Re-inventing the wheel

[Tropical hydro]

Hi all,

      I am writting to float an idea I have about overshot wheels. We all know they are very efficient at converting water force into turning force but have a problem of having too low RPM. Alternators need higher revs to make electricity. I was thinking of putting magnets all the way around the wheel and then having small banks of coils, each being it's own alternator set up around the top of the wheel. This would artificialy increse the revs.

Do the coils care if it's not the same magnet passing it? A friend said that it would suffer from "leaking flux" (sounds painfull) Also he said I would have a hard time keeping the air gap constant. Well I can fix the engineering problems but the flux is a mystery to me.

The biggest problem with using an over shot wheel is the losses in efficiency due to the gearing required to spin a generator at an efficient speed. The 80% efficiency of the water wheel is pretty attractive but we lose so much efficiency in the gearing it's not worth the effort.

I propose using the wheel it's self as the rotor to overcome this. I will explain.

I have a small 12volt permanent magnet generator capable of producing approx 10 amps.

I pulled it apart and found that it had  4 magnets, each with 3 effective pulling sections.

This gives the generator 12 points where there is magnetic pull. Effectively 12 magnets.

The stator has 12 coils.

At 1000rpm 12000 magnets pass each of these coils producing 10 amps.

The big question is, does the stator care if they are the same magnets?

Does the stator have to be in a circle? Can it be in a semi circle?

If you put 401 magnets around a three metre wheel and had a group of 12 coils,

On a three meter wheel at 4 rpm, you would have an effective RPM of 1604.( If I have read my notes right)

 You could have many groups of 12 coils, each able to be moved in to service or out to stop them charging when flows are low. You would fit 35 groups of coils on this wheel giving 350amps (minus losses)  Of course you wouldn't be able to really get that but you could just keep adding groups of coils to suit the maximum flow rate.

I have based this on 100lsec.

The wheel is very powerful. In fact with just three buckets full at 100litres per bucket( it has an arm of 1.5 meters) you get a moment of 3254.98 ftlb. That converts to over 4.4kw or 5.78 hp.

I am keen to find out if this has been tried before and if I have missed some fatal flaw in my workings which will render this idea as just plain stupid.

Any ideas would be great....Thanks        KEN

 

[MountainMan]

Tropical,

Your basic thesis is sound.  Larger rotor makes for more magnet speed across the field of the coils per revolution.  No question there.

Also no problem not having coils "all the way around" your huge rotor.  Your suggested design would have several coils with magnets flying past them all the time.  Should work just fine.

The one problem I see is that you will be in the poor house buying all those magnets.  To make it work well, you should have several coils in the field of one or more magnets at all times, otherwise you will have dropouts in your final DC waveform (not a horrible thing, but not preferred).  That would mean your magnets still need to be relatively frequent along the enormous diameter of your rotor/wheel.  That's a lot of magnets if you use rare-earth.  Maybe you could find some cheaper ceramics, but then you loose much of your voltage gain by having less flux.

Your magnets do need to be on a solid steel backing of some sort.  If your wheel is wood or some such, it would not provide a magnetic path from magnet, through stator, and back around to the next magnet.  This is probably what your friend was talking about with flux leakage.  If you can deal with the rust problems of having a thick plate of steel around the diameter of your wheel to put  your magnets on, then I think you idea can work well.  But only if you can shell out several hundred dollars for lots of good magnets.

jp

[willib]

Ken  i agree with mtn man it would work with the right size coils to the size of the magnet .. BUT, it would indeed use a lot of magnets ..why not put an axial flux gen. on the central shaft??

Does it look anything like this?





i took this pic ,this past summer in North Carolina ..

[Tropical hydro]

Thanks for the input guys! very helpfull. I haven't built a wheel to try this on yet. I am still looking for someone with the right creek who is crazy enough to let me build this thing when there is other proven types on the market. Yes the magnets will be the biggest cost but,..... if it works..........all that efficiency! I currently live off a Hydro with a turgo wheel giving over 20kwhrs a day with only 60%efficiency. A wheel won't suit my place but it might suit other places if I can do away with the belts/gears etc.

I was planning on a ring of thick steel to attatch the magnets and spacing them the same as they were in the original alternator. If I made a disc as a separate item from the wheel itself, I could have adjustable braces going back to the shaft to allow for adjustment of airgap. It's still in the "kicking around" ideas stage but from your comments I think it may be worth trying! Thanks again....Ken

[Tropical hydro]

Thanks for the input guys! very helpfull. I haven't built a wheel to try this on yet. I am still looking for someone with the right creek who is crazy enough to let me build this thing when there is other proven types on the market. Yes the magnets will be the biggest cost but,..... if it works..........all that efficiency! I currently live off a Hydro with a turgo wheel giving over 20kwhrs a day with only 60%efficiency. A wheel won't suit my place but it might suit other places if I can do away with the belts/gears etc.

I was planning on a ring of thick steel to attatch the magnets and spacing them the same as they were in the original alternator. If I made a disc as a separate item from the wheel itself, I could have adjustable braces going back to the shaft to allow for adjustment of airgap. It's still in the "kicking around" ideas stage but from your comments I think it may be worth trying! Thanks again....Ken

PS: what is the alternator you speak of?

[willib]

I was trying to talk you out of it !!

the purpose of having a large wheel is to get the most torque from the central shaft , not the perimiter..this will also  allow you to concentrate your magnets and coils..what is the expected RPM of this machine?

[finnsawyer]

You could place the twelve magnets 30 degrees apart on the inner side of a large ring.  On the inner ring you could have several coils (sets) in that 30 degrees.  You then rectify the output from each set and feed it to a capacitor.  Add the outputs from all the capacitors.  Of course, you'll get better results with more magnets.

[Gary D]

Depending on the size waterwheel you use, there is an interesting link on a permanent magnet alternator that our hosts built... The alt. could be hooked directly to the shaft no gearing needed if you keep above 38 rpm's. They used a different unit to power the alt tho. I'll try posting the link below. (hope you don't mind Dan)  Gary D.

http://www.otherpower.com/scotthydro1.html

[Clifford]

Sounds like a lot of magnets...

But, of course, you could add as big of a stator assembly as you wanted (as long as you could adequately protect from "the elements").

You would have to make sure the wheel was rigid enough that it would be true, and you could maintain a minimal gap.

What about doing a single reduction gear off of the outside of the main wheel....

Say...  10' to 2" would give you about a 60 to 1 gear reduction.  

No idea where you could find a huge gear that wouldn't be succeptable to water damage though.

[Oso]

I do see certain flaws in your reasoning/calculations. Whether they are fatal or not is another question.

Before I go into those, I will just mention that I love waterwheels. There is a certain majesty to them that I do not find in a turgo or pelton runner.  However, a Turgo runner should have an efficiency of over 90 percent if properly installed in an appropriate application. (your 60 percent figure most likely is a system figure, turgo plus alternator) So, I am not sure how you expect to design a system based on a 80 or 85 percent efficient water wheel and beat the turgo.

Your horsepower figure of 5.78hp is too high. 100L per second is 1585 gallons a minute. One of my references (for an 85 % efficient wheel ) lists 3.22 hp for a 10 foot (3 meter) wheel using 1500 gpm. According to their figures, you would need an 18 foot (5.4 meter) wheel to produce 5.79 horsepower with that flow.

I think your error was in assuming that you could get 3 buckets worth of water at full radius.  I don't know how many buckets you are assuming, but let us use 12 as an example.  Draw a clock face.  Your water comes in from the left and spills into the bucket at 1 o'clock.  The wheel will rotate clockwise.  Draw a radius line from the center of the wheel/clock to the bucket in the 3 o'clock position. This is the only full bucket with a full 1.5 meter radius lever arm. Now draw a vertical line connecting the 2 and 4 o'clock buckets. This vertical line will intersect the radius at some point roughly 1 meter from the center point. That is the effective lever arm for those two buckets.  If you draw another vertical line from the 1 and 5 o'clock buckets, it will have an effective lever arm of roughly .5 meter and the weights are reduced because 1 is filling and 5 is spilling. So it is essentially impossible to get 3 buckets to full radius lever arm, on a wheel of this size.  

(I also am missing the correlation of how based on a 100 liters a second and a 10 foot wheel, you developed the figures of 4 rpm or a bucket size of 100 liters, but will base my calcs on those figures.)

The largest error is your counting magnets and coils to leap to an "effective 1604 rpm".

If your small motor/alternator that is turning 1000 rpm has a 5 inch diameter rotor, the circumference is 15.70795 inches. At 1000 rpm it is 15, 707.95 inches a minute or 261.7992 inches a second.  This is your "coil/magnet passing speed". Assuming that you mount the magnets/coils near the rim on your 10 foot water wheel, I came up with 24.72071 inches per second "passing speed".

If you would rather count impulses, in the alternator you get 12 magnets under 12 different coils or 144 impulses per revolution. 1,000 rpm is 16.6667 revolutions a second so 144x16.6667 equals 2400 impulses a second. On the water wheel you have 12 coils and 401 magnets. In one second, you will only get about 26 magnets passing under each coil. 26 times 12 equals 312.

So whether you look at "passing speed" or counting impulses, the water wheel mounted coils/magnets are not going to have an "effective 1604 rpm" or come anywhere close to output of the small alternator turning at the high rpm.

Mountainman made the comment that larger rotor equals higher passing speed.  That is true if you are driving an alternator from the shaft, at X rpm.  With waterwheels they are driven on the circumference. So as a rule of thumb, two different sized but otherwise identical wheels driven by equal water flows, should have identical "passing" or rim speeds. So, with identical rim speed, the smaller wheel (smaller circumference/radius) will have a higher rpm than the larger wheel.

I would actually encourage you to build a water wheel or several of them. The experience of building/operating water wheels should give you a wonderful appreciation of several thousand years of seat of the pants engineering. They are a much more refined machine, and far more complex than their simple appearance.  

The best reason for building a water wheel driving a generator is simply because you like a wheel. It is going to be less efficient than the more modern turbines. But with enough water flow, it can produce enough electricity. So, assuming that the excess flow is available, to offset the lower efficiency, do we really care ?

The "wheel" has that majesty or a certain ambiance that a turbine just doesn't have.  

[willib]

i have other pics of the water wheel if ya want to see them?

[Tropical hydro]

Thanks again guys,

                  just got home from work....I will try to digest the info from Oso tomorow and reply properly then as I am a bit vague at the mo....Great stuff though.....Thanks

[Gary D]

Hi again, is there a purpose for having a specific number of 'pulses' per second? If you are charging batteries, a 3 phase charging system will work fine. It doesn't matter how fast the alternator is spinning- to a point. Dan B. used only 24 magnets (12 per rotor) to get to charging voltage, but with the power available, a larger rotor with 24 magnets (48 total) probably would be needed (and twice the coils). OSO is correct about the shaft horsepower on a 10 foot wheel at 1500 gallons per minute... given the wheel is 85% effecient. What I like about waterwheels is they don't require different nozzels for differing flows. If the water level rises feeding the wheel, any increase in speed will increase the amps. produced. No clogged nozels etc. I'll post a link to The waterwheel factory where I confirmed his figures. If you check around the site(home link at the bottom of page) he has plans for sale (with parts list) to build your own.... Good luck on your quest. Gary D.

http://waterwheelfactory.com/HP%20Table.htm

[zukecanoe]

So, I think I understand reasoning here... but, it somehow doesn't sink in... I have never built an alternator or a waterwheel so my comments here are "unencumbered" by experience.  In an example {admittedly the waterwheel I want to build},  If I had 2" magnets at 1" apart all the way around a 17ft diameter rim {640" circumference} I would have 200 some odd magnets  {I know cost prohibitive, but this is theory not practice}.  a Fitz-type metal waterwheel has a peripheral velocity of about 6 feet per second on most any wheel over 8ft in diameter...{supposedly 93% efficient at full and partial gate} so, that would be 18 pole changes a second...per coil...  I think... anyway intuitively that seems like a bunch of impulses since you have lots of space for coils... no?

I can imagine dozens of other problems that prevent the practical application of doing this beyond cost: air gap adjustment over a 17ft disk, iron in the silt/sand, coil heat, ice, water, rust, ... and lots of other problems, but it seems like, a theoretical alternator 17ft in diameter would have enough "rpm" even with the shaft at of 8 rpm to 10 rpm... no?

zukecanoe

[DBuller]

Hey Tropical,

   I have a new form of wheel, it gets around the gearing problem by using a spiral pump to change the slow motion into pressure that spins a turbine. Take a look at http://WildWaterPower.com

-Dennis

[alancorey]

I'm wondering what progress is being made on this, if any, after all this time.

I also like waterwheels and I've been weighing many of the same questions.  I have a brook with about 150 gal/min flow, and about 140 feet of head if I stretch a penstock for about 1400 feet, but I can get about 50 feet of head in the steepest 400 feet so I'm thinking about a turbine in this part and some waterwheels elsewhere.  I've also got a small seasonal brook with about the same head and maybe 1/10 that flow.

I've been pondering the expense of magnets around a large wheel (probably only about 6 feet diameter though) and alternatives other than gearing.  I just finally hit me that if you put only 4 magnets on the wheel at 90 degree spacing and a set of coils covering 90 degrees of arc you could get almost the same effect as full coverage but have the coils contained in a smaller area that would be easier to keep weather out of.  If you had coils at something like 4 inch spacing you could get enough cycles per wheel revolution that the ripple frequency should still be fairly high and easy to filter out.  Each coil would probably need its own bridge rectifier or at least a pair of diodes.  If you wanted something more like 3 phase you could put 12 magnets on the wheel.  I've been inspired by the magneto on a Briggs & Stratton flywheel, with 1 magnet passing 1 coil.

I don't want to invest a lot in magnets, at least for experiments that might not work out.  Lately there's a push toward lots of high-powered expensive magnets, but I see the windmill experimenters have started to hit the point where their mills won't turn their alternators.  I've been scrounging some hard drive magnets and found them to be strange little beasts with apparently 3 poles in the length of each.  That's from some 3-1/2 inch hard drives, but the nicest ones I've seen come from some older 5-1/4 SCSI full height high-end drives meant for servers.

Turbines have their place, but I also like the big, slow wheels.  I'm interested in larger diameter axial flux alternators for these and Savonius rotor wind machines.

  Alan