakes.

[kitestrings]

I agree with Adriaan.  Even though failures may be rare, unfortunately they are usually catastrophic.

I worked for some time on grid-tie, induction turbines (up to about 60 kW).  They were downwind turbines and used dynamic braking for high-wind shut down.  Both the start up and high wind shut down were controlled by an anemometer signal for windspeed.  It generally worked very well, but for safety, they also employed tip-brakes.  In this case they were different from those presented earlier.  They basically were aluminum plates, with a mid-mounted hinges at the tips of the blades.  There was a latch that released the plate if centrifugal force exceeded design level.  These were independent of controls, electronics and motors, and work very well if other systems failed; and they did.  Unfortunately, the tip brakes also added another device that had to be regularly maintained, in a rather difficult location off the tower.

[mbouwer]

So let's focus on modern technique fail safe designs.

[Adriaan Kragten]

About three years ago I came in contact with the Dutch company Dutch Wind Design (DWD). This company had the intention to design a medium size 3-bladed wind turbine with a rotor diameter of 18 m. I have made several design reports of the rotor for them. This company was aware that big wind turbines have a safety system which is not fail safe. So their wind turbine should get a safety system with pitch control for which the blades go back to the vane position (with a blade angle of about 95°) when something goes wrong in the electric or hydraulic system which steers the pitch control. Although I designed the rotor geometry, I didn't design this pitch control system. That was their job.

During the first year of their activities they put a lot of effort in the design of the direct drive PM-generator. Several times I have asked how they made progress with the design of the safety system but this appeared to be difficult and I never got a good answer. About one year ago they no longer reacted on my e-mails and their e-mail adress is no longer existing. So I think that they have stopped their activities. This demonstrates that designing a rather big wind turbine with a fail safe pitch control safety system isn't as easy as it seems to be.

I have designed a pitch control safety system for a 2-bladed rotor with a rotor diameter of 15 m and a design tip speed ratio of 8. This design is explained in public report KD 437.

I have also designed a safety system for a 3-bladed rotor with fixed blades which can be used up to rather big rotor diameters. It is the pendulum safety system with a torsion spring as described in my public report KD 439. This system is used for the VIRYA-10 wind turbine as described in my public report KD 715. This wind turbine uses a double vane to keep the rotor in the wind. I think that this system can be used up to a rotor diameter of about 20 m but for such a big rotor, it seems better to keep the rotor in the wind by a servo rotor or by an electric motor steered by a wind vane.

[mbouwer]

On their site DWD shows this design.

Good find:
Blades do not extend to the middle of the rotor; I think that mounting the bladeroots near around the shaft and the nacelle does not bring much anyway ( but a lot of swirls )

[mbouwer]

The spokes will also give swirls.
Perhaps a cone in front of it would be better.

[Adriaan Kragten]

The generated power is proportional with the swept area. The swept area within the ring generator is only about 4 % of the swept area of the whole rotor for the given diameter of the rotor and the generator. So doing a lot of effort to generate this little power is mostly not worth while. To keep the inside of the generator open has as advantage that the generator is well coolded at high powers.

[mbouwer]

That is why the Lagerwey design is so nice where the lamination package as cooling fins forms the exterior.

[mbouwer]

The Piggot generator design is also strong in cooling.
A nice base to build on.

[mbouwer]

[mbouwer]

Happy New Year
May knowledge areas find each other on this forum in the design of a small self-build windturbine.

[mbouwer]

As a self-builder there is always the chance to put all your energy and creativity into impossible constructions.
Let's start from professional proven designs.

[mbouwer]

As an example setup I made a working model of a Nordex.
The blade adjustment is here manual and the blades should actually be a bit slimmer.

[mbouwer]

On      I now read:

Small wind turbines in Bayern up to 15 m high without permission

[mbouwer]

It's also great to see on the site of https://kitex.tech/
how they are building small turbines.

[mbouwer]

Nice features we could discuss like large blade diameter, low speed, light transmission, blade adjustment and control.
I also wonder which generator is used.

[mbouwer]

What the rotor also seems to want to emphasize is that you can get very little energy from the center of the blade diameter.

[kitestrings]

Okay, I confess, I like the name 

This is actually pretty cool though.  They've got some interesting things going on here, and I've often thought there might be a knitch for a small, semi-portable wind turbine.  They also seem to be realistic about the rated power and expected output.  All too often we see start-ups with over the top projections, and lots of glossy marketing.  None of that here from my cursory look anyway.

On the down-side, I don't like the belt-drive gearing.  This seems like a big power suck to me.  And, I wonder if much of this hardware and moving parts would endure in our climate.  It was about -10 degF (-23C) here this morning.

None the less, a very interesting effort.  I wish them luck.  Thanks for sharing mbouwer.  ~ks

[kitestrings]

I just realized that they have a larger 11 M, 5 kW version (at approximately $20k US).  It will be interesting to see how they do with it all.

[mbouwer]

After all not a bad idea to use guy wires to tie the blades.

[mbouwer]

On the down-side, I don't like the belt-drive gearing.  This seems like a big power suck to me.  And, I wonder if much of this hardware and moving parts would endure in our climate. 

Better is to build in the drive train and protect it against the weather.

[mbouwer]

In  中国中车 I read about their new 20 Megawatt wind turbine.
Curious about the generator and the drivetrain.

[mbouwer]

Also impressive is this floating Minyang of more than 16 MW with 2 rotors on one mast.

[kitestrings]

Colossal.  It's hard to get a sense of this scale.  There are two wind farms near me, one with 21- 3 MW units.  This capacity could be me with three of these.  With a ballpark estimate for the rotor speed and capacity factor, I think one of these would crank out about 30-40 kWh per rotation.

[mbouwer]

@ Kitestrings,
The question is what would give you more yield.
The turbine you have now or two smaller diameters on your mast.

[kitestrings]

Interesting question mbouwer.  As I get older, I'm starting to wonder if a smaller, lighter unit (or multiple units) on shorter, tilt-up towers might be easier to manage.

Of course it is not linear, right.  Our turbine is 5.6 m (15').  At 8 m/s (18 mph) we're seeing about 1,500 watts at the rotor.  If we had a 2.3 m turbine, same conditions, we might see 450 watts, so even with two, maybe 2/3rds the output.

I do wonder about using lighter weight materials.  The Sencenbaugh that we ran here for nearly three decades used some really nice cast aluminum housings & structural components, and the whole thing was only about 82 kg (180#).  It was 3.6 m and rated at 1 kW.

Our turbine is ~180 kg (400#).  The tube/adapter another 78 kg (170#), and required a much heavier tower.

[Adriaan Kragten]

There are two aspects of scaling on the generated power. If you have a certain windturbine + tower and you scale all dimensions with a factor 2, the swept rotor area increases by a factor 4 but the mass increases by a factor 8. As the generated power is proportional with the swept rotor area, the prime effect is that the amount of energy per mass generated at a certain wind speed, decreases if you scale up. However, the tower height also increases by a factor two and at higher heights you have higher wind speeds. As the power increases by the cube of the wind speed, the positive effect of a higher tower can be stronger than the negative effect of the lower power per mass at a certain wind speed. But the increase of the wind speed for increasing height is logaritmic and once you have reached a tower height of about 50 m, the wind speed is increasing only very little above 50 m. This cubic increase of the mass can be compensated by using hollow blades or by using materials like carbon fibre which can have a large stress for a low weight.

The second effect is the Reynolds number. The local Reynolds number for a certain blade section is given by formula 5.5 of my public report KD 35. The Reynolds number is proportional with the chord c. If you scale up a certain rotor by a factor 2, the chord is also scaled by a factor 2 and so the Reynolds number at a certain wind speed is also scaled by a factor 2. The drag coefficient of normal airfoils becomes lower as the Reynolds value is higher and so the losses due to aerodynamic drag are smaller for big than for small rotors. This is the reason why big rotors have a maximum Cp of about 0.5 and why small rotors have a maximum Cp of only about 0.4. If the Reynolds value is smaller than about 10^5, normal airfoils stall already at low angles of attack and the maximum Cp can therefore even be much lower than 0.4 for very small chords. For very small rotor diameters one can better use the 7.14 % cambered sheet airfoil because the sharp nose of this airfoil makes the boundery layer turbulent and this prevents stalling.

[kitestrings]

Thanks for this Adriaan.  It's helpful to me better understand the variables.  Can you expand on one thing for me, that is to describe a bit more the Reynolds number.  Broadly I understand that larger diameter turbines have higher Reynolds, therefore lower drag, and I assume a more efficient transfer of power.  The fluid, the air in this case doesn't change though, so what are the factors that influence the Reynolds number.  Is it only the profile and chord thickness of the airfoil, or do things like the surface smoothness effect Reynolds?

And related, I've read portions of KD 35, I'm confused by the units for Reynolds.  It's in the range of  5 x 10^5 for a small turbine, but in in the definitions it shows (-).  I'd thought somehow this just meant the result was a negative number, but I notice for example the number of blades it is also (-), so perhaps this has another meaning.

Mbouwer, looking at the floating 2-rotor design, I assume there could be some additional torsion loads, and possibly vibration/harmonics introduced?

~ks

[Adriaan Kragten]

Thanks for this Adriaan.  It's helpful to me better understand the variables.  Can you expand on one thing for me, that is to describe a bit more the Reynolds number.  Broadly I understand that larger diameter turbines have higher Reynolds, therefore lower drag, and I assume a more efficient transfer of power.  The fluid, the air in this case doesn't change though, so what are the factors that influence the Reynolds number.  Is it only the profile and chord thickness of the airfoil, or do things like the surface smoothness effect Reynolds?

And related, I've read portions of KD 35, I'm confused by the units for Reynolds.  It's in the range of  5 x 10^5 for a small turbine, but in in the definitions it shows (-).  I'd thought somehow this just meant the result was a negative number, but I notice for example the number of blades it is also (-), so perhaps this has another meaning.

Mbouwer, looking at the floating 2-rotor design, I assume there could be some additional torsion loads, and possibly vibration/harmonics introduced?

~ks

Formula 5.5 of KD 35 gives the local Reynolds value for a certain airfoil station, so for a place at which the local speed ratio is lambda r design. The derivation of this formula is given at page 30 of KD 35. This derivation starts with the general formula of Reynolds
Re = W * c / gamma (I can't type Greek letters at this forum). W is the relative velocity in m/s which is felt by the the airfoil, c is the chord in m, gamma is the kinematic viscosity of air in m^2/s. The kinematic viscosity of air is about 15 * 10^-6 m^2/s. The relative velocity W depends on the wind speed V and on the local speed ratio lambda r design (see formula 5.12). If you follow the steps at page 30 of KD 35, you see how I came from the general formula for Reynolds to the specific formula 5.5 for a certain blade section.

The surface roughness of the airfoil is not incorporated in the Reynolds number but it can have a similar influence on the drag coefficient especially for spheres or round pipes. In figure 10 of my report KD 213, I give the influence of the Reynolds number on the drag coefficient of of a smooth pipe. Curves for other roughnesses are given in report R 999 D but this report is no longer available.

The Reynolds value is dimensionless and so the dimension is (-). If you go to the original formula of Reynolds given at the top of page 30, you find that the dimension is m/s * m / m^2/s = m^2/s / m^2/s = (-). You can't use formula 5.5 for a dimension analysis because the kinematic viscosity of air, gamma, is incorporated in the coefficient 0.667 * 10^5.

In my report KD 598, I describe a floating water turbine. The rotor of this turbine is designed using the same formula's as used for wind turbines. However, for the calculation of the Reynolds value you have to use the kinematic viscosity of water.