A Different Alternator - Part 1

[SamoaPower]

This diary will cover quite a span and will be posted in installments. I'm not going to cover the complete machine design in this diary, but focus on the alternator because I believe I have some useful things to offer.

After discovering this site around the beginning of the year, I've been a frequent visitor to learn as much as I can about the axial flux alternator. I hadn't been aware of this configuration previously, even being involved with RE for 30 years (guess I didn't keep up).

It appears that Hugh Piggott has done us a huge favor by developing this configuration into a practical DIY design. Further work, by others notably DanB, have brought the design to a fairly high level and it's often recommended that his work be emulated. The advent of neodymium magnets made it all practical, of course.

Further reading showed some problems still remained to be reckoned with. There seemed to be a fairly high failure rate among machines as shown by a recent poll, which indicated about  23%. Whether due to design, execution or whims of the gods can't be determined.

Being located on a small island (American Samoa) in the middle of the Pacific, presents its own set of problems, particularly in obtaining materials to work with. Also, services available elsewhere, such as water-jet and laser cutting are non-existent here. The cost for shipping these in from elsewhere is very expensive. One rotor disk might cost US$100, just for shipping. Clearly, some design changes were necessary to allow building a practical machine.

Problems I saw that needed attention were:

    Low alternator efficiency with the attendant stator heating problems

    Stator cooling

    The whole resin-casting bit, magnet rotors and stator

    Excessive alternator weight

    Practical magnet selection

    Coil design

    Choice of materials

    Testing procedures.

The basic design objectives are:

    To produce one kilowatt in a 12 volt battery charging system at 16 mph wind speed (my site energy peak) at an efficiency of =>75% with my existing 16 foot, variable pitch air rotor.

    To have NO stator overheat (failure) issues at the 25 mph feathering point (>3 kW) (not furling, this is a variable pitch, down-wind machine, but about the same as).

    NO resin casting

    An alternator weight of <150 lb

    Minimal magnet cost

    Locally available materials, mostly

.

Now, some of this may seem like fantasy to some of you, but I believe I've done the homework to show the feasibility. I'll report success or failure along the road.

The next installment will report on coils since I believe that the actual coil winding results impact the physical design of the rest of the alternator. First, it was necessary to design a special motorized winding jig to achieve the results I wanted. The bits and pieces for the jig are on the way as well as 56 lb of #14 Square Hi-Temp wire. As a teaser, the 12 coils are 13, single layer turns of 7-in-hand #14 square wire, well bonded.

Until next time.

[pepa]

hi SamoaPower, i will be following your post very close. there are a lot of people that will tell you not to reinvent the wheel but there has been a lot of improvements to the wheel sinse it first rolled down the hill. the reasons you listed for improvement of current designs are close to the end golds of all of us here. the unproven design that i am working on has some of these improvements, but will have to be tested. keep us informed, pepa.

[Old F]

The failures you see are not so much of a design or system  problem.

It has to do with the learning curve any learning curve is the perfect platform for Murphy's Law  to plant both feet on.  

An on any project there will always be a learning  curve . Old Fs  rules  one an two  :  )

The are really only two must haves when building the current  design the wire and magnets.

Every thing else is fair game an depends on what kind of tools and toys you have in the shop and what back ground you have.

  An any time you have to adapt a part or build some thing you never tried be for  there will be a learning curve.  See rules one an two :  )

    A part no matter how available they are in your neck of the woods will always be scares as  hens teeth some were else. I guess this is rule three :  )

Look likes you have a fun project a head of you  keep us informed . And most of all have fun.

If your not having fun you are doing some thing wrong.   Rule four ?

Old F

Having so much fun it should be illegal

[SamoaPower]

Oops, I forgot that many of you like to have pictures. These few don't have much to do with the alternator directly but they are part of what lead up to what is coming.

This is looking toward the direction of the prevailing trade winds at 450 feet MSL.

It all goes up on this at about 60 feet when done.

The 16 foot diameter rotor aluminum blades on the paint rack.

This is the airfoil used. Looking at the blade end less molded tips.

Some parts of the variable pitch hub after coming back from the galvanizing shop.

[SamoaPower]

Sounds like your rules come from a man of extensive experience. I certainly agree with every one.

Half the fun comes from designing to solve the problems and probably picking up a few more along the way. The other half comes from the execution with the end results probably being anti-climatic.

Thanks for the comments.

[WXYZCIENCE]

SamoaPower, real nice wings you have there. What is the weight of one. I will be following your future posts.


My variable pitch hub works very well. I was expecting a challenge and I received one. Monkeys have slowed my progress the tower is lots of physical work.
Joe.

[willib]

what we need is a control theory expert on this board all that PID stuff make my eyes glaze over .

my point is building a varible pitch mechanism is great , and not an easy task at all , but proper control is key to a functional system , and there lies the problem

http://www.engin.umich.edu/group/ctm/examples/pitch/PIDpitch.html

lol the first eq. on the page is an example of eyes glazing over . maybe we just need a math expert with an interest in RE.

[finnsawyer]

If you really mean to critique the alternator then you need to determine its power producing characteristics independent of the prop.  This would require a protocol that others could duplicate.  It would mean, for instance, driving the alternator with a two or three horsepower variable speed motor so current, voltage, and efficiency could be measured for various loads (and various incarnations of the alternator?).  I mention this because in spite of my intentions I seem to have entered the realm of alternator design, as a result of my participation on this board.  I would like to build my design using the same magnets and coils (wound the same way) as a known 3/4 design for comparison purposes and to save money.  Once built and tested the design could be matched to a prop size.  The problem is that if I am correct about the greater power output from my design (maybe 6 times as much), than your 1 kw unit, it would require a larger prop then what I would like to do as a first case.  I would like to know the particulars about your coils and magnets, in any case.  Good luck.    

[altosack]

Hello willib,

I have a little experience with control theory (not an expert), a bit more with math and a lot of interest in RE; however, I'm not sure the page you referred to is a good starting point. BTW, when I looked, and saw that it was a tutorial on pitch control, I thought, WOW, just what I've been looking for !, but then I clicked and saw the picture of the airplane... Oh, well.

Anyway, if you would detail what you're trying to achieve, I wouldn't mind having a look at it. However, a warning, I am predisposed to using a microcontroller rather than discrete components for the control; I prefer math in software rather than hardware ! Not to say I wouldn't try to help, but I gather that others could help more with that than I could if this is the way you want to go.

Best,

Dave

[SamoaPower]

willib,

I'm trying to figure out how your comment relates to this post, paticularly since your link is an aircraft example.

[willib]

i am too , i will be programmong a PIC 16F877A

Sorry samoa , my apologies , this was one of those unavoidable diversions.

Dave we could continue this in my diary two steps down

sorry for the interruption everyone ,please carry on

[SamoaPower]

Joe,

They're 11.1 lb each with a fresh seven-coat paint job.

These have already had four years flying time and have proved themselves well. When I stripped them down, I checked very carefully for any fatigue cracks and found nothing.

The planform is a constant 13" chord and the airfoil starts at 25% of the mounted radius. Using variable pitch lets you simplify things somewhat without much penalty.

Glad to hear your VP worked out well. Maybe we can get a few converts.

I'd be interested in hearing about your pitch control algorithm.

[SamoaPower]

GeoM,

"If you really mean to critique the alternator then you need to determine its power producing characteristics independent of the prop."

Yes indeed, I certainly intend to do this. I've already set aside a 5 Hp engine and am working on the design of a high power shunt regulator to simulate a constant voltage battery bank. Testing into a resistive load or a battery bank that varies in voltage during the test doesn't supply as useful data.

My design approach is the opposite of yours. I'm starting with an existing air rotor of measured efficiency (42%) so the alternator is designed to match that rotor.

I may have led some astray by calling this a 'different alternator'. Its electrical/magnetic 12/16 design is pretty conventional axial flux. Its uniqueness is in the materials and construction techniques.

I rate the power of my machines differently than the convention. Rating a machine at some high end wind speed that I might see <1% of the time makes no sense to me. The commercial manufacturers numbers don't tell me much about my real world. For example, I call my Air-X an 80 watt machine rather than 400 watt since that is about what it can do at my energy peak wind speed of 16 mph. My goal for this alternator design at 16 mph is 1 kW. In more conventional terms it would be a 3+ kW machine.

This is not a small alternator. The stator is 24" in diameter and the magnet rotors are about 20". There's better than 30 lb of copper and 96 in^3 of magnet. A brief description of the coils was made above. This is a low RPM design that will be limited to about 160 RPM.

Perhaps, this design might not be the best one for you to compare to.

I might mention here that there are two additional design objectives I failed to note in the original post. One is that no machine shop services will be required (and I don't have one myself) and the other is that I have a three-year service interval goal.

 

[finnsawyer]

"Perhaps, this design might not be the best one for you to compare to."

   I have to agree with that statement.  It's a massive machine and I'd have to go to a larger rotor/stator.  Well, there will be others.

[BigBreaker]

Microcontrollers are so good today it's sick.  The ATMEL ones come with built in PWMs, D/A inputs for tach, voltage, current, etc and a serial interface for reprogramming, logging and monitoring from a full fledged PC.  You get all that for about $10.  Discrete components are a terrible idea.  I've advocated a separate section on the site for electronics but especially microcontrollers.  They open up control schemes that would be impossible otherwise.  Assuming that you have tach going into your micro it is trivial to use FETs for recitification - a nice voltage savings right there.