potting individual coils

[kitestrings]

Greetings,

Howdy,

I've been experimenting a bit with with different epoxies and strategies for the coil mold.  Our mold is intended to give us one of the twelve identical sectors that will form the stator.  T& G on the sides.

If I was pourin' concrete - something I've done alot of - I'd consider using mesh or fiber-strand, or both.  With potting a coil in this fashion - something I haven't done alot of - is there a coarse, reinforcing fiber-mesh that could be used at near the top and bottm surfasces that might improve the strength, particularly in the area beyond the coil?

~kitestrings

[gww]

In what way would toung and groove be helpfull?
gww

[kitestrings]

gww,

The concept is that the stator is only fastened at the perimeter of the stator.  The T&G just keeps all of the sectors in one plane.  All the wiring is external and the stator, or portions of it, can be removed without having to take the machine down.

~kitestrings

[gww]

So if you had nine coils you would have to have a stator bracket with nine points to secure the stator?

[kitestrings]

No.  There will be a mounting ring around the and beneath the perimeter of the sectors that, in turn are fastened to the stator brackets.  I believe there is only six brackets.

You might rember that Bob Shau post a similar project.  his had a ring at the center also.  We are hoping to avioid that (by using the T& G).

~ks

[gww]

So when stators burn out do you usally only lose one coil? (do you trust the rest of the coils that are part of a burnt out stator?)
If you went from nine coils to twelve would the angle of the toung and grove allow a larger rotor?  Is the reason for doing this for easy changing of different wired stators?
How would having this set up be helpful.
Thanks
gww

[kitestrings]

Gww,

Regarding the logic for individual coils, I'd say the main reasons are:

better cooling
longevity - there seems to be a a fair number of examples of stator cracking, signs of weather deterioration, perhaps due, in part, to expansion and contraction
no unexposed wire connections
consistency - all coils will be identical
ability to replace one or more coils without taking the machine down - consider for example a 'rub' due to a failed bearing.

I'm not suggesting this approach is better; I'm sure we'll learn all the unintended downsides as well.

As far as the angles, I'm not sure I understand your querstion, but we do have 12 coils.  The layout and size is pretty standard it is just that there is a T& G joint between each 30-degree sector.

" alt="" class="bbc_img" />

Here's the link that I couldn't find earlier:

http://www.fieldlines.com/index.php/topic,138337.0.html

~kitestrings

[gww]

Wow impressive older post.  I learn more posibilities everyday that I pay attention.  As for as your toung and groove,  It would seem to me that some of the cooling benifits of the other stator may be lost as the space between the coils would no longer be empty.  One thing I have never understood was if a rectifier is placed on a peice of aluminum and that cooled it,  Why would a totally cast stator not have more surface area and therefor also cool better?  I know most think it doesn't cool better but I don't really understand why.

I don't know enough to help you in anyway, but apretiate your responces and reasonings for the direction you are headed.

I also like the coil spacing uniformity.

I would think exsposed wiring would be a double edged sword.  Easy wiring configuration but more connections exsposed.

I am also sorry if I sidetracked you original quetion of what to use to stengthen your casting.

Thanks
gww

PS
to clareify my question;  If you took out three coils and went from twelve to nine,  would the coil angle still fit but make a smaller circle?

[windvision]

Kitestrings:

  Interesting concept! Would it be possible to take the coils to a rewind shop to dip them in varnish? My worry would be the wires would start vibrating under a heavy load causing the insulation to wear off and short the coil. This would solve that problem. Keep up the good work!

[Flux]

I am inclined to think that some form of glass strand is the best thing to give strength to the T & G region.  Chopped strand mat is readily available but something with shorter strands may be easier to deal with, you could probably cut it for the small quantities you will need.

A good general mineral filler is also good for strength and also aids the heat transfer. We used marble flour for mouldings but slate powder, ATH and various other things should also do the job.

Some people seem worried that this thick mixture will not penetrate the coil, if it is well wound I don't see this as an issue but if you are worried it may be best to impregnate the coil with a thin penetrating epoxy and then use the thick filled resin for mechanical strength. I don't like the idea of mixing systems so I wopuldn't use winding varnishes and then encapsulate in epoxy.

If you can mould good quality sectors then this idea seems to have some merit and it is worth getting the moulding right to produce accurately fitting sectors. I am inclined to suggest you get as much mineral filler in as you can work with, it will aid stability, heat flow and probably mould release ( without quality metal moulds I see an issue there with epoxy)

Flux

[kitestrings]

gww,

Wow impressive older post

I agree.  I thought this was a very impressive build.  Haven't seen much follow-up from him though.  Too bad, I'd like to know how it turned out for him.

Regarding cooling - In Bob's design the sides of the sectors taper back as they near the perimeter, so the cooling is probably better.  In ours the 'island' will be open, but we're opting for more material at the perimeter to add strength - a trade off - but still should be better than a stator that has no voids.

One thing I have never understood was if a rectifier is placed on a peice of aluminum and that cooled it,  Why would a totally cast stator not have more surface area and therefor also cool better?

I suspect the answer to this lies in the materials.  With a rectifier you have a high concentration of heat that you are trying to dissipate from the bridge or diodes thru a heat exchanger - presumably an efficient air-cooled material, like copper or aluminum (neither of which we really want in the stator).

With the potted coil(s), the potting material is not a good heat exchanger, probably a better insulator in some cases.  In theory the copper coils would dissipate heat very well alone, if we could just hold them in place somehow.  There have been quite a few discussions here on the "non-cast stator".  What we found in researching & experimenting with epoxies is that some have reasonably good thermal properties, but they tend to be thicker - so there can be a trade off between viscosity and the best thermal properties.  The one we're planning to use is 3x more better thermally (12 BTU in/hr degF ft^2)than some of the first we tried, but has nearly 5x higher viscosity (10,000) - though it can be improved a bit with heat and thinner.

windvision, I don't think the varnish would be strong enough.  In motors/generators/transformers, things are usually mechanically supported thru other means.

Flux, thanks, this gets to my original question.  I think the matting sounds good.  I was picturing that we could pre-cut and insert the material.  I'm not too concerned about penetrating all of the strands.  A combination of 3% thinner, a small amount ogf heat and vabrating the mold (air needles scaler) seems to help alot.  The thinner trades strength if used in too high a %.

If anyone has a brand-name product they like, please share.

thanks, ~ks

[kitestrings]

If you took out three coils and went from twelve to nine,  would the coil angle still fit but make a smaller circle?

I missed this one - if we went form 12 to 9 coils, I'd assume that we would have 9 sectors with a 40 degree angle.  The mold would be different, but the concept the same.  Probably the coils would be more trapazoidal (if there's such a word); narrower at the center; wider at the perimeter.

Regarding connection - we plan to use heat-shrink ring terminals with rubber boots, so picture something similar to a spark plug boot in the end.
http://www.ebay.com/itm/BLUE-SEA-SYSTEMS-p-n-4009-Cable-Cap-Stud-Insulator-10pc-/260558280442

[DanG]

I made up a chamber large enough to do a whole stator only to have the glass dome chip and crack from inattention when moving it.

The only way to ensure void free wound coil castings is a vacuum chamber - a one coil form would boil up and continue to foam and bubble until the air pockets has dispersed as the vacuum becomes harder. Any air that is left behind will only occupy a tiny percentage of its 'under-vacuum' shape when pressure is reintroduced so you really don't need a perfect 29-inches vacuum.

If you can modify each coil mold to include a 'tank' design with high walls to stop the resin from escaping, and have a strong enough chamber to accept 14-20 inches vacuum minimum. The homemade micro air pumps will not be continuous duty - having a vacuum cleaner and then a portable air tank evacuated to reduce the run times of DIY air pumps may make a shoe-string budget workable. The little medical nebulizer air pumps can make a vacuum but only to 12 or 14 inches. Cheap 12V air compressors can get better than that but will likely overheat unless you go in stages and have a leak-free chamber.

[kitestrings]

Dan,

The only way to ensure void free wound coil castings is a vacuum chamber

Is this necessary is what I'm questioning?  The wire is enameled, so we don't really need it for electrical insulation, or thermal protection.  It seems to me we are really just trying to pot the coils piece-meal as apposed to one casting.  The limitation is probalby more structural.

I cast a small section, with the femal portion of the T&G, mainly to see how it released from the mold.  After a baked post-cure it seemed pretty bullet-proof.

Having said all this, we do have access to a 7.5 hp vacuum system for our maple sugaring operation.  30" of vacuum is quite possible, but I want to make sure the benefits outway the complexity.

Thanks for the input, ~ks

[gww]

ks
Thank you for taking the time to adress my questions.
gww

[kitestrings]

One other thought for gww,

There's possibly one other side benefit from this, or a split stator, approach when you think about the assembly/disassembly.  With the conventional design the only way the alternator goes together, or comes apart, is by jacking the magnet rotor plates apart, removing the upwind rotor and then the stator.  Something that I dare say you'd never attempt at 100+' in our case.

If on the other hand the stator can be removed independent of the magnet rotors, it's conceivable that we could remove the stator, then the rotor/hub assembly, leaving the yaw assemby, tail and spindle behind.  This might allow some repairs like bearings, seals, paint/corrosion mitigation, without taking the hole unit to the ground - and maybe done with a lighter weight gin pole.

We've worked on many turbines that seemed to have little forethought to service in the future, mostly without cranes...

~ks

[fabricator]

If you took out three coils and went from twelve to nine,  would the coil angle still fit but make a smaller circle?

I missed this one - if we went form 12 to 9 coils, I'd assume that we would have 9 sectors with a 40 degree angle.  The mold would be different, but the concept the same.  Probably the coils would be more trapazoidal (if there's such a word); narrower at the center; wider at the perimeter.

Regarding connection - we plan to use heat-shrink ring terminals with rubber boots, so picture something similar to a spark plug boot in the end.
http://www.ebay.com/itm/BLUE-SEA-SYSTEMS-p-n-4009-Cable-Cap-Stud-Insulator-10pc-/260558280442

You can get those right off the hook at West Marine stores.

[fabricator]

I'm wondering if you even need any kind of cloth in the castings, the reason for glass fibers in one piece stators is because it is a monolithic pour and much more prone to breakage than  a single coil.
if you had somebody mill you a mould out of aluminum with a top half and a bottom half getting the mould apart would be a piece of cake.
30 inches of vacuum is a perfect vacuum and almost impossible even with extremely expensive laboratory grade pumps, besides I don't think you need it if you can saturate your coil relatively well, IMHO you would be better of with a small pneumatic vibrator bolted to the bottom half of the mould, fill the mould and have a small reservoir machined into the top mould at the fill hole then fire up the vibrator, I think you would end up pretty air pocket free.

[DanG]

The only way to ensure void free wound coil castings is a vacuum chamber

Is this necessary is what I'm questioning?  The wire is enameled, so we don't really need it for electrical insulation, or thermal protection.

I worked as electrician-mechanic on 750VDC subway cars and take the most basic 'coil' seriously. How seriously? Say the continuous duty aspect of wind power, the high amperage, all-weather operation, the unattended nature of their work, the costs and hazards of incidents and down time, the effort and materials for repairs...

Have you ever heard a dual rotor units growling at cut-in for hours at a time? The impulses throwing down the wires pulse everything enough to shake the assemblies and resonate the whole structure. Copper wire is malleable even at small lengths when generating power and will yield eventually and wind up (pun) escaping partially supported clamping . Even consumer-grade disposable motors have a lot of hidden engineering in them, if the coils were commercially wound they have tricks and methods to lay the wire down so there is no slack or undue stresses built-in, can we wind 12 or 15 pounds of wire by hand so there is zero strand flex in the core of the windings?

Thinking about abrasion causing a short circuit isn't just one winding loop but jumping winding layers - then 10, 15, 30 times a second throwing an arc that moves metal - now you have a whole flock of windings damaged, an imbalanced stator with malformed waveform and constrictions in conductor size making even hotter hot spots. Even with single potted coils reducing the nuisance factor and labor costs its still higher risk on the hours-between-failure risk chart. Can we talk about open circuits and run-away blade sets?

So, in this sparky's opinion, making sure the two inches of swept area on each leg of a coil is bubble free - think long toothpick shaped bubbles - for an advanced design is worth the effort. And its not much effort, no need for interstellar vacuum pressures, just enough to make sure every strand is wetted.

[jvnn]

So, in this sparky's opinion, making sure the two inches of swept area on each leg of a coil is bubble free - think long toothpick shaped bubbles - for an advanced design is worth the effort. And its not much effort, no need for interstellar vacuum pressures, just enough to make sure every strand is wetted.

That's correct, you just need to get the major bubbles out and medium vacuum will do that quite nicely.

The process you guys are talking about is pretty much the same as people use to make fiber composite structures.
There's a lot of materials and know-how out there, for example here's a how-to video on vacuum bagging a glue-up.
http://www.youtube.com/watch?v=VodfQcrXpxc

[Tritium]

It would be relatively simple to remove the air from the mold containing the coil and THEN inject the encapsulation compound into the air free mold area. As long as the chosen compound is one that has favorable low out-gassing properties it should produce a nearly bubble free casting.

Thurmond

[Flux]

For high voltage machines it is important to not have air voids as you get a corona discharge in the voids and they result in rapid deterioration of the insulation .

I assume this is moderate or low voltage so the problems are not so great. Size is also a factor and for a small wind alternator as most build the forces are really quite small and unless you have significant lengths of unsupported copper there is probably not enough force to cause any form of damage. By far the most drastic force you will ever see is if you use a brake switch at full speed. If you intend to do this then more care is needed but even so it shouldn't be a big issue.

This is just a personal opinion and some may disagree, but I don't think it is that important that you get perfect penetration between turns of the coil. Similarly I cant see that significant air voids in the moulding will do much harm either as long as the weather doesn't get in.

On the other hand it is probably worth taking at least some trouble to de gas the resin as it is near impossible to mix it without introducing air.

As said. a perfect vacuum is not needed in this case.  Industrial VPI plant have very elaborate vacuum systems but it is a different problem with high voltage coils.  We did a lot of potting of small electronic modules with just a normal laboratory vacuum pump but we did have fairly constant problems with the pump seizing up from resisn that seems to get carried over as vapour during the constant pumping, no liquid resisn ever got through so it was not a direct contamination problem.

I can't advise how near perfection you need to aim for but if you take no care at all I don't see it being any consequence. Things like growling at cut in really are tiny forces exciting resonances and I can't see them being important, the only force ever likely to do anything is during a short and this is so infrequent that I can't see it being a fatigue issue.

If you can solve the mould release problem I think you are virtually home and dry and you should get good workable coils.

Flux

[Flux]

I see that Tritium posted before I did. What he is proposing is in fact the method used in commercial VPI plant and it is the correct way to go for a perfect coil but even then you would need a very thin penetrating resin to be absolutely sure you got completely between turns.

Commercially the coils are pulled down to near a perfect vacuum, the heated and de gassed resin is then introduced and finally the chamber is pressurised to something up to 150psi for a few hours to force the resin in as well as it can be done.

It is highly unlikely the pressure part of the cycle is practical on a small scale but the vacuum part is probably better than applying the vacuum to the resin when in the mould, although that is likely more than good enough if you can deal with the stuff frothing up and spilling everywhere.

How near perfection do you want to go? I think it is a good basic scheme anyway and worth a try, if it turns out to be a far better way to build a stator then maybe then is the time to go for perfect potting.

Flux

[fabricator]

It would be relatively simple to remove the air from the mold containing the coil and THEN inject the encapsulation compound into the air free mold area. As long as the chosen compound is one that has favorable low out-gassing properties it should produce a nearly bubble free casting.

Thurmond

As in the videos above I believe you need a certain amount of time to evacuate every air bubble while the resin is flowing, that pot they have between the vac pump and the mould is simply a drop out chamber most likely with some pretty fine filters in it to keep any resin vapor out of the pump.

[kitestrings]

Howdy,

It may not be clear form the photos, but the mould as we currently have it configured is set up to be broken down piece-meal after potting.  The basic approach is to set the coil in over the center island, attach the terminals thru the top (held with washers and nuts), add the cover and squeeze it all together.

After the potting, the cover is removed, then the top, and male portion of the T& G (left side, forming the female groove in the potted sector).  The island is pin located, but has two small jacking screws should we need them.

Tolerances are pretty close, but if we we're to evacuate the mold, is there a 'form-a-gasket' or some such that allows one to temporarily seal between mating surfaces?  I was assuming just silicone around the terminals.

Thanks very much for the advise and contributions.  Great diversity of experiences, and lot to consider here.  I'll thank you all if it works well, and leave out al names if the attempt falls flat.

~ks

[fabricator]

Any of the silicone sealers will work, and will come off easily after you take the parts apart, what are the mould parts made of?

[kitestrings]

Aluminum is what we're planning.  1060 sound right?

[neilho]

You're probably thinking 6061. The T6 version is commonly available, machines and welds nicely, all good properties for mold material.

neil

[fabricator]

Sounds good, I was just hoping you were not going to try to use some kind of wood. It's almost certainly 6061 T6.

[fabricator]

I just talked to a guy who uses this http://www.grainger.com/Grainger/SPRAYON-Insulating-Varnish-1D276?Pid=search stuff a lot, it's pretty low low viscosity, you can get it in quarts or gallons too, what if you dipped each coil several times until they were basically solid then potted them?
On my last stator I cut shaped pieces of pink styrofoam in the center and after casting just pushed it out after a little deburing around the edges I wiped the left over foam with acetone and it went away real fast, might be something to think about.

[gww]

fabricator
nice
gww

[fabricator]

I know Flux won't like the tape. But it is Scotch super 88 high temp.

[gww]

I got lot of cheep eletric tape in all mine.  I know it acts funny with bondo.  Oh well I guess I should have read more before I built.  Now I'm just going to run with it and see where it takes me.  Nice stator though.
gww