Potting Individual Stator Coils

[kitestrings]

Hi Folks,

I thought I'd share my recent efforts on this stator.  It is for a 15' er; 16-pole alternator.  Coils are 2-in hand #14, 52 turns.  Slightly different approach on the stator design.  It should allow the stator to be assembled or removed without separating the magnet rotors.

The encapsulant is a 2-part, thermally conductive epoxy with a ‘resbond' thinner to reduce the viscosity.

The process goes briefly like this:

(vacuum) ‘charge' portable tank,
warm mold and coil,
apply mold release,
mix epoxy,
position coil,
terminals and spacer,
cover/seal,
draw vacuum on mold, inject epoxy,
plug vacuum port,
invert mold, top off fill,
cure 24-hrs,
post cure at 250 & 350 degF

The molding process is pretty new to me, so I welcome any tips.  Still learning.  Thanks.

~kitestrings

[fabricator]

Looks pretty good to me.

[Boondocker]

Looks very nice and neat!

Did you do the machining for the mold?
Also are you planning on some kind of outer ring frame for support and rigidity?

[Flux]

I remember someone discussing this some time ago. I assume it must be you as the things look similar or identical, I remember asking about the g & T support.

Looks excellent, the biggest problems normally with epoxy is mould release but you look to have solved that. I take it that you are curing at near room temperature ( it was cure at high temperature that caused us so much release trouble).

I am a little puzzled by the two syringes, I originally thought you were injecting the epoxy and hardener separately and couldn't see that working, then I guessed it was just to increase the quantity of mix.

Anyway it is a really nice bit of engineering and well done for the effort. Should work out well.

Flux

[DanG]

Novel vacuum pump?

[kitestrings]

Did you do the machining for the mold?

No.  I mostly did the design with input form Neilho & my son, Mitch.  I made a mock-up in wood.  Mitch did the CAD work, and I had a associate at a local shop make it.  Nice workmanship.

Also are you planning on some kind of outer ring frame for support and rigidity?

Yes.  All the coil sectors will be identical, but we're planning on every other one connecting to the SS stator bracket assembly.  The bracket connects to the spindle shaft with a keyless bushing.  I was envisioning a poly-carb ring, and every other set would have a delrin spacer/reducer and just a 1/4" thru-bolts squeezing things together.  I want to see how ridgid the coils are alone.  The T & G is very close tolerance (+/- .001-.0015 as I recall), and the ring may not be necessary.




Flux, yes I recall discussing T & G (or G & T) with you and the crazy way we quantify vacuum in the US.  The syringes allow a pretty controlled fill of the mixed epoxy.  They also push much of the entrapped air from mixing  do to the orientation (which the vacuum pulls form the mold cavity.  I've been 'charging' the portable tank to about 27" Hg (I'll let you do the conversion), but it works well.  Lastly, the syringes allow easy of measurement - each sectr takes about 230 mL - and allow you to pressurize the cavity slightly at the end of the fill, forcing it into more of the voids.

The last one came out much better.  I took it out after 19-hours, air-cured another 5-hours, then post-cured.  It's nice material.  I spilled two drops on the shop floor.  Can't get them off, so I've decided to patch the concrete spawling on my floor with any excess.

Novel vacuum pump?

Dan, I'm sorry I don't follow.  Is this a manufacturer, or does it refer to a pump without storage?  We used our maple sap pump which is a noisy thing - 7.5 hp AirBlo (CA).  More here if interested:

Some related earlier discussion if interested:
http://www.fieldlines.com/index.php/topic,146368.msg1001486.html#msg1001486
http://www.fieldlines.com/index.php/topic,146931.0.html

Thanks for the help and encouragement.

~ks

[fabricator]

I think Dan was thinking you might be using the syringes to pull a vacuum.

[kitestrings]

Ahh, got it.

No, the tank is a regular ~10 gallon portable air tank that I PR off and added a ball valve and gage.  It is yellow and in one of the pics above.

Thanks Fab

[tecker]

 Good plan for a burn out recovery . How well it stays linear for dual rotors is a factor

[kitestrings]

Tecker,

That's clearly been a consideration.  We've tried to make it as monolithic as possible, once assembled, by holding a very close fit on the T &G, and by limiting the clearance on the mounting holes.  I used tempered C440 SS shim stock spacers, rather than just steel bushings (or not using anything).  Comparatively they have a about +.0002 tolerance on the ID as I recall.

The T &G has come out as close as I'd hoped.  Hopefully this pays off.



With regard to burn out recovery - there's that, though I think we're at no higher (nor lower) risk than other axial's.  Consider though, that there's a bearing scheduled for, or in need of replacement, a rub, or that the magnet rotors need a fresh coat of paint (hopefully all several years out).  With the stock designs, the machine has to come down.  This is may not be an issue with a 10'ter, but with a heavier machine it is IMO.

If this works as planned for assembly the rotors can be mounted to the hub, gapped and positioned on the spindle; then the stator goes in.  During disassembly, it is conceivable the blades are removed, the stator, and the alternator and hub is lower to the ground leaving the yaw assembly, tail-boom, hinge, vane all up top.  A big "if" I know, but this has been my motivation - to have potentially better serviceability, perhaps better cooling, logevity etc.

regards, ~ks

[fabricator]

Looks really nice so far.

[GoVertical]

Hi, great idea and great work. I could never cast a stator larger than 10 inch OD.  I look forward to seeing the rotors.

[kitestrings]

Here are the magnet rotors so far.  They are just primed, and need to have finish painting.  This is why the retainers are spaced off the plates at this pioint (so we can get the paint seal around the magnets).  I had to drill a hole in DW rotor to access the grease zerk, that's why there are metals filings in the one pic.  The SS spacers will hold the DW rotor off the back of the hub casting.  The blades will go on the upwind side of the (upwind) rotor.

Thanks for the comments.  ~ks

[coldspot]

Using the
  "way back time machine"
There was talk about something similar using ceramic
as a coil holding form or mold

[Mary B]

I had talked about using heat conductive coil epoxy for better cooling, this is 3 steps beyond that.

[Boondocker]

KS

Nice thinking, magnet retainer & alignment all-in-one.
Setting those magnets must have taken some extra maneuvering. 

[karlb]

This is great stuff and you have a good idea here.
Although I can see that you loose some winding space due to the 'slots'. Is this significant or does it not impact your output?

[Flux]

Interesting question.

Everything with wind power is a compromise, this is no exception. With direct charging without mppt you need significant electrical losses to get a reasonable match to the prop so loosing a bit of winding space is not a big issue. The main thing is keeping stator temperature within safe limits. If the improved heat transfer from this construction lets you push the wire harder, the lack of a bit of winding space may be of no consequence. Probably we shall not know how the compromises stack up until it has run and more information is available.

The whole issue of coil and magnet spacing is rather complicated and would take a lot of computer modelling to reach a definite conclusion. Looking at the pictures I would guess that this design uses the available space very well, and the uniformity and precision may give a better use than most one off windings.

If you want the ultimate best use of winding space i really think radial machines have the edge, but the axial design has simplicity and is very forgiving as you can correct a lot of design error by changing the air gap.

How it works out technically only time will tell, but I can't help admiring the elegance and attention to detail.

Flux

[neilho]

Lookin' good!

"The encapsulant is a 2-part, thermally conductive epoxy with a ‘resbond' thinner to reduce the viscosity."

What brand/part # epoxy did you settle on?

Neil

[kitestrings]

Hi Neil,

We experimented with a few different epoxies and techniques, though I could see where more thorough testing could get expensive.

We selected this one; the Durapot 865:
http://cotronics.com/vo/cotr/pdf/860.pdf
Note, I discovered an inconsistency between the thermal conductivity rating on this pdf (12), and the catalog (20).  I confirmed with the manuf. that 20 is correct (Btu in/hr. degF ft^2)

The biggest trade-offs seem to be that the best epoxies thermally tend to be thicker - 5 to 15 times thicker in some cases.  You can add a thinner, up to about 5% reportedly, before it starts to impact maximum service temperature and strength.  We used 3%.  The other 'trick' the manufacturer suggested was to heat the mold and coil slightly, and keep the resin & hardener at least at room temperature.  This seemed to make a big difference in how well it flowed and filled the smaller voids.

The other trade off was that ultimately I decided against using mesh (which I'd purchased and cut), or chopped strand.  The concern was that anything gained in strength might be more than offset if we didn't get thorough penetration or had voids, particularly in the T &G area.

The post cure reportedly improves thermal properties, chemical resistance and strength.  The latter is pretty self-evident between the 'green' material fresh out of the mold and the post-cured sector.

Regarding the loss of winding space, Flux has it pretty well covered.  The T& G is about 3/16".  The coils are wound such that they are tight at narrowest spot between adjacent coils.

From the outset we've been designing around using MPPT (with contributions from folks here).  Our alternator is somewhat overbuilt relative to the amount of prop we have (15').  There are some direct-tied units with about the same amount of copper (and magnets) with 17' rotors.  The idea is MPPT should let us start at a bit lower cut-in, and be efficient in low to moderate winds.  If we forego a bit on the top end because the alternator is more robust than we have blades to drive it, I'm okay with that.

As Flux points out, time will tell if this all works as planned, but I thought I'd describe the thought process to this point.  Still much to learn, and many spots to fail...

~kitestrings

[kitestrings]

Here's an update on this alternator.  We painted the rotors first with an epoxy primer with the retainers up, to seal around the magnets, then sucked down tight to the plates.


Jacking the rotors is the opposite of what most folks seem to be doing.  Here, the upwind rotor plate is mounted to the hub.  The downwind plate is jacked down onto SS spacers on the back of the hub and gapped.  The stator bracket will go in before the final mount on the spindle, then the stator will go in.


Rather than machine the back of the hub, I added or removed SS shims of the same ID/OD as the spacers.  Started out off a bit, after two adjustments I think it is within about +/-.005".  The gap is about .96-.97, so there should be about .10-.12" mechanical clearance.  May have to adjust some.






(my chainfall operator)


then the alternator is mounted on the spindle


Here's the fitted stator


Here I'm drilling the poly-carbonate stator rings.


We worked mostly on the yaw assembly over a long weekend for me.  Starting to take shape.

Does anyone have a strong opinion on what bearing grease to use in a climate that ranges from -30 to say 95 degF?

~kitestrings

[fabricator]

Molybdenum Disulfide grease is the best there is.

[kitestrings]

Getting a little closer.  Most of the metal fab is now done.  This is the stator bracket assy that holds the coil sectors in place.


I wanted to confirm the final location on spindle shaft and decided to spin it up with one of the sector/coils in place.  We ran 6.0 VAC at about 90 rpm, which should be close to the cut-in goal.



Here is the final fit-up of the alternator assy.  The inner blade bolts are held in place with stainless retainers, as they have to go in before the magnet rotors are gapped.


This is the yaw-head.  it is a 9.5" offset.  There is a matching flange welded on the tower stub, and a bronze/graphite bearing fits in between the two.  I also tapped two grease zerks on the on the yaw tube for the radial contact.  Here the taiboom is against the lower stop (unfurled postion).


Furled position


I fitted up the tail-mast, vane & strut.  It can be shortened or lengthened by 6" from this 'calculated' setting.  The tail is a bit heavier than anticipated.  I neglected to add the weight of the strut (behind the vane), and it needed to be heavier gauge stock (#11 ga.) for stiffness.  It's another ~6.5#, so I may have to adjust things.  Mary, my wife, offered some much needed ballast, and supervision.


Now I'm trying to work out the details of the manual furling.  I believe a cable will work, but it has to share space with an SO twist-cable (no slpip-rings on this one).  Still an element I'm not real pleased with - ideas are welcome.


This is close to the layout.  sandblasting and painting this week.  Hope this weather holds.

[fabricator]

looking really nice, do you have room to put a small diameter piece of pipe with a long radius on it in the down tube for your manual furling cable?

[midwoud1]

Highclass machine.

No sliprings .  A drop cable 3 phase  and your furling line with guiding tube like fab. mentioned.
Down the tower a plug and socket to untwist ( system Sparweb and Hugh P. )
Furling line Dyneema  anti friction and strong.
Good luck with your progress.

 - Frans -

[kitestrings]

Thanks guys.

Fab, I want to make sure I follow your thought here.  Right now there is a hole in the top of the yaw-head for a strain-relief (1.3125").  The wire will be a SO cord (SOW, I think they're now designated now).  In the next to the last pic you can see where I've put a small pulley in the center.  I was picturing a small pull cable, or line as Frans suggests, routed thru a tube - like a miniature gin pole - to a small winch, or capstan.

The only down tube is the 21' length of 3" schedule 40.

If the cord could serve as the means to manuall furl it - reinforced with Dyneema, or similar, or actuating (pumping) a small hydraulic cylinder, I'd prefer it.

Frans, yes, years ago I worked on some downwind induction units up to about 40 kW with just a twist-cord.  There's been several posts on whether the benefits of slip-rings outway the draw backs.  Done right, I still favor them, but we agreed on this one to try it without them.

Lastly, is there a consensus as to whether a separate ground conductor is neede.  My Sencenbaugh has 3-phase leads, and two field wires, and that's it.  Lot's of attention to tower grounding and MOV's & such, but no dedicated ground to the machine (other than thru its bond to the tower).

~ks

[kitestrings]

Fab,

[Molybdenum Disulfide/quote]

Is this "Moly-B", or something different?  Are there greases with varying amounts (%) of Molybdenum Disulfide?  And, what about temperatures - are there different weights like with motor oil?  We have some pretty cold winter lows here.  I'm aobut 20 mi. form the Canadian border.

Is there such a thing as a dry-lube (graphite-based) grease?  We're using a bronze/alloy/graphite thrust bearing.  Not that grease would hurt anything, but it does attach dirt, and probably is not as good in low temps.

~ks

[fabricator]

Thanks guys.

Fab, I want to make sure I follow your thought here.  Right now there is a hole in the top of the yaw-head for a strain-relief (1.3125").  The wire will be a SO cord (SOW, I think they're now designated now).  In the next to the last pic you can see where I've put a small pulley in the center.  I was picturing a small pull cable, or line as Frans suggests, routed thru a tube - like a miniature gin pole - to a small winch, or capstan.

The only down tube is the 21' length of 3" schedule 40.

If the cord could serve as the means to manuall furl it - reinforced with Dyneema, or similar, or actuating (pumping) a small hydraulic cylinder, I'd prefer it.

Frans, yes, years ago I worked on some downwind induction units up to about 40 kW with just a twist-cord.  There's been several posts on whether the benefits of slip-rings outway the draw backs.  Done right, I still favor them, but we agreed on this one to try it without them.

Lastly, is there a consensus as to whether a separate ground conductor is neede.  My Sencenbaugh has 3-phase leads, and two field wires, and that's it.  Lot's of attention to tower grounding and MOV's & such, but no dedicated ground to the machine (other than thru its bond to the tower).

~ks

Most guys use a 90 degree long radius tube to hold the furling cable, there is no chance for it to run off the pulley that way.

[fabricator]

Fab,

[Molybdenum Disulfide/quote]

Is this "Moly-B", or something different?  Are there greases with varying amounts (%) of Molybdenum Disulfide?  And, what about temperatures - are there different weights like with motor oil?  We have some pretty cold winter lows here.  I'm aobut 20 mi. form the Canadian border.

Is there such a thing as a dry-lube (graphite-based) grease?  We're using a bronze/alloy/graphite thrust bearing.  Not that grease would hurt anything, but it does attach dirt, and probably is not as good in low temps.

~ks

Yeah molyb is the same thing, prolly not necessary with your bronze bearing, and yeah there are several different kinds of dry lube products.

[kitestrings]

Thanks for the clarification Fab.

I spent quite a bit of time on this detail over the weekend.  Here's what I've come up with.  The pulley is the same as what we have on our farm (PTO) winch.  The rope/cord comes up through a tube mounted on the back of pulley, and it  can then rotate to face the direction of the line.  I was able to "actuate" the furling with just my foot thru a loop in the line.  I've got to check out this Dyneema line.  It sounds promising.


Here I set it up with just a block of wood with a counterbored thrust-washer clamped close to the final location on the yaw-head.  I since have drilled an oversized hole in the top next to the strain relief.  I have a steel bushing fitted through the top, and this will get welded in place to form the bearing for the pulley.


I aslo added rubber bumpers on the back of the spindle shaft and in the resting stop at the lower position.  The tail is still a bit too heavy, and I'd rather heir in the other direction, so I'm looking at removing a bit of material from both the strut (the source of most of the overage) and the vane.  It looks like 5-6# should be do-able.


Everything got sandblasted and painted last week, we'll be setting it up initially on this ~23' pipe and stub.  Endless details...


Regards, ~kitestrings

[kitestrings]

Well, the test tower/stub went up over the weekend.  It is only about 23-24', but should allow us to keep an eye on things and make any adjustments.



I added a small sheave in the base for the manul furling and picked up some 4 mm Dyneema line as Frans had suggested.


The tower crew.


Started the final assembly - I've torqued the stator bracket to the spindle & packed the bearings.  There's got to be a trick for seating the grease seal(?).  I considered putting it in first, but was concerned the seal might end up rolled.


Here you can see the new bushing and pulley for the manual furling


~ks

[Other David]

Hello kitestrings your workmanship is of the highest quality thank you for sharing with everybody. stator assembly is very interesting I'm hoping to adopt this approach, thank you very much.
David

[kitestrings]

This post is originally from 2013.  We made quite a few changes between then and now.  I've added a few related links, in hopes that no one has to suffer all the same mistakes; there are plenty to go around, so let's each contribute our own ;>]

Related specifically to the individually potted coils, we eventually changed the stator rings from polycarbonate to Garolite, with spacers added.  I would also note that while this design has worked well for us, if I had it to do over I might consider still having removeable coils, but breaking it into only maybe three separate sections.

https://www.fieldlines.com/index.php/topic,149894.0.html

We also changed the tail hinge and hinge-pin, and the vane was changed from stainless steel to aluminum.

https://www.fieldlines.com/index.php/topic,149071.msg1038351.html#msg1038351

Lastly, we added power furling:

https://www.fieldlines.com/index.php/topic,149230.msg1040709.html#msg1040709