Christmas Windmill Time

[taylorp035]

I re-ran the points and the results were a bit better.  I'm still dissapointed that the "10A" BMS unit is really only good for about 2A continuous.  It might be damaged, but I already blew a MOSFET(?) off the other BMS unit I had.   I'm thinking I need to have something else, like a relay or something that shuts off at 4.2 volts that capable to something closer to 25A continuous.  I'm glad I decided to bench test this today, else it would of been a pain to fix out in the cold since it would of failed almost immediately.

The power output is pretty good in my opinion compared to what I was hoping for.  I bet it will be good for 50-100 Ah easy on a mildly breezy day.

[SparWeb]

Since it's a spreadsheet, I won't be able to see inside your calculations, so I'll simply say that the power calculations that refer to the blades look too low.
Your blades have about 6 feet diameter to work with, right?

In a 10m/sec (22 mph) wind they will have about 1.5 kW coming at them, of which up to 1 kW can be converted to mechanical energy (Betz limit) though more like 800W in practical terms.  That much mechanical power must be met with an electrical generator and balance of system that can deal with the power.

[taylorp035]

7 foot diameter = 1.0668 m radius.

I only had my wind velocity squared and not to the 3rd power like I should have.  Working with the air drag equation so much for the supermileage car made me confuse the two equations...

Now I have 0.3 efficiency * 1/2 * 1.2 air density * PI * 1.0668 * 1.0668 * wind speed ^3

The numbers look more reasonable now.

[SparWeb]

Yeah, the wind prediction looks better.
I'm still not lining up "28.428" Watts Electrical power with the blade power column where you put in a knock-down for efficiency.  The statement made by the title of the column makes sense - that sounds like a good simple way to approach it, but if that's what you mean, then the numbers have to line up for the system to balance.  Otherwise you've got a baby and an elephant on the see-saw.

This is why we all have furling mechanisms, you see...

[bigrockcandymountain]

I think i have it figured but as sparweb said, you would be aiming for value 1 in the "Blade Power / 50% Gen Efficiency" column.

At 12, you are almost definitely going to burn that motor with no furling.  And the case only gets worse at higher wind speeds over 23 mph.

Higher voltage would help the situation a lot. 

[taylorp035]

Today I broke out the torque wrench and put 84 ft*lbs on the blade bolts and 50 ft*lbs on the bolts in between the blades.  I'm figuring this achieved about 30,580 lbf of clamp load on each blade, which is about half the compressive yield of the ash blades.  You could see the hubs squishing in to the wood by about 1/32" at most.  I'll have to do some calculations later to see how many RPM I think it should hold up to.  I plan on taking another pass at the bolts after a few days to see if they have loosened up at all.

After mounting the bearings, motors, and shaft clamps, I spun the blades to see how balanced they were.  I was quite surprised how balanced it was with the relatively free spinning bearings.  So I then hooked up higher and higher voltage batteries to do a spin test to see if it would shake at all at speed.  I probably made it to 600-700 mph @ 24 volts on the work bench before deciding that it was windy enough in my basement and it was balanced enough for an outdoor test.

I also wired up my new 40A relay to replace my abysmal 10A 1S li-ion BMS units.  It worked for 1 cycle and then I found out that I had bought an AC output solid state relay, which doesn't work with DC.  After much research, I found out it's difficult to find a DC / DC SSR and one that works with an output below 5 volts.  But I eventually found one and it should be here next week.  A normal coil relay would be easier, but it would also consume around 6 Ah / day just powering the coil.  Once I get the new relay, I'll do another test on the drill press, only this time hopefully pushing it to 1500-2000 RPM and maybe 20A.   I've pushed that motor to 15A for 30 minutes straight before with zero air flow with a Briggs and Stratton hooked up to it.  We also pushed it to 4400 RPM @ 120 VDC and 20+ Amps in 100 RPM increments in a 10HP CNC mill.... she was smoking by that point.  Best $30 I've ever spent.

[bigrockcandymountain]

Ok so here is what i see.  Your motor and turbine have huge potential but are limited by the 5v output.  Can you increase this somehow?

[taylorp035]

Ok so here is what i see.  Your motor and turbine have huge potential but are limited by the 5v output.  Can you increase this somehow?

It's a long story, but this windmill is really just a toy.  The goal was never to make the maximum amount of power or energy from it.   I've made about 10 windmills in the past that had nice sets of blades (hand carved, sheet metal, CNC'ed), but I've never made one that can make power in low winds (less than ~15 mph).  My motor also has to spin too fast before getting to 12 volts to charge, so I've had this long standing dream to make something that could charge a lower voltage battery (a bunch of 1.2v NiMh's) and then switch the cells around to run a 12v inverter.  At one point I spent a ton of time making a 3 bladed savonious vertical axis blade set, only to find that it didn't have enough torque to spin my motor.  After some pondering and deciding to spend more than $10 on the project, I decided I would make a 4.2v li-ion setup since these batteries are so cheap, and I happen to have 18 of them sitting around.  I since purchased another 40 cells that were 3.2 Ah each.   A long time ago, I had this motor up on a pole for 18 months straight with no furling mechanism and it survived.  It's only ~ 7 feet off the ground and surrounded by woods on the east and west sides.  I'm in approximately a class 4 wind site if I were up above the 60' trees.


I did some more measurements today on the blades so I can do some better calculations on the forces.   Center of mass of the blade is 0.355m.  The mass is close to 1 kg ( a bit less actually ).
Assuming fully unloaded is TSR of 14.  I have a feeling my blades are a bit chunky and won't spin as fast as some of my previous blade sets.
53.55 mph @ TSR of 14 = 3,000 RPM = 314.159 rad/s
F = ma = m* (omega^2*r)
F = 1 kg * (314.159^2*0.355) = 35,037N = 7,876.6 lbf
The blade is clamped now on two sides instead of one.
Clamp force assuming 0.15 friction coefficient = 7,876.6/(2*0.15) = 26,255 lbf
Clamp force assuming 0.2 friction coefficient = 7,876.6/(2*0.2) = 19,692 lbf

Speed corrected for 14 TSR, 30,580 lbf and 0.15 friction coefficient = 57.8 mph
Speed corrected for 14 TSR, 30,580 lbf and 0.2 friction coefficient = 66.7 mph

It seems like it would be difficult to make a windmill much larger than 7' in diameter without having an excessive amount of Cf force @ 60+ mph un-furled unless you made them out of carbon fiber.  Certainly not something that you would want to risk with a real tower and something powering your house or something that would burn out your alternator.

[taylorp035]

I received my DC-DC 40A solid state relay and wired it up.  Though it claimed to work down to 0V on the 40A output circuit, that turned out not to be the case.  Luckily the 0.9v drop with the diode that I needed was enough to make it operate as intended.

After almost a week of testing the setup in the drill press, I was able to sort out a few things on the motor and shaft coupling.  Turns out I didn't have the shaft coupling tight enough on the first data set, so it was slipping quite a bit above 6 amps.  I took it up to 1280 RPM as indicated on the drill press, though I bet the RPM sagged a bit under load.  I also had messed with the brushes and the timing advance of the coils vs. the brushes between the first and second runs, so they probably need to wear in a bit.  I was quite pleased on the first round with 2 amps at 250 RPM.  With my old 10A 1S BMS unit, it was only about 0.1 A or less.  If you compare the results to my previous graphs, you can see just how much of an improvement this was.

I also finished the battery, where I now have 14 12-gauge wires going to a 6 pole, 2 double throw 20A switch that goes between my 3S BMS unit for running my inverter and the 1S input wires from the windmill.  It's a 100x better than my old setup where I was constantly connecting to all the leads and in fear of ripping the wires out of my batteries.  This is so much faster, safer and easier.  Inside there are 3 40A fuses just incase something goes wrong.  The density of wires and plugs is quite high in the box, as seen in the photos.

I added the appropriate wind speeds as a secondary x-axis on the graph assuming a TSR of 7.  There's a good chance she will spin a lot faster than that since it will be so lightly loaded.  Having a cut in speed of around 5-7 mph seems about right from what I remember last spring.  Hopefully it still starts up in those low of winds.

[SparWeb]

Hi Taylor,

Still need to warn you about this - the blades are drastically overpowered for the small generator you are using.  By 10X or so.

In any of the situations you calculate above, you have a generator that needs a few dozen watts to be operated, coupled to blades that can produce kiloWatts for it.  The excess energy will ACCELERATE the machine drastically.

I think I see the assumption driving the concern.  TSR is not an independent variable.  You have no direct control over the TSR of the blades.  It's OK to estimate the blade speed, but that estimate only works when it predicts that the wind power is balanced with the generator load.  If the generator is too small, the blades will run fast, or vice-versa.  In your case, your generator is so small it is guaranteed that the blades will run very fast with no way to make them stop.

You will have no trouble with cut-in.  The blades with a free-hand turn will respond to even the slightest wind.  However, in any gust, they will be turning so fast that (a) you will have no way to stop them and (b) they will bend the shaft when they yaw.

If you choose to fit a brake, you will also need to have a keyed shaft.  That is missing from your current shaft and hub.  Your mention of slipping just makes my concern greater.

I wouldn't be so stern about it, but you could get hurt, and this system is likely to fly apart without a much more robust generator on it.

[bigrockcandymountain]

Here is a picture as a warning.  My turbine is in the background and a piece of blade (white thing) is in the foreground.  This was my first set that hit the tower.  The distance is probably 300 or 400 yds.  My house is well within that radius.  After that incident, i built a coupe overlapping shutdown mechanisms in. 

[taylorp035]

Since I haven't mounted the nacelle yet... I guess I could do this rather easily.   How far off the tower axis should I start with for the motor axis to do this reasonably?  A convenient answer would be 6-7".  My tail will probably be 2 square feet and about  5-ish feet behind the tower axis.

I definitely understand the dangers... I had a 8' diameter windmill explode 2 feet in front of me in ~30-40 mph winds.  My concern on this windmill is more of the blade tips going super sonic and breaking apart, causing the whole thing to be unbalanced.

[SparWeb]

For the sake of seeing it spin, about 15 years ago I carved a 2x6 board into a rough propeller, mounted it on a hub, and let it go freely.



The next day I picked up the pieces.

[bigrockcandymountain]

6 to 7" sounds good.  That is quite a bit of offset for a 8', but you will need it to furl pretty early so that will work out good i think. 

[kitestrings]

Or, you could take a stab at calculating the offset using Hugh’s method.  His approach, which we used, looks at calculating the rotor thrust and comparing it to the restoring moment of the tail.  You want the restoring moment (mkg) to be not more than 65-75% of the rotor thrust.  If you have any of his books, you may have it there, but I believe it is:

Thrust (mkg):
[ [ d (m)^2 * v (mps)^2 ] / 33] * offset (m)

Restoring moment (mkg):

Tan Angle (off vertical) * COG (m) * weight (kg)

[bigrockcandymountain]

I definitely understand the dangers... I had a 8' diameter windmill explode 2 feet in front of me in ~30-40 mph winds. 

Wow how did that happen? Sounds like a good story. 

[taylorp035]

I definitely understand the dangers... I had a 8' diameter windmill explode 2 feet in front of me in ~30-40 mph winds. 

Wow how did that happen? Sounds like a good story. 

When you have a 8 foot diameter blade set on a 6' aluminum step ladder and there's a huge gust of wind... eventually the thrust force from the blades gets to be a very large amount and you and your brother decide it's time to turn in out of the wind while you're holding the ladder.  While making the evasive maneuver, a blade struck the front leg of the ladder and the blades split down the centers.  This caused the blades to fly outwards, some of which went far above the high tension power lines and landed ~60 yards behind us.  It took a long time to find all the blades.

Just in time for Christmas again...

After a lot of hard work over the last two weeks finally getting the relay to function the way I wanted it to, I finished the rest of the windmill.  Put a lot of nice finishing touches on it, like a viewing window for the power meter, an extra large storage box below for battery storage and a hinged door at the back with a latch.   Today was the day to put it up given the weather (windy + 50F temps, right before it dropping into the 20's + over a foot of forecasted snow).   It only took about 90 minutes to take it from the basement and then have it fully set up.  I should give a word of advice to always re-torque your hub bolts, since I got another half a turn going to 84 ft*lbs on the blade bolts.

The tower turned out better than anticipated.  The base of it is 3' 9" down in the ground with some big rocks.  The ~30" long 1 3/8" hex shaped bar used in the rotating assembly for the tower worked and fit perfectly in the schedule 40 PVC 4" pipe.  Also drove 4 large stakes in the ground and put 100' of 3/8" polyester rope around them with 4 strands on each leg.  The tower doesn't sway much at all, even with the off center mass of the windmill (~100-150 lbs). 

The forecast said 20 mph winds all day, but it was probably only 5-8 mph at best for the hour I stood there and watched it.  It spun up a few times to the point where the airfoils really hit their sweet spot to prove my blades were balanced, but unfortunately it never recorded any power generation.  I'm not sure what's wrong at the moment because it should of made some power.... I even swapped the positive and negative from the treadmill motor to make sure I wasn't loosing my marbles from my drill press bench testing.  I tested the electronics no less than 4 times in the last 48 hours and my relay was on, so I'll have to do some investigating.  It definitely would of charged at those RPM's last spring...  On the plus side, my furling mechnism worked like I knew what I was doing.  It furled fully at about 10-15 mph, so I added a rubber tie down strap to the tail to make it a bit more aggressive.  I don't know how well aimed the blades are into the wind, so I will have to get a ribbon or something.  I do need a hair more clearance to the PVC pole since it's rubbing a bit on my battery box.  I'm going to wait a few days before deciding if my blades are now too flat for low speed start up since they seemed marginal today.

Tonight it is super windy, so I took a bungee cord and tied a blade to the tower.  I want to see it spin up to 400-500+ rpm before I decide to do more electrical diagnosis on it.

7' diameter.  1" main shaft and 10" diameter hubs.  ~4" wide x ~2 1/4" thick Ash blades with a negative curvature airfoil trying to achieve a higher amount of lift.

[MagnetJuice]

Is that shaft securely attached to the hub?



Ed

[taylorp035]

Yeah, it's not going anywhere.  The hubs are a decent fit on the shaft and there is a 1/4" x 7/8" dowel pin co-axially with the shaft.  Then a strong 1/2" wide hose clamp to make sure the hub nor the dowel pin slide forwards.

[taylorp035]

So I just walked back to the windmill and took the bungee straps off the blades. This morning there was a nice 20 mph breeze going straight down my back yard, so I could see how it behaves better.

1) The furling mechanism works perfectly.  I'm guessing the added weight upped the furling speed to about 20 mph.

2) The windmill doesn't point into the wind very well.  Based on me throwing some grass a few times, it seems like the blades are over powering the offset tail by about 45 degrees.

3) It's definitely not charging

4) Blades look super balanced, as I see no shaking.  The tower doesn't bend much more than 1/2" under the highest wind gusts.

5) I think I need to add a 1/2 degree or so to the blades again since it doesn't start up or get into the sweet spot of the lift coefficient curve real well like it used to.

[SparWeb]

First steps OK then...

I would check that the length is about the same length as one blade, and the area of the tail vane to be about 1/10 the swept area of the prop.
With a prop 5 feet in diameter, then the swept area is about 20 square feet.  Is the tail more than 2 square feet in area?  Is the center of the tail about 5 feet back from the tower axis?

Have you been measuring prop speed?  If it's greater than 600 RPM then an abrupt yaw could overwhelm the stability of the ropes.

[taylorp035]

So I managed to fixed everything...

2) - I made the tail twice as long and has twice as much area.  It seems like the tail is winning just a little bit, but it's real close.  The tail is about 15" x 55" if I had to guess, so the the very back edge is about 6' from the center of the tower.

3) - I decided to plug the battery into the motor to see how fast it would spin the blades.  It was really slow and weak.  After 10 seconds or so, I saw some smoke come out of the commutator area of the motor.  So I went to the trouble of removing the motor and took it inside.  When inside, I found the metal spring that held the positive carbon brush on the commutator was shorting against the case of the motor... which would of happened since I clocked the motor after I had it in the wooden structure to give me better clearance to the 4 bolts that hold the motor down. So I found a happy medium and tested it.  It now makes 1.5-2 volts at 60 -ish rpm, much better than the 0.1-0.2v I was getting previously.  It also means the motor was completely shorted out for the last week while it was spinning away outside.

4)  -  I decided today to try and push on the tower as hard as I could to see how stiff it was.  Once the ropes engages, it doesn't move.  I'm betting there would be less than 1" of movement at 30+ mph.

5)  - Now that the windmill is pointing into the wind, it seems to be starting up better.   Tomorrow it's supposed to pickup around 4AM and peak at 23 mph at 11 AM, so it will be good to watch.  The battery probably needs 30-40 Ah to charge up, so we shall see.

[SparWeb]

One through five: Good!

It now makes 1.5-2 volts at 60 -ish rpm

You're aiming for about 3.7V right?  Cut-in about 120 RPM...

In a 20mph wind, the blades will have about 500W to work with -> 300W into the shaft -> 150W through the wires of the generator.
150W / 4V = approx 40 Amps

Your wires and battery can handle that?

[taylorp035]

Cut is about 0.8v above the cell voltage.  When the pack was fully charged at 4.2v, cut in using the drill press suggested ~180 RPM.  Historically the motor was about 34-37 RPM/volt, so the math lines up.  I currently have the voltage divider that triggers the relays to be a hair on the conservative side, so it will probably shutoff when the battery hits 4.1 - 4.15volts.  The whole charging circuit and power meter draw a continuous 0.088A or 2.1 Ah/day.

1200 RPM (per the v-belt speed chart) was about 14-15A when hooked up to the drill press, which it could do all day.  The motor was rated for 20A originally at 120v.  It was out of a very expensive treadmill from what I was told.

I used 12 gauge wire and I made sure all the connections were of high quality.  The switch is a 20A 6 pole, double throw switch, so that is the weakest link when used in series mode.  In parallel in the windmill, it should have 3 paths at 20A each assuming the cells are equally sharing the load.  All the connectors are 45A Powerpoles and I've put 200+ A through them before for short periods of time.  The relay is 40A continuous and is mounted to a large aluminum sheet for heat dissipation, as well as the 150A of diodes.  The power meter is 10 gauge wire and hopefully the 200A rating can handle 40A.  The three batteries each have 40A fuses going to them.  The cells themselves are supposedly rated 4A a piece for max charge rate, but I wouldn't want to go much past 2A... which shouldn't be an issue with 39 cells in parallel.

If I wanted to use more of the shaft power, I might consider doing an additional project down the road to make a mini 4V alternator or get my hands on a better motor.  I first want to see how livable the current design is and how much it makes and also get it tuned in for maximum production.  I think if I go much more than the current treadmill motor, there's going to be some significant costs to update everything else.  It would also likely be most beneficial to make a second battery that is larger before getting more power out of the windmill.

[SparWeb]

Oh, you mean you have v-belt pulley speeding up the generator.  That will soak up some energy from the blades. 
You say that the treadmill motor was rated for 20A, so then (20A)*(4.2V) = 84W.
See if you can get the furling to kick in at about 15mph wind so that the blades don't overpower everything. 
At least, not until you've had a chance to witness it run in that much wind, and see all the other pieces are working OK.

[taylorp035]

"Oh, you mean you have v-belt pulley speeding up the generator.  That will soak up some energy from the blades. " - It has a 2' long 1" diameter shaft, then a flexible shaft coupling straight to the 5/8" shaft on the treadmill motor.  No v-belts.

So yesterday's power generated was 80 Wh or 25 Ah.  This raised the voltage on my battery from 3.62v to 3.88v.  It was near furling during all of the sunlight hours, but quickly died off once it was dark.  Adding mass to the tail twice raised the peak amps from 4.6 to 6.6.  Furling after the mass was added was at about 4 amps or ~400 RPM.  6.6A should be close to 600 RPM    After watching it extensively, the windmill wasn't pointing into the winds as much as I thought, with it operating about 45 degrees to the wind when it was 10-20 mph.  At 20-25+ mph (estimate) it was probably 70 degrees from the wind as the tail was mostly folded.  It was doing well in my opinion protecting itself in the 30+ mph gusts.

I need to straighten the tower a bit.   I bet I'll increase power production significantly if it's pointing into the wind better.  I want to offset the tail a bit, change the angle and maybe even make it slightly steeper for the furling.  I think letting it get to 800 RPM would be a good max or about 10A.  I think I could easily double the power generated if it was pointed into the wind better, which would then be closer to my goal.

During a 2.5 hour stint, it averaged 2.6 amps while right on the edge of furling.

Here's a video of it in action with my extended tail.  This was earlier in the day when it was maybe 15-18 mph.

[taylorp035]

Slowly making improvements to the wind turbine so it makes more power.  I ended up spinning my living room couch around so I could watch it from the back of my house all day long while I worked from home yesterday while there were up to 16 mph winds from the south.

I noticed that it still struggles to point into the wind.  In a medium breeze, it would stay about 45 degrees from the wind without any furling.  My larger ~6" offset of the tower vs. the shaft seems to be much too large.  I would not recommend such a large distance for anyone else.  So tonight I took the tail off to make some more modifications.  In addition to offsetting it another 4" farther away from the nacelle two weeks ago, I added more angle to help compensate for the force from the blades.    I have also extended the tail another few inches and put the two pieces of plywood in-line with each other instead of overlapping so the pressure doesn't bleed off.  This should make it a decent bit more effective without changing my furling speed too much.

I also straightened the tower some.  I probably need to do it again as much as I just changed it to get it perfect.  I've been really happy with how sturdy it has been. There also haven't been too many vibrations in the system, especially as I have lowered the natural frequency of the tail with more mass at the end.

I did burnout my DC-DC converter that powers my power meter, so it's harder to track the power production.  I was so excited to get it up that I forgot to wire in the switch to isolate the DC-DC converter when the battery is disconnected.  Now that i think about it, I should probably find a mini dump load when the battery fills up.  I'll have to think about how to wire that up...

With the incremental changes, it has been starting up in lower winds.  When I first built it, it needed about 15 mph to run and now that's down to maybe 11-12.  I think once I get it pointed in the wind, it will get closer to 8 mph.

I've noticed how important wind direction is when it is mounted effectively in a forest.  I've had several days when the wind isn't blowing in the SW to S direction that there is no wind back there, but it's 15+ mph above the trees.  Sometimes the wind doesn't move the 60 foot tall trees and you see the clouds still racing by.   It's been a strange winter... certainly less windy, though Erie finally is back in the lead for the 100k city snowfall total race.  Way warmer than normal the last few winters and this one is similar.

[taylorp035]

Question for the more experienced people on here:

If I wanted to do something easy and add more magnets to my motor in the empty spaces between the two original ceramic magnets to make the motor improve its voltage at a given RPM, what configuration of magnets should I use?  I can buy just about any sized magnet I want in Neodymium.    Even if I saw a 20% improvement in the RPM/volt, that would be huge.

The current motor has two brushes 180 degrees away from each other on the communtator, so I'm assuming the motor is effectively a two pole motor.  I was thinking the options would be:
1) - Halbach array with the N/S going tangential to the case with my new magnets.
2) - Or just continuing the current magnets to fill in the space.   

I tried to draw a diagram that's fairly close geometrically to show what I have to work with.  The magnets look like they are almost 1/2" thick, but may a bit less.  I think there's a good 1 1/4" gap to play with.

[SparWeb]

The current motor has two brushes 180 degrees away from each other on the commutator, so I'm assuming the motor is effectively a two pole motor.

Not really.  The brushes just provide current In/Out to make a complete circuit through the rotor's armature.  The armature can have 2 poles, or 4 or 6 or any even number.  To get the number of poles, count the number of commutator bars and divide by the number of brushes, IIRC.  But that's nothing but a nit-pick point because
the wiring on the rotor, commutator, and brushes will stay constant unless you consider re-winding, and I didn't see you suggest that anyway. 

I'm actually not sure how much you can change anything by replacing the ceramic magnets with Neo's in a DC motor like that.  You're showing about 30 degrees of arc without magnet in them, but if you filled it with a magnet, with poles in any direction, would that add flux lines to link through the rotor armature at all?  I am not sure.  The crescent ceramics will not "conduct" much more field density than they already have, so if you tried to squeeze in a Neo arc with tangential poles, that field might not increase the field intensity across the left-to-right main field gap, but rather make a tiny short-circuit around the surface of the Neo.  The field lines might not even pass through the rotor's armature.

You could consider sleeving another set of magnets over the existing ones.  Or make a new external housing of a larger diameter.  Get another pair of magnets that "nest" round the old ones, or just get new bigger ceramic/neo crescent magnets.  Still, that depends on the wire in the rotor, and the back-emf is proportional to the field.  I haven't worked out the details, so there's a chance that what I just suggested could be pointless, if increased back-EMF just counteracts any increase in field intensity.

What's the wire gauge on the armature?

[taylorp035]

What's the wire gauge on the armature?

Without going outside and measuring it, I think it's 16 gauge.  I've run it at ~15 amps for a half hour straight with no cooling and it seemed okay.  20A will start go get pretty hot after a few minutes.  The label on the motor says it's good for 16A continuous.

I think the commutator has about 16-20 slots.

What if I were to mill out some of the steel on the case and put some magnets in the middle, but on the outside?  Maybe a 1" wide x 1/2" thick (or thicker) x length(4" I think?)

[SparWeb]

OK I didn't think of that....  Maybe round magnets would do well enough and it would be easier to make with an end-mill on a vertical milling machine.  Don't want to cut off too much steel housing or the thing comes apart.  A few holes shouldn't hurt but I don't know how many is too many.

Have you ever considered getting FEMM to model this stuff?  It's way damn easier to use than Ansys like you were messing with in university.  Once you figure out the basics, you can cycle through a model and run an analysis in about 10 minutes for a case like this.

[Adriaan Kragten]

"Oh, you mean you have v-belt pulley speeding up the generator.  That will soak up some energy from the blades. " - It has a 2' long 1" diameter shaft, then a flexible shaft coupling straight to the 5/8" shaft on the treadmill motor.  No v-belts.

So yesterday's power generated was 80 Wh or 25 Ah. 

How can you write this. Power is in W or kW, energy is in Ws, Wh or kWh and battery capacity is in Ah.

[SparWeb]

I think it's an inadvertent error in the typing, Adriaan.  I understood it as "energy".