A Case For the Non-Cast Stator II

[SamoaPower]

Thanks to the excellent work by Boondocker, we now have some better test data to work with to try and analyze the effects of casting the stator coils of an axial flux alternator in resin. This popular technique developed and promoted by Hugh Piggott and DanB is probably the most copied approach to making a stator assembly as judged by posts on this board.

Also appearing on this board is an increasing number of reports of failed cast stators on machines of 10' diameter and up. The current poll on the question of "Have you built a wind generator?"  reveals that of the current 2441 responders, 332 have/had a flying machine and of those, about 33% (109) have failed for one reason or another. We can only wonder how many of those were due to stator failure.

Intuitively, to my way of thinking, surrounding a heat-producing element (coil) with a thermal insulator is contra-indicated when substantial power is being dissipated. The element temperature will increase considerably compared to one with little or no thermal insulation.

Two limitations to coil temperature present themselves when looking at the problem. The first is the temperature rating of the magnet wire insulating varnish. This can be from 100C to about 220C so it pays to know before you buy. The second is the effect of higher temperatures on the structural integrity of the casting resin since the resin is what holds things together mechanically. Most resins are known to soften at higher temperature and at some point will deform with stress. Since the stator is quite close (0.125" or less typically) to the magnet rotors, any physical contact would be of concern and usually cause for destruction. Unfortunately, this issue is difficult to quantify for the various resins and additives used but has been mentioned in numerous failure reports.

It would be useful to derive an estimate of the increase in coil temperature due to casting. Let's take a coil of 30 in^2 total copper surface area with that area equally divided between the front face, rear face and the remaining sides (inside and outside). Now we cast this coil in polyester resin with 0.063" of resin over the front and rear faces. Obviously, the thermal path for the sides is longer, from 1 to 4 times as long for 0.5" thick coils.

Assume that 90 watts is being dissipated in the coil resulting in near equal surface temperatures over the surface in free air. There will be a temperature gradient from the center of a coil leg to the surface due to its own thermal resistance.

We now need to calculate the thermal resistance of the resin which is given by

R = t/kA where:

R = thermal resistance ^C/W

t = material thickness = 0.063"

k = thermal conductivity for polyester resin = .0063 W/inch-^C

A = area of thermal path = 10 in^2 (one face)

This results in 1.0 ^C/W. Assuming even thermal power distribution over the coil surface results in 30 watts for either face. This is a temperature gradient of 30^C for the thin resin or saying it another way, we've increased the coil temperature by 30^C by casting it. It's actually worse than this because the sides, which represent 1/3 of the area, have a much longer thermal path length through the resin. The temperature rise with casting is probably closer to 40^C.

I've assumed a rectangular coil leg cross section. I've seen pictures posted here that indicate a more oval cross section for some. This will also increase the temperature rise.

In http://www.fieldlines.com/story/2007/6/25/12033/9605 Boondocker presented test data for his alternator, which uses a stator cast with vinyl ester resin with ATH as an additive and is considered by some to be the best current approach. He showed stator surface (opposite a coil) temperature vs. time for three different output power levels. Among other things, his charts show that it does take sustained power for some time (20-30 minutes) for the temperature to stabilize at maximum.

Using his temperature data along with his efficiency data, the following chart, relating stator surface temperature and estimated coil core temperature to stator power dissipation, was derived. An ambient temperature of 70F, a slope of 0.3F/Watt and a resin thermal gradient of 104F (40C) were assumed.

If we select 300F (149C) core temperature as a reasonable safe limit for 200C (392F) wire to allow for higher ambient temperatures and a safety factor, stator dissipation should be limited to about 420W total. Of course, this says nothing about what effect 300F would have on the resin structural integrity.

 Since Boondocker used a nine-coil stator, this works out to about 47W per coil of sustained power. For the same limit, a non-cast stator would be about 84W/coil, an improvement of about 79%.

I shudder when I read of those machines putting out 2-4kW at only 50% efficiency using cast stators. Probably, the main reason we don't hear of even more failures is due to the vagaries of the wind, which doesn't sustain the power for long periods. But, are you willing to bet a stator replacement on this?

I think the lesson is clear. We need to strive to produce more efficient alternators and not to stress components where it isn't necessary. Stator casting is not necessary, it's only one way to skin the cat. Also, limiting power output by various means to a reasonable number is better than the "all I can get" approach and will enhance reliability.

Even considering the assumptions, estimates and data roughness, I sincerely believe that casting a stator in resin is the wrong thing to do for larger machines (=> 10').

Disclaimer: I have been accused by some of being in some sort of competition with the Hugh/DanB design. Such is far from the truth. My only objective is to bring to light what I perceive to be a major fallacy in their design for larger machines based on solid principles of physics. Hopefully, this will save some from the grief, expense and trouble of a stator failure.

I fully expect to receive a rocket from TomW for presenting formulas, numbers and theory, which he doesn't seem to think is appropriate for this board and, of course, I disagree. I believe understanding should come before construction.

At least, I put it in a diary, Tom.

[Volvo farmer]

All right, I'll jump into the fire.

It appears that Boondocker was testing his alternator with a 3HP motor. I would argue that your temperature chart is meaningless in a real-world situation where there would be 15+ MPH winds blowing over the stator to help cool it.

I can appreciate what you're trying to do here, I.E., suggest improvements to the tried-and-true axial flux alternator. However, I think you would have a much more convincing argument if you actually built a non cast stator, put some blades on it and hoisted it 40' in the air for a period of time to test how much better it is than a cast stator.

I could argue that my Volvo would be better with a Dodge Cummins turbodiesel motor in it and list all the reasons why it would be better than the Volvo motor from here behind the keyboard. However, I'm not sure that such a diatribe would lead to any fruitful development of the idea, since I'm not willing to do it myself and show people it works, and works better.

The neat thing about cast stators is that if I want to build one, I can buy a book, do it, and make electricity (my goal is to make electricity). I think your idea might have merit but do you have a step-by-step guide as to how to make a non-cast stator? I'd prefer a design that has a year or two of field tests, like Dan's and Hugh's stators. The fact that Dan's 20 footer survived that huge wind storm a while back and actually failed the blades but not the stator makes me question whether or not there really is a problem with cast stators. Matt's brother told me he saw 80+ amps out of his 24V alternator in that same storm, and I believe it's been flying close to three years on the same stator.

In the hot-rod world, they describe this activity you've started here as bench racing. Which is faster, a '69 Dart with a 440 and a Quadrajet? or an '82 Delorean with a flux capacitor?

[harrie]

Hey, Im with you on this for sure, The more cooling the better, Most of the stators I have built, are poured with one type of resin or another, but have left the centers of the coils open. does it help, I think its anyones guess. It would be great if we didnt have to use resin, but I think we need something that will be ridged enough to hold up, expecially when we short the mill out in high winds.

I guess if we dont like to use resin, we need to change the whole design of the genney, and go more for a drum style, using laminates to hold the coils?? anyway, if you come up with a way, I for sure would be interested. Thanks for keeping this alive.

[TomW]

Farmer of Volvos;

Good point about the wind cooling effect. It pretty much invalidates the data WRT wind driven units. But some folks will accept it as supporting proof of their pet theory.

Just eyeball a wind chill chart sometime to see how dramatic wind induced cooling is.

I have no data to support my point, but anyone who has stood outside in a 15 MPH wind will have experienced it first hand.

Just an opinion.

Cheers.

TomW

[DanB]

Well... of course I have to bite on this one.  I think you have a 'case' but I disagree - I like cast stators just fine.  Any way you do it has pros and cons.  I have no doubt that you could do things differently and have much better cooling.  

I believe that a cast stator can work fine - usually (and this depends a bit on the design) all the way down to about 50% efficiency, which is about as low as you'd want to go anyhow (I think any sane wind turbine should be furling by that time).  People have stator burnouts when the furling is not setup right - or when they try to put too small an alternator on too big a machine.  (not enough magnets - and lots of resistance to keep the speed low is a recipe for burnout)

I wouldn't argue with anybody wanting to do it another way - but I think the problem of stator burnouts is usually more the result of other problems.  If we keep efficiency at 50% or better before we furl I think the cast stator is fine.

Samoa - for somebody like you, whos planning on MPPT - Im surprised you're even that worried about it.  MPPT alone could solve the problem all together I should think.

I've had a couple stator burnouts where I knew the furling was right and the machine was well matched to the load.  I see too possible ways to prevent that from happening...  (without decreasing the swept area) - we could try not casting the stator, or we could build with larger magnets and much lower resistance.  

The cast stator is simple, strong and very workable even in larger machines.  You think its more of a problem with larger machines but I disagree...  when I look at the surface area of the copper near the surface of the stator (which Ill admit - is the only area that really does much for cooling) on my larger machines vs smaller ones it seems the larger ones have an even better chance.  

We can go either way and there will be pros and cons to both.  I think the bigger problem than cast stators though is getting everything else within reason.  I'm stuck on cast stators for now till I see - or figure out myself something better.  Right now I'd rather throw a few more bucks at magnets and keep the efficiency higher - and if it stalls the blades I'll add the resistance back somewhere else.  

You talk about the poll where we see a pretty high failure rate.  I wouldn't be surprised if it's actually higher to be honest.  It's not easy to build something that holds together.  I expect most the the stator failures were cases of alternators too small - or furling systems not setup right - or mismatched loads.

All that said - there is no harm in doing it without casting, it definitey has obvious advantages - how much it would help Im not sure (be fun to test).  I do think your estimate might be accurate.  I dont see why being in the wind - or not being in the wind would change things.  But again - I think you're worried about solving the  issue of alternators burning out when they're at - or below the 50% efficiency mark and it'd be nice not to have to go there in the first place.  

[oztules]

I have to disagree with this to some extent Tomw. If the mag disk is resin cast and in close proximity to the stator, the intervention of the outside air into the air gap may be minimal.

 Perhaps some help can be gained from the outer part of the coils exposed on the perifery of the stator sticking proud of the disk area being in the "free air", but this air is compromised in part from the prop blades and the air disturbance they produce.

It would be good to actually measure the difference to see if there is much to gain..... but I don't think the "wind chill" will be of much benefit down in the inter disk-stator space.

The ground test will move the same amount of air through this space as being up the tower.

I't currently blowin about 60kph here... sticking my head and ears outside the door tells me that wind chill sure works on me though....Of course I've been plum wrong before too.

........oztules

ps, I'm building a resin stator anway, but will make a variable pitch prop for it to help mitigate the heat problem....and then panic anyway

[oztules]

Samoapower

Thanks for your analysis. I happen to agree with the thrust of your diary, but have yet to satisfy myself of a better way to deal with the mechanics of a good mechanically easy to fabricate solid stator without the resin (and it's inherent insulating effects).

I've decided to go the variable pitch route to hide the problem.... and the load matching techniques that Flux propounds. Hopefully I will then not be in the wrong region of your graph.

Well written, and thoughtful diary.

........oztules

[Mary B]

I am casting each coil individually with this http://www.mgchemicals.com/products/832ht.html the thermal properties are a lot better than standard resins, and it is high temp Maximum Service Temperature

275 ºC (527 ºF) ,  Thermal Conductivity ASTM-E-1530-99 0.210 W/mºK. See how it goes in a few weeks when I get my tower back up. In the meantime I am going to stick a coil on the big 50 amp power supply and see where it self destructs.

[Flux]

I can't add a lot to this.

It is perfectly obvious that the cooling will be better with a stator that is not cast in resin, it really comes down to convenience and ease of construction against a possible improvement in performance.

There is no doubt that wind cooling helps greatly but I suspect it is not the big factor that seems obvious at first sight.

I am convinced that the peak to mean power distribution of wind power is the main reason for these things performing far better than could reasonably be expected.

Industrial motors and alternators are rated for continuous power at a temperature that keeps the windings in a safe region for long service life. Probably you can push them to 150% and still have a decent life.

It seems as though Zubbly is pushing motor stators well beyond their normal ratings with no problem and the less than 100% duty cycle of the wind power input must surely be the reason.

No commercial motor winding will stand running at 50% efficiency on a continuous load test even with its cooling fans at full rated speed.

I have no idea what the effective output is under wind conditions and it may be far more on some sites than others. In the poor wind area that I have experience of, a 1kW alternator is probably producing less than 500W long term average. A machine that would survive 500W would fry in a few minutes if driven continuously at 1kW by a motor.

I agree with Dan about the actual failures, I am absolutely convinced that most of the failures have been due to lack of EFFECTIVE furling. It also seems that the furling even when it does work as intended is only effective with stall limited machines. If someone uses an undersized alternator that will let the blades run up to peak power rather than being heavily stalled then the furling speed will need to be reduced an order of magnitude and the conventional plywood tail and 20deg pivot angle will not manage it.

I notice that Dan has come to my conclusion that you can't run at very low efficiencies in high winds and dissipate the power in the stator. For survival the alternator has to stay reasonably efficient but to match the blades you have to keep the system efficiency low. It makes a lot of sense to make the alternator too efficient for load matching and produce the heat externally ( and even to use it for heating if you can).

If you can manage to produce a stator that is structurally sound and can stand the environmental conditions without casting it, then you can run at lower efficiency than with a cast one and still stick to the winding temperature limit. In a damp marine atmosphere it would be far more challenging than in a dry region. At present I see no simple way that is not labour intensive and doesn't use industrial glass/epoxy that is possibly not easy to obtain and is certainly costly.

I notice here that people seem to have different ideas of what they find difficult and sometimes the methods they revert to to stay within their comfortable limitations

brings them into the realms of things that I wouldn't consider. That is a way of life and we have to respect these things.

I agree that the cast stator has limitations but it certainly doesn't need to fail, that is an entirely different issue.

Boondockers tests are interesting, they are the first I have seen with real figures for temperature. It would be very interesting to see a repeat with a fan blowing the equivalent of a 30 mph wind over the stator. It is almost impossible to imagine what wind does really blow over a stator in real life, but I suspect that when furled and at an angle to the wind it is higher than in normal operation. Again I have a feeling that the wind passing the centre of a nearly stalled prop is higher than one running at peak power point. Many areas for much needed research.

Flux

[fungus]

Interesting stuff, there would certainly be less thermal rise in relation but the heat still has to be got ridden of. Also, how do you suggest to make a non-cast stator?

On a separate note;

Just a thought for better cooling...

How about drilling a large hole in the center of the blades (beyond the support bolts), with open magnet rotors the air would have a direct path in through this hole and spread outwards through the generator assembly. A centrifugal fan could be used to help it. This would mean that there would be an easy path for air through the alternator meaning better cooling. At least I think?

[pepa]

one thing that i have not seen posted is that when the machine is furled, the wind is blowing directly between the coils and the magnets. heat always goes to cold so the air will take a lot of the generated heat with it as it passes through. pepa.

[pepa]

let the machine furl early, with the machine sideways, the air passes through where its needed, espically if there are holes left in the coil centers. even with a poured stator, there is three fourhs of the coil only lightly covered and exposed to the wind. pepa

[dinges]

Flux,

It's good to see you mention motors and conversions. During this entire thread I've been wondering why everyone is staring himself blind on the axial flux.

The -only- advantage I see of an axial flux is its low wind performance. The mechanical factors, cooling included, aren't to my liking. Everything out in the open, large thin plates, bending, cast stator, in general, 'fragile'.

TomW has had some interesting experiences running a Zubbly conversion way beyond its ratings. I recall figures of 40 A for a 1.5 hp (2 hp? not sure) in a 24 V system. And surviving. IIRC, he doesn't even furl. Don't try that with an axial flux. Maybe he could do a writeup on the ways he abused that thing. It would make good reading.

The mechanical benefits of motorconversions (closed housing, good cooling, strong bearings, easy to replace bearings, in general 'industrial quality' as opposed to the common 'homebrew cr@p axial fluxes' (sorry to be so blunt) outweighs the disadvantages (for me) of stator losses. A little less output seems like a small price to pay for continuoud reliability.

Personally I keep wondering why people invest so much time and effort in axial fluxes that are, at best, when built 'properly', still mediocre w.r.t. life and endurance.

Not to get this thread more off-topic, but I know of only one failed motorconversion, due to the magnets not being bonded properly (using polyester instead of epoxy). Enough said.

Anyone that laid an industrial motor next to a DIY axial flux will immediately spot the difference. A motor you can throw around, an axial flux has to be handled carefully.

I'm sure I've stepped on quite a few people's toes here. So be it. But after building two small axial fluxes (more like playthings), I realized I did NOT want a 10 ft generator based on that design. Let alone anything larger.

I've read about ceramic stators and stator coils held in place with a few tie-wraps without being cast in resin. I get the shivers just thinking about it. Yes, for the axial flux casting in resin is a necessary evil. So why not see the light and convert ?

Just my humble opinion.

[oztules]

Pepa,

Thanks for the words of encouragement, but as Flux points out, it is the vagaries of the wind in most locations that allow the resin poured stator to have a less than 100% duty cycle.

I see you have read the African wind story, you will note from that that the AWP is a 1.3kw (or thereabouts) alternator, and it spends a great deal of it's life well over the 1kw mark. A lot of the time when i'm up there, its putting out over 25A into a 500ah 48v battery. (around 1400W@56v)continiously... and furled to some extent.... we rebuilt the furling angles and overdid it a bit, but the turbine puts up with it with no trouble... well not from the coils anyway.

In this environment, it's duty cycle is 100% a lot of the time, for weeks on end it seems to be all of the time...then a few days of 2-10Amps then the steady winds start again.

 @50%efficiency thats effectively a 1.4kw radiator in the stator with no let up, and it's only an 11 footer.

The stator is very heavy steel, and there is no sign of overheating. I'm worried about my upcoming resin stator, because even if I get some wind across the resin, it's a lot of heat to dissapate... hence the variable speed regulated prop... (when i can locate some s/steel for the arms). Also if I match the load via boost converter, my resistance may drop enough to keep it out of trouble....

Fingers crossed though.

I will take on board your advise on the holes in the front of the rotor, as this may swing it my way. The blades will be held away from the rotor and should give access to the air.

If i didn't have so much invested in magnets and resin and wire, I think I would do a motor conversion.

......oztules

[TomW]

Peter;

Not sure about a writeup but you are correct the conversion is a 2 meter [~6.5 foot] conversion and it flies with no furling. It has seen sustained 40 mph winds for most of one day. I have seen multiple instances of 40+ amps into a battery bank hovering around 26 volts. That is getting over 1 KW.

It has flown in winds that were downright scary on the ground and it is at 60 feet.

Damned tough unit. It is a  1 HP motor originally.

The copper buried in the iron core has a lot of cooling effect.

A fairly large photo of it in the air

For scale in that photo the gray box is a standard outdoor outlet box [3" by 4.5"].

Maybe the Z man will pop in with the actual specs from the motor.

RonB hunted down the diary I did when I got it in December of 2005:

http://www.fieldlines.com/story/2005/12/26/184650/27

Cheers.

TomW

[Flux]

Peter you make some good points and I suspect that it comes back to the comment I made about what people find within their capability and what they find difficult.

Some are completely afraid of windings and using a standard motor removes much of this worry. Quite how well they cope with messing about with connections and finding star points I don't know, I would prefer to wind from scratch than dig into impreganated windings.

Yes motor conversions have good bearings but that should not be an issue with an axial except for those forced to use some commercial hub. A good well designed hub with adequate bearings will stand up as well as any motor. It is easier to attach a prop to an axial but I would never do it the way that seems standard. I see no problems with magnet discs, they need never flex or magnets need never come off if designed properly.Magnets rubbing the stator shouldn't happen either.

At low voltage I don't think the open axial stator is any worry. (At high voltage I tend to agree with you).

Most seem content to build axials in a primitive way, skimping on steel for the discs and using poor mechanical design, such people almost certainly would not manage a motor conversion that needs basic sound engineering. If you have an engineering background you can build a reliable machine either way.

Motor conversions work out cost effective because scrap motors are cheap, if you built from scratch then they would be costly and very labour intensive, I know, I have done it that way. In terms of the labour and facilities needed, some of the high quality machines from China are coming in at such a price that that makes the whole thing questionable, but you have to pick the good ones and import significant quantities to see the benefit.

I really have an open mind on this, there is no doubt that in a poor wind area the axial will do better. In a high wind area the motor conversion will be as good and be more rugged, it's efficiency will be no better and to be cost effective it will most likely be faster and perhaps noisier. Either way can be completely reliable.

I have generally adopted a middle of the road compromise, using the Bergey/ Awp type construction with outer magnet rotor and a laminated smooth iron core wound with overlapped coils. It has lower ( a lot lower) iron loss than a motor conversion, it can't cog, it has superior cooling to an axial but it is still a partly open construction but better protected than a simple axial. It uses a quantity of magnet mid way between an axial and a motor conversion, but it has two of the disadvantages of the motor conversion.

It needs decent facilities to build and you have no control over the air gap, mess up the design and you are stuck with it, that is a big advantage of the axial where you can easily change the air gap.

My view is that you use the method you feel most confident with, do it properly and it will be fine.

Flux

[BigBreaker]

As I recall the AWP stator has iron.  Iron saturation will protect it against any over speed.  It would be a design flaw if it didn't.  Your blades may run away in high wind but the stator will be fine.  It sounds like you addressed over speed the right way... early furling.

[DanB]

Hi Peter - this is all running a bit off topic but..

I agree about some of the advantages with the conversion.  If you can solve the cogging issue (which is evidently not difficult) and match it well to the blades then I think you have something.

I don't really see the axial flux machine as terribly fragile - yes... if it falls too far good chance you'll break the stator but other than that there's no good reason to think it cant last a very long time.  Scottys hydro plant has been sitting in the creek for 3 years now - its fine, and (other than that hydro plant) I've never seen bearing problems or anything.

The obvious advantages of the axial flux machine... no iron losses - it's likely going to work a bit better/startup better in very low winds.  It's easy to build and its easy to adjust (thats a big one).  The ability to change the airgap - or quickly change teh stator is nice.  A 10% change in the speed of an alternator can have a drastic effect on the machines overall performance - in other words, things need to be matched up really well!  If you're lucky I think maybe you could find a motor to convert that's got the 'perfect windings' in it to start with, more likely you'll either need to design the blades around the motor, or rewind the motor.

Rewinding the motor probably isn't difficult for many - but it does need to be just right.  It would be nice some time to gather up some very detailed information about which motor to use - how to rewind it if necessary - and what blades will be an excellent match to it.  Lots of people are sticking blades on various motors/alternators and homemade axial flux machines - I think that logs of them are poorly matched.  Most folks probably never realize this - they see power coming in and they're happy...  people tend to be optimistic.  I find there is not a lot of room for slop though with regard to matching the alternator to the blades and getting that nailed with a motor conversion is tricky in my opinion.  At this point I know Zubbly has posted some things that work well - and Jerry has done well with some things.

With an axial flux machine I can look at past projects and make pretty good guesses about where to go with a new design - Im not sure there is data out there that works well for induction motor conversions.   For example...  say you wanted to build a 20' machine, with a cutin speed at 60 rpm and you need about 5KW @ 250 rpm.  Where would you start? - what motor would you buy - what magnets and how many -  would you have to rewind it? - If so... how?  For me that data is missing for the conversions.

Converting induction motors is a brilliant way to build a good alternator  - but the trouble is too many people think they can just take any more, pop some magnets in and go.

The first things I made up here when we started this site were conversions and they seemed to work OK.  They didn't furl either and they held together.  The first axial flux machines I made were also non-furling (always 7' diameter or under) and they survived as well.  I tend to think that any non-furling machine that survives must either be running in stall - or have severe alternator efficiency problems and running in overspeed.  

Overall I think either way you go is great if you build it well and match it well to the blades and the load and it's good to see different approaches.  I have not been converted yet though ;-)

[BigBreaker]

Solution for cast stator:

Cast thin copper tubes, running bottom to top, along the left and right sideof the stator, running bottom to top.  Press the tubing right up against the coils.  They will not attract large eddy currents as the profile presented to the magnets is similiar to the magnetic wire (and it does not loop).  Now connect the top of the tubes to a condenser (a car radiator will do) then to a small reservoir / accumulator with plenty of air space, then plumb a drain to the bottom of the tubes.  A pressure relief valve is a good idea.

Presto, the outside legs of the coils CANNOT go above 100C.  Copper has very good thermal conductance to the rest of the coil should also stay cool.  The heat transport of a loop heat pipe (what we've designed with the tubes) is insane - way over 2k watts once the water starts boiling.  The relief valve can do double duty by throwing a shutdown switch for added protection.  The pressure in the reservoir will provide plenty of force.  No moving parts and the heat problem can now be solved OUTSIDE the stator.

[Flux]

Having seen the problems with water cooled windings on large hydro generators and knowing the things likely to fail on wind generators in general I refrain from comment.

Flux

[cyplesma]

would / does the rotor having slots in it to allow air flow to the coils (whether cast or not) help to keep the coils from burning up?

I myself don't know, I'm just asking. 8 )

.

[Dave B]

Great to read all this in a diary or rants and opinions instead of the main page, thank you. I'm with Dan, the cast axial design is simple and is proven to perform very well. Like any design if you push it beyond it's limits it will fail. Design a different stator and push that beyond it's limits and it will fail also. If all is engineered in proportion to the required output then the cast stator is not necessarily a design flaw for larger machines either and the "builders" here rather that the talkers know this. You could build a 10 KW axial machine with a cast stator and never burn it up. Creative ideas and thoughts are the fuel for design, my hat's off to those who "just do it" besides.  Dave B.  

[Ungrounded Lightning Rod]

Additional point:

If some of the dust in your area is magnetic (like iron ore) and things are dry enough that it is occasionally windblown (like in a desert or drought-prone area), building your PMA as a sealed unit is a must.  Otherwise the dust will build up on the magnets until it closes the gap and starts scraping.

Lots of sealed motors available for conversion.  I have yet to figure out how to seal an axial flux machine.

[wooferhound]

Try rubbing a magnet on a copper pipe. You will change your mind about the lack of Eddy Currents. You could use Plastic though, and then you will need slip rings for your plumbing.

[wooferhound]

Here is how I sealed up my machine . . .

[wooferhound]

When the Genny is making power there is a resistance to the magnet rotor which has a braking effect. This exact braking force is seen on the stator too, trying to rotate the stator along with the magnets. Well this force is also seen on each individual wire as this is where the force is actually occurring. This results in the wires vibrating together. it is a rather weak force on each wire because the individual wire is producing a small amount compared to all the wires in total. Well these small vibrations add up over time and can possibly rub the thin insulation off of the magnet wire creating shorts inside of the individual coils and reducing output.

Casting the stator will prevent this from happening. I have seen someone on this forum that wound their coils while adding a layer of glue to the wire to prevent this vibration. I can't help but think that the coils would be deforming themselfs and getting fatter as the wires migrate around do to the vibrations caused when generating power.

Aren't motors made with a coating of varnish on the windings to prevent this ?

[rossw]

"Bench Racing" for a moment.....

If the magnet disc(s) had fins or grooves, and slots or holes near the inner radius, these would act as centrifugal fans to move air DIRECTLY across the surface of the stator, right exactly where you want it.

At low speed, they would present little wind load (drag) and offer little cooling effect, but then thats when the stator needs least cooling anyway.

At higher speeds the "fan" would have more effect, but there is lots more power from the prop anyway.

With some ingenuity, the front "cone" could perhaps even be more like a scoop and let the wind "push" in to the centre of the mag disk to provide some forced-cooling directly across the stator face.

Ok, you can laugh now, and return to your scheduled programming....

[SamoaPower]

Thanks to all for the comments. The responses were pretty much as expected with a lot of common sense displayed.

A couple of important points were brought out concerning people's capabilities and what they are comfortable with. Certainly it's applicable to the issue.

The point about proper system design is probably most poignant. I strongly suspect that many systems are `designed' piecemeal and suffer the results of that approach. Let's build an alternator today and worry about the rotor and load tomorrow - that sort of thing.

To the main topic and aside from the axial vs. radial controversy, I have a few comments in additional to the thermal issue that no one seems to dispute. Cast stators run hotter, period.

The main points brought out in favor of casting seem to be simplicity, strength and the lack of a better design. Some cite it's history but I'll let the failure reports speak for themselves.

Simple is a relative term. Again, individuals experience and capabilities come into play. I'm not a stranger to glass and resin work so I have no biases. Building molds is something I usually reserve for multiple copies of something. For someone building one or possibly two of an item, a mold is just an added burden that is cast aside after completion of the project.

Material availability can be difficult in out-of-the-way places. It certainly is here. Obtaining better resins, glass and additives can be an added frustration to those of us in the boonies.

Coils need to be prepared as a separate assembly before casting. One mistake in a joint or coil polarity, once cast, results in a difficult, if indeed possible, repair job. Attention also has to be paid to how the wiring exits the resin. A broken wire at the exit point isn't fun to repair.

The process itself is less than straightforward. Proper catalyst amount and mixing is significant. How much of what additive and what effect on the cure? How it's poured and trapped air can be an issue. A proper curing environment is important. If cracks result, they can cause stress concentration that compromises structural integrity. I have to wonder how many novices turn out a good one on their first try?

Simple? Not by my definition.

Strength in the typical cast stator is an illusion. Because it's massive and heavy, many will assume it's also strong. Such is not the case. In a fiberglass lay-up, strength is obtained from the glass cloth, not the resin. In the typical stator, a layer of cloth is added above and below the coils. This small amount compared to the total volume will add little strength. It certainly won't prevent a casting from deforming under thermal and mechanical stress. And, it doesn't take much to make contact with the spinning magnets.

The lack of a proven alternative design is probably the largest deterrent to breaking out of the cookie-cutter mold that most tend to follow.

Come on, fellas and gals! Don't become one of the sheep! There's a lot of inventive, capable minds out there to solve what is basically a pretty simple problem. I challenge all of you (even you Dan) to rise to the need of a better stator!

I do have a design for and have built a non-cast stator that I have confidence it will prove itself to be superior to the cast version and is cost competitive. I have mentioned it in past posts and comments but will not promote it any further until proven because I don't want to seem to be in competition with anyone. I just want a better machine for you and me.

[vawtman]

Samoa

 I think one of the flaws in the system is that the plans are there but it takes time to absorb how it REALLY works.

 People seem to try the easy way out to get power and fail(I KNOW)And wouldnt think a properly designed solid stator would be a problem but it takes time too understand.

 So many variables come into play but a challenge for us.

[DanB]

Hi Samoa - this is a good posting and I really respect you're efforts and contributions....

I tend to think that perhaps your the type that likes to make things a bit more complex than necessary - and maybe I tend to resist change and make things too simple sometimes.  Probably the best way lies some place in the middle.  I also like to play the devils advocate sometimes so forgive me if you think I'm arguing - I see you points sometimes but I like to put fourth another point of veiw for folks to think about (for better or for worse).

"The point about proper system design is probably most poignant. I strongly suspect that many systems are `designed' piecemeal and suffer the results of that approach. Let's build an alternator today and worry about the rotor and load tomorrow - that sort of thing."

Yes - that often seems to be the case.  It's the sort of trial and error approach that I and lots of other folks take.  These days it's a bit less trial and error for me it seems  - partly due to all the past trial and error, partly due to lots of good ideas from other folks.

" Cast stators run hotter, period."

No doubt about that.. how much hotter Im honestly not sure but your posting is a nice effort at putting some numbers on it.

"The main points brought out in favor of casting seem to be simplicity, strength and the lack of a better design. Some cite it's history but I'll let the failure reports speak for themselves."

Again - any machine will burnout.  A cast stator is usually fine if you keep it above 50% efficiency  - and if you're going below that then the stator is not the problem, the design of the whole machine is.  It needed a bigger alternator...  Done right, a cast stator could last forever and it's not difficult to get it right in my opinion.

"Simple is a relative term. Again, individuals experience and capabilities come into play. I'm not a stranger to glass and resin work so I have no biases. Building molds is something I usually reserve for multiple copies of something. For someone building one or possibly two of an item, a mold is just an added burden that is cast aside after completion of the project."

Well - the damned things burn out so often you want to keep the mold ;-)

Actually - the stator mold is not complicated, it's a very quick 1 hour or less project in my opinion.  Any sort of form to hold coils and be part of the stator is surely more complicated.

"Material availability can be difficult in out-of-the-way places. It certainly is here. Obtaining better resins, glass and additives can be an added frustration to those of us in the boonies."

Polyester resin is pretty easy to get.  I expect if you cant find that you'll never find magnets or copper wire.  I don't use polyester anymore but it works fine.

"Coils need to be prepared as a separate assembly before casting. One mistake in a joint or coil polarity, once cast, results in a difficult, if indeed possible, repair job."

Yes, true.  That used to be my argument for bringing all the wires out of the stator and making connections on the outside.  But geez... it's pretty simple stuff - kind of like stacking up D batteries in a flash light.  Trick is get it right and double check your work before you cast it.  It's not difficult...

" Attention also has to be paid to how the wiring exits the resin. A broken wire at the exit point isn't fun to repair."

Its not that bad - use a chizel or a dremel tool and go fix it - we've had to do that  sort of thing a couple times.  Although if you're careful it should never happen.

"The process itself is less than straightforward. Proper catalyst amount and mixing is significant. How much of what additive and what effect on the cure? How it's poured and trapped air can be an issue. A proper curing environment is important. If cracks result, they can cause stress concentration that compromises structural integrity. I have to wonder how many novices turn out a good one on their first try?"

Depends what you call a 'good run'.  A workable stator can be pretty ugly.  The first stators I ever made were quite ugly and worked out fine - two are still working well and making good power after 4 years.  Air bubbles are not a problem and the amount of  catalyst is pretty simple - if you follow the simple instructions it almost always works out fine.  I've seen cracks in large castings but they were nothing that couldn't be repaired.  You could write a book on it - there is a lot to that stuff, but it doesn't need to be that complicated to workout quite well in my opinion.

"Simple? Not by my definition."

I find it very simple and it seems that most attempts by folks who've never done it before work out fine.  I'm always 'tweaking' my own process trying to improve things but for the most part the improvements are only cosmetic-  or to save time.

"Strength in the typical cast stator is an illusion. Because it's massive and heavy, many will assume it's also strong. Such is not the case. In a fiberglass lay-up, strength is obtained from the glass cloth, not the resin."

yes, very true...

 "In the typical stator, a layer of cloth is added above and below the coils. This small amount compared to the total volume will add little strength."

There are not forces trying to bend or break the stator - it doesn't need to be that strong.  Except for one that got smashed in shipping once - I've never seen one break.  Strength is a non-issue.

 "It certainly won't prevent a casting from deforming under thermal and mechanical stress. And, it doesn't take much to make contact with the spinning magnets."

I find the inside of the coils (the copper) fails before the resin deforms.  By the time the resin starts deforming it's allready too late.  The machine should've had a more efficient alternator or furled earlier.

"The lack of a proven alternative design is probably the largest deterrent to breaking out of the cookie-cutter mold that most tend to follow."

Yes - I agree.  When better ideas come fourth I expect folks will move in that direction.  And... there's a good chance that better ways of doing things exist.  I'm always trying to improve things about the machine here - always making things stronger.  I did get concerned about stator efficiency this year (out of about 30 machines flying we had 2 burnouts which should not have happened.)and decided to  start using larger magnets  - that reduced stator resistance by about 50%  - and so far so good.  

"Come on, fellas and gals! Don't become one of the sheep! There's a lot of inventive, capable minds out there to solve what is basically a pretty simple problem. I challenge all of you (even you Dan) to rise to the need of a better stator!"

I'm sorry ;-) - I like our stators just fine!  I do hope something better comes along, but I like what we have now, it works well.  It takes about 2 hours to make one - they're flat - strong and they work fine.  I know there are other ways - Jerry's doing something similar to you (although hes still pouring resin in the holes I think so no improvement for cooling) - there are surely many possibilities.

"I do have a design for and have built a non-cast stator that I have confidence it will prove itself to be superior to the cast version and is cost competitive."

I hope so - Im sure it will run cooler.  Again though - if you have MPPT like you've talked about in the past I really don't know why you're worried about it.

" I have mentioned it in past posts and comments but will not promote it any further until proven because I don't want to seem to be in competition with anyone. I just want a better machine for you and me."

There's no real competitiion here Samoa - I looked at your diary about how you made your coils with great interest and I've been chomping at the bit to see the rest of it all - I expect everyone else has too.

At any rate...  thank you for a good posting/a nice discussion!

[wooferhound]

How do you keep the wires from vibrating against each other and migrating around and the coils get fatter and rubbing the magnets ?

[BigBreaker]

The copper tubing would be a "ribbon" of 4 to 6 small tubes in a row that would not be much thicker individually than the coil wire.  You could use plastic tubing but the heat conduction is not as good.  The tubing would be on the outside of the coils, so it would be somewhat removed from the direct flux path.

I don't know why you would need slip rings...  the water condenser sits on the mill, not the ground.  The water just circulates through the stator.

Apparently Flux is aware of some additional failure modes, but I think the motionless stator of the axial flux format cures a lot of them.  Water is a great working fluid but others might be better.  Mixing in some alcohol would lower temperatures.

[dinges]

"The tubing would be on the outside of the coils, so it would be somewhat removed from the direct flux path."

Then it would also be removed from the place where (most of the) heat is generated.

Personally I like the KISS approach.