Author Topic: Going towards axial flux (Read 3896 times)

imsmooth · « **on:** February 14, 2009, 07:06:34 PM »

Well, I took apart my F&P turbine only to find that the neo-conversion rotor had cracked from the impact. It no longer stays perfectly aligned with the stator and wobbles a little bit which is unacceptable. I am now thinking of trying the axial flux and using one of Otherpower's kits to build a 10 footer.

Here is my question and it has a few parts...

I did not get my Power-One inverter, but it was going to have MPPT and have a voltage range of 50-600v. Since I am going to redesign the stator, is this a good way to go, or are there cheaper grid-tie inverters using lower voltage ranges. The Power-One allows me to program the power curve, so if other inverters don't allow this than I will still go with my initial plan.

That being said, will the axial flux design perform well enough if I use smaller gauge wire for more turns to get the higher voltages? I have no plans to charge batteries; I am going to tie this to my grid.

Finally, and this is probably for someone familiar with the OtherPower kits, I have a question about the rotor plates (part 6610). I have seen some people grind out indentations to allow the magnets to stay in place before being cast in fiberglass or epoxy. Is this necessary? Will the magnets definitely stay in place with just a good casting?

TomW · « **Reply #1 on:** February 14, 2009, 12:46:04 PM »

imsmooth;

It is the 14th of the month. You have posted 12 fresh threads so far this month. Did you ever consider that it might be a LOT more productive if you just made it one or two continuous thread?

It just seems inefficient to post a fresh thread for every new thought you have on this ongoing project. Don't get me wrong, I just think you would be better served with it less spread out.

It would certainly be easier for those trying to help you get things sorted out.

Think about it.

Tom

Airstream · « **Reply #2 on:** February 14, 2009, 01:18:18 PM »

View imsmooth's diary "User Information" link is EMPTY!

We who read this site hourly will still find you through the miracle of recent activity ...

Keeping the non-diary stories to a minimum sure helps those who are genuinely interested in every post. There may be a dozen mini-threads available in one long diary post but viewing replies as nested keeps the replies in some order.

I sure am not saying making your own year-book diary is easy, choosing where new chapters start, but you have the space granted to you!

On the "Post a comment" form the small 'subject' box text does not need to be the default "Re: insert post title here"; one can use that to note a change of focus that the recent activity screen will display...

vawtman · « **Reply #3 on:** February 14, 2009, 03:37:20 PM »

Boy has i recall you spent mega bucks for the custom order chinese mags.Did the neos initiate that much drag that your givin up without attempting a repair of some sorts?

imsmooth · « **Reply #4 on:** February 14, 2009, 05:52:39 PM »

I don't see how I can get them out without destroying them. The rotor's center is cracked and won't sit correctly on the shaft. It is a shame. There was no cogging; it had good power generation. At this point I would rather make a rotor out of something more durable than hard plastic.

wpowokal · « **Reply #5 on:** February 15, 2009, 05:54:32 AM »

And so say all of us! Australia being a third world country we have very poor internet outside of the metropolitan areas, it cost me dearly to open each thread repeatedly, so now I try not to review too much or comment too often. All in one thread works for me.

allan down under

DanB · « **Reply #6 on:** February 15, 2009, 09:40:26 AM »

Howdy.

My experience is - with a good casting, the magnets should stay put. I've run these rotors up to 700 rpm with no casting/glue of any kind, and things stayed put - but - in the event that the magnets lose their magnetism (overheating from rubbing on the stator) or - should the machine for some reason get much over 700 rpm it's risky and I've seen a few machines where the casting failed/magnet flew out. So there are lots of ways to prevent that, some folks pin them, some folks bolt them in, some folks machine little pocket... I like to band the rotor. Odds are a good casting is fine but I prefer to add a bit more safety factor.

imsmooth · « **Reply #7 on:** February 15, 2009, 11:27:52 AM »

Thanks, Dan. I am planning on getting your metal kit to make one. I am not sure if I am going to use the 1" x 2" x 1/2" magnets and more coils, or go with more of the 1" x 3/8" round magnets and more coils. I want to go for 100v at 150 rpm for grid-tie.

As far as machining pockets: I like the idea and have seen others do this. What took can I hook up to my drill press to do this? I guess if this is what I want to do I would have to go with the round magnets and drill a small 1" pocket into your rotor disc.

Flux · « **Reply #8 on:** February 15, 2009, 12:38:27 PM »

I really can't see the point of lots of complicated machining to put the magnets into pockets. If you go deep you reduce the flux. If shallow they can only just stop the things moving radially, they aren't going in any other direction. You can do this as effectively with rectangular magnets by turning a small groove in the disc for them to rest against if you must. The band that Dan uses absolutely prevents then flying out and would work even without casting ( pockets probably wouldn't).

If they are stuck on with structural glue they don't go anywhere anyway but extra safety is no bad thing. Forget superglue it is a liability. Epoxy potting is great, I haven't used vinyl ester but it should help. Polyester is no glue but is perfectly adequate to keep the magnets in place as long as you make provision for the resin block not to move ( band round the outside or a few screws between the magnets for the resin to anchor to).

The magnet attraction holds them on, just do something to stop them moving radially. If they are demagnetised most of these methods fail but so what? what use is a set of non magnetic lumps nicely screwed on or retained in some complicated contraption.

100v at 150 rpm presents no problem, the current will be less so thinner wire to get the turns is fine. For a grid tie inverter with load tracking you will not run stalled so the larger round magnets ( ideally on a larger disc) may be a better choice as you can use the higher efficiency. Furling will not behave as for the stalled battery charger so bear that in mind. You decide how you limit volts to that acceptable to the inverter, if you don't do that you will kill it.

Flux

imsmooth · « **Reply #9 on:** February 15, 2009, 01:56:42 PM »

"100v at 150 rpm presents no problem, the current will be less so thinner wire to get the turns is fine. For a grid tie inverter with load tracking you will not run stalled so the larger round magnets ( ideally on a larger disc) may be a better choice as you can use the higher efficiency"

Can you elaborate on that point? If I have an oval coil of wire 1" x 2" in shape, how is the voltage/current different if I have 1" neo disc cutting through compared with a 1" x 2" rectangular block cutting through the coils. I know that only the magnetic lines cutting perpendicular induce a voltage, so is it fair to say that the rectangular block (same thickness, and grade) would have more flux and induce a higher voltage?

Second question: I have read discussions about the stator burning out, which did not seem to be a problem for the F&P as the coils were open to air. Why hasn't anyone left the center of the stator coils open to allow better cooling? Pour resin around the outsides of the coils, but leave the inside open. The coils will stay put and allow better heat exchange.

(I'm trying to keep all my thoughts in one thread, guys)

Flux · « **Reply #10 on:** February 15, 2009, 02:41:19 PM »

I don't know where to start.

"Can you elaborate on that point? If I have an oval coil of wire 1" x 2" in shape, how is the voltage/current different if I have 1" neo disc cutting through compared with a 1" x 2" rectangular block cutting through the coils. I know that only the magnetic lines cutting perpendicular induce a voltage, so is it fair to say that the rectangular block (same thickness, and grade) would have more flux and induce a higher voltage?"

If you start going down this cutting lines perpendicular route you will get completely bogged down. Just think of flux linkage and life is easier. If you link all the flux then area is the thing that matters.

Assuming that the 2" diameter magnets are 1/2" thick same as the rectangular ones then the 2 " dia have an area of 3.14 sq inch. The rectangular have an area of 2 sq inch. that is why the round ones do better. I don't particularly like round ones and they are significantly more costly than the rectangular ones. I personally would use more of the rectangular ones but as you are intending to use an existing design then the round ones are a better option.

For round magnets I cant help suspecting that the best shape for the coils is round but when you just replace the rectangular magnets with round ones on the same disc size the magnets are too close together to use round coils. Squeezing the things oval gets you more copper in and the loss of flux is tiny, the bits where the magnets nearly touch have little area and there is a lot of leakage flux so not linking that bit effectively makes little difference. The magnets are not used to best advantage but Dan has found that the round magnet version is significantly more powerful than the rectangular one so the use can't be that bad.

For the same thickness the shape matters little, to get more out you need larger area magnets or more of them, either way more magnet area means bigger discs if you want the full advantage.

The stator burning out is an issue of man, not the type of alternator. Your conditions are different as you will have some form of mppt but just for the record with the normal directly connected battery scheme if you go for a very low cut in and make the alternator efficient it will stall in high winds. The only way to keep the low cut in and get away from the worst of the stall is to make the overall efficiency less than 50%. This way the power you gain from matching the prop curve far outweighs the loss from low electrical efficiency.

If you choose to cause this loss within the stator it will get hot and how hot will ultimately depend on how well you get the heat out. You can play tricks with not potting things, play with air flow, fans, holes in coils or whatever you want but the thing inherently is not going to cool as effectively as a machine with coils wound on iron poles where you have thermal conductivity from copper to iron to an exposed surface. That may mean that you can push an iron cored machine harder for less risk of burn out but the system efficiency is usually lower for the iron cored machines.

The axials don't burn out if you keep within sensible ratings, most don't furl safely. If you want more output then just make the alternator more powerful and add the electrical losses outside, not in the stator.

In your case the heating will not be such a problem as your voltage will rise with load. What you are hoping to do is exactly the same as Halfcrazy is doing with the mppt classic controller. The standard 10ft machine is doing near 3kW when its realistic rating for direct battery charging is about 600W ( 700 -800 pushing it)

Winding for the higher voltage is just a case of using more turns of thinner wire. If you want to stick with the standard tried and tested design use the round magnet version in preference to the rectangular unless you want to go to N50/52 at far greater cost and reduce the turns and use thicker wire relative to the N35 design.

For a given gap length the flux depends on area and magnet grade. The other option is use thicker magnets a wider air gap and then you have more winding space for the same number of turns. There are so many choices but it is wiser in the end to stick to something tried and tested unless you really understand the whole concept, so many things interact and just changing one thing usually messes up several others.

Flux

imsmooth · « **Reply #11 on:** February 15, 2009, 05:51:46 PM »

I am not set on round or rectangular. I am going to order the steel disc on Dan's website. I just haven't decided on the number of magnets, coils and turns.

Let's assume I am going to use a 2" x 1" x 1/2" N50 magnet. Would there be a significant loss of flux if I use a 2" x 1" x 1/4" N50? The surface area is the same.

Flux · « **Reply #12 on:** February 16, 2009, 12:45:20 AM »

To get the same flux with 1/4" thick magnets you need half the air gap. That means less than half stator thickness and hence less than 1/2 wire csa for the same turns so your output will be well down.

If you keep the same air gap the flux density will fall and the end result will be that you need about twice the turns, again thinner wire and more resistance.

I may have confused you by saying that it all depends on area but I did clearly state that this was true for the same air gap.9or at least magnet thickness) In the end if you change all the parameters it comes down to magnet volume and grade.

I suspect that 3/4 thick N40 would be more cost effective than N50 at 1/2" but don't forget you are altering the air gap and you may be causing other mechanical issues depending on how you get the extra gap ( more overhang on allthered? comes to mind).

Flux

TomW · « **Reply #13 on:** February 16, 2009, 04:18:28 AM »

smooth;

After I posted this another user cited this story of a start to finish thread from about 6 months back:

http://www.fieldlines.com/story/2008/8/25/12258/0492

Just FYI on how effective putting it all in one thread can be.

Tom

imsmooth · « **Reply #14 on:** February 17, 2009, 07:11:21 AM »

A Neo grade35 has 12100; a grade48 has 14500. If everything else is the same, how can one use this information to determine the increased voltage a coil will produce?

Flux · « **Reply #15 on:** February 17, 2009, 11:45:02 AM »

In a closed magnetic circuit with the figures you give the N48 gives a 20% increase in flux. when working at the best air gap the gain in voltage will again be about 20%.Things are complicated a bit by the fact that the best gap changes with magnet grade but no by that much so roughly a 20% increase in voltage, which lets you reduce turns and use thicker wire so in the end you gain from reduced resistance.

Flux

imsmooth · « **Reply #16 on:** February 17, 2009, 12:38:09 PM »

So is it fair to say that the gauss ratio approximates the ratio of induced voltage?

gizmo · « **Reply #17 on:** February 20, 2009, 04:58:20 PM »

Before you give up on the F&P and build a whole new machine, why dont you try using a standard ceramic hub and do the capacitor conversion. It will give near to the power you had from the neo conversion, and be reliable.

I've heard of several F&P neo conversion's cracking or warping, that plastic just isn't up to the job I think.

Glenn

imsmooth · « **Reply #18 on:** February 20, 2009, 05:30:41 PM »

The crack wasn't from warping. It was from the impact of the tail on the blades when it was doing 400-500 rpm.

imsmooth · « **Reply #19 on:** February 20, 2009, 08:02:17 PM »

I'll try here.

Is there a difference with the voltage, current or heating issues with these coil comparisons? A and B are the same diameter, but the inner diameter of B is greater than A. Same turns. Magnetic flux goes from L to R across center.

D is taller than C, but has same turns.

I am guessing that there is little difference with A and B. I would think that with more copper mass, D heats up less.

News:

Author Topic: Going towards axial flux (Read 3896 times)