Overcoming high starting torque

[Gary]

I am curious about solving the high starting torque requirement caused by cogging, the force of one or even worse more than one magnet's attraction to the steel core of a coil winding when stopped.   It seems that serious comprises are made in permanent magnet alternators.  Overcoming a force to start motion in machinery is not unique to generating windmills.
Although on a completely different scale and differing requirements several internal combustion engine starting methods have been used.  Tiny airplane model engines – Turn the propeller backwards winding a spring, then releasing it.  Big radial engines – mechanic inserts a crank in the side of the cowl and apparently winds up a free-wheeling flywheel to high speed then engages it like our common automotive starters.  A more interesting method was inserting something like a blank shotgun shell and firing it into a gas turbine powered starter.  Then something called a “pony” motor for some unknown reason was used on large crawler tractors, Caterpillars.   A small gasoline powered engine with a rope pulled starter was started and warmed up then engaged to the engine starter.  In a somewhat similar mechanism was the magneto ignition system.  To generate a high energy pulse for the spark plugs from a slow turning starting engine a mechanism in the magneto wound up a spring and released it at the proper time causing a winding to pass a permanent magnet at high speed.
In each case low energy was stored over time and released quickly.  It looks like that's exactly what is needed (and available) in wind and water powered machines to overcome cogging.  Being more mechanically than electrically inclined I have been looking for a way to achieve this with hardware.  Perhaps a small independent freewheeling propeller and flywheel could accelerate until reaching a desirable speed then centrifugal force extends a “tooth” on the flywheel that engages with the main propeller and gives it the needed “kick” to start.  It would then disengage and repeat until wind is fast enough to keep the main propeller turning.
I am sure it could be done electrically also.  A miniature propeller driving a very small generator developing enough energy to kick start the main rotor. 
Any thoughts? 

[electrondady1]

build an air core alternator. use two magnet rotors with the stator in between them. no cogging

[Adriaan Kragten]

The starting wind speed depends on two things, the peak on the sticking torque of the PM-generator and the starting torque coefficient of the rotor.

The starting wind speed can be calculated with formula 8.6 of my public report KD 35. The sticking torque of the generator depends on the friction torque of the bearings and the seal on the generator shaft and on the iron losses of the stator if the stator contains iron. Generators with no iron in the stator have the lowest sticking torque but if there is no iron in the stator the magnetic flux flowing through the coils is rather low and big thick magnets are needed to get an acceptably strong flux. So most electric motors and many PM-generators have stators with an iron stamping made out of thin sheets to minimize strong eddy currents. However, a steel stator stamping may result in a large peak on the sticking torque if the armature isn't designed well. You get the highest peak on the sticking torque if the armature pole number is equal to the stator pole number and if the armature poles and the stator poles have the same shape. In my public report KD 341, I give several methods how the peak on the sticking torque can be reduced.

The starting coefficient of the rotor depends very much on the design tip speed ratio. The higher the design tip speed ratio, the lower the starting torque coefficient. The starting torque coefficient of a certain rotor can be calculated with formula 6.12 of KD 35. All public KD-reports can be copied for free from my website: www.kdwindturbines.nl at the menu KD-reports.

The starting wind speed Vstart is the wind speed for which the rotor starts rotating. The cut-in wind speed Vcut-in is the wind speed for which the Pel-V curve of the wind turbine starts. It is useless to have a Vstart which is lower than Vcut-in because this means that the rotor is turning but that the rotational speed is too low to generate energy. This is specially important for battery charging wind turbines as charging only occurs if the open DC voltage is larger than the battery voltage. But if Vstart is higher than Vcut-in, there is hysteresis in the Pel-V curve because power is only generated for wind speeds in between Vstart and Vcut-in, if the wind speed has earlier been larger than Vstart.

Starting of the rotor is a dynamic behaviour. If the rotor torque is larger than the sticking torque, the difference is used to accelerate the rotor. It depends on the mass moment of inertia of the rotor and the available nett torque how long it takes until the rotor runs that fast that the rotational speed is high enough to start charging. So if the wind speed is only a little higher than Vstart, it may take a long time for the rotor to reach the required rotational speed.

[topspeed]

Doesn't the CUT IN speed relate to this phenomena ?

[Adriaan Kragten]

Doesn't the CUT IN speed relate to this phenomena ?

No, the cut-in wind speed is the wind speed for which the Pel-V curve starts. So the cut in wind speed is the wind speed for which an unloaded rotor runs that fast that the open DC voltage is the same as the battery voltage. But this requires a rotating rotor and the required rotational speed is only reached if the rotor has started earlier. The rotor starts at the starting wind speed which is the wind speed for which the rotor torque becomes equal to the peak in the sticking torque of the generator. For some generators, the sticking torque can rise faster than the rotor torque at the starting wind speed and in this case the rotor will rotate only very slowly at the starting wind speed. The sticking torque rises rather fast for PM-generators with an iron stator stamping and for a winding which is rectified in delta (see measurements KD 78). This is because higher harmonic currents can circulate in the winding. So for such a generator, the real starting wind speed can be somewhat higher than the value which is calculated using formula 8.6 of KD 35.

The starting torque coefficient for a HAWT is determined by the blade chord of all blades together and by the blade angle. The total blade chord of all blades decreases about quadratic with the design tip speed ratio. The blade angel decreases about linear to the design tip speed ratio. The final result is that the starting torque coefficient decreases about with a factor 8 if the design tip speed ratio is increased with a factor 2. Big modern wind turbines have a rather high design tip speed ratio of about 8 and the starting torque coefficient is therefore very low. The sticking torque of the rotor bearings and of the generator is rather high, especially if a gear box is used. So for these rotors, the starting torque coefficient is increased by increasing of the blade angle. If the rotor is accelerating,  the blade angle is slowly reduced up to the optimal value for the design tip speed ratio. But this requires a rather complex pitch control and steering mechanism and most small wind turbines have a rotor with fixed blades. Therefore generally the design tip speed ratio for that kind of rotors is chosen not higher than about 6.

The design tip speed ratio is chosen as high as possible because this gives the highest rotational speed at a certain wind speed and rotor diameter. The generator dimensions and so the generator costs decrease if the generator is used at a higher rotational speed and so at a lower torque level for a certain power.

[Gary]

Thank you.  My question is not what it is or how much it is BUT HOW TO OVERCOME IT ELECTRICALLY or mechanically.  So when the blade is stopped a device repeatedly applies a starting torque to start it spinning. Like someone standing there hand starting it every time it stops.  I know most of the time there won't be enough wind to keep it turning but when there is I want it starting.

[Adriaan Kragten]

I think that increasing the blade angle, like it is done for big wind turbines, is the only way to increase the starting torque coefficient of the rotor. For Darrieus rotors, which have a negative starting torque coefficient at low tip speed ratios, the generator is used as motor to start the rotor. This can also be used for a HAWT, but every time the rotor starts this way, some energy is used. It is also rather complicated to use a 3-phase PM-generator as motor as you need a 3-phase inverter with a variable frequency. It is much simpler to use a generator with a sticking torque which is that low and a rotor with a starting torque coefficient which is that high that the rotor starts at an acceptably low wind speed.

[bigrockcandymountain]

Overcoming high starting torque is best done by designing a generator with low starting torque.

On an iron core coil, the magnets can be skewed slightly, resulting in a very low or non existent cogging effect.

Search for decogging tutorial written by dinges.  That document explains it way better than i can. It is focussed on decogging induction motor conversions, but the principles apply to pretty much anything. 

I followed it for my motor conversion and ended up with almost no cogging and good startup performance with a less than ideal blade profile. 

[bigrockcandymountain]

http://www.otherpower.com/images/scimages/3538/decogging_tutorial_V1.pdf

[mab]

I suppose you could have some sort of wind speed sensor (anenometer i guess) which could trigger a starter circuit:- maybe six relays driven by a micrcontroler which could close momentarily in sequence to connect the battery side of the rectifiers to each of  the three generator phases in turn for a cycle or two. Don't know how well it would work in practice - you wouldn't know where the generator had stopped so you might be starting the relays at the wrong phase and start it moving the wrong way initially.

You'd probably need series resistors as you'd only want enough current to overcome the sticking torque; that combined with the positive torque contribution from the wind might make the momentary reverse start irrelevant.

Not sure how easy it would be to implement though - i'd probably follow the advice of the others and go for a low-stick generator if it were me.

[Gary]

Thanks for working on it gentleman.

[richhagen]

Two options that I can think of, both already mentioned here.  First is skewing the magnets, this may lower the peak efficiency in high winds, but most folks with turbines are more worried about the low end of the spectrum.  Look up Zubbly's files on motor conversions on this site.  Sadly he has passed away, - has it been 12 years already? - but my Canadian friend Wayne Abraham posted very detailed instructions for motor conversions which made adjustments to avoid cogging.  Look at the Dan's or Scoraig Winds designs of those.  Most folks are better off with, and would rather have power most of the time when winds are low, and when it is super windy, most folks are looking for ways to get rid of the extra power when it is very windy anyway.  Second solution is an air core alternator, Requires more magnet and is generally more costly, but offers no startup cogging and also no iron losses in the stator.  If SparWeb is around today, he has a good motor conversion flying up in Alberta - named Spirit of Zubbly - and can summarize his real world experience with it as he has had it up for years and just put new wood blades on it this summer.  The issues that I see with a "mechanical" solution that you inquired about are the addition of more parts to break with their added complexity and the fact that you still will have cogging losses at slow speeds where you can not really afford them.  Rich

[JW]

Look up Zubbly's files on motor conversions on this site.

https://www.google.com/search?sitesearch=https%3A%2F%2Fwww.fieldlines.com&q=Zubbly%27s+files+on+motor+conversions

https://www.google.com/search?sxsrf=ACYBGNQC4GDH886CpV40VI9wakOd7V163A:1577597484777&q=Zubbly%27s+files+on+motor+conversions+on+this+site.+site:https://www.fieldlines.com&tbm=isch&source=univ&sa=X&ved=2ahUKEwii0rbNkNrmAhVPaM0KHd9xDzcQsAR6BAgKEAE&biw=1331&bih=835

[Gary]

Thanks JW but I spent considerable time looking through the search results with no joy.  It's easy to find "I'm trying this" but no "I finished and this is the result".  I guess it's in there somewhere but I didn't find it.

[Gary]

The issues that I see with a "mechanical" solution that you inquired about are the addition of more parts to break with their added complexity

I use a tractor regularly that is 69 years old and got to see a clock that has been running a lot since since built in 1386 at Salisbury cathedral.  Not unusual to see cars with 230,000 mi.  It doesn't necessarily follow that  "mechanical" equates to more parts to break.

[Karana]

I have been concerned about the same issue of torque and needing a good push to get the machine turning. Once it does, there is generally no problem. I have carved a number of blades with more severe angles to ensure a good start every time - sometimes while sacrificing the ability to have a faster blade. But in the end, it works out. I realized that my blades were turning just fine at a slower speed and still giving me the needed output. I know it can still be improved upon, but my very clunky looking blades are great at starting now.  I am assuming that you are using - Alton's measuring tool to do the calculations. alton-moore dot netwind_calculations) You can select the tsr and with enough experimentation you can see what angles will provide the kind of initial inertia you need. Like Adriaan said, once the rotor has that initial inertia it will keep moving. I am building a iron-core unit now and Hugh Piggott advised that I allow for a greater distance than if I had an air core. Or rather that I could allow for a greater air gap. Perhaps that will also help with you initial push - short of training some pigeons to give it a shove.

[SparWeb]

Hi Gary
Sorry I missed this for so long.

I've been running WT's with motor-conversion alternators (3 of them so far) and I haven't had any serious start-up problems due to cogging.  I've put magnets on the rotors in a skewed orientation and every way I've done it, the cogging hasn't been so bad that it interfered with starting.  Plenty of time is spent by my WT's just spinning in hardly enough wind to rise up to cut-in.  I don't get the cogging down to zero, but it's low enough that the machine is fine.  I would have spent more time investigating this if I'd ever had a serious problem as a result of it.

Take a look at the blades I've made and my methods of converting the motors: www.sparweb.ca 

I use a blade dimensions & twist calculator like Alton Moore's, just like Karana just recommended (sorry about the link restriction Karana, it's just for newbies ) )  http://www.alton-moore.net/wind_turbines.html